Evaluation of Basketball Shooting Technique Based on Video Analysis

In this paper, an Open Adversarial Pose Vector Machine (OAPVM) is proposed, which combines OpenPose, generative adversarial network (GAN) and support vector Machine (SVM) to evaluate basketball shooting technique. First of all, OpenPose was used to extract the key bone information of athletes in real time from the video to accurately capture the dynamic characteristics of the shooting action. Then, GAN is used to enhance the extracted key data, generate diversified shooting action samples, expand the data set and improve the generalization ability of the model. Finally, SVM is used to model and classify the key features after processing, and evaluate the normalization and quality of the shooting action. The experimental results show that BCEA performs well in the basketball shooting technique assessment task, and can distinguish high-quality and low-quality shooting movements more accurately than traditional methods. This fully verifies the effectiveness and feasibility of this method in the evaluation of basketball shooting technology based on video analysis, and provides a new technical means for realizing more intelligent sports analysis.