Search

Efficient and fast stereo matching is a challenging task due to the presence of occlusion and low texture areas. In stereo matching, the correspondence between left and right images may be difficult owing to the lack of matching information. The cost metrics proposed before are not robust enough or are computationally expensive. In this work, we propose a novel generic tree structure, Pyramid-tree, which improves the single mode of traditional tree, and can achieve cross-regional connection between different regions with similar colors and similar depths. This unique structure can achieve cross-regional cost smoothing, significantly reducing the possibility of mismatch due to occlusion or lack of corresponding matching information, and has stronger robustness to occlusion and low texture regions. In addition, we also propose a new bearings-only cost metric, Log-angle, which is not affected by occlusion, low texture, illumination and other factors. Log-angle combines with traditional metrics can show better performance. We show that the Pyramid-tree structure and Log-angle are very important as it efficiently expands the state-of-the-art stereo matching methods and leads to significant improvements. Qualitative and quantitative experiments on Middlebury data sets verify the superior performance of the algorithm, and very effective compromise between the accuracy and computation load is achieved.

Recently, single image super-resolution (SISR) has been widely applied in the fields of multimedia and computer vision communities and obtained remarkable performance. However, most current methods ignore to utilize multi-granularity features of low-resolution (LR) image to further improve the SISR performance. And the channel and spatial features obtained from original LR images are treated equally, resulting in unnecessary computations for abundant uninformative features, thereby hindering the representational ability of super-resolution (SR) models. In this paper, we present a novel Multi-Granularity Pyramid Attention Network (MGPAN) which fully exploits the multi-granularity perception and attention mechanisms to improve the quality of reconstructed images. We design a multi-branch dilated convolution layer with varied kernels corresponding to receptive fields of different sizes to modulate multi-granularity features for adaptively capturing more important information. Moreover, a novel spatial pyramid pooling attention (SPPA) module is constructed to integrate the channel-wise and multi-scale spatial information, which is beneficial to compute the response values from the multi-scale regions of each neuron, and then establish the accurate mapping from low to high dimensional solution space. Besides, for long-short-term information preservation and information flow enhancement, we adopt the short, long, and global skip connection structures to concatenate and fuse the states of each module, which can improve the SR network performance effectively. Extensive experiments on several standard benchmark datasets show that the proposed MGPAN can provide state-of-the-art or even better performance in both quantitative and qualitative measurements.

Although recently a series of deep‐learning based methods have promoted the performance of face super resolution (FSR), most of these methods cannot recover essential face attributes accurately, especially to super‐resolve a very low‐resolution (LR) face image (16×16 pixels) to its (8×) high‐resolution (HR) version. To address this issue, a novel alternating optimisation algorithm to estimate facial attributes and restore facial images in a single network is presented . Specifically, two convolutional neural modules (denoted as Restorer and Corrector) are constructed and these two modules are alternated repeatedly to form an end‐to‐end trainable network. The Restorer module reconstructs face images based on estimated facial attributes, while the Corrector module corrects the estimated attributes with the help of restored face image. Since Corrector can utilise information from previous estimated attributes and the FSR image, the estimated attributes are iteratively corrected and gradually approach the ground truth. Moreover, a new attribute transformation scheme is designed to introduce attribution information into Restorer, in which the facial attribute vectors as the control conditions can guide the face image restoration explicitly. By conducting extensive experiments on the well‐known CelebA dataset, it is demonstrated that the proposed method can provide superior FSR performance in both quantitative and qualitative measurements.

We consider visual tracking in numerous applications of computer vision and seek to achieve optimal tracking accuracy and robustness based on various evaluation criteria for applications in intelligent monitoring during disaster recovery activities. We propose a novel framework to integrate a Kalman filter (KF) with spatial-temporal regularized correlation filters (STRCF) for visual tracking to overcome the instability problem due to large-scale application variation. To solve the problem of target loss caused by sudden acceleration and steering, we present a stride length control method to limit the maximum amplitude of the output state of the framework, which provides a reasonable constraint based on the laws of motion of objects in real-world scenarios. Moreover, we analyze the attributes influencing the performance of the proposed framework in large-scale experiments. The experimental results illustrate that the proposed framework outperforms STRCF on OTB-2013, OTB-2015 and Temple-Color datasets for some specific attributes and achieves optimal visual tracking for computer vision. Compared with STRCF, our framework achieves AUC gains of 2.8%, 2%, 1.8%, 1.3%, and 2.4% for the background clutter, illumination variation, occlusion, out-of-plane rotation, and out-of-view attributes on the OTB-2015 datasets, respectively. For sporting events, our framework presents much better performance and greater robustness than its competitors.

Salient object detection is getting more and more attention in computer vision field. In this paper, we propose a novel and effective framework for salient object detection. Firstly, we develop a robust background-based map by using spatial prior to remove the foreground noises of image boundary regions. The proposed background-based map and Objectness map are integrated to obtain a coarse saliency map. Then, an effective saliency propagation mechanism is utilized to further highlight salient object and suppress background region by defining a novel graph model, each node connects to its more similar neighbors and nodes with low saliency values in the proposed graph. As a result, the coarse saliency map is optimized to the refined saliency map by novel graph based saliency propagation. Finally, we construct a novel integration framework to further integrate two saliency maps for performance improvement. Experiments on three benchmark datasets are tested, experimental results show the superiority of the proposed algorithm than other state-of-the-art methods.

Low-dose CT (LDCT) images have been widely applied in the medical imaging field due to the potential risk of exposing patients to X-ray radiations. Given the fact that reducing the radiation dose may result in increased noise and artifacts, methods that can eliminate the noise and artifacts in the LDCT image have drawn increasing attentions and produced impressive results over the past decades. However, recent proposed methods mostly suffer from noise remaining, over-smoothing structures or false lesions derived from noise. In this paper, we propose a generative adversarial network (GAN) with novel architecture and loss function for restoring the LDCT image. Firstly, the inception-residual block and residual mapping are incorporated in the U-Net structure. The modified U-Net is applied as the generator of the GAN network so that the noise feature can be eliminated during the forward propagation. Secondly, a novel multi-level joint discriminator is designed by concatenating multiple convolutional neural networks (CNNs) where the output of each deconvolutional layer in the generator is compared with the corresponding down-sampled ground truth image. The adversarial training can be sensitive to noise and artifacts in different scales with this discriminator. Thirdly, we novely define a loss function consisting of the least square adversarial loss, VGG based perceptual loss, MSE based pixel loss and the noise loss, so that the differences in pixel, visual perception and noise distribution are comprehensively considered to optimize the network. Experimental results on both simulated and official simulated clinical images have demonstrated that the proposed method can provide superior performance to the state-of-the-art methods in noise removal, structure preservation and false lesions elimination.

Recently, single image super-resolution (SISR) has been widely applied in the field of remote sensing image processing and obtained remarkable performance. However, existing CNN-based remote sensing image super-resolution methods are unable to exploit shallow visual characteristics at global receptive fields, which results in the limited perceptual capability of these models. Furthermore, the low-resolution inputs and features contain abundant low-frequency information, which are weighed in channels and space equally, hence limiting the representational ability of CNNs. To solve these problems, we propose a non-locally up-down convolutional attention network (NLASR) for remote sensing image super-resolution. First, a non-local features enhancement module (NLEB) is constructed to obtain the spatial context information of high-dimensional feature maps, which allows our network to utilize global information to enhance low-level similar structured texture information with effect, overcoming the defects of deficiency perceptual ability of shallow convolutional layers. Second, an enhanced up-sampling channel-wise attention (EUCA) module and enhanced down-sampling spatial-wise attention (EDSA) module are proposed to weight the features at multiple scales. By integrating the channel-wise and multi-scale spatial information, the attention modules are able to compute the response values from the multi-scale regions of each neuron and then establish the accurate mapping from low to high resolution space. Extensive experiments on NWPU-RESISC45 and UCMerced-LandUse datasets show that the proposed method can provide state-of-the-art or even better performance in both quantitative and qualitative measurements.

Image clustering is a complex procedure that is significantly affected by the choice of the image representation. Generally speaking, image representations are generated by using handcraft features or trained neural networks. When dealing with high dimension data, these two representation methods cause two problems: i) the representation ability of the manually designed features is limited; ii) the non-representative and meaningless feature of a trained deep network may hurt the clustering performance. To overcome these problems, we propose a new clustering method which efficiently builds an image representation and precisely discovers the cluster assignments. Our main tools are an unsupervised representation learning method based on Deep Mutual Information Maximization (DMIM) system, and a clustering method based on self-training algorithm. Specifically speaking, to extract the informative representation of image data, we derive the maximum mutual information theory and propose a system to learn the maximum mutual information between the input images and the latent representations. To discover the clusters and assign each image a clustering label, a self-training mechanism is applied to cluster the learned representations. The superiority and validity of our algorithm are verified in a series of real-world image clustering experiments.

The state-of-the-art trackers using deep learning technology have little special strategy to gain the bounding box well when the target suffers drastic geometric deformation. In this paper, we take full use of the convolutional neural network (CNN) features of the deepest layer to represent the semantic feature model, and affine transformation to be as the space information model. A tracking method based on geometrical transformation region CNN is proposed. Firstly, affine transformation is applied to predict possible locations of a target, and the candidate bounding boxes obtained by affine transformation sampling can locate the possible geometric regions of the target more effectively before extracting features from CNN. Furthermore, RoI pooling with different sizes and shapes are designed to describe the geometric deformation region of the target. Then, multi-tasks loss function including the affine transformation regression is designed to refine the affine bounding box. Finally, the affine transformation NMS (Non-maximum suppression) is used to ensure the tracking bounding box having the largest IoU value. Extensive experimental results show that the proposed algorithm performs favorably against the compared methods in the public benchmarks.

For current learning-based single image deraining methods, deraining networks are usually designed based on a simplified linear additive rain model, which may not only cause unreal synthetic rainy images for both training and testing datasets, but also adversely affect the applicability and generality of corresponding networks. In this paper, we use the screen blend model of Photoshop as the nonlinear rainy image decomposition model. Based on this model, we design a novel channel attention U-DenseNet for rain detection and a residual dense block for rain removal. The detection sub-network not only adjusts channel-wise feature responses by our novel channel attention block to pay more attention to learn the rain map, but also combines the context information with the precise localization by the U-DenseNet to promote pixel-wise estimation accuracy. After rain detection, we use the nonlinear model to get a coarse rain-free image, and then introduce a deraining refinement subnetwork consisted of the residual dense block to obtain a fine rain-free image. For training our network, we apply the nonlinear rain model to synthesize a benchmark dataset called as RITD. It contains 3200 triplets of rainy images, rain maps, and clean background images. Our extensive quantitative and qualitative experimental results show that our method outperforms several state-of-the-art methods on both synthetic and real images.

With the rapid development of the Internet of Things (IoT) and smart cities, more and more types of applications have been emerging. In fact, different applications have different features and different requirements on services. In order to satisfy users' Quality of Service (QoS) requirements, the application-awareness technique should be leveraged to distinguish different applications for providing the differentiated services. However, the traditional Internet only can obtain the local network view, which belongs to the offline awareness mode and cannot adapt to the dynamical network environment. At the right time, Software-Defined Networking (SDN) has been accepted as a new networking paradigm thanks to its network awareness on the global status information, which can greatly facilitate the online application-awareness. At present, three ways, i.e., port number, depth packet inspection and deep learning can be used for the application-awareness. To the best of our knowledge, the deep learning based application-awareness method is the most cutting-edge technique. In spite of this, the previous related schemes fail to effectively guarantee the correctness and stability. To this end, this paper proposes a Convolutional Neural Network (CNN) based deep learning mechanism to do the application-awareness, including three phases, i.e., traffic collection, data pre-processing and application-awareness. The SDN environment is implemented based on the MiniNet and the simulation experiments are made based on the TensorFlow. The experimental results show that the proposed application-awareness mechanism outperforms three benchmarks on recall ratio, precision ratio, F value and stability.

For visual tracking, key factors that affect the performance of trackers are related to whether it can effectively extract the appearance information and spatial information of a target. And most of state-of-the-art trackers either do not model the appearance information and spatial information separately or do not design special strategies to deal with the strong geometric deformation of the target. In this paper, we design an appearance information model and a spatial information model separately, and then combine them to obtain complementary benefits. Firstly, because the features from deeper layers of a convolutional neural network (CNN) can better describe the semantic information of a target while the spatial information becomes less, we adopt the features from the deepest layer as the appearance information model. Secondly, we focus on tracking the target with drastic geometric deformation through utilizing a projection transformation group (SL(3) group) to model the geometric transformation of the target, where SL(3) group can describe the geometric deformation more accurately. Furthermore, a standard discriminative correlation filter is used to develop the effect of convolutional features and is more efficient than other methods used for CNN. Extensive experiments results show that our tracker outperforms all the compared trackers.

Purpose A robotic wheelchair system was designed to assist disabled people with disabilities to walk. Design/methodology/approach An anticipated sharing control strategy based on topological map is proposed in this paper, which is used to assist robotic wheelchairs to realize interactive navigation. Then, a robotic wheelchair navigation control system based on the brain-computer interface and topological map was designed and implemented. Findings In the field of robotic wheelchairs, the problems of poor use, narrow application range and low humanization are still not improved. Originality/value In the system, the topological map construction is not restricted by the environment structure, which helps to expand the scope of application; the shared control system can predict the users’ intention and replace the users’ decision to realize human-machine interactive navigation, which has higher security, robustness and comfort.

Salient object detection, which aims at locating the most important object in a scene (or image), has been extensively studied in various tasks, such as robot vision. In this paper, we present an effective salient object detection framework based on a novel bottom-up semi-supervised learning algorithm, which obviously outperforms the existing works in complex scenes. Given an input image, it is firstly segmented into a fixed number of non-overlapping image patches as basic units. A novel segmentation-based sampling method is developed to select a subset of all image patches as training samples. Then, all samples are divided into labeled and unlabeled groups based on multiple prior cues. The labels of all the unlabeled data are inferred by a novel label propagation mechanism. As a result, a complete training set can be obtained and used to train a classifier to classify all image patches into salient object and background. In addition, we also use neighbor-constraint smoothness function to further boost the saliency map. We compare the proposed method with the state-of-the-art approaches on two datasets. Experimental results demonstrate the effectiveness and superiority of the proposed method.

Following the development of autonomous underwater vehicles (AUVs), multiple trajectory-based submarine target information collection constitutes one of the key technologies that significantly influence underwater information collection ability and deployment efficiency. In this paper, we propose an underwater information collection AUV, O-AUV, that can perceive the omnidirectional area and could detect a larger area than the traditional AUV. A 3D sensing model for the O-AUV is proposed to describe the complex underwater information collection spaces. Thereafter, a cube-based environment model involving candidate observation point calculation methods are suggested to adapt the O-AUV model. A voyage cost map is also built according to the multi-AUV path planning for a common submarine mission that must traverse numerous mission targets in complex environments through the R-Dijkstra algorithm. Specifically, the voyage planning problem is solved through a critical algorithm called ANSGA (accelerated NSGA-II algorithm), which in turn, is developed by modifying the non-dominated sorting genetic algorithm (NSGA-II) to accelerate the optimization rate for the Pareto solution. Experiments are carried out in MATLAB, and the results verify the validity of the proposed O-AUV+ANSGA algorithm framework.

With the continuous development of computer vision technology, deep learning technology has been well applied in human pose estimation. However, the latest human pose estimation algorithms can not solve the problem of interdependence and occlusion between joints, so the key point detection in human pose estimation is still a challenging task. In this paper, a method based on masked pre-training model and lightweight SHNet combined method, random occlusion is performed on the key points of the human body in the picture, and the pre-training model is used to predict them, so that the problems of mutual occlusion and interdependence can be solved. Experiments show that the algorithm is effective in key point detection of MPII and COCO 2017 datasets.

Event-based vision sensors are a paradigm shift in the way that visual information is obtained and processed. These devices are capable of low-latency transmission of data which represents the scene dynamics. Additionally, low-power benefits make the sensors popular in finite-power scenarios such as high-speed robotics or machine vision applications where latency in visual information is desired to be minimal. The core datatype of such vision sensors is the ‘event’ which is an asynchronous per-pixel signal indicating a change in light intensity at an instance in time corresponding to the spatial location of that sensor on the array. A popular approach to event-based processing is to map events onto a 2D plane over time which is comparable with traditional imaging techniques. However, this paper presents a disruptive approach to event data processing that uses a tree-based filter framework that directly processes raw event data to extract events corresponding to interest point features, which is then combined with a Harris interest point approach to isolate features. We hypothesise that since the tree structure contains the same spatial information as a 2D surface mapping, Harris may be applied directly to the content of the tree, bypassing the need for transformation to the 2D plane. Results illustrate that the proposed approach performs better than other state-of-the-art approaches with limited compromise on the run-time performance.

With the development of computer vision, high quality images with rich information have great research potential in both daily life and scientific research. However, due to different lighting conditions, surrounding noise and other reasons, the image quality is different, which seriously affects people's discrimination of the information in the image, thus causing unnecessary conflicts and results. Especially in the dark, the images captured by the camera are difficult to identify, and the smart system relies heavily on high-quality input images. The image collected in low-light environment has the characteristic with high noise and color distortion, which makes it difficult to utilize the image and can not fully explore the rich value information of the image. In order to improve the quality of low-light image, this paper proposes a Heterogenous low-light image enhancement method based on DenseNet generative adversarial network. Firstly, the generative network of generative adversarial network is realized by using DenseNet framework. Secondly, the feature map from low light image to normal light image is learned by using the generative adversarial network. Thirdly, the enhancement of low-light image is realized. The experimental results show that, in terms of PSNR, SSIM, NIQE, UQI, NQE and PIQE indexes, compared with the state-of-the-art enhancement algorithms, the values are ideal, the proposed method can improve the image brightness more effectively and reduce the noise of enhanced image.

Three-dimensional depth perception using structured light has been widely researched and developed due to its robustness, high precision, and density. An open research issue exists with current structured light technology to obtain depth maps of scenes with objects in different motion states, while ensuring high accuracy and resolution. In this paper, a new algorithm for depth acquisition of multi-state targets is designed. A novel projection pattern is designed based on sinusoidal, speckle and line patterns. The foreground targets are separated according to the background segmentation technology, and regions with different motion states are divided into static, slow motion or fast motion regions. Then, the different motion states are used to select different pattern decoding schemes. In the experiments, the depth of multi-state objects are successfully calculated by our technique.

Fractional-order flight control has a history of nearly 10 years. Fractional-order controllers (FOCs) have been proved to be better in rising time, overshoot and robustness against plant variation. However, there are still not many real applications of FOC in industry. More case studies need to be carried out to accelerate the maturation of FOC. Quadrotor vision-based control often has a large and time-varying sampling period due to drone’s resource limitation. Therefore, our research has been focused on a specific case for drone vision-based control to investigate the benefits of FOC. In this paper, FOC has been first discovered to be able to tolerate larger sampling period than integer-order controllers. This fact has been proved both theoretically and numerically. First of all, the speed model was identified from real flight tests. Then an integral-order proportional, integral and derivative (IOPID) controller and a fractional-order proportional–derivative (FOPD) controller were designed. After that, a stability criteria, optimization method, graphic method and parallel computing techniques were employed to theoretically prove that the largest sampling period of the designed FOC (0.933 s) is much larger than that of the designed integral-order controller (0.546 s). Later, Simulink simulation with identified linear model proved that the FOC can tolerate larger sampling period. Finally, flight tests showed that the designed FOC has a nearly 20% better precision on drone vision-based hovering than the IOC.

Traditional myoelectric prostheses that employ a static pattern recognition model to identify human movement intention from surface electromyography (sEMG) signals hardly adapt to the changes in the sEMG characteristics caused by interferences from daily activities, which hinders the clinical applications of such prostheses. In this paper, we focus on methods to reduce or eliminate the impacts of three types of daily interferences on myoelectric pattern recognition (MPR), i.e., outlier motion, muscle fatigue, and electrode doffing/donning. We constructed an adaptive incremental hybrid classifier (AIHC) by combining one-class support vector data description and multi-class linear discriminant analysis in conjunction with two specific update schemes. We developed an AIHC-based MPR strategy to improve the robustness of MPR against the three interferences. Extensive experiments on hand-motion recognition were conducted to demonstrate the performance of the proposed method. Experimental results show that the AIHC has significant advantages over non-adaptive classifiers under various interferences, with improvements in the classification accuracy ranging from 7.1% to 39% ( ${p} ). The additional evaluations on data deviations demonstrate that the AIHC can accommodate large-scale changes in the sEMG characteristics, revealing the potential of the AIHC-based MPR strategy in the development of clinical myoelectric prostheses.

Underwater object recognition in sonar images, such as mine detection and wreckage detection of a submerged airplane, is a very challenging task. The main difficulties include but are not limited to object rotation, confusion from false targets and complex backgrounds, and extensibility of recognition ability on diverse types of objects. In this paper, we propose an underwater object detection and recognition method using a transformable template matching approach based on prior knowledge. Specifically, we first extract features and construct a template from sonar video sequences based on the analysis of acoustic shadows and highlight regions. Then, we identify the target region in the objective image by fast saliency detection techniques based on FFT, which can significantly improve efficiency by avoiding an exhaustive global search. After affine transformation of the template according to the orientation of the target, we extract normalized gradient features and calculate the similarity between the template and the target region, which can solve various difficulties mentioned above using only one template. Experimental results demonstrate that the proposed method can well recognize different underwater objects, such as mine-like objects and triangle-like objects and can satisfy the demands of real-time application.

Sunshine-based trajectory analysis plays a significant impact on comprehensive scenarios, including driving safety and routing planning for mobile solar collection. To the best of our knowledge, however, a little work analyzes the impact of sunshine on mobile objects on different trajectories. This paper investigates the relationship between sunshine and the trajectories of mobile objects. We introduce a sunshine orientation model to integrate sunshine information with trajectories. A sunshine-based trajectory simplification model is proposed to remove redundant points in trajectories, which carry a little sunshine-related information. The sunshine-based trajectory simplification model is indeed versatile. We apply this model to help users efficiently query the sunshine information of their trips. Our design is evaluated with extensive simulations based on three-month general transit feed specification data from August to October in 2017 in five USA cities, including Washington, DC, USA, and Minneapolis. Furthermore, we develop a sunshine query system for mobile scenarios by embedding our sunshine-based trajectory simplification model and conduct a field study to show the validity and efficiency of this model. The evaluation results show that our design reduces 87.4% position points in trajectories while only sacrificing 9.6% of sunshine information in the field tests.

Glaucoma is an eye disease which is one of the most common causes of blindness. Accurate optic disc (OD) and optic cup (OC) segmentation play a critical role for detecting glaucoma. Considering that the existing approaches can’t effectively integrate the multi-view information deriving from shape and appearance to sufficiently describe OD and OC for segmentation, Locally Statistical Active Contour Model with the Information of Appearance and Shape (LSACM-AS) and Modified Locally Statistical Active Contour Model with the Information of Appearance and Shape (MLSACM-AS) are proposed in this paper. The main contributions are as below: (1) we introduce the Locally Statistical Active Contour Model (LSACM) to address the commonly occurred intensity inhomogeneity phenomenon caused by imperfection of image devices or illumination variations. (2) In order to overcome the common effects caused by pathological changes (i.e., peripapillary atrophy (PPA)) and vessel occlusion in OD and OC segmentation, we integrate the local image probability information around the point of interest from a multi-dimensional feature space into our model to preserve the integrity of the OD and OC structures. (3) Since the segmentation objects have the similar ellipse shape structure, we incorporate the shape priori constraint information into our model to further improve the robustness of the variations found in and around objects regions. To evaluate the effectiveness of the proposed models, an available publicly DRISHTI-GS database is employed in this paper. Seen from the abundant experiments, the proposed models outperform the state-of-the-art approaches in terms of the obtained qualitative and quantitative results.

The state-of-the-art trackers using deep learning technology have no special strategy to capture the geometric deformation of the target. Based on that the affine manifold can better capture the target shape change and that the higher level of Convolutional Neural Network (CNN) can better describe semantic information of objects, we propose a new tracking algorithm combining affine transformation with convolutional features to track targets with dramatic deformation. First, the affine transformation is applied to predict possible locations of a target, then a correlative filter is designed to compute the appearance confidence score for determining the final target location. Furthermore, a standard discriminative correlation filter is used to develop the effect of convolutional features, which is more efficient than other methods used for CNN Networks. Comprehensive experiments demonstrate the outstanding performance of our tracking algorithm compared to the state-of-the-art techniques in the public benchmarks.

There have been active studies on optimized dithering techniques to improve 3D shape measurement quality with image sensor and defocusing projector. These techniques can be classified into intensity-based optimization technology and phase-based optimization technology. However, those phase-based optimization methods are sensitive to the amount of defocusing while intensity-based optimization methods cannot reduce phase errors efficiently. This paper presents an optimization framework. This framework combines structural similarity index measurement and intensity residual error optimization. By applying this optimization framework to patches and tiling the best patch, high quality fringe pattern can be generated for three-dimensional measurement. Both simulation and experimental results show that this proposed algorithm can achieve phase quality improvements and it is robust to various defocusing levels.

Convolutional autoencoders are making a significant impact on computer vision and signal processing communities. In this work, a convolutional autoencoder denoising method is proposed to restore the corrupted laser stripe images of the depth sensor, which directly reduces the external noise of the depth sensor so as to increase its accuracy. To reduce the amount of training data and avoid overfitting, a patch size of the laser stripe image is determined, on the basis of which a small-scale dataset called Laser Stripe Image Patch (LSIP) is created. Also, a 14-layers convolutional autoencoder is constructed to reduce the noise of the image patches, which can learn the most salient features on the LSIP dataset. Moreover, the trained convolutional autoencoder is applied to an omnidirectional structured light system. Experimental results demonstrate that the proposed method obtains useful features and superior performance both visually and quantitatively on denoising tasks, and significantly improves the accuracy of the structured light system. Keywords: Convolutional autoencoder, Deep learning, Depth perception, Image denoising, Laser stripe, Structured light

This paper presents a novel Reducing-Over-Time (ROT) binary tree structure for event-based vision data and subtypes of the tree structure. A framework is presented using ROT, that takes advantage of the self-balancing and self-pruning nature of the tree structure to extract spatial-temporal information. The ROT framework is paired with an established motion classification technique and performance is evaluated against other state-of-the-art techniques using four datasets. Additionally, the ROT framework as a processing platform is compared with other event-based vision processing platforms in terms of memory usage and is found to be one of the most memory efficient platforms available.

Image clustering is a complex procedure, which is significantly affected by the choice of image representation. Most of the existing image clustering methods treat representation learning and clustering separately, which usually bring two problems. On the one hand, image representations are difficult to select and the learned representations are not suitable for clustering. On the other hand, they inevitably involve some clustering step, which may bring some error and hurt the clustering results. To tackle these problems, we present a new clustering method that efficiently builds an image representation and precisely discovers cluster assignments. For this purpose, the image clustering task is regarded as a binary pairwise classification problem with local structure preservation. Specifically, we propose here such an approach for image clustering based on a fully convolutional autoencoder and deep adaptive clustering (DAC). To extract the essential representation and maintain the local structure, a fully convolutional autoencoder is applied. To manipulate feature to clustering space and obtain a suitable image representation, the DAC algorithm participates in the training of autoencoder. Our method can learn an image representation that is suitable for clustering and discover the precise clustering label for each image. A series of real-world image clustering experiments verify the effectiveness of the proposed algorithm.

As we know, the video transmission traffic already constitutes 60% of Internet downlink traffic. The optimization of video transmission efficiency has become an important challenge in the network. This paper designs a video transmission optimization strategy that takes reinforcement learning and edge computing (TORE) to improve the video transmission efficiency and quality of experience. Specifically, first, we design the popularity prediction model for video requests based on the RL (reinforcement learning) and introduce the adaptive video encoding method for optimizing the efficiency of computing resource distribution. Second, we design a video caching strategy, which adopts EC (edge computing) to reduce the redundant video transmission. Last, simulations are conducted, and the experimental results fully demonstrate the improvement of video quality and response time.

In this paper, we propose a retinex-based decomposition model for a hazy image and a novel end-to-end image dehazing network. In the model, the illumination of the hazy image is decomposed into natural illumination for the haze-free image and residual illumination caused by haze. Based on this model, we design a deep retinex dehazing network (RDN) to jointly estimate the residual illumination map and the haze-free image. Our RDN consists of a multiscale residual dense network for estimating the residual illumination map and a U-Net with channel and spatial attention mechanisms for image dehazing. The multiscale residual dense network can simultaneously capture global contextual information from small-scale receptive fields and local detailed information from large-scale receptive fields to precisely estimate the residual illumination map caused by haze. In the dehazing U-Net, we apply the channel and spatial attention mechanisms in the skip connection of the U-Net to achieve a trade-off between overdehazing and underdehazing by automatically adjusting the channel-wise and pixel-wise attention weights. Compared with scattering model-based networks, fully data-driven networks, and prior-based dehazing methods, our RDN can avoid the errors associated with the simplified scattering model and provide better generalization ability with no dependence on prior information. Extensive experiments show the superiority of the RDN to various state-of-the-art methods.

It is important for underwater wireless sensor networks (UWSNs) to satisfy the diverse monitoring demands in harsh and perilous three-dimensional underwater environments. After the monitoring missions and demands transform, a large number of underwater event coverage holes will appear. Traditional network repair strategies cannot be applied to the ever-changing underwater monitoring missions and the harsh multi-constrained three-dimensional underwater environments. Multiple autonomous underwater vehicles (multi-AUVs) have strong adaptability and flexibility in perilous and harsh three-dimensional underwater environments. First, an underwater event coverage hole (UECH) repair model under various constraints is proposed. Next, a multi-agent event coverage hole repair algorithm (MECHR), which combines multi-agent strategy with diversity archive strategy, is proposed to repair UECHs in UWSNs. The presented algorithm symmetrically completes subtasks through information exchange and interactive operations with other agents. Unlike existing repair strategies, the MECHR algorithm can effectively repair a large number of UECHs resulted by the transformations in underwater monitoring scenes and demands. The MECHR algorithm can adapt to a wide range of harsh scenes and multi-constrained three-dimensional underwater environments. Eventually, the effect of the MECHR algorithm is verified through underwater repair simulation experiments, which can adapt to the constantly changing three-dimensional underwater monitoring environments.

Event-based image processing is a relatively new domain in the field of computer vision. Much research has been carried out on adapting event-based data to comply with established techniques from frame-based computer vision. On the contrary, this paper presents a descriptor which is designed specifically for direct use with event-based data and therefore can be considered to be a pure event-based vision descriptor as it only uses events emitted from event-based vision devices without transforming the data to accommodate frame-based vision techniques. This novel descriptor is known as the Post-stimulus Time-dependent Event Descriptor (P-TED). P-TED is comprised of two features extracted from event data which describe motion and the underlying pattern of transmission respectively. Furthermore a framework is presented which leverages the P-TED descriptor to classify motions within event data. This framework is compared against another state-of-the-art event-based vision descriptor as well as an established frame-based approach.

As the key technology for Internet of things, wireless sensor networks have received more attentions in recent years. Mobile localization is one of the significant topics in wireless sensor networks. In wireless sensor network, non-line-of-sight propagation is a common phenomenon leading to the growing non-line-of-sight error. It is a fatal impact for the localization accuracy of the mobile target. In this article, a novel method based on the nearest neighbor variable estimation is proposed to mitigate the non-line-of-sight error. First, the linear regression model of the extended Kalman filter is used to obtain the residual of the distance measurement value. After that, the residual analysis is used to complete the identification of the measurement value state. Then, by analyzing the statistical characteristics of the non-line-of-sight residual, the nearest neighbor variable estimation is proposed to estimate the probability density function of residual. Finally, the improved M-estimation is proposed to locate the mobile robot. Experiment results prove that the accuracy and robustness of the proposed algorithm are better than other methods in the mixed line-of-sight/non-line-of-sight environment. The proposed algorithm effectively inhibits the non-line-of-sight error.

The traffic block port monitors and manages the road traffic by shooting and recording the motor vehicles. However, due to the complex factors such as shooting angle, light condition, environmental background, etc., the recognition rate of license plate is not high enough. High light and low light under complex lighting conditions are symmetry problems. This paper analyzes and solves the low light problem in detail, an image adaptive enhancement algorithm under low light conditions is proposed in the paper. The algorithm mainly includes four modules, among which, the fast image classification module uses the deep and separable convolutional neural network to classify low-light images into low-light images by day and low-light images by night, greatly reducing the computation burden on the basis of ensuring the classification accuracy. The image enhancement module inputs the classified images into two different image enhancement algorithms and adopts the idea of dividing and ruling; the image quality evaluation module adopts a weighted comprehensive evaluation index. The final experiment shows that the comprehensive evaluation indexes are all greater than 0.83, which can improve the subsequent recognition of vehicle face and license plate.

This paper presents an adaptive filtering algorithm for monocular depth estimation. This is a geometric calculation method under the assumption that the inverse depth conforms to the Gaussian distribution. In the inverse depth update, the average value of the historical stable points is introduced to smooth the output, and the depth uncertainty of the unit pixel offset is calculated to eliminate the estimation error. In the similarity search along the epipolar line, the normalized cross-correlation and gradient are used as joint metrics. Finally, the effectiveness of the algorithm is verified through experiments, and it can get a better trade-off between performance and time.

In order to solve the problems of face features extraction and the long time consuming in network training, we propose a facial expression recognition network based on multi-level feature fusion structure. The network structure consists of the following three parts: The multi-level feature fusion block; The multi-granularity feature extraction unit; The global feature fusion structure composed of global residual connection. We evaluated our proposed algorithm on the CK+ dataset and FER2013 dataset, and finally achieved 94.07% accuracy and 65.4% accuracy respectively. Experimental results show that our algorithm can effectively improve the accuracy of expression recognition tasks.

Automated segmentation of brain tumor is of great importance for the diagnosis and treatment. Although manual segmentation could achieve high accuracy, it is time and labor consuming, and not available batch processing. Recently, deep neural networks represented by U-Net have been successfully applied on medical image segmentation and achieved state-of-the-art performance. However, there is no standard segmentation method that can produce desirable results for all images, various segmentation methods are optimized to deal with specific areas and imaging modes, which have strong pertinence. In this paper, we propose a U-Net like by integrating residual blocks, convolutions with dense connection and auxiliary outputs for multi-class brain tumor segmentation. Firstly, residual blocks are used in the contracting and expanding path to extract deep semantic information of images. Secondly, densely connected convolutions are applied in the last layer of contracting path to strengthen feature propagation and feature reuse. Three auxiliary outputs are then employed as deep supervision before each upper sampling to decrease the probabilities of vanishing gradient and avoid overfitting. Finally, we use weighted cross entropy and generalized dice loss as a fused loss function to address the class imbalance problem in brain tumor data. The proposed model is evaluated on the BraTS 2018 training dataset and achieves state-of-the-art performance, with average dice scores of 0.884, 0.806, 0.702 for the whole tumor, tumor core and enhancing tumor core, respectively.

In this paper, a two-degree-of-freedom (2-DOF) speed control scheme for permanent magnet synchronous machines is developed. Both the reference-tracking and disturbance rejection performance are taken into account and only a single parameter related to expected bandwidth determines the dynamics. To improve the performance of speed control, a novel identification method for mechanical parameters is also proposed. The effectiveness of the proposed method is verified by simulations.

Neural coding schemes are powerful tools used within neuroscience. This paper introduces three different neural coding scheme formations for event-based vision data which are designed to emulate the neural behaviour exhibited by neurons under stimuli. Presented are phase-of-firing and two sparse neural coding schemes. It is determined that machine learning approaches, i.e. Convolutional Neural Network combined with a Stacked Autoencoder network, produce powerful descriptors of the patterns within events. These coding schemes are deployed in an existing action recognition template and evaluated using two popular event-based data sets.

The phase ambiguity is the main difficulty in the phase unwrapping of three-step phase-shifting. In this paper, a novel three-step phase-shifting phase unwrapping method based on sparse depth map is proposed. The sparse depth map is an auxiliary in phase unwrapping for depth acquisition. Firstly, a novel random grid pattern is proposed based on the random speckle and Delaunay triangulation in this algorithm. Secondly, the sparse depth can be calculated based on the proposed random grid. Finally, the phase unwrapping sequence is designed based on proposed quality of relative phase map (quality map). The calculated sparse depth is used as an auxiliary to solve the absolute phase according to designed phase unwrapping sequence. The depth can be acquired based on proposed phase unwrapping algorithm. In the experiment, the reliability of the proposed phase unwrapping algorithm has been verified in simple and complex scenes respectively. The phase ambiguity can be avoided by this method. Experiments show that this method has advantages in both accuracy and resolution.