基于计算机视觉技术的乒乓球技术动作AI评分系统研究与设计

Research and Design of AI Scoring System for Table Tennis TechniquesBased on Computer Vision Technology

鉴于目前乒乓球教学中存在学生挥拍动作不规范、个性化施教不足及教学效率相对较低的三大问题,基于AI(人工智能)计算机视觉技术设计乒乓球技术动作AI评分系统。该系统结合广东科学技术职业学院乒乓球精品课程建设,以乒乓球八个基本技术动作构建AI模型,使用目标检测、姿态估计、动作识别算法对学生的实际动作进行AI评分。实验结果表明,使用乒乓球技术动作AI评分系统有利于开展个性化教学,提高了教学效率,还能支持远程教学,可以在乒乓球教学中推广应用。

In view of the three major problems in current table tennis teaching,non-standard student swing movements,insufficient personalized teaching,and relatively low teaching efficiency,this paper proposes to design an AI(Artificial Intelligence)scoring system for table tennis technical movements based on AI computer vision technology.Based on the construction of top-quality table tennis courses in Guangdong Polytechnic of Science and Technology,an AI model with eight basic technical movements of table tennis is built,and target detection,posture estimation,and action recognition algorithms are used to score students'actual movements.The experimental results show that using the AI scoring system for table tennis technical movements is beneficial for carrying out personalized teaching,improves teaching efficiency,and supports distance teaching.Therefore,this proposed system can be promoted and applied in table tennis teaching.