記述言語：英語   掲載種別：研究論文（学術雑誌）

Building an accurate map is essential for autonomous robot navigation in the environment without GPS. Compared with single-robot, the multiple-robot system has much better performance in terms of accuracy, efficiency and robustness for the simultaneous localization and mapping (SLAM). As a critical component of multiple-robot SLAM, the problem of map merging still remains a challenge. To this end, this article casts it into point set registration problem and proposes an effective map merging method based on the context-based descriptors and correspondence expansion. It first extracts interest points from grid maps by the Harris corner detector. By exploiting neighborhood information of interest points, it automatically calculates the maximum response radius as scale information to compute the context-based descriptor, which includes eigenvalues and normals computed from local structures of each interest point. Then, it effectively establishes origin matches with low precision by applying the nearest neighbor search on the context-based descriptor. Further, it designs a scale-based corresponding expansion strategy to expand each origin match into a set of feature matches, where one similarity transformation between two grid maps can be estimated by the Random Sample Consensus algorithm. Subsequently, a measure function formulated from the trimmed mean square error is utilized to confirm the best similarity transformation and accomplish the coarse map merging. Finally, it utilizes the scaling trimmed iterative closest point algorithm to refine initial similarity transformation so as to achieve accurate merging. As the proposed method considers scale information in the context-based descriptor, it is able to merge grid maps in diverse resolutions. Experimental results on real robot datasets demonstrate its superior performance over other related methods on accuracy and robustness.

DOI： 10.1145/3403948

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85116216543&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

With the construction and improvement of 5G infrastructure, more devices choose to access the Internet to achieve some functions. People are paying more attention to information security in the use of network devices. This makes lightweight block ciphers become a hotspot. A lightweight block cipher with superior performance can ensure the security of information while reducing the consumption of device resources. Traditional optimization tools, such as brute force or random search, are often used to solve the design of Symmetric-Key primitives. The metaheuristic algorithm was first used to solve the design of Symmetric-Key primitives of SKINNY. The genetic algorithm and the simulated annealing algorithm are used to increase the number of active S-boxes in SKINNY, thus improving the security of SKINNY. Based on this, to improve search efficiency and optimize search results, we design a novel metaheuristic algorithm, named particle swarm-like normal optimization algorithm (PSNO) to design the Symmetric-Key primitives of SKINNY. With our algorithm, one or better algorithm components can be obtained more quickly. The results in the experiments show that our search results are better than those of the genetic algorithm and the simulated annealing algorithm. The search efficiency is significantly improved. The algorithm we proposed can be generalized to the design of Symmetric-Key primitives of other lightweight block ciphers with clear evaluation indicators, where the corresponding indicators can be used as the objective functions.

DOI： 10.1145/3417296

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85116286304&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

The development of various fields of computer is constantly improving people's lives. In recent years, the most obvious is the development of Artificial Intelligence (AI). Computer vision is applied to face recognition, natural language processing is applied to speech recognition, and machine learning is applied to data. Analysis and other typical application scenarios (He et al., 2019; Hosny et al., 2018; Yu et al., 2018 [1,2,3]). As the software part of robotics technology, artificial intelligence has become more and more mature. Recent research on robotics technology mainly centers on drones and assembly robots in intelligent manufacturing (Zhong et al., 2017; Egger and Masood, 2020; Chen, 2020 [4,5,6]). In addition to the above application scenarios, unmanned driving as the most important part of intelligent transportation is also a product of the combination of artificial intelligence and robots. In recent years, both academic and industrial aspects have conducted in-depth and extensive research on unmanned driving (Chen et al., 2019; Zhao et al., 2019 [7,8]). Moreover, with the gradual commercial use of the fifth-generation mobile communication technology (5G), theoretical research on the sixth-generation mobile communication technology (6G) has also begun. Stronger mobile communication technology will make the information transmission in the intelligent transportation system more stable. It also improves the reliability of Intelligent Transportation Systems (ITS), which is the most important property. We believe that 5G/6G and artificial intelligence will be the two core technologies of the future intelligent transportation system, and these two fields will also receive extensive attention, which is very likely to produce breakthrough research results, so this paper The purpose is to summarize the current development of these two fields, and summarize the cross-research progress between them, and then discuss the bottleneck of the current development of intelligent transportation system and point out the future research direction.

DOI： 10.1016/j.compeleceng.2021.107372

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85115800825&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

In this paper, we propose an instance segmentation method uses a single-stage detector. Compared to the two-stage method, the single-stage method is simpler and easier to train. Not rely on the traditional region proposal, it directly uses pixels, which reduces the complexity of the network and significantly increases the speed. Our segmentation method is based on anchor boxes, which performs multi-scale detection by setting anchors of different sizes on multi-scale feature maps. We add a new branch to the prediction head to generate prototype masks and mask coefficients, then linearly combine them to generate mask. In our experiments, the proposed model had better performance, we got 35.12 fps on a single NVIDIA GEFORCE GTX 2080 GPU, which proves that our method is simple, effective, and faster.

DOI： 10.1016/j.compeleceng.2021.107464

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85115304542&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

In the intelligent transportation system, the session data usually represents the users' demand. However, the traditional approaches only focus on the sequence information or the last item clicked by the user, which cannot fully represent user preferences. To address this issue, this paper proposes an Multi-aspect Aware Session-based Recommendation (MASR) model for intelligent transportation services, which comprehensively considers the user's personalized behavior from multiple aspects. In addition, it developed a concise and efficient transformer-style self-attention to analyze the sequence information of the current session, for accurately grasping the user's intention. Finally, the experimental results show that MASR is available to improve user satisfaction with more accurate and rapid recommendations, and reduce the number of user operations to decrease the safety risk during the transportation service.

DOI： 10.1109/TITS.2020.2990214

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85110861948&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（学術雑誌）

Despite the great performance in layout analysis tasks made by semantic segmentation, they usually need a large number of annotated images for training and are difficult to learn a new category which is absent in the training categories. Meta-learning and few-shot segmentation have been developed to solve the above two difficulties. In this paper, we propose a novel method dubbed Few-Shot Prototype Alignment Regularization Network (FS-PARN). The FS-PARN method is inspired by recent studies in both metric learning and few-shot segmentation, which just need a few annotated images to solve the above two difficulties. Our FS-PARN method can make better use of the information of the support set by metric learning and have a better effect on image segmentation. It learns classification prototype within an embedding space and then completes pixel classification by matching each pixel on the query image with the learned prototype. In addition to obtaining high-quality prototypes through metric learning methods, our FS-PARN method also introduces prototype alignment regularization between support and query sets to make segmentation better. Notably, our FS-PARN model achieves the mean-IoU score of 28.8% and 31.7% on the practical document image datasets, i.e. PASCAL-5i, DSSE-200, and Layout Analysis Dataset, for 1-shot and 5-shot settings respectively.

DOI： 10.1016/j.patcog.2021.107882

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85101343657&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Understanding the surrounding environment is crucial for autonomous vehicles to make correct driving decisions. In particular, urban scene segmentation is a significant integral module commonly equipped in the perception system of autonomous vehicles to understand the real scene like a human. Any missegmentation of the driving scenario can potentially result in uncontrollable consequences such as serious accidents or the exception of the perception system. In this article, we investigate the vulnerability of the popular scene segmentation models designed with the backbones of deep neural networks (DNNs), which have been shown to be sensitive to adversarial attacks. Specifically, we propose an iterative projected gradient-based attack method that can effectively fool several DNN-based segmentation models with a remarkably higher attacking successful rate, and much smaller adversarial perturbations. Moreover, we also develop an adversarial training algorithm with min-max optimization style to enrich the robustness of the scene segmentation models. Extensive experiments on the Cityscape benchmark dataset consisting of large-scale urban scene images for autonomous vehicles demonstrate the effectiveness of our proposed attack method, as well as the benefit of the adversarial training scheme for the scene segmentation models.

DOI： 10.1109/TII.2020.3024643

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85102389729&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Road scene semantic segmentation often requires a deeper neural network to obtain higher accuracy, which makes the segmentation model more complex and slower. In this paper, we use shallow neural networks to achieve semantic segmentation for intelligent transportation system. Specifically, we propose a lightweight semantic segmentation model. First, the image features are extracted by using a simple superimposed convolutional layer and the three branches of ResNet and optimized by the attention mechanism. Then element multiplication and feature fusion are performed. Finally, the segmentation mask is obtained. Fewer convolutional layers and ResNet will not take up a lot of resources, we use the main resources to calculate the fusion between features. Experiments show that our method achieves high accuracy and comparable speed on the Cityscapes and CamVid datasets. On the Cityscapes dataset, our method achieves 75.0% mIoU, which is 0.2% higher than the better-performing BiSeNet.

DOI： 10.1016/j.compeleceng.2021.107155

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85105296221&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Currently, advanced communications and networks greatly enhance user experiences and have a major impact on all aspects of people's lifestyles in terms of work, society, and the economy. However improving competitiveness and sustainable vehicle network services, such as higher user experience, considerable resource utilization and effective personalized services, is a great challenge. Addressing these issues, this paper proposes a virtual network resource management based on user behavior to further optimize the existing vehicle communications. In particular, ensemble learning is implemented in the proposed scheme to predict the user's voice call duration and traffic usage for supporting user-centric mobile services optimization. Sufficient experiments show that the proposed scheme can significantly improve the quality of services and experiences and that it provides a novel idea for optimizing vehicle networks.

DOI： 10.1109/TITS.2020.2991766

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85104993954&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Semantic segmentation is widely applied in autonomous driving, in robotic picking, and for medical purposes. Due to the breakthrough of deep learning in recent years, the fully convolutional network (FCN)-based method has become the de facto standard in semantic segmentation. However, the simple FCN has difficulty in capturing global context information, since the local receptive field is small. Furthermore, there is a problem of low image resolution because of the existence of the pooling layer. In this paper, we address the shortcomings of the FCN by proposing a new architecture called WideSegNeXt, which captures the image context on various spatial scales and is effective in identifying small objects. In addition, there is little loss of position information, since there are no pooling layers in the structure. The proposed method achieves a mean intersection over union (MIoU) of 72.5% and a global accuracy (GA) of 92.4% on the CamVid dataset and achieves higher performance than previous methods without additional input datasets.

DOI： 10.1109/JSEN.2020.3008908

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85104614498&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（学術雑誌）

To better understand the road condition and make correct driving decisions, traffic sign recognition becomes a crucial component commonly equipped in the vision system of modern autonomous cars. The state-of-the-art traffic sign recognition models are designed with the backbones of deep neural networks (DNNs) since DNNs are powerful to extract more effective visual features that benefit recognition performance. As the recent studies on adversarial attacks have shown that DNNs are easy to be fooled by perturbed images and lead to misclassification, in this article, we explore the vulnerability of the DNN-based traffic sign recognition model. Most existing adversarial attack methods limitedly focus on the white-box attack on the recognition models whose underlying configurations (e.g., network architectures and parameters) are accessible. Differently, we propose a novel attacking method dubbed adaptive square attack (ASA) that can accomplish the black-box attack, i.e., bypassing the access the configurations of the recognition models. Specifically, the proposed ASA method employs an efficient sampling strategy that can generate perturbations for traffic sign images with fewer query times. Extensive experiments on the benchmark data set German traffic sign recognition benchmark with large-scale traffic sign images for autonomous cars show that our proposed ASA method is advanced to perform the black-box attack with high efficiency. Although the generated adversarial traffic sign images by the proposed ASA method are visually similar to the raw images with almost imperceptible differences, they can successfully lead to the misclassification of the state-of-the-art recognition model.

DOI： 10.1109/JIOT.2020.3016145

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85104075516&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

With the development of cloud computing, the storage and processing of massive visual media data has gradually transferred to the cloud server. For example, if the intelligent video monitoring system cannot process a large amount of data locally, the data will be uploaded to the cloud. Therefore, how to process data in the cloud without exposing the original data has become an important research topic. We propose a single-server version of somewhat homomorphic encryption cryptosystem based on confused modulo projection theorem named CMP-SWHE, which allows the server to complete blind data processing without seeing the effective information of user data. On the client side, the original data is encrypted by amplification, randomization, and setting confusing redundancy. Operating on the encrypted data on the server side is equivalent to operating on the original data. As an extension, we designed and implemented a blind computing scheme of accelerated version based on batch processing technology to improve efficiency. To make this algorithm easy to use, we also designed and implemented an efficient general blind computing library based on CMP-SWHE. We have applied this library to foreground extraction, optical flow tracking, and object detection with satisfactory results, which are helpful for building smart cities. We also discuss how to extend the algorithm to deep learning applications. Compared with other homomorphic encryption cryptosystems and libraries, the results show that our method has obvious advantages in computing efficiency. Although our algorithm has some tiny errors (10-6) when the data is too large, it is very efficient and practical, especially suitable for blind image and video processing.

DOI： 10.1109/JIOT.2020.3015032

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85099558019&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

The deployment of a very large number of readers in a limited space may increase the probability of collision among radio-frequency identification (RFID) readers and reduce the dependability and controllability of Internet-of-Things (IoT) systems. Intelligent computing technologies can be used to realize intelligent management by scheduling resources to circumvent collision issues. In this article, an improved RFID reader anticollision model is constructed by modifying the measure index, introducing a constraint function, and simultaneously considering collisions among readers and between readers and tags. The dense deployment of large numbers of readers increases the number of variables to be encoded, resulting in a high-dimensional problem that cannot be effectively and efficiently solved by traditional algorithms. Accordingly, distributed parallel cooperative co-evolution particle swarm optimization (DPCCPSO) is proposed. The inertia weight and learning factors are adjusted during evolution, and an improved grouping strategy is presented. Moreover, various combinations of random number generation functions are tested. For improved efficiency, DPCCPSO is implemented with distributed parallelism. Experimental verification shows that the proposed novel algorithm exhibits superior performance to existing state-of-the-art algorithms, particularly when numerous RFID readers are deployed.

DOI： 10.1109/JIOT.2020.3033473

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85101715875&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（学術雑誌）

DOI： 10.1016/j.compeleceng.2021.107055

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85101921245&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：記事・総説・解説・論説等（その他）

DOI： 10.1007/s11036-020-01724-y

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85100068480&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

In the real world, 3D point cloud data is generally obtained by LiDAR scanning. However, objects in the real world are occluded from each other, which will cause the point cloud scanned by LiDAR to be partially missing. In this paper, we improve PF-Net (a learning-based point cloud completion network), which is better to obtain the feature of the point cloud. Specifically, our improved network is an encoder-decoder-discriminator structure, which can directly take the missing point cloud data as input without additional preprocessing. In the encoder, we use the ALL-MLP (ALL-Multi Layer Perceptron) method to extract features from the point cloud. It combines the features obtained by each convolution in the feature extraction process, and finally sends it to the decoder. The decoder generates a prediction for the missing part of the point cloud, and the discriminator feeds back the generated result to the decoder to produce a more realistic effect. Our experiments show that the improved network has better accuracy in most categories than the state-of-the-art methods, and generates a relatively complete point cloud with achieving the purpose of complementing missing point cloud data.

DOI： 10.1007/s11280-021-00938-8

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85114376063&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（学術雑誌）

Recently, 3D object detection based on deep learning has achieved impressive performance in complex indoor and outdoor scenes. Among the methods, the two-stage detection method performs the best; however, this method still needs improved accuracy and efficiency, especially for small size objects or autonomous driving scenes. In this paper, we propose an improved 3D object detection method based on a two-stage detector called the Improved Point-Voxel Region Convolutional Neural Network (IPV-RCNN). Our proposed method contains online training for data augmentation, upsampling convolution and k-means clustering for the bounding box to achieve 3D detection tasks from raw point clouds. The evaluation results on the KITTI 3D dataset show that the IPV-RCNN achieved a 96% mAP, which is 3% more accurate than the state-of-the-art detectors.

DOI： 10.1109/TITS.2021.3071790

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85107209244&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Vehicle logo detection (VLD) is a special and significant topic in object detection for vehicle identification system applications. Nevertheless, the range of the research and analysis for VLD are seriously narrow in the real complex scenes, although it's a critical role in the object detection of small sizes. In this paper, we make further analysis work toward vehicle logo recognition and detection in real-world situations. To begin with, we propose a new multi-class VLD dataset, called VLD-45 (Vehicle Logo Dataset), which contains 45000 images and 50359 objects from 45 categories respectively. Our new dataset provides several research challenges involve in small sizes object, shape deformation, low contrast and so on. Meanwhile, we use 6 existing classifiers and 6 detectors to evaluate our dataset and show the baseline performance. According to the result, our dataset has very significant research value for the task of small-scale object detection. The dataset source: https://github.com/YangShuoys/VLD-45-B-DATASET-Detection

DOI： 10.1109/TITS.2021.3062113

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85103250557&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Hashing methods for efficient image retrieval aim at learning hash functions that map similar images to semantically correlated binary codes in the Hamming space with similarity well preserved. The traditional hashing methods usually represent image content by hand-crafted features. Deep hashing methods based on deep neural network (DNN) architectures can generate more effective image features and obtain better retrieval performance. However, the underlying data structure is hardly captured by existing DNN models. Moreover, the similarity (either visually or semantically) between pairwise images is ambiguous, even uncertain, to be measured in the existing deep hashing methods. In this article, we propose a novel hashing method termed deep fuzzy hashing network (DFHN) to overcome the shortcomings of existing deep hashing approaches. Our DFHN method combines the fuzzy logic technique and the DNN to learn more effective binary codes, which can leverage fuzzy rules to model the uncertainties underlying the data. Derived from fuzzy logic theory, the generalized hamming distance is devised in the convolutional layers and fully connected layers in our DFHN to model their outputs, which come from an efficient xor operation on given inputs and weights. Extensive experiments show that our DFHN method obtains competitive retrieval accuracy with highly efficient training speed compared with several state-of-the-art deep hashing approaches on two large-scale image datasets: CIFAR-10 and NUS-WIDE.

DOI： 10.1109/TFUZZ.2020.2984991

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85098868387&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Autonomous driving relies on trusty visual recognition of surrounding objects. Few-shot image classification is used in autonomous driving to help recognize objects that are rarely seen. Successful embedding and metric-learning approaches to this task normally learn a feature comparison framework between an unseen image and the labeled images. However, these approaches usually have problems with ambiguous feature embedding because they tend to ignore important local visual and semantic information when extracting intra-class common features from the images. In this paper, we introduce a Semantic-Aligned Attention (SAA) mechanism to refine feature embedding and it can be applied to most of the existing embedding and metric-learning approaches. The mechanism highlights pivotal local visual information with attention mechanism and aligns the attentive map with semantic information to refine the extracted features. Incorporating the proposed mechanism into the prototypical network, evaluation results reveal competitive improvements in both few-shot and zero-shot classification tasks on various benchmark datasets.

DOI： 10.1109/TITS.2021.3127632

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85120556998&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（学術雑誌）

Since legal data contains both structured data and unstructured data, it is a great challenge to implement machine reading comprehension technology in empirical analysis of law. This paper proposes a multi-tasking reading for intelligent legal services, which applies statistical analysis and machine reading comprehension techniques, and can process both structured and unstructured data. At the same time, this paper proposes a machine reading comprehension model that can perform multi-task learning, LegalSelfReader, which can solve the problem of diversity of questions. In the experiment of the legal reading comprehension dataset CJRC, the model proposed in this paper is far superior to the two classic models of BIDAF and Bert in three evaluation indicators. And our model is also better than some models published by HFL(Harbin Institute of Technology and iFly Joint Lab), and has also achieved lower consumption in training costs. At the same time, in the experiment of visualizing the attention value, it also demonstrates that the model proposed in this paper has a stronger ability to extract evidence.

DOI： 10.1016/j.future.2020.07.001

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85087901615&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Compared with traditional ineffective methods, such as acquiring more spectrum and deploying more base stations, edge caching is a highly promising solution for increased data flow needs and has attracted considerable attention. However, owing to the lack of careful consideration of cached data, existing related methods neither reduce network load nor improve the quality of experience. In this study, we propose a proactive sequence-aware content caching strategy (PSAC). Specifically, for general content at the network edge and content with sequential features, PSAC_gen (based on a convolutional neural network) and PSAC_seq (based on an attention mechanism that can automatically capture sequential features), respectively, are proposed to implement proactive caching. Experiments demonstrate that the proposed deep learning content caching method can effectively improve user experience and reduce network load.

DOI： 10.1109/TNSE.2020.2990963

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85095338501&origin=inward

記述言語：英語   掲載種別：記事・総説・解説・論説等（その他）

DOI： 10.1007/s11036-019-01428-y

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85078588045&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

Big data-driven deep learning methods have been widely used in image or video segmentation. The main challenge is that a large amount of labeled data is required in training deep learning models, which is important in real-world applications. To the best of our knowledge, there exist few researches in the deep learning-based visual segmentation. To this end, this paper summarizes the algorithms and current situation of image or video segmentation technologies based on deep learning and point out the future trends. The characteristics of segmentation that based on semi-supervised or unsupervised learning, all of the recent novel methods are summarized in this paper. The principle, advantages and disadvantages of each algorithms are also compared and analyzed.

DOI： 10.1109/COMPSAC48688.2020.00-84

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85094163880&origin=inward

記述言語：英語   掲載種別：記事・総説・解説・論説等（その他）

DOI： 10.1007/s11036-019-01265-z

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85065399005&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Owing to characteristics of the flow and the variability, it is extremely difficult to achieve the stability and the transparency of the underwater tele-operation system. In practice, the accurate force may not easily be acquired due to model uncertainties, the time delay and external disturbances. In order to enhance the stability and the transparency of the underwater tele-operation, an adaptive neural fuzzy inference system disturbance observer-based impedance control is proposed to both the master side and slave side. The learning algorithm of the adaptive neural fuzzy inference system network and the disturbance observer may simultaneously suppress model uncertainties of the nonlinear system and disturbances of external underwater environment. Concerning the time delay, the stability is analyzed by Lyapunov theorem. Numerical simulations are performed and results demonstrate the effective performance of the proposed method.

DOI： 10.1016/j.asoc.2020.106194

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85080149160&origin=inward

記述言語：英語   掲載種別：記事・総説・解説・論説等（その他）

DOI： 10.1007/s00521-020-04886-8

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85082547838&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Artificial bee colony has received much attention in recent years as a competitive population-based optimization algorithm. However, its slow convergence speed and one-dimensional search strategy limit it from demonstrating advantage in separable functions. To address these concerning issues, this paper introduces a coevolution framework into ABC and designs a global best leading artificial bee colony algorithm with an improved strategy to accelerate its convergence and conquer the dependency of dimension separately. A set of classical and Congress on Evolutionary Computation 2015 benchmark functions are adopted for validating the efficiency of our algorithm. In addition, in order to show the practicality of our algorithm, a robot path-planning problem is tested, and our algorithm still achieves superior results.

DOI： 10.1016/j.asoc.2019.106037

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85077511891&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

In the process of network interaction, CAPTCHA (completely automated public Turing test to tell computers and humans apart) is usually required. At present, the main method for differentiating between humans and machines is to use CAPTCHA, which includes images and text. Users need to interface with CAPTCHA for verification, which prolongs the interaction step. The interactive experience is poor, and the process of verification is not concealed. To simplify the verification process, optimize the interaction behavior and increase the security of verification, this paper proposes a CAPTCHA method based on a user's mouse operation behavior, and carries out a study on cursor trajectory recognition. This method can reduce the manual procedure and improve the user interactive experience. In this article, we propose an ensemble learning algorithm model based on sliding sampling to classify and recognize cursor trajectory data. Experiments show that the classification performance of the model is reliable for recognition based on human-machine cursor trajectories, which is a new concept for CAPTCHA.

DOI： 10.1109/MCE.2019.2936631

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85076352908&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

In this article, a method to reduce artifacts on temporal difference images is introduced. The proposed method uses a nonrigid registration method for ground glass opacification (GGO), which is light in concentration and difficult to detect early. In this method, global matching, local matching, and three-dimensional (3D) elastic matching are performed on the current and previous images, and an initial temporal subtraction image is generated. After that, we use an Iris filter, which is the gradient vector concentration degree filter, to determine the initial GGO candidate regions and use supervoxel and graph cuts to segment region of interest in the 3D images. For each extracted region, a support vector machine is used to reduce the oversegmentation. The voxel matching is applied to generate the final temporal difference image, emphasizing the GGO regions while reducing the artifact.

DOI： 10.1109/MCE.2019.2941468

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85076374234&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Vehicle detection plays an effective and important role in traffic safety, which has attracted extensive attention from both academic and industry. Deep learning has made significant breakthroughs in vehicle detection application. The Single Shot Detector (SSD) algorithm, which is one of the object detection algorithms, is used to detect vehicles. However, its main challenge is that the computing complexity and low accuracy. In this paper, an improved vehicle detection algorithm based on SSD is proposed to improve accuracy, especially for small vehicles detection. We add an Inception block to the extra layer in the SSD before the prediction to improve its performance. Then we use a new method that is more suitable for vehicle detection to set the scales and aspect ratios of the default bounding boxes, which benefits position regression and maintains the fast speed. The validity of our algorithm is verified on KITTI and UVD datasets. Compared with SSD, our algorithm achieves a higher mean average precision (mAP), while maintaining a fast speed.

DOI： 10.1007/s12652-020-02085-w

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85087342104&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Image segmentation with language referring expression can complete object segmentation based on expression text. Existing image segmentation methods show good results on high-performance computers, but most robot systems need real-time and high accuracy. At present, most methods cannot meet these requirements well. Therefore, we propose a high-precision and real-time deep learning network that integrates the two tasks of image segmentation with language referring expression and referring expression comprehension and then treats them as two branches. Specifically, the proposed network first merges the two tasks. The feature maps of different scales extracted by each branch are fed back to the two branches to obtain prediction results. These two tasks promote and restrict each other. Experiments show that our method has better real-time performance and higher accuracy than existing methods.

DOI： 10.1109/JSEN.2020.3041046

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85097962167&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Image super resolution is a vital research topic in the field of computer vision. It aims to reconstruct high resolution images from low resolution images. Although the conventional image super resolution methods have achieved good performance and effect, there are still have some issues, e.g., the high-frequency details information is insufficient, and the reconstruction process will bring additional noise, and most basic interpolation techniques produce blurry results. To settle the problems mentioned above, we consider combining the deep learning method with the frequency domain fusion method. In this paper, a novel single image super resolution method based on SRFeat network and wavelet fusion is proposed. First, the training image is taken as the input of the backbone SRFeat network, then the generative adversarial network training is carried out. Then, the up-sampling is utilized to obtain the coarse super resolved image. Finally, the output image after the network training is combined with the up-sampling image of the low-resolution image by Wavelet fusion to obtain the final result. Without increasing the depth of the network and the redundant parameters, the proposed method can achieve better reconstruct result. The experimental results show that the proposed method can not only reduce the probability of image distortion, but recover the global information of the reconstructed image and remove the noise brought by the reconstruction process. The PSNR value of the proposed method is improved 0.3 dB, and the SSIM is improved 0.02.

DOI： 10.1007/s12652-020-02065-0

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85084968696&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Recently, cloud storage and processing have been widely adopted. Mobile users in one family or one team may automatically backup their photos to the same shared cloud storage space. The powerful face detector trained and provided by a 3rd party may be used to retrieve the photo collection which contains a specific group of persons from the cloud storage server. However, the privacy of the mobile users may be leaked to the cloud server providers. In the meanwhile, the copyright of the face detector should be protected. Thus, in this paper, we propose a protocol of privacy preserving face retrieval in the cloud for mobile users, which protects the user photos and the face detector simultaneously. The cloud server only provides the resources of storage and computing and cannot learn anything of the user photos and the face detector. We test our protocol inside several families and classes. The experimental results reveal that our protocol can successfully retrieve the proper photos from the cloud server and protect the user photos and the face detector.

DOI： 10.1007/s00500-019-03834-6

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85061804024&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（学術雑誌）

With the development of AI technology, the application of AI will greatly change and influence people's daily lives. While AI technology brings great convenience to people's lives, people have shifted their focus from the physical world to the spiritual world, so there is an increasing demand for emotional services. As a result, emotional AI systems and emotional calculation are favored by many scholars nowadays. However, the existing emotional AI work mainly focuses on improving the accuracy of emotion recognition, lacking personalized emotional services for users. Therefore, in this article, the authors propose AI-EmoCom, which casts emotion as a communication medium in the network and makes the emotional communication system more intelligent by combining it with AI technology. We applied the AI-enabled emotional communication system to the field of unmanned driving, and proposed “people-centered” hybrid driving to reduce the incidence of traffic accidents to a greater extent. We also apply AI-enabled emotional communication to emotional social robots to provide users with personalized service emotion. Then the system architecture for the AI-enabled emotion communication is introduced in detail, and the no-tag learning model of dataset labeling and processing as well as the AI algorithm model for emotion recognition are elaborated in detail, and experiments are done to verify the interactive delay in the AI-enabled emotion communication system and accuracy of emotion recognition.

DOI： 10.1109/MNET.001.1900070

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85096960826&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（学術雑誌）

With the global popularity of Internet of Things (IoT) technology, increasingly numbers of digital mobile products have been developed, and they have increased the productivity of people's daily lives. These electronic products are used in all aspects of life, such as medical care, office life, home services, and sports. However, most of these products are designed for healthy people with high literacy rates. For disabled people, these products cannot be widely used. In this paper, new, differently shaped touch sensors are proposed for body sensor network-based devices. This touch sensor can be formed into any shape because of the use of conductive fabric adhesive tape as a switch. That property is why the sensor can change positions in the body sensor network in which the human body is used as a trigger to safely activate the touch switch. The number of switch sensors can easily be increased or decreased without changing the wiring of the central controller. The number of sensors in a switch sensor system is greater than that in other touch switch systems, and the accuracy is higher.

DOI： 10.1016/j.comcom.2019.07.019

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85069818135&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

At present, the aging population is growing in Japan. Along with that, the expectation for the utilization of welfare equipment is increasing. Electric wheelchair, a convenient transportation tool, is popularized rapidly. On the other hand, accidents have occurred, and the dangers for driving are pointed out. Therefore, it needs to improve accident factors, reduce accidents and improve the convenience of electric wheelchair by automation. Environmental recognition is necessary for the development of autonomous electric wheelchair. Environmental recognition includes self-position estimation, recognition of sidewalks, crosswalks and traffic lights, moving object prediction, etc. In order to solve these various problems, this paper examines the segmentation of sidewalks, crosswalks and traffic lights. We develop the WideSeg that is one of semantic segmentation algorithms applying convolutional neural networks (CNN).

DOI： 10.23919/ICCAS47443.2019.8971608

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85079103064&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

Recently, a new research field 'Computational Ethology' is attracting much attention from not only biologists but also computer scientists because advances in computer technology enabled to automate the measurement and the analysis of animal behavior. Especially, analyzing communications performed by honeybee workers in their hive is one of the most important and interesting issue in ethological research area to reveal a mechanism of honeybee's language. However, these analyses have been usually conducted by manually extracting honeybee's walking trajectories from recorded long-Time video data. For a systematic and theoretical analysis of honeybee's communication, we have developed an automatic tracking algorithm of multiple honeybees using image processing and constructed an automatic recording system for long-Term tracking of honeybee behaviors with Radio Frequency Identification (RFID) sensors and high-resolution camera modules using multiple small-size single board computers, Raspberry Pi. Using this system, we conducted recording experiments from 6:30 am to 7:30 pm during one month once or twice per a year from 2015 to 2018. In this paper, first we show the overview of a behavior monitoring system of honeybee in hive. Next, we explain the simultaneous tracking algorithm we proposed. Finally, we show the experimental results and confirm the system capabilities.

DOI： 10.1109/JCSSE.2019.8864160

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85074229586&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

The scientific and technological revolution of the Internet of Things has begun in the area of oceanography. Historically, humans have observed the ocean from an external viewpoint in order to study it. In recent years, however, changes have occurred in the ocean, and laboratories have been built on the sea floor. Approximately 70.8 percent of the Earth's surface is covered by oceans and rivers. The Ocean of Things is expected to be important for disaster prevention, ocean resource exploration, and underwater environmental monitoring. Unlike traditional wireless sensor networks, the Ocean Network has its own unique features, such as low reliability and narrow bandwidth. These features will be great challenges for the Ocean Network. Furthermore, the integration of the ocean network with artificial intelligence has become a topic of increasing interest for oceanology researchers. The cognitive ocean network (CONet) will become the mainstream of future ocean science and engineering developments. In this article, we define the CONet. The contributions of the article are as follows: a CONet architecture is proposed and described in detail; important and useful demonstration applications of the CONet are proposed; and future trends in CONet research are presented.

DOI： 10.1109/MWC.2019.1800325

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85061026040&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

As it combines AI and IoT, autonomous driving has attracted a great deal of attention from both academia and industry because of its benefits to the economy and society. However, ultra-low delay and ultra-high reliability cannot be guaranteed by individual autonomous vehicles with limited intelligence and the existing architectures of the Internet of Vehicles. In this article, based on a cloud/fog-computing pattern and the IoT AI service framework, we propose a cross-domain solution for auto-driving. In contrast to existing studies, which mainly focus on communication technologies, our solution achieves intelligent and flexible autonomous driving task processing and enhances transportation performance with the help of the Cognitive Internet of Vehicles. We first present an overview of the enabling technology and the architecture of the Cognitive Internet of Vehicles for autonomous driving. Then we discuss the autonomous driving Cognitive Internet of Vehicles specifically from the perspectives of what to compute, where to compute, and how to compute. Simulations are then conducted to prove the effect of the Cognitive Internet of Vehicles for autonomous driving. Our study explores the research value and opportunities of the Cognitive Internet of Vehicles in autonomous driving.

DOI： 10.1109/MNET.2019.1800339

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85068858755&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Recent progress has witnessed continued attention in discriminative correlation filter (DCF) tracking algorithms due to its high-efficiency. However, the existing DCF inevitably introduces some cyclic repetitions in learning and detection, which might lead to the heavy drift problem encountered in significant appearance variants owing to occlusion, deformation and motion blur. In this paper, we propose a dilated-aware discriminative correlation filter framework for visual tracking, which fully exploits multi-scale receptive contextual information of correlation filter to mitigate the impact of unwanted boundary and model degradation. On the premise of nondestructive filtering structure, our method adopts a simple formulation based on Kronecker product over discriminative correlation filter. By hands of multiple dilated factors perceive the multi-level spatial receptive map on objects. The framework learns a reliable response map by the residual understanding of multiple factor-dilated correlations filters. Furthermore, experiment results in a recent comprehensive tracking benchmark demonstrate a promising performance of the proposed method subjectively and objectively compared with several state-of-the-art algorithms.

DOI： 10.1007/s11280-018-0555-4

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85045733986&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Image denoising is one of important tasks in image processing. Diffusion filter-based methods have been widely used for image denosing. In order to solve the staircase effect introduced by second-order diffusion filter, different fourth-order diffusion filters were proposed, which can achieve good performance on image denoising. However, fourth-order diffusion filter often suffers over-smoothness. In this paper, we propose an anisotropic fourth-order diffusion filter for astronomical image denoising, in which the eigenvalue of Hessian matrix is introduced for preserving edge information. The experimental results of denoising on several astronomical images show that the proposed method can achieve better performance than several other methods.

DOI： 10.1016/j.optlastec.2018.08.043

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85052867360&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

A Free-Ascending Tripod (FAT) was deployed at a water depth of 2100 m to measure the currents and sediment movement at the seafloor. FAT is used to better understand how and where deep-seafloor sediment moves and accumulates. We also use FATs to study deep-sea biology. In the images obtained by the camera, biological animals can hardly be distinguished. In this paper, we use image processing technology to uncover the real deep-sea scene. We propose four methods for improving the underwater image quality. First, we use the deep-sea optical imaging model to determine the properties of water in different sea areas and then remove the haze from underwater images using the underwater dual dark channel model. Next, we remove the footprint of artificial light through halo-estimation devignetting. Then, we obtain the real deep-sea scene color based on the color temperature of the camera and the inherent optical properties of water. Finally, we propose a semi-self-similarity-based super resolution for super-resolving the low-quality images. The experimental results demonstrate that the proposed algorithm outperforms the state-of-the-art methods.

DOI： 10.1016/j.optlastec.2018.05.034

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85047556931&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（学術雑誌）

In recent years, underwater image processing has been a focus of many studies. Most underwater image processing technologies are focused on descattering, absorption correction and reflection correction. For a deep sea mining machine, dust seriously affects visual acuity. To correct for this, a two-part dust removal approach is proposed. Underwater red-green minimum channel prior descattering is used to remove fine dust in the first stage. However, the impact of dust streaks on images is always undesirable. Consequently, we propose a further deep convolutional neural-network-based dust removal method. The experimental results show that the proposed method performs better in removing haze-like scatters and dust-like scatters.

DOI： 10.1016/j.optlastec.2017.09.017

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85029658908&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Feature tracking as one of the most important components of 3D reconstruction based on structure from motion (SFM) has attracted a wide range of attention from computer vision community. However, existing feature tracking methods often suffer from image distort, scale-change and varying lighting, then resulting many incorrect matches, namely outliers. Meanwhile, these methods are very costly to calculate. To defend this drawback, a fast and robust feature tracking (FRFT) is proposed for 3D reconstruction with SFM. Firstly, to save computational cost, the feature clustering method is used to cluster a big volume of image collection to some small ones to avoid some undesirable feature matches. Secondly, the union find set (UFS) method is used to achieve fast feature matching, this can furtherly save computation time of feature tracking. Thirdly, a geometry-constraint method is proposed to remove outlier from tracks produced by feature tracking method. Finally, a comprehensive evaluation is conducted to assess the proposed FRFT and the state-of-the-art methods. Experimental results show that the proposed FRFT method has the best performance on both efficiency and effectiveness.

DOI： 10.1016/j.optlastec.2018.05.036

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85048502703&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

3D reconstruction based on structure from motion is one of the most techniques to produce sparse point-cloud model and camera parameter. However, this technique heavily relies on feature tracking method to obtain feature correspondences, then resulting in a heavy computation burden. To speed up 3D reconstruction, in this paper, we design a novel GPU-accelerated feature tracking (GFT) method for large-scale structure from motion (SFM)-based 3D reconstruction. The proposed GFT method consists of GPU-based Gaussian of image (DOG) keypoint detector, RootSIFT descriptor extractor, k nearest matching, and outlier removing. Firstly, our GPU-based DOG implementation can detect thousands of keypoints in real-time, whose speed is 30 times faster than that of the CPU version. Secondly, our GPU-based RootSIFT descriptor can compute thousands of descriptors in real-time. Thirdly, our GPU-based descriptor matching is 10 times faster than that of the state-of-the-art methods. Finally, we conduct thorough experiments on different datasets to evaluate the proposed method. Experimental results demonstrate the effectiveness and efficiency of the proposed method.

DOI： 10.1016/j.optlastec.2018.08.045

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85053678405&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Deconvolution method has been widely used for spectral resolution enhancement. In order to preserve the detailed information and suppress noise better, trimmed φHS regularization and weighted φHS regularization are proposed in this paper. Then the semi-blind deconvolution methods with trimmed φHS regularization (SBD-THS) and with weighted φHS regularization (SBD-WHS) are presented. The results of deconvolving simulated degraded spectra and real experiment spectra demonstrate that SBD-THS and SBD-WHS can enhance spectral resolution effectively while estimating the parameter of blur kernel accurately. In particular, SBD-WHS can produce great performance on preserving local details and suppressing noise.

DOI： 10.1016/j.optlastec.2018.01.046

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85041019470&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

This paper addresses the problem of single image super-resolution (SR). In recent years, sparse representation based SR methods have been proposed and achieved great success. Traditional sparse representation based SR methods learn a unified high-resolution (HR) and low-resolution (LR) dictionary pair. All LR patches share the same dictionary pair to conduct reconstruction to get their corresponding HR patches. The reconstruction process introduces reconstruction error, which limits the performance of SR results. In this paper, to minimize the reconstruction error, we utilize a K-Nearest-Neighbours (KNN) based dictionary pair for each LR patch instead of the unified dictionary pair to conduct the reconstruction process. The KNN based dictionaries are selected among a huge dictionary which contains hundreds of millions of patch pairs. To speed up the KNN retrieval process, we adopt a binary encoding method which preserves local information for the LR patches, and retrieve the KNN of each LR patch in the Hamming space. Besides, since a large patch contains more structured information than a small patch, we utilize large patches instead of small ones as the atoms of dictionaries, which further improves the SR results. Experimental results demonstrate that our method outperforms the existed state-of-the-art methods, especially when the magnification factor is large or the image is blurred.

DOI： 10.1016/j.optlastec.2018.01.043

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85044367953&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Many hospitals are suffering from ineffective use of big data analytics with electronic health records (EHRs) to generate high quality insights for their clinical practices. Organizational learning has been a key role in improving the use of big data analytics with EHRs. Drawing on the knowledge-based view and big data lifecycle, we investigate how the three modes of knowledge can achieve meaningful use of big data analytics with EHRs. To test the associations in the proposed research model, we surveyed 580 nurses of a large hospital in China in 2019. Structural equation modelling was used to examine relationships between knowledge mode of EHRs and meaningful use of EHRs. The results reveal that know-what about EHRs utilization, know-how EHRs storage and utilization, and know-why storage and utilization can improve nurses' meaningful use of big data analytics with EHRs. This study contributes to the existing digital health and big data literature by exploring the proper adaptation of analytical tools to EHRs from the different knowledge mode in order to shape meaningful use of big data analytics with EHRs.

DOI： 10.1109/ACCESS.2019.2939158

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85077962015&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Neural architecture search (NAS) has significant progress in improving the accuracy of image classification. Recently, some works attempt to extend NAS to image segmentation which shows preliminary feasibility. However, all of them focus on searching architecture for semantic segmentation in natural scenes. In this paper, we design three types of primitive operation set on search space to automatically find two cell architecture DownSC and UpSC for semantic image segmentation especially medical image segmentation. Inspired by the U-net architecture and its variants successfully applied to various medical image segmentation, we propose NAS-Unet which is stacked by the same number of DownSC and UpSC on a U-like backbone network. The architectures of DownSC and UpSC updated simultaneously by a differential architecture strategy during the search stage. We demonstrate the good segmentation results of the proposed method on Promise12, Chaos, and ultrasound nerve datasets, which collected by magnetic resonance imaging, computed tomography, and ultrasound, respectively. Without any pretraining, our architecture searched on PASCAL VOC2012, attains better performances and much fewer parameters (about 0.8M) than U-net and one of its variants when evaluated on the above three types of medical image datasets.

DOI： 10.1109/ACCESS.2019.2908991

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85064574280&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（学術雑誌）

To monitor the sedimentary process and morphological evolution in the South China Sea, free-ascending deep-sea tripod (FDT) has been developed. This FDT was equipped with a deep-sea camera and landed on the sea floor at a depth of 2100 m. Although the FDT was equipped with an artificial light, the battery capacity limited the duration and intensity of light. Therefore, enhancing such low-illumination images to obtain clear visual effects is an important advancement for analyzing the geological evolution process. In this paper, an adaptive bright-color channel-based low-light underwater image-enhancement method and a denoising method are proposed to enhance such images and remove noise and artifacts. The experimental results demonstrated that the proposed method outperformed state-of-the-art methods.

DOI： 10.1109/ACCESS.2019.2897691

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85064689140&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（学術雑誌）

High-definition display technology for underwater images is of great significance for many applications, such as marine animal observation, seabed mining, and marine fishery production. The traditional underwater visual display systems have problems, such as low visibility, poor real-time performance, and low resolution, and cannot meet the needs of real-time high-definition displays in extreme environments. To solve these issues, we propose an underwater image enhancement method and a corresponding image super-resolution algorithm. To improve the quality of underwater images, we modify the Retinex algorithm and combine it with a neural network. The Retinex algorithm is used to defog the underwater image, and then, the image brightness is improved by applying gamma correction. Then, by combining with the dark channel prior and multilayer perceptron, the transmission map is further refined to improve the dynamic range of the image. In the super-resolution process, the current convolutional neural network reconstruction algorithm is only trained on the Y channel, which will lead to problems due to the insufficient acquisition of the color feature. Therefore, an image super-resolution reconstruction algorithm that is based on color features is proposed. The experimental results show that the proposed method improves the reconstruction effect of the image edges and texture details, increases the image clarity, and enhances the image color recovery.

DOI： 10.1109/ACCESS.2019.2925209

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85068923735&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（学術雑誌）

Image segmentation is a fundamental process in image processing that has found application in many fields, such as neural image analysis, and underwater image analysis. In this paper, we propose a novel fast level set method (FLSM)-based underwater image segmentation method to improve the traditional level set methods by avoiding the calculation of signed distance function (SDF). The proposed method can speed up the computational complexity without re-initialisation. We also provide a fast semi-implicit additive operator splitting (AOS) algorithm to improve the computational complex. The experiments show that the proposed FLSM performs well in selecting local or global segmentation regions.

DOI： 10.1504/IJCSE.2019.101879

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85072112667&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

Recently, consistent container transportation of roads and ships is mainstream of international freight transport. Because of various factors, automation of cargo handling work is required at the container terminal. Various causes are decrement of future labor force population by an increasing trend of container moving amount and declining birthrate and aging population. Therefore, this study presents the relative position of hanger and container measurement technology using Single Shot Multibox Detector (SSD) for the purpose of improvement of cargo handling work efficiency and unmanned container terminal. In the case of undetected by SSD, it will be detected using AKAZE feature. The proposed method is applied to 368 images of container gripping taken by a camera installed in a container crane. As a result, Interaction of Union (IoU) targeted for container gripping is 87.79%, and a detection rate is 94.57%.

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85060451358&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

In the task of object recognition, convolutional neural networks (CNNs) have achieved high performance. In addition, these CNNs are also applied to the field of semantic image segmentation. However, applying the classification models to semantic segmentation tasks has a problem, lack of global context and reduction in resolution. In this work, we propose global context module and high resolution path in order to solve above problems. By simply combining them with an existing classification model (wide residual networks), our methods yield high-accuracy segmentation models. Our proposed approaches produce competitive results, the mean intersection over union (IoU) 67.6% and global accuracy 91.1%, on CamVid test set.

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85060445374&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

In this paper, we propose a method to reduce artifacts on temporal difference images by improving the conventional method using a non-rigid registration method for ground glass opacification (GGO), which is light in concentration and difficult to detect early. In this method, global matching, local matching, and 3D elastic matching are performed on the current image and past image, and an initial temporal difference image is generated. After that, we use an Iris filter, which is the gradient vector concentration degree filter, to determine the initial GGO candidate regions and perform segmentation using SuperVoxel and Graph Cuts in which a superpixel is extended to three dimensions for each region of interest. For each extracted region, a support vector machine (SVM) is used to reduce the over-segmentation. Finally, in the method that greatly reduces artifacts other than the remaining GGO candidate regions, Voxel Matching is applied to generate the final temporal difference image, emphasizing the GGO regions while reducing the artifact. The resulting ratio of artifacts to lung volume is 0.101 with an FWHM of 28.3, which is an improvement over the conventional method and shows the proposed method’s effectiveness.

DOI： 10.1007/s11036-018-1111-2

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85053445997&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Large-scale 3D reconstruction based on structure from motion (SFM) has attracted much attention from the computer vision community. Local feature extraction, which is typically used to locate correspondences between images, is one of the most important components of SFM. In this paper, we propose a fast and robust local feature extraction method, called OOD, for SFM-based 3D reconstruction. First, the adaptive and generic accelerated segment test (AGAST) detector is used to detect keypoints, and the image moment is used to define an orientation for these keypoints. Second, a novel descriptor based on the oriented difference of Gaussians is proposed to describe the keypoints, which can be computed directly from the difference of Gaussian image. Finally, a comprehensive evaluation is conducted using several benchmark datasets. Our experimental results indicate that the OOD method outperforms state-of-the-art methods.

DOI： 10.1016/j.compeleceng.2018.08.012

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85052625995&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Convolutional networks are currently the most popular computer vision methods for a wide variety of applications in multimedia research fields. Most recent methods have focused on solving problems with natural images and usually use a training database, such as Imagenet or Openimage, to detect the characteristics of the objects. However, in practical applications, training samples are difficult to acquire. In this study, we develop a powerful approach that can accurately learn marine organisms. The proposed filtering deep convolutional network (FDCNet) classifies deep-sea objects better than state-of-the-art classification methods, such as AlexNet, GoogLeNet, ResNet50, and ResNet101. The classification accuracy of the proposed FDCNet method is 1.8%, 2.9%, 2.0%, and 1.0% better than AlexNet, GooLeNet, ResNet50, and ResNet101, respectively. In addition, we have built the first marine organism database, Kyutech10K, with seven categories (i.e., shrimp, squid, crab, shark, sea urchin, manganese, and sand).

DOI： 10.1007/s11042-017-4585-1

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85015699675&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

Human wearable helmet is a useful tool for monitoring the status of miners in the mining industry. However, there is little research regarding human emotion recognition in an extreme environment. In this paper, an emotional state evoked paradigm is designed to identify the brain area where the emotion feature is most evident. Next, the correct electrode position is determined for the collection of the negative emotion by the electroencephalograph (EEG) based on the international 10-20 system of electrode placement. And then, a fusion algorithm of the anxiety level is proposed to evaluate the person's mental state using the θ, α, and β rhythms of an EEG. Experiments demonstrate that the position Fp2 is the best electrode position for obtaining the anxiety level parameter. The most visible EEG changes appear within the first two seconds following stimulation. The amplitudes of the θ rhythm increase most significantly in the negative emotional state.

DOI： 10.1109/IWCMC.2018.8450382

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85053937610&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Unmanned aerial vehicles (UAVs) are used in many fields including weather observation, farming, infrastructure inspection, and monitoring of disaster areas. However, the currently available UAVs are prone to crashing. The goal of this paper is the development of an anomaly detection system to prevent the motor of the drone from operating at abnormal temperatures. In this anomaly detection system, the temperature of the motor is recorded using DS18B20 sensors. Then, using reinforcement learning, the motor is judged to be operating abnormally by a Raspberry Pi processing unit. A specially built user interface allows the activity of the Raspberry Pi to be tracked on a Tablet for observation purposes. The proposed system provides the ability to land a drone when the motor temperature exceeds an automatically generated threshold. The experimental results confirm that the proposed system can safely control the drone using information obtained from temperature sensors attached to the motor.

DOI： 10.1109/JIOT.2017.2737479

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85028968216&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Spectral resolution enhancement is essential for spectral analysis and assignment. In this study, a regularization term in form of a convex function φHS is included in the spectral deconvolution model to enhance spectral resolution. Based on the regularization term, a non-blind deconvolution (NBD) method is proposed, and to improve feasibility in practice, a semi-blind deconvolution (SBD) method is also presented. Simulation and experimental results demonstrate that both methods enhance spectral resolution effectively. When the blur kernel is known accurately, NBD achieves better performance than SBD. In other cases, the latter achieves better results.

DOI： 10.1016/j.compeleceng.2018.02.004

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85041954720&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（学術雑誌）

In recent years, unmanned aerial vehicle (UAV) technologies have rapidly developed. Drones, which are one type of UAV, are used in many industrial fields, such as photography, delivery and agriculture. However, a commercial drone can fly for only approximately 20 min on one charge. Furthermore, drones are prohibited from flying in some areas, and cannot be operated in bad weather. Due to the development of drone technologies, we must reduce energy consumption and achieve long-range movement. To overcome these limitations, we develop a new air–land amphibious car drone that can fly and requires less power consumption in land mode; this extends the range of mobility of the drone. Moreover, land mode can be used to pass through restricted areas or bad weather conditions by sliding. Furthermore, we develop a Convolutional Neural Network (CNN)-based algorithm for detecting the road in a captured scene. To more accurately segment the road region based on images from the equipped camera of the drone, we propose atrous spatial pyramid pooling (ASPP) ResNet blocks, instead of Resblocks, which were proposed by DeepLab. The experimental results demonstrate that the proposed method improves the pixel accuracy (PA) to 85.6% and achieves a mean Intersection over Union (mIoU) of 55.8%.

DOI： 10.1016/j.future.2018.02.036

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85043254725&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（学術雑誌）

In this paper, an identifying and classifying algorithm is proposed to solve the problem of recognizing objects accurately and effectively. First, via image preprocessing, initial images are obtained via denoising, smoothness, and image erosion. Then, we use granularity analysis and morphology methods to recognize the objects. For small objects identification and to analyze the objects, we calculate four characteristics of each cell: area, roundness, rectangle factor, and elongation. Finally, we segment the cells using the modified active contour method. In addition, we apply chromatic features to recognize the blood cancer cells. The algorithm is tested on multiple collected clinical cases of blood cell images. The results prove that the algorithm is valid and efficient when recognizing blood cancer cells and has relatively high accuracy rates for identification and classification. The experimental results also certificate the effectiveness of the proposed method for extracting precise, continuous edges with limited human intervention, especially for images with neighboring or overlapping blood cells. In addition, the results of the experiments show that this algorithm can accelerate the detection velocity.

DOI： 10.1007/s11042-017-4529-9

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85014610096&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Underwater optical images are usually influenced by low lighting, high turbidity scattering and wavelength absorption. To solve these issues, a great deal of work has been performed to improve the quality of underwater images. Most of them use the high-intensity LEDs for lighting to obtain the high contrast images. However, in high turbidity water, high-intensity LEDs cause strong scattering and absorption. In this paper, we propose a light field imaging approach for solving underwater imaging problems in a low-intensity light environment. As a solution, we tackle the problem of de-scattering from light field images by using deep convolutional neural networks with depth estimation. Furthermore, a spectral characteristic-based color correction method is used for recovering the color reduction. Experimental results show the effectiveness of the proposed method by challenging real-world underwater imaging.

DOI： 10.1016/j.future.2018.01.001

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85040510117&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（学術雑誌）

This paper describes a novel method of enhancing deep-sea optical images using guided local Laplacian filter. Absorption and are two major distortion issues for deep-sea optical imaging. While light traveling through the water, light rays are scattered and absorbed depending on the wavelength. Scattering is caused by large suspended particles, as in turbid water that contains abundant particles, which causes the degradation of the captured image. Absorption corresponds to the varying degrees of attenuation encountered by light traveling in water at different wavelengths that causing ambient underwater to be dominated by a bluish tone. Our key contributions are proposed include a novel deep-sea imaging model to compensate for the attenuation discrepancy along the propagation path and an effective underwater scene enhancement scheme. The recovered images are characterized by a reduced noised level, better exposure of the dark regions, and improved global contrast where the finest details and edges are enhanced significantly.

DOI： 10.1007/s11042-017-5300-y

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85055744325&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Infrared moving point target detection is a challenging task in infrared tracking systems. In this paper, an anisotropic spatial-temporal fourth-order diffusion filter (ASTFDF) is proposed for background prediction from infrared frame images. Using it, we can detect infrared point targets by subtracting the predicted background from the original image. Experiments in detecting targets from several different infrared image frame sequences indicated that the proposed ASTFDF can predict the background effectively and that the target can be detected precisely while achieving a higher background suppression factor (BSF) and signal-to-clutter ratio gain (SCRG) compared with state-of-the-art methods.

DOI： 10.1016/j.compeleceng.2018.05.009

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85046872186&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Detection of Alzheimer’s disease (AD) from magnetic resonance images can help neuroradiologists to make decision rapidly and avoid missing slight lesions in the brain. Currently, scholars have proposed several approaches to automatically detect AD. In this study, we aimed to develop a novel AD detection system with better performance than existing systems. 28 ADs and 98 HCs were selected from OASIS dataset. We used inter-class variance criterion to select single slice from the 3D volumetric data. Our classification system is based on three successful components: wavelet entropy, multilayer perceptron, and biogeography-base optimization. The statistical results of our method obtained an accuracy of 92.40 ± 0.83%, a sensitivity of 92.14 ± 4.39%, a specificity of 92.47 ± 1.23%. After comparison, we observed that our pathological brain detection system is superior to latest 6 other approaches.

DOI： 10.1007/s11042-016-4222-4

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85048115686&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（学術雑誌）

Optical underwater images often demonstrate low contrast, heavy scatter, and color distortion. Contrast enhancement methods have been proposed to solve these issues. However, such methods typically do not consider high-level inhomogeneous scatter removal and do not focus on real-scene color restoration. We proposed a hierarchical transmission fusion method and a color-line ambient light estimation method for image de-scattering from a single input image. Our proposed method can be summarized into three steps. Firstly, we take the dark channel as prior information to estimating the preliminary transmission and ambient light. In the second step, we then use color lines to estimate the refined ambient light in selected patches. The refined transmission is obtained by hierarchical transmission maps using maximum local energy-based fusion at different turbidity levels. We then use a joint normalized filter to obtain the final transmission. Finally, a chromatic color correction method and de-blurring algorithm are used to recover the scene color. Experimental results demonstrate that the accurate estimation of the depth map and ambient light by the proposed method can recover visually appealing images with sharp details.

DOI： 10.1007/s11036-017-0933-7

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85029692278&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Artificial intelligence (AI) is an important technology that supports daily social life and economic activities. It contributes greatly to the sustainable growth of Japan’s economy and solves various social problems. In recent years, AI has attracted attention as a key for growth in developed countries such as Europe and the United States and developing countries such as China and India. The attention has been focused mainly on developing new artificial intelligence information communication technology (ICT) and robot technology (RT). Although recently developed AI technology certainly excels in extracting certain patterns, there are many limitations. Most ICT models are overly dependent on big data, lack a self-idea function, and are complicated. In this paper, rather than merely developing next-generation artificial intelligence technology, we aim to develop a new concept of general-purpose intelligence cognition technology called “Beyond AI”. Specifically, we plan to develop an intelligent learning model called “Brain Intelligence (BI)” that generates new ideas about events without having experienced them by using artificial life with an imagine function. We will also conduct demonstrations of the developed BI intelligence learning model on automatic driving, precision medical care, and industrial robots.

DOI： 10.1007/s11036-017-0932-8

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85029688332&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Structure from motion (SFM) is an effective approach for reconstructing large-scale 3D scene frommultiple images. In this field,many local featuremethods have been proposed to detect feature point and compute descriptor. For designing a robust SFM system, how to select a good feature from existingmethods is an important problem. In this paper, we aim to help different users for making decision by an experimental way for large-scale 3D reconstruction where many high resolution images are captured. To this end, we make a comprehensive evaluation of several local features on the ground truth datasets. Experimental results show that SIFT and SURF have a better performance than that of some binary features such as ORB and BRISK.

DOI： 10.1007/s11042-018-5864-1

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85055753999&origin=inward

記述言語：英語   掲載種別：記事・総説・解説・論説等（その他）

In the original publication, the Fig. 3 was incorrectly presented. The white background of B44 in Fig. 3b should be removed. This figure was corrected in the original.

DOI： 10.1007/s11042-017-4391-9

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85014909931&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

In order to develop an artificial intelligence and computer-aided diagnosis system that assists neuroradiologists to interpret magnetic resonance (MR) images. This paper employed Hu moment invariant (HMI) as the brain image features, and we proposed a novel predator-prey particle swarm optimization (PP-PSO) algorithm used to train the weights of single-hidden layer neural-network (SLN). We used five-fold stratified cross validation (FFSCV) for statistical analysis. Our proposed HMI + SLN + PP-PSO method achieved a sensitivity of 96.00 ± 5.16%, a specificity of 98.57 ± 0.75%, and an accuracy of 98.25 ± 0.65% for the DA-160 dataset, and yields a sensitivity of 97.14 ± 2.33%, a specificity of 97.00 ± 0.34%, and an accuracy of 97.02 ± 0.33% for the DA-255 dataset. Our method performs better than six state-of-the-art approaches.

DOI： 10.1007/s11042-016-4242-0

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85006372795&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

In recent years, unmanned aerial vehicle (UAV) technologies are developing rapidly. Drone, one type of the UAVs, is used in many industrial fields, such as photography, delivery and agriculture. However, the commercial drone can flying only about 20 minutes at one charge. Furthermore, the drone prohibits flying at the limited area, and it also can’t work in bad weather. Due to the development of drone technologies, we must reduce energy consumption, and realize high range movement. In order to solve these limitations, we develop a new type of drone, which has the function of flight and vehicle can move less power consumption. It extends high range of mobility to drone. Moreover, it can be used to pass through the limitation area or bad weather condition by sliding.

DOI： 10.4108/eai.28-9-2017.2273334

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85051034970&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Hyperspectral image (HSI) has been widely used in target detection and classification. However, various kinds of noise in HSIs affect the applications of HSIs. In this paper, we propose a low-rank (LR) tensor recovery model to remove noise. Considering that the HSI is a 3-D HSI data, and the underlying LR tensor property is used in the model. Then, according to the similarity of adjacent bands images, the regularization on the difference of adjacent bands images is considered. The experiments of removing noise from different noisy HSIs show that our method can achieve better performance on removing sparse noise, especially for strips removal.

DOI： 10.1109/ACCESS.2018.2876038

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85055017796&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

A signature is a useful human feature in our society, and determining the genuineness of a signature is very important. A signature image is typically analyzed for its genuineness classification; however, increasing classification accuracy while decreasing computation time is difficult. Many factors affect image quality and the genuineness classification, such as contour damage and light distortion or the classification algorithm. To this end, we propose a mobile computing method of signature image authentication (SIA) with improved recognition accuracy and reduced computation time. We demonstrate theoretically and experimentally that the proposed golden global-local (G-L) algorithm has the best filtering result compared with the methods of mean filtering, medium filtering, and Gaussian filtering. The developed minimum probability threshold (MPT) algorithm produces the best segmentation result with minimum error compared with methods of maximum entropy and iterative segmentation. In addition, the designed convolutional neural network (CNN) solves the light distortion problem for detailed frame feature extraction of a signature image. Finally, the proposed SIA algorithm achieves the best signature authentication accuracy compared with CNN and sparse representation, and computation times are competitive. Thus, the proposed SIA algorithm can be easily implemented in a mobile phone.

DOI： 10.1155/2018/5412925

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85049841734&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

Aerial unmanned vehicle is widely used in many fields, such as weather observation, framing, inspection of infrastructure, monitoring of disaster areas. However, the current aerial unmanned vehicle is difficult to avoid falling in the case of failure. The purpose of this article is to develop an anomaly detection system, which prevents the motor from being used under abnormal temperature conditions, so as to prevent safety flight of the aerial unmanned vehicle. In the anomaly detection system, temperature information of the motor is obtained by DS18B20 sensors. Then, the reinforcement learning, a type of machine learning, is used to determine the temperature is abnormal or not by Raspberrypi processing unit. We also build an user interface to open the screen of Raspberrypi on laptop for observation. In the experiments, the effectiveness of the proposed system to stop the operation state of drone when abnormality exceeds the automatically learned motor temperature. The experimental results demonstrate that the proposed system is possibility for unmanned flight safely by controlling drone from information obtained by attaching temperature sensors.

DOI： 10.1007/978-3-319-69877-9_32

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85033698798&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

Underwater optical images are usually influenced by low lighting, high turbidity scattering and wavelength absorption. In order to solve these issues, a great deal of work has been used to improve the quality of underwater images. Most of them used the high-intensity LED for lighting to obtain the high contrast images. However, in high turbidity water, high-intensity LED causes strong scattering and absorption. In this paper, we firstly propose a light field imaging approach for solving underwater depth map estimation problems in low-intensity lighting environment. As a solution, we tackle the problem of de-scattering from light field images by using deep convolutional neural fields in depth estimation. Experimental results show the effectiveness of the proposed method through challenging real world underwater imaging.

DOI： 10.1007/978-3-319-69877-9_33

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85033692294&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

In view of the consumption and distribution of resources in the world, the necessity of deep-sea mining video image processing is illustrated. Based on a large number of relevant literatures, this paper summarizes the research status of image processing methods for deep-sea mining observation video system, introduces the advantages of the improved median filter algorithm, the improved dark channel prior algorithm and the improved nonlocal mean denoising method. Some problems of the original methods are analyzed. These aim to provide reference for the optimization and improvement of image processing methods for deep-sea mining video observation.

DOI： 10.1007/978-3-319-69877-9_24

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85033683998&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Underwater cameras are widely used to observe the sea floor. They are usually included in autonomous underwater vehicles (AUVs), unmanned underwater vehicles (UUVs), and in situ ocean sensor networks. Despite being an important sensor for monitoring underwater scenes, there exist many issues with recent underwater camera sensors. Because of light’s transportation characteristics in water and the biological activity at the sea floor, the acquired underwater images often suffer from scatters and large amounts of noise. Over the last five years, many methods have been proposed to overcome traditional underwater imaging problems. This paper aims to review the state-of-the-art techniques in underwater image processing by highlighting the contributions and challenges presented in over 40 papers. We present an overview of various underwater image-processing approaches, such as underwater image de-scattering, underwater image color restoration, and underwater image quality assessments. Finally, we summarize the future trends and challenges in designing and processing underwater imaging sensors.

DOI： 10.1007/s11036-017-0863-4

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85018366927&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

There have been increased developments in ocean exploration using autonomous underwater vehicles (AUVs) and unmanned underwater vehicles (UUVs). However, the contrast of underwater images is still a major issue for application. It is difficult to acquire clear underwater images around underwater vehicles. Since the 1960s, sonar sensors have been extensively used to detect and recognize objects in oceans. Due to the principles of acoustic imaging, sonar-imaged images have many shortcomings, such as a low signal to noise ratio and a low resolution. Consequently, vision sensors must be used for short-range identification because sonars yield to low-quality images. This thesis will concentrate solely on the optical imaging sensors for ocean observing. Although the underwater optical imaging technology makes a great progress, the recognition of underwater objects also remains a major issue in recent days. Different from the common images, underwater images suffer from poor visibility due to the medium scattering and light distortion. First of all, capturing images underwater are difficult, mostly due to attenuation caused by light. The random attenuation of the light mainly causes the haze appearance along with the part of the light scattered back from the water. In particular, the objects at a distance of more than 10 m are almost indistinguishable because of absorption. Furthermore, when the artificial light is employed, it can cause a distinctive footprint on the seafloor. In this paper, we will analysis the recent trends of ocean exploration approaches.

DOI： 10.1007/978-3-319-46245-5_1

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84994651981&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

Cluster analysis is important in scientific and industrial fields. In this study, we proposed a novel chaotic biogeography-based optimization (CBBO) method, and applied it in centroid-based clustering methods. The results over three types of simulation data showed that this proposed CBBO method gave better performance than chaotic particle swarm optimization, genetic algorithm, firefly algorithm, and quantum-behaved particle swarm optimization. In all, our CBBO method is effective in centroid-based clustering.

DOI： 10.1109/PIC.2016.7949472

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85024503335&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

Wound area changes over multiple weeks are highly predictive of the wound healing process. A big data eHealth system would be very helpful in evaluating these changes. We usually analyze images of the wound bed for diagnosing injury. Unfortunately, accurate measurements of wound region changes from images are difficult. Many factors affect the quality of images, such as intensity inhomogeneity and color distortion. To this end, we propose a fast level set model-based method for intensity inhomogeneity correction and a spectral properties-based color correction method to overcome these obstacles. State-of-the-art level set methods can segment objects well. However, such methods are time-consuming and inefficient. In contrast to conventional approaches, the proposed model integrates a new signed energy force function that can detect contours at weak or blurred edges efficiently. It ensures the smoothness of the level set function and reduces the computational complexity of re-initialization. To increase the speed of the algorithm further, we also include an additive operator-splitting algorithm in our fast level set model. In addition, we consider using a camera, lighting, and spectral properties to recover the actual color. Numerical synthetic and real-world images demonstrate the advantages of the proposed method over state-of-the-art methods. Experimental results also show that the proposed model is at least twice as fast as methods used widely. Copyright © 2016 John Wiley & Sons, Ltd.

DOI： 10.1002/cpe.3927

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84980340447&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

Computer-generated Holograms (CGHs) are widely used for constructing optical wave fronts from numerically objects. In this paper, we propose a novel type of speckle elimination method for Lohmann-type CGHs, which is called Iterative Speckle Reduction Algorithm (ISRA). This algorithm is proposed for the speckle reduction in the reconstruction of digital holograms. The speckles can be reduced by avoiding the isolated zeros among the sampled points of the reconstructed image. The results show that the speckle elimination method highly improved the traditional Lohmann-type method for image reconstruction.

DOI： 10.1109/TENCON.2016.7848111

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85015430566&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Smart touch sensors are used in many applications, such as the iPhone and some smart home systems. Recent touch sensors perform well in consumer electrical devices; however, there are some drawbacks. For example, most touch sensors have low accuracy for detecting human movements. Many require significant power and have a fixed shape. We propose a flexible, highly accurate, and energy-conserving touch sensor. Our primary contributions are as follows: (1) an energy-conserving touch sensor is developed and tested experimentally; (2) a flexible and arbitrarily shaped touch sensor is designed; and (3) the manufacturing cost is very low and up to 200 touch sensors can be connected to the system. As a result, this energy-conserving touch sensor can be fabricated using common manufacturing processes for consumer electronic devices.

DOI： 10.18494/SAM.2017.1458

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85020809645&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Underwater images are degraded due to scatters and absorption, resulting in low contrast and color distortion. In this paper, a novel self-similarity-based method for descattering and super resolution (SR) of underwater images is proposed. The traditional approach of preprocessing the image using a descattering algorithm, followed by application of an SR method, has the limitation that most of the high-frequency information is lost during descattering. Consequently, we propose a novel high turbidity underwater image SR algorithm. We first obtain a high resolution (HR) image of scattered and descattered images by using a self-similarity-based SR algorithm. Next, we apply a convex fusion rule for recovering the final HR image. The super-resolved images have a reasonable noise level after descattering and demonstrate visually more pleasing results than conventional approaches. Furthermore, numerical metrics demonstrate that the proposed algorithm shows a consistent improvement and that edges are significantly enhanced.

DOI： 10.1109/ACCESS.2017.2648845

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85018527963&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Underwater optical cameras are widely used for security monitoring in ocean, such as earthquake prediction and tsunami alarming. Optical cameras recognize objects for autonomous underwater vehicles and provide security protection for sea-floor networks. However, there are many issues for underwater optical imaging, such as forward and backward scattering, light absorption, and sea snow. Many underwater image processing techniques have been proposed to overcome these issues. Among these techniques, the depth map gives important information for many applications of the post-processing. In this paper, we propose a Kinect-based underwater depth map estimation method that uses a captured coarse depth map by Kinect with the loss of depth information. To overcome the drawbacks of low accuracy of coarse depth maps, we propose a corresponding reconstruction architecture that uses the underwater dual channels prior dehazing model, weighted enhanced image mode filtering, and inpainting. Our proposed method considers the influence of mud sediments in water and performs better than the traditional methods. The experimental results demonstrated that, after inpainting, dehazing, and interpolation, our proposed method can create high-accuracy depth maps.

DOI： 10.1109/ACCESS.2017.2690455

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85028360504&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Image contrast enhancement for outdoor vision is important for smart car auxiliary transport systems. The video frames captured in poor weather conditions are often characterized by poor visibility. Most image dehazing algorithms consider to use a hard threshold assumptions or user input to estimate atmospheric light. However, the brightest pixels sometimes are objects such as car lights or streetlights, especially for smart car auxiliary transport systems. Simply using a hard threshold may cause a wrong estimation. In this paper, we propose a single optimized image dehazing method that estimates atmospheric light efficiently and removes haze through the estimation of a semi-globally adaptive filter. The enhanced images are characterized with little noise and good exposure in dark regions. The textures and edges of the processed images are also enhanced significantly.

DOI： 10.1007/s11042-015-2977-7

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84944710433&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

In daily life, people often forgotten their luggage or things by some mistakes. It makes people's life much more terrible. In order to prevent such incidents from happening again, we designed a lost luggage reminder system, which based on electrically conductive fiber antenna (ECFA) tag and acceleration sensors. The information of IC tag marked objects (such as books or wallet) are collected by ECFA tag, and send it to RFID reader. Meanwhile, utilize the acceleration sensors for distinguish the status of objects. If the objects are lost, system will notice the owner. The proposed system overcomes the drawbacks of traditional reminder systems (e.g. it can be applied in any occasion, can be distinguished within a very short time, and can be detected the luggage with a long distance). During abundant experiments, we evaluate the proposed system both with high accuracy and high speed.

DOI： 10.1109/IS3C.2016.17

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84986216643&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

Vision-based underwater navigation and object detection requires robust computer vision algorithms to operate in turbid water. Many conventional methods aimed at improving visibility in low turbid water. In this paper, we propose a novel contrast enhancement measurement for different enhancement methods' assessment. As a rule to compare the performance of different image enhancement algorithms, a more comprehensive image quality assessment index Q is proposed. The index combines the benefits of SSIM index and colour distance index. Experimental results show that the proposed approach statistically outperforms state-of-the-art general purpose underwater image contrast enhancement algorithms.

DOI： 10.1109/IS3C.2016.19

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84986224250&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

This paper describes a novel underwater observing system for Taiping floodgate mentoring in Taiping river of Yangzhou, China. While light traveling through the water, light rays are distorted depending on the wavelength. That is, absorption, scattering and color distortion are three major distortion issues for underwater optical imaging. In this paper, our key contributions are proposed include a novel underwater imaging model to compensate for the attenuation discrepancy along the propagation path and an effective underwater scene enhancement scheme. The recovered images are characterized by a reduced noised level, better exposure of the dark regions, and improved global contrast where the finest details and edges are enhanced significantly.

DOI： 10.1109/IS3C.2016.191

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84986212676&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

Vision-based underwater navigation and object detection requires robust computer vision algorithms to operate in turbid water. Many conventional methods aimed at improving visibility in low turbid water. In this paper, we propose a novel contrast enhancement to enhance high turbid underwater images using descattering and color correction. The proposed enhancement method removes the scatter and preserves colors. In addition, as a rule to compare the performance of different image enhancement algorithms, a more comprehensive image quality assessment index u is proposed. The index combines the benefits of SSIM index and color distance index. Experimental results show that the proposed approach statistically outperforms state-of-the-art general purpose underwater image contrast enhancement algorithms. The experiment also demonstrated that the proposed method performs well for image classification.

DOI： 10.1109/ICIP.2016.7532708

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85006819417&origin=inward

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（学術雑誌）

Vision-based underwater navigation and object detection requires robust computer vision algorithms to operate in turbid water. Many conventional methods aimed at improving visibility in low turbid water. High turbid underwater image enhancement is still an opening issue. Meanwhile, we find that the de-scattering and color correction of underwater images affect classification results. In this paper, we correspondingly propose a novel joint guidance image de-scattering and physical spectral characteristics-based color correction method to enhance high turbidity underwater images. The proposed enhancement method removes the scatter and preserves colors. In addition, as a rule to compare the performance of different image enhancement algorithms, a more comprehensive image quality assessment index Qu is proposed. The index combines the benefits of SSIM index and color distance index. We also use different machine learning methods for classification, such as support vector machine, convolutional neural network. Experimental results show that the proposed approach statistically outperforms state-of-the-art general purpose underwater image contrast enhancement algorithms. The experiment also demonstrated that the proposed method performs well for image classification.

DOI： 10.1016/j.compeleceng.2016.08.008

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84989891983&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

In recent years, sonar has been widely used for restoring the underwater terrain. Sonar imaging has the benefits such as long-range photographing, robust for turbidity water. However, it is not suitable for short-range imaging. Meanwhile, it also cannot meet the need of mining machine. Therefore, it is important to develop a 3D reconstruction method for short-range imaging. In this paper, we propose a Kinect-based underwater 3D image reconstruction method. To overcome the drawbacks of low accuracy of depth maps, we propose a novel super-resolution (SR) method, which uses the underwater dark channel prior dehazing, weight guided image SR, and inpainting. The proposed method considered the influence of mud sediments in water, it performs better than the traditional methods. The experimental results demonstrated that, after inpainting, dehazing and the super-resolution, it can obtain high accuracy depth maps.

DOI： 10.1109/ICPADS.2016.0168

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85018521829&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

This paper describes a novel method for enhancing optical images using a weighted guided trigonometric filter and the camera's spectral properties in turbid water. Absorption, scattering, and artificial lighting are three major distortion issues in underwater optical imaging. Absorption permanently removes photons from the imaging path. Scattering is caused by large suspended particles found in turbid water, which redirect the angle of the photon path. Artificial lighting results in footprint effects, which cause vignetting distortion in the captured image. Our contributions include a novel deep-sea imaging method that compensates for the attenuation discrepancy along the propagation path, and an effective underwater scene enhancement scheme. The recovered images are characterized by a reduced noise level, better exposure of dark regions, and improved global contrast such that the finest details and edges are significantly enhanced. Our experiments showed that the average Peak Signal to Noise Ratio (PSNR) improved by at least 1 dB when compared with state-of-the-art-methods.

DOI： 10.1016/j.jvcir.2016.03.029

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84962911203&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

Wound area changes over multiple weeks are highly predictive of the wound healing process. Big data processing is considered as one of the main solutions for it. We usually analysis the images of wound bed. Unfortunately, accurate measurements of wound area changes are difficult. In this paper, we propose a novel level set model (LSM) for pre-processing the intensity inhomogeneity images before encoder. State-of-The-Art LSMs can segment objects. However, most of these methods are time-consuming and inefficient. The proposed fast level set model (FLSM) is based on the piecewise constant and piecewise smooth Chan-Vese model and additive operator splitting algorithm. Different from conventional approaches, the proposed model integrates a new signed energy force function that can efficiently detect contours at weak or blurred edges. It ensures the smoothness of the level set function and reduces the computational complexity of re-initialization. Numerical synthetic and real world images demonstrate the advantages of the proposed method over state-of-The-Art methods. Experimental results also show that proposed model is at least twice as fast as the widely used models.

DOI： 10.1109/CBD.2015.37

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84966559690&origin=inward

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）

The deep-sea tripod systems designed and built at the U.S. Geological Survey (USGS) Pacific Coastal and Marine Science Center (PCMSC) in Santa Cruz, California. They were recovered in late September 2014 after spending about half a year collecting data on the floor of the South China Sea. The Free-Ascending Tripod (FAT) was deployed at 2,100 meters water depth-roughly 10 times as deep as most tripods dedicated to measuring currents and sediment movement at the seafloor. Deployment at this unusual depth was made possible by the tripod's ability to rise by itself to the surface rather than being pulled up by a line. Instruments mounted on the tripod took bottom photographs and measured such variables as water temperature, current velocity, and suspended-sediment concentration. FAT is used to better understand how and where deep-seafloor sediment moves and accumulates. Besides of this, we also use them to study the deep-sea biology. The obtained the images from the camera, the biology animals are hardly to be distinguished. In this paper, we are concerned to use the image processing technology for recovering the deep-sea scene.

DOI： 10.23919/oceans.2015.7401872

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84963949087&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

This paper describes a novel method for ocean scene 3D reconstruction. While light is travelling through water, light rays are distorted depending on the wavelength. That is, absorption, scattering and colour distortion are three major distortion issues for underwater optical imaging. Scattering is caused by large suspended particles, as in turbid water that contains abundant particles, which causes the degradation of the captured image. Colour distortion corresponds to the varying degrees of attenuation encountered by light travelling in water at different wavelengths, causing ambient underwater environments to be dominated by a bluish tone. Our key contributions proposed here include a novel deep-sea imaging model to compensate for the attenuation discrepancy along the propagation path and an effective underwater scene 3D reconstruction method. The recovered 3D images are characterised by a reduced noise level, better exposure of the dark regions, and improved global contrast where the finest details and edges are enhanced significantly.

DOI： 10.1504/IJCSE.2016.076950

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84974814386&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

We describe a method for removing non-Gaussian image noise in natural images, underwater images and biomedical images, based on a statistical model of the decomposed contourlet coefficients. This method utilises the non-Gaussian multivariate shrinkage (NGMS) probability density function (PDF) to model neighbourhood contourlet coefficients. Then, according to the proposed PDF model, we design a maximum a posteriori (MAP) estimator, which relies on a Bayesian statistics representation for the contourlet coefficients of noisy images. There are three obvious virtues of this method. Firstly, contourlet transform decomposition prior to curvelet transform and wavelet transform by using ellipse sampling grid. Secondly, NGMS model is more effective in presentation of the noisy image contourlet coefficients. Thirdly, the NGMS model takes full account of the correlation between coefficients. Some comparisons with the best available results will be presented in order to illustrate the effectiveness of the proposed method.

DOI： 10.1504/IJCSE.2016.076210

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85019882126&origin=inward

記述言語：英語   掲載種別：研究論文（学術雑誌）

Absorption, scattering, and color distortion are three major issues in underwater optical imaging. Light rays traveling through water are scattered and absorbed according to their wavelength. Scattering is caused by large suspended particles that degrade underwater optical images. Color distortion occurs because different wavelengths are attenuated to different degrees in water; consequently, images of ambient underwater environments are dominated by a bluish tone. In the present paper, we propose a novel underwater imaging model that compensates for the attenuation discrepancy along the propagation path. In addition, we develop a fast weighted guided normalized convolution domain filtering algorithm for enhancing underwater optical images. The enhanced images are characterized by a reduced noise level, better exposure in dark regions, and improved global contrast, by which the finest details and edges are enhanced significantly.

DOI： 10.1587/transinf.2014EDP7405

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84953324138&origin=inward