Video Analysis Method of Basketball Training Based on Deep Learning

Abstract

In order to solve the problem of poor data clustering effect of the original basketball shooting track capture and antisense capture method, a basketball shooting track automatic capture method based on background difference method is designed. Then the sensor network is designed to capture the trajectory. By using this method to improve the accuracy of basketball shooting trajectory automatic capture, ensure the integrity of the data. In the process of design, it paid attention to the installation position of the sensor and the processing process after data acquisition. Inertial sensors are used to ensure the real-time and effectiveness of data acquisition. The trapper is used to complete the data processing and trajectory capture. The background difference method is used to detect and process the target trajectory. Realize the automatic capture of shooting trajectory. This method can greatly improve the accuracy of track capture. After its popularization and application, it can effectively improve the track capture effect in basketball