Analysis of Volleyball Video Intelligent Description Technology Based on Computer Memory Network and Attention Mechanism

There are some problems in the process of video intelligent description and analysis of volleyball, such as poor effective information extraction rate and poor dynamic tracking effect. Based on this, combined with long-term and short-term memory network and attention mechanism, this paper designs an intelligent description model of volleyball video based on deep learning algorithm and studies how to improve the extraction rate of volleyball video information through intelligent detection hardware and image recognition technology. This paper first introduces the application of image recognition technology and deep learning algorithm in the intelligent description of volleyball video, then designs the volleyball video and image recognition model based on deep learning algorithm according to the requirements of volleyball video intelligent description, and selects three correlation factors related to the impact indicators of volleyball skills. This study selects three characteristic parameters associated with volleyball video analysis indexes, namely, take-off, bounce, and hand movement, combined with image sensing hardware assisted sensor network to realize real-time monitoring of action state in volleyball video analysis system. The experimental results show that, compared with the current mainstream sports video intelligent analysis and image recognition methods with data analysis as the core, the intelligent volleyball sports video intelligent description and image recognition system based on the integration of deep learning algorithm and sensor hardware assistance has the advantages of good detection effect, high data effectiveness, low cost, and high efficiency of volleyball sports video analysis. It can effectively improve the efficiency of volleyball video intelligent description.

Value Analysis of Video Image Multiprocessing Based on Vision Sensor in the Field of Football Scientific Research

The development and application of mobile portable video capture and image processing systems, as well as the development and improvement of computer multimedia courseware technology have made image processing technology widely used in football project scientific research services. This research mainly discusses the value of video image multiprocessing based on vision sensors in the field of football scientific research. The development and conduct of this research work require the application of sports video image processing technology software to conduct scientific experimental research and analysis on the athletes’ technical and tactical level, evaluate the athletes’ technical and tactical performance through qualitative and quantitative tests, and give different sports items at the same time. Humans can exchange information with the outside world through vision, hearing, and language and can express the same meaning in different ways. However, current intelligent machines or computers require programs to be written in strict accordance with various machine languages. Only in this way can the machine run. In order to enable more people to use complex machines, it is necessary to change the past situation where people adapt to machines. Instead, let the machine adapt to people’s habits and requirements and exchange information with people in the way people are used to, that is, let the machine have the ability to see, hear, and speak. At this time, the machine must have the ability of logical reasoning and decision-making. Use sports video image processing technology and use examples to study and explore practical methods and means that are most in line with the scientific work of the project, and build the most scientific and useful sports video image processing system.. The CCD visual image sensor is used for signal acquisition and image processing when the motion video image processing system is established. Through logical analysis and induction, discuss and prove the intuitiveness, practicability, efficiency, and scientificity of the application of video image processing technology in the field of football scientific research. Finally, in the mathematical statistics method, the statistical software Excel 2003 in the office system is used to perform statistical analysis on all data results and analyze and compare the key kinematic parameters of some sports events. Analyzing the video images of the 2016 European Cup, the average possession rate of the Welsh team is 48%, and the average pass success rate per game is 82.7%. This research helps to improve the ability of football players to analyze, research, and evaluate sports skills.

Intelligent Sports Video Classification Based on Deep Neural Network (DNN) Algorithm and Transfer Learning

Traditional text annotation-based video retrieval is done by manually labeling videos with text, which is inefficient and highly subjective and generally cannot accurately describe the meaning of videos. Traditional content-based video retrieval uses convolutional neural networks to extract the underlying feature information of images to build indexes and achieves similarity retrieval of video feature vectors according to certain similarity measure algorithms. In this paper, by studying the characteristics of sports videos, we propose the histogram difference method based on using transfer learning and the four-step method based on block matching for mutation detection and fading detection of video shots, respectively. By adaptive thresholding, regions with large frame difference changes are marked as candidate regions for shots, and then the shot boundaries are determined by mutation detection algorithm. Combined with the characteristics of sports video, this paper proposes a key frame extraction method based on clustering and optical flow analysis, and experimental comparison with the traditional clustering method. In addition, this paper proposes a key frame extraction algorithm based on clustering and optical flow analysis for key frame extraction of sports video. The algorithm effectively removes the redundant frames, and the extracted key frames are more representative. Through extensive experiments, the keyword fuzzy finding algorithm based on improved deep neural network and ontology semantic expansion proposed in this paper shows a more desirable retrieval performance, and it is feasible to use this method for video underlying feature extraction, annotation, and keyword finding, and one of the outstanding features of the algorithm is that it can quickly and effectively retrieve the desired video in a large number of Internet video resources, reducing the false detection rate and leakage rate while improving the fidelity, which basically meets people’s daily needs.

Image Quality Evaluation of Sanda Sports Video Based on BP Neural Network Perception

In the special sports camera, there are subframes. A lens is composed of multiple frames. It will be unclear if a frame is cut out. The definition of video screenshots lies in the quality of video. To get clear screenshots, we need to find clear video. The purpose of this paper is to analyze and evaluate the quality of sports video images. Through the semantic analysis and program design of video using computer language, the video images are matched with the data model constructed by research, and the real-time analysis of sports video images is formed, so as to achieve the real-time analysis effect of sports techniques and tactics. In view of the defects of rough image segmentation and high spatial distortion rate in current sports video image evaluation methods, this paper proposes a sports video image evaluation method based on BP neural network perception. The results show that the optimized algorithm can overcome the slow convergence of weights of traditional algorithm and the oscillation in error convergence of variable step size algorithm. The optimized algorithm will significantly reduce the learning error of neural network and the overall error of network quality classification and greatly improve the accuracy of evaluation. Sanda motion video image quality evaluation method based on BP (back propagation) neural network perception has high spatial accuracy, good noise iteration performance, and low spatial distortion rate, so it can accurately evaluate Sanda motion video image quality.