Video Image Acquisition Research Articles

A Review and Comparative Study of Explainable Deep Learning Models Applied on Action Recognition in Real Time

Video surveillance and image acquisition systems represent one of the most active research topics in computer vision and smart city domains. The growing concern for public and workers’ safety has led to a significant increase in the use of surveillance cameras that provide high-definition images and even depth maps when 3D cameras are available. Consequently, the need for automatic techniques for behavior analysis and action recognition is also increasing for several applications such as dangerous actions detection in railway stations or construction sites, event detection in crowd videos, behavior analysis, optimization in industrial sites, etc. In this context, several computer vision and deep learning solutions have been proposed recently where deep neural networks provided more accurate solutions, but they are not so efficient in terms of explainability and flexibility since they remain adapted for specific situations only. Moreover, the complexity of deep neural architectures requires the use of high computing resources to provide fast and real-time computations. In this paper, we propose a review and a comparative analysis of deep learning solutions in terms of precision, explainability, computation time, memory size, and flexibility. Experimental results are conducted within simulated and real-world dangerous actions in railway construction sites. Thanks to our comparative analysis and evaluation, we propose a personalized approach for dangerous action recognition depending on the type of collected data (image) and users’ requirements.

Design and implementation of moving object detection based on Zynq

In recent years, video image processing technology has made rapid progress. Some algorithms related to moving target detection and tracking, relying on the rich resources of computers, make the implementation of these algorithms based on the PC platform has been relatively mature, but there are problems such as poor real-time performance, high power consumption, resource consumption. The Zynq hardware platform has both FPGA, which is good at parallel operation, and ARM core, which is good at control, which can effectively implement the algorithm transplantation of moving target detection and effectively improve the above shortcomings. The system takes Zynq hardware platform as the core unit, which is composed of video image acquisition module, image processing module, information transmission module, DDR cache module and display module. The PL side of Zynq implements video image acquisition, gray conversion, frame difference algorithm, expansion corrosion algorithm. The PS terminal mainly completes the configuration of IIC camera and VDMA. The output module finally realizes the output of HDMI through protocol conversion. This system has wide application value in moving target detection.

Clinical Efficacy of Medical Dextrose Tincture Liquid in the Treatment of Facial Photoaging.

Exogenous aging mainly refers to photo-aging, which is caused by environmental factors including ultraviolet exposure. Dextran is a homopolysaccharide composed of glucose as monosaccharide, and glucose units are connected by glycosidic bonds. The purpose of this study was to explore the clinical efficacy of medical dextrose tincture liquid (medical dextrose tincture) in the treatment of facial photoaging. Thirty-four volunteers were included in the randomized double-blind study. According to the random number table method, the subjects were randomized into control and treatment groups. The subjects in the control and treatment groups were treated with medical hyaluronic acid gel and medical dextrose tincture, respectively. They received mesotherapy therapy three times with an interval of 28 days between treatments. Video image acquisition was performed before treatment and 28 days after treatment. Skin moisture content, glossiness, heme content, collagen density, and elasticity were tested. The subjective evaluations of subjects and doctors before and after treatment were compared. Compared with the pre-treatment baseline, medical dextran tincture significantly increased skin moisture retention, skin gloss, and skin collagen density (p<0.001). Additionally, the skin retraction time was significantly reduced, and the skin retraction time was also markedly decreased after treatment with medical dextran tincture (p<0.001). The effects of medical dextran tincture were more significant in comparison with medical hyaluronic acid gel (p<0.05). The subjective evaluation results of doctors showed that after 84 days of treatment, the overall skin photoaging score was significantly reduced (p<0.001). The subjective evaluation results of volunteers showed that the various skin problems of more than 50% of volunteers were improved after treatment. Medical dextran tincture has obvious effects of moisturizing, increasing luster, improving skin redness, increasing skin collagen content and enhancing skin elasticity.

Acquisition and Processing of Medical Ultrasonic Video Images

Acquisition and Processing of Medical Ultrasonic Video Images

Motion Video Image Based on Sensor and Multiprocessing Technology in Track and Field Sports

In order to use motion video image processing technology to better guide and evaluate athletes’ technical and tactical ability, this paper first constructed a motion video image acquisition system. Secondly, a fuzzy kernel estimation method for uniform linear motion based on cepstrum property is proposed. Then, the motion blur areas to be processed are selected, and only these areas are deblurred. Finally, the effective removal of local motion blur is realized, and a clear scene image is obtained. Then, the maximum value of the total entropy of the image is calculated by using the information entropy theory, and the particle swarm optimization algorithm is introduced to find the maximum threshold of image segmentation. Finally, small wave optical flow estimation algorithm and rectangular window scanning algorithm across scales of motion image target detection algorithm are proposed, to not only solve the traditional optical flow estimation for fast moving object detection accuracy but also improve the efficiency of the optical flow computation. Compared to many kinds of algorithms, this paper proposed an algorithm that can improve the accuracy of moving target detection and measurement accuracy. And, the detection accuracy of the proposed algorithm is up to 86.5%. The estimated accuracy was as high as 65%. The segmentation accuracy is up to 95%.

Research on Intelligent Evaluation System of Sports Training based on Video Image Acquisition and Scene Semantics

This paper proposes to check the travel target of the dynamic background in the video surveillance with a fixed camera. A travel target detection method based on video picture acquisition and scene semantics for surveillance video was proposed. First, on the basis of combing the concepts and methods of picture recognition, the semantic information of the scene was fused to eliminate the interference factors in the unnecessary detection area. Secondly, a remote sensing picture visual feature representation method containing a semantic recognition method of remote sensing picture scenes and CSIFT features based on PLSA was presented. 10 types of typical remote sensing picture scenes are used for tests, and the visual vocabulary extraction method remains the same. The fixed visual vocabulary was 600, and the potential semantic subjects changes between 8∼50. The test results indicated that the highest average recognition rate was obtained when the latent semantic topics were 20. Inappropriate latent semantic topics will lead to a decline in recognition rates. The effectiveness of this method was fully verified.

Terminal velocity and drag coefficient models for disc-shaped particles based on the imaging experiment

Terminal velocity and drag coefficient are the core parameters of particle motion, but their predictive models and measuring methods have rarely been focused. In this work, terminal velocity and drag coefficient models for disc-shaped particles are proposed based on the imaging experiment. Experimental devices include particle sedimentation and video image acquisition systems. A total of 120 disc-shaped particles sedimentation are investigated. Results show that terminal velocity closely relates to the shape, size and density of the particles. Thus, particle shape factors should be considered in the drag coefficient model. The proposed models fitted by the LM + UGO optimization algorithm, have more accuracy and stability than the published models. The present study is applicable for the disc-shaped particles in the ranges of sphericity φ = 0.56–0.77, flatness ratio da /dv = 0.65–1.21, circularity ψ = 0.59–1, and Reynolds number Re = 323–10,800. Furthermore, the imaging measurement method can be extended to other non-spherical particles.

Abstract 12481: Automated Assessment of Video Quality and Ejection Fraction in Emergency Department Point-of-Care Echocardiograms

Introduction: Bedside point of care ultrasound (POCUS) has greatly improved bedside emergency medicine through improved cost-effectiveness, prompt availability to augment physical exams, and the ability to evaluate unstable patients at the bedside. While most emergency rooms have ultrasound machines, there is severe underutilization of this bedside imaging modality, in part due to the lack of formalized ultrasound training in most emergency room physicians. Accuracy of bedside assessment of cardiac function remains variable and operator dependent. Methods: To aid with interpreting and improving the quality of emergency room transthoracic echocardiograms (TTEs), we train a deep learning algorithm to take POCUS videos and identify patients with a reduced ejection fraction. Our algorithm also identifies which POCUS videos are low quality and need to be retaken. We curated all patients who received POCUS echocardiograms in the apical 4 chamber view from Stanford ED between July 2020 and November 2020, including 927 videos from 574 patients recorded from three different machines (Phillips, GE venue, and Mindray). Each video is annotated with their ejection fraction (775 normal, 152 reduced) and their quality (768 sufficient, 159 low). Results: Using these videos, we trained and evaluated our model using 10-fold cross-validation over patients, achieving an AUROC of 0.92 (0.89 - 0.94) for predicting ejection fraction on videos of sufficient quality and an AUROC of 0.81 (0.78 - 0.85) for predicting video quality. We use our model to identify videos with qualities ranked in the top quartile and find that on these videos, our model achieves an AUROC of 0.94 (0.91 - 0.97). Conclusions: This algorithm can be run on commodity hardware in real-time, aiding in ultrasound education and significantly decreasing the barrier to entry into ED POCUS for TTEs. the quality predicted by the model can be given as feedback to physicians to aid with improving video quality and image acquisition.

A New super-resolution restoration method with Generated Adversarial Network for underground video images in coal mines

The computer can be used in Super-resolution reconstruction (SR) to process low-resolution images to obtain high-resolution images. Aiming at solving problems of complex underground video image acquisition environment, uneven brightness, blurred images etc, this paper adopts the idea of deep learning to perform super-resolution restoration of underground video images in coal mines, and proposes a generational confrontation network to super-resolution underground video images in coal mines. The experiment proves that Generated Adversarial Network (GAN), while being compare with Super-resolution Deep Convolutional Neural Network (SRCNN), Efficient Sub-Pixel Convolutional Neural Network (ESPCN), Deeply Recursive Convolutional Network (DRCN) the effect of GAN method is better, because it can better realize the super-resolution restoration of underground video images in coal mines and provide preliminary support for the subsequent and further application research of underground images in coal mines.

Research on the Application of Rigid-flexible Compound Driven Fruit Picking Robot Design in Realizing Fruit Picking

This article mainly designs an efficient fruit picking robot that combines rigidity and flexibility. The design of the robot adopts a mechanical arm with 5 degrees of freedom for the mecanum wheel. The mechanical arm is equipped with a gripper made using artificial muscle technology. The ZED2 binocular vision system is used to obtain spatial coordinates to realize the positioning and positioning of the mechanical arm. Video image acquisition. In order to realize the automatic recognition and picking of strawberries, this paper creates a strawberry training set, and uses the yolov3 algorithm to train the data and obtains a strawberry recognition model, which has a recall rate of larger strawberries and immature strawberries and the precision rate reached 90%, and the detection precision and precision rate of small target strawberries reached more than 80%. It can Meet the requirements of picking in real life.

Design of video image acquisition and SDI display in large field of view

Design of video image acquisition and SDI display in large field of view

Application of SVGA video real-time transmission technology in music education information communication

The current music education information has certain information loss in the real-time transmission process, which leads to poor educational effect. The research content of this thesis is based on FPGA video image acquisition and processing system. At the same time, this research mainly uses FPGA as the platform to realize and simulate video acquisition, transformation, storage, display and transmission. This research solves the problem of long-distance real-time transmission of high-definition video stream, and in order to improve the subjective quality of the recovered video at the receiving end, the algorithm processing is added to reduce the blockiness phenomenon in the video displayed at the receiving end. Finally, the validation test was designed to validate the research perspective. The experimental results show that the system works stably and realizes the functions of video image acquisition, conversion, display and transmission, and achieves the design goal. However, due to time constraints and other conditions, the entire design needs further improvement, so that it can provide a basis for the development of subsequent music education information communication technology.

Low-light image enhancement algorithm based on an atmospheric physical model

Under low illumination, the colour constancy of human vision can be used for correctly determining the colour of objects according to the fixed reflection coefficient of external light and objects. However, video image acquisition equipment does not implement the colour constancy characteristic of the human visual system. Under low illumination, only a small amount of light is reflected from the surface of the imaged object; as a result, the captured image is underexposed. After statistical analysis of low-light images, these inverted underexposed images appear foggy. Inversion is a uniform and reversible operation that is performed on the entire image. Hereby, a method is proposed for resolving low-light images using conventional physical models. First, a low-light image is inverted for obtaining a foggy image. Subsequently, a pyramid-type dense residual block network and a dark channel prior K-means classification method are applied to the foggy image, to calculate the transmission and atmospheric light. Finally, the parameters obtained from this solution are incorporated into the low-light imaging model to obtain a clear image. We subjectively and qualitatively analysed the experimental results, and used information entropy and average gradient for objective quantitative analysis. We demonstrate that the algorithm improves the overall brightness and contrast of the imaged scenes, and the obtained enhanced images are clear and natural.

Video Image Acquisition Technology of Underwater Vehicle Based on 3D Sensor

ABSTRACT Guo, G., 2019. Video image acquisition technology of underwater vehicle based on 3D sensor. In: Hoang, A.T. and Aqeel Ashraf, M. (eds.), Research, Monitoring, and Engineering of Coastal, P...

Design of Real Time Video Image Acquisition and Processing System Based on FPGA

Based on characteristics of instantaneously, large capacity and miniaturization of digital image acquisition and processing technology, a real-time video image acquisition and processing circuit system was designed, FPGA was used as the center of controlling the whole system and the center of processing image data. With DDR2 SDRAM the core component of high speed storage module, and CMOS 7670 the video image acquisition device. Through the Quratus II and Modelsim software for edge detection algorithm mand the control process, each module of the hardware design and simulation, the realization of the video images from the collection and storage to the treating and display, which shows smooth, clear and real-time.

Design and Application of Insulator Detection Robot System for UHVDC Transmission Line

In recent years, with the vigorous construction of the ultra-high voltage grid, the line operating voltage level has become higher and higher. In order to ensure the safe operation of the transmission line, the line insulator string has a length of several tens of meters, and traditional manual testing has been unable to meet the requirements. With the continuous maturity of wireless communication technology, electromagnetic compatibility technology, sensor technology and other comprehensive technologies, it is necessary to develop an insulator string detection robot to replace artificial insulator detection work on UHV transmission lines. By means of wireless sensor communication technology, electromagnetic compatibility technology, video image technology and many other technologies, the insulator detection robot designed in this paper has realized automatic detection of zero resistance value of insulator of UHV transmission line. In addition, the real-time monitoring of insulator resistance value can be carried out through wireless communication to achieve the purpose of timely alarm. In addition, the robot's motion status and the insulator's field status can be observed in real time through video image acquisition. The insulator detection robot designed in this paper realizes the automatic detection of the zero value of insulator of the UHV DC transmission line. It can adapt to the UHV AC DC transmission line with different voltage levels by changing the length of the climbing arm, and at the same time adapt to the climbing detection of the string insulator of tension tower and the vertical insulator of the straight tower. The problem that the insulator string of the UHVDC transmission line can not be automatically detected by electric charge has been solved. It greatly improves the efficiency of insulator detection and the automation level of power transportation maintenance, reduces the cost of human transportation, and creates higher economic and social benefits.

Development of an Automatic Crack Inspection System for Concrete Tunnel Lining Based on Computer Vision Technologies

This paper presents an automatic crack inspection system for concrete tunnel lining based on image processing technique and computer vision. The system includes a video image acquisition device, an image stitching software, and a crack detection software. Firstly, the video image acquisition device uses six video cameras and three illuminators mounted on a steel framework which is capable of sliding from side to top of the inspecting vehicle to shoot the full surface of the tunnel lining. Secondly, the image stitching software based on image matching technique was developed to create layout panorama from the tunnel lining surface images, making it easier to visualize a large and detailed section of the tunnel lining. Finally, we propose a semi-automatic crack detection software relied on a combination of image processing technique and interactive genetic algorithm (iGA) to crack detection for concrete surface images of the tunnel lining. Experimental results demonstrate the effectiveness of the proposed method.

Video image target monitoring based on RNN-LSTM

Traditional image object classification and detection algorithms and strategies cannot meet the problem of video image acquisition and processing. Deep learning deliberately simulates the hierarchical structure of human brain, and establishes the mapping from low-level signals to high-level semantics, so as to achieve hierarchical feature representation of data. Deep learning technology has powerful visual information processing ability, which has become the forefront technology and domestic and international research hotspots to deal with this challenge. In order to solve the problem of target space location in video surveillance system, time-consuming and other problems, in this paper, we propose the algorithm based on RNN-LSTM deep learning. At the same time, according to the principle of OpenGL perspective imaging and photogrammetry consistency, we use 3D scene simulation imaging technology, relying on the corresponding relationship between video images and simulation images we locate the target object. In the 3D virtual scene, we set up the virtual camera to simulate the imaging processing of the actual camera, and the pixel coordinates in the video image of the surveillance target are substituted into the simulation image, next, the spatial coordinates of the target are inverted by the inverse process of the virtual imaging. The experimental results show that the detection of target objects has high accuracy, which has an important reference value for outdoor target localization through video surveillance images.

A New Learning Control System for Basketball Free Throws Based on Real Time Video Image Processing and Biofeedback

Shooting free throws plays an important role in basketball. The major problem in performing a correct free throw seems to be inappropriate training. Training is performed offline and it is often not that persistent. The aim of this paper is to consciously modify and control the free throw using biofeedback. Elbow and shoulder dynamics are calculated by an image processing technique equipped with a video image acquisition system. The proposed setup in this paper, named learning control system, is able to quantify and provide feedback of the above parameters in real time as audio signals. Therefore, it yielded to performing a correct learning and conscious control of shooting. Experimental results showed improvements in the free throw shooting style including shot pocket and locked position. The mean values of elbow and shoulder angles were controlled approximately on 89o and 26o, for shot pocket and also these angles were tuned approximately on 180o and 47o respectively for the locked position (closed to the desired pattern of the free throw based on valid FIBA references). Not only the mean values enhanced but also the standard deviations of these angles decreased meaningfully, which shows shooting style convergence and uniformity. Also, in training conditions, the average percentage of making successful free throws increased from about 64% to even 87% after using this setup and in competition conditions the average percentage of successful free throws enhanced about 20%, although using the learning control system may not be the only reason for these outcomes. The proposed system is easy to use, inexpensive, portable and real time applicable.

Design of the Hotel Monitoring System for the Image and Video Collection

Abstract With the more attention of the security of the hotels, an effective hotel monitoring system was required. Based on the Samsung S3C2440, the proposed system uses the Linux system to accomplish the intelligent video surveillance. Using the image acquisition process and the embedded BOA server, designed an intelligent video image acquisition system. The system capturing image by the USB camera and building an embedded Web server, which complete the image acquisition, image processing, and the image transmission function. It can see the acquire result on to the LAN at the PC via the webpage. It could complete the function of image acquisition, image processing, image transmission, the realization of ultimate in local area net PC through a Web view the result.