Student Physical Fitness Test System and Test Data Analysis System Based on Computer Vision

Computer vision technology is one of the main research directions artificial intelligence. With rapid growth image or video data scale and improvement computing power, computer has achieved unprecedented development in recent years widely used a variety scenes. This study mainly discusses design student physical fitness test system analysis based on vision. motion attitude determination algorithm to identify motion. In hardware configuration, key CPU GPU. The model realizes large-scale matrix computation parallel power provided by GPU uses realize reading preprocessing. assessment controller responsible for transmission instructions status information controls operation entire pitch system. It control center ZigBee wireless communication adopted as method human posture measurement terminal controller. input preprocessed through scaling standardization. scaled resolution 224 × when input, which performed training. was changed means random horizontal flip, rotation, color change achieve effect expanding dataset. Then, evaluation module evaluate various indexes body. During sit-up test, nine out 10 sit-ups can be accurately counted recognition rate reaches 90 percent. results show that designed this high accuracy good performance, students.