Recognition of Wrong Sports Movements Based on Deep Neural Network

Abstract

During physical education (PE), the teaching quality is severely affected by problems like nonstandard technical movements or wrong demonstrative movements. High-speed photography can capture instantaneous movements that cannot be recognized with naked eyes. Therefore, this technology has been widely used to judge the sprint movements in track and field competitions, and assess the quality of artistic gymnastics. Inspired by three-dimensional (3D) image analysis, this paper proposes a method to recognize the standard and wrong demonstrative sports movements, based on 3D convolutional neural network (CNN) and graph theory. Firstly, a 3D posture perception strategy for demonstrative sports movements was constructed based on video sequence. Next, the authors provided the framework of the recognition system for standard and wrong demonstrative sports movements. After that, a 3D CNN was stablished to distinguish between standard and wrong demonstrative sports movements. The proposed method was proved effective and superior through experiments. The research results provide a good reference for the application of 3D image analysis in the recognition of other body behaviors and movements.

Recognition and Difference Analysis of Human Walking Gaits Based on Intelligent Processing of Video Images

“…The generation of vector graph for human walking features is a prerequisite of gait recognition. First, the part affinity fields (PAFs) [18] were extracted by preprocessing gait images. Then, the PAFs of successive frames were combined into the vector graph, which serves as the input of feature extraction and classification network.…”

Information Extraction and Spatial Pattern Analysis of Sports Events Based on Deep Learning

“…All sorts of sports event information are delivered via information communication media, such as sports newspapers, sports television channels, sports websites, and sports new media [5][6][7][8][9][10]. e analysis of sports event information can disclose the public's concern over di erent types of sports events, re ect the spatial pattern of sports event information, and guide the organization of such events [11][12][13][14][15][16]. e information extraction and spatial pattern analysis of sports events greatly promote the formulation of scienti c development strategies for sports events, the optimization of the spatial distribution of sports events, and the dissemination of sports culture.…”