Search

This paper uses adaptive BP neural networks to conduct an in-depth examination of eye movements during reading and to predict reading effects. An important component for the implementation of visual tracking systems is the correct detection of eye movement using the actual data or real-world datasets. We propose the identification of three typical types of eye movements, namely, gaze, leap, and smooth navigation, using an adaptive BP neural network-based recognition algorithm for eye movement. This study assesses the BP neural network algorithm using the eye movement tracking sensors. For the experimental environment, four types of eye movement signals were acquired from 10 subjects to perform preliminary processing of the acquired signals. The experimental results demonstrate that the recognition rate of the algorithm provided in this paper can reach up to 97%, which is superior to the commonly used CNN algorithm.

This paper proposes a novel method of extracting roads and bridges from high-resolution remote sensing images based on deep learning. Edge detection is performed on the images in the road area along with the road skeleton line, and the result of the detected binary edge is vectorized. The interference of protective belts on both sides of the road, road vehicles, road green belts, traffic signs, etc. and the shadow interference of the bridge itself are eliminated to determine the parallel sides of the road. The bridge features on the road are used to locate the detected bridge and obtain information such as the location, length, width, and direction of the bridge, verifying the experimental results of the Shaoguan Le point images. In addition, in order to learn higher-level road feature information, the algorithm in this paper introduces the hollow convolution and multicore pooling modules. Secondly, the residual refinement network further refines the output of the prediction network to improve the ambiguity of the prediction network results. In addition, in view of the small proportion of road pixels in remote sensing images, the network also integrates binary cross entropy, structural similarity, and intersection ratio loss function to reduce road information loss. The applicability of the proposed study was tested, and the results show that the algorithm is very effective for the extraction of road and bridge targets.

Monocular SLAM has attracted more attention recently due to its flexibility and being economic. In this paper, a novel metric online direct monocular SLAM approach is proposed, which can obtain the metric reconstruction of the scene. In the proposed approach, a chessboard is utilized to provide initial depth map and scale correction information during the SLAM process. The involved chessboard provides the absolute scale of scene, and it is seen as a bridge between the camera visual coordinate and the world coordinate. The scene is reconstructed as a series of key frames with their poses and correlative semidense depth maps, using a highly accurate pose estimation achieved by direct grid point-based alignment. The estimated pose is coupled with depth map estimation calculated by filtering over a large number of pixelwise small-baseline stereo comparisons. In addition, this paper formulates the scale-drift model among key frames and the calibration chessboard is used to correct the accumulated pose error. At the end of this paper, several indoor experiments are conducted. The results suggest that the proposed approach is able to achieve higher reconstruction accuracy when compared with the traditional LSD-SLAM approach. And the approach can also run in real time on a commonly used computer.

At present, the human visual perception system is the most effective, accurate, and fast image processing system in the world. This is because human eyes have some special visual features, but the features closely related to image enhancement include color constancy and brightness constancy. This paper presents a new image enhancement framework and computational model which can better simulate human visual features. It is based on the analysis of color constancy and luminance constancy and Retinex theory. And, this is a new image enhancement method in the compressed domain based on Retinex theory. In Retinex theory, DCT coefficients consist of incident components (DC coefficients) and reflection components (AC coefficients). By adjusting the dynamic range of DC coefficients, carefully adjusting AC coefficients, and using the threshold method for block suppression, the compressed domain image can be enhanced. On the basis of Retinex theory, the incident light and reflected light components are considered synthetically, the dynamic range (DC coefficient) of the incident light component and the details of the reflected light component (AC coefficient) are adjusted, and then the incident light component is reexamined. Moreover, it achieves a better image enhancement effect and avoids the blocking effect.

With the rapid development of the information age, the efficiency of image information dissemination has been improved and the way of information dissemination has gradually moved from text information to image information. In the process of using equipment to take pictures, because of some objective reasons, the images taken are different from the ideal images taken by the equipment, so the interference brought by these objective factors to the images is eliminated, thus presenting a more realistic image process. In the process of network propagation, degraded images show different characteristics in the network. In the process of propagation, images degenerate again, which makes it difficult for images to be authentic or restored. In this paper, an SIR model is selected from three classical infectious disease models to simulate and reflect the propagation path and influence of degraded images and the influence of degraded images on propagation is evaluated by extracting the moderate and degree distribution of undirected network. In addition, the distribution and integration between nodes are evaluated to distinguish the average road sources. Based on the SIR propagation model, a propagation model of information timeliness is constructed. By describing the update of subjective attitude values of nodes and then defining the probability function of state transition between different nodes, the model has higher fitting and adaptability. Finally, using BA, WS, Facebook, and Sina Weibo as the base map and setting the network environment parameters, based on the SIR model, the propagation of degraded images in different network environments is analyzed and the influence results of degraded images in network propagation are obtained.

In this paper, we study a task of slope collapse detection (SCD) for river embankment and formulate it as the tasks of motion detection and image recognition. Specifically, we introduce an SCD method based on motion detection and image recognition technologies to help inspector attendants detect the slope collapse. In this method, we use the foreground motion detection algorithm to identify the slope collapse of the scene of the river embankment. Since the moving targets in the foreground may not only be the slope collapse but also maybe some biology, we further use the image feature extraction and image recognition technology to recognize the foreground motion area, thus eliminating the influence of the biology on the detection results. Experimental results on the relevant scene data show that the proposed method can identify the slope collapse in real-time, and can effectively eliminate the motion interference of the biology, which has a high practical value.

For image registration, feature detection and description are critical steps that identify the keypoints and describe them for the subsequent matching to estimate the geometric transformation parameters between two images. Recently, there has been a large increase in the research methods of detection operators and description operators, from traditional methods to deep learning methods. To solve the problem, that is, which operator is suitable for specific application problems under different imaging conditions, the paper systematically reviewed commonly used descriptors and detectors from artificial methods to deep learning methods, and the corresponding principle, analysis, and comparative experiments are given as well. We introduce the handcrafted detectors including FAST, BRISK, ORB, SURF, SIFT, and KAZE and the handcrafted descriptors including BRISK, FREAK, BRIEF, SURF, ORB, SIFT, KAZE. At the same time, we review detectors based on deep learning technology including DetNet, TILDE, LIFT, multiscale detector, SuperPoint, and descriptors based on deep learning including pretrained descriptor, Siamese descriptor, LIFT, triplet network, and SuperPoint. Two group of comparison experiments are compared comprehensively and objectively on representative datasets. Finally, we concluded with insightful discussions and conclusions of descriptor and detector selection for specific application problem and hope this survey can be a reference for researchers and engineers in image registration and related fields.

The automatic segmentation and classification of an unknown motion data stream based on given motion classes constitute an important research problem with applications in computer vision, animation, healthcare, and sports sciences. In this paper, the scenario of trampoline motions is considered, where an athlete performs a routine consisting of sequence of jumps that belong to predefined motion classes such as somersaults. The purpose of this study was to make theoretical discussions on the turning starting time and starting technique of trampoline somersault based on image recognition and point out that the appropriate turning starting time of trampoline somersault is the event when the spring net of the trampoline recovers and applies force to the human body, and the overturning start exists in the latter half of the take-off action. It is considered that how to obtain the ideal full reaction force of the net facing the human body is the flip starting technique. This work analyzes the key steps and events for trampoline somersaults and the application of artificial intelligence for the recognition of actions in the healthcare and sports fields. The effectiveness of the proposed study is shown through experimental results. The study can facilitate the process of recognition of trampoline somersault.

The optical images collected by remotely operated vehicles (ROV) contain a lot of information about underwater (such as distributions of underwater creatures and minerals), which plays an important role in ocean exploration. However, due to the absorption and scattering characteristics of the water medium, some of the images suffer from serious color distortion. These distorted color images usually need to be enhanced so that we can analyze them further. However, at present, no image enhancement algorithm performs well in any scene. Therefore, in order to monitor image quality in the display module of ROV, a no-reference image quality predictor (NIPQ) is proposed in this paper. A unique property that differentiates the proposed NIPQ metric from existing works is the consideration of the viewing behavior of the human visual system and imaging characteristics of the underwater image in different water types. The experimental results based on the underwater optical image quality database (UOQ) show that the proposed metric can provide an accurate prediction for the quality of the enhanced image.

Three-dimensional extension of the high efficiency video coding (3D-HEVC) is an emerging international video compression standard for multiview video system applications. Similar to HEVC, a computationally expensive mode decision is performed using all depth levels and prediction modes to select the least rate-distortion (RD) cost for each coding unit (CU). In addition, new tools and intercomponent prediction techniques have been introduced to 3D-HEVC for improving the compression efficiency of the multiview texture videos. These techniques, despite achieving the highest texture video coding efficiency, involve extremely high-complex procedures, thus limiting 3D-HEVC encoders in practical applications. In this paper, a fast texture video coding method based on motion homogeneity is proposed to reduce 3D-HEVC computational complexity. Because the multiview texture videos instantly represent the same scene at the same time (considering that the optimal CU depth level and prediction modes are highly multiview content dependent), it is not efficient to use all depth levels and prediction modes in 3D-HEVC. The motion homogeneity model of a CU is first studied according to the motion vectors and prediction modes from the corresponding CUs. Based on this model, we present three efficient texture video coding approaches, such as the fast depth level range determination, early SKIP/Merge mode decision, and adaptive motion search range adjustment. Experimental results demonstrate that the proposed overall method can save 56.6% encoding time with only trivial coding efficiency degradation.

Capturing the body motion of fish has been gaining considerable attention from scientists of various fields. In this paper, we propose a method which is able to track the full-body motion of multiple fish with frequent interactions. We firstly propose to model the midline subspace of a fish body which gives a compact low-dimensional representation of the complex shape and motion. Then we propose a particle swarm-based optimization framework whose objective function takes into account multiple sources of information. The proposed multicue objective function is able to describe the details of fish appearance and is also effective through mutual occlusions. Excessive experimental results have demonstrated the effectiveness and robustness of the proposed method.

Aiming at the multiple target recognition problems in large-scene SAR image with strong speckle, a robust full-process method from target detection, feature extraction to target recognition is studied in this paper. By introducing a simple 8-neighborhood orthogonal basis, a local multiscale decomposition method from the center of gravity of the target is presented. Using this method, an image can be processed with a multilevel sampling filter and the target’s multiscale features in eight directions and one low frequency filtering feature can be derived directly by the key pixels sampling. At the same time, a recognition algorithm organically integrating the local multiscale features and the multiscale wavelet kernel classifier is studied, which realizes the quick classification with robustness and high accuracy for multiclass image targets. The results of classification and adaptability analysis on speckle show that the robust algorithm is effective not only for the MSTAR (Moving and Stationary Target Automatic Recognition) target chips but also for the automatic target recognition of multiclass/multitarget in large-scene SAR image with strong speckle; meanwhile, the method has good robustness to target’s rotation and scale transformation.