石家庄学院学报

针对传统健美操评分过程中评委主观性评价标准影响过大的特点,采用目标图形监测和朴素贝叶斯最优分类器理论,设计了一种健美操力度与动作的自动评分系统.当目标处于最优动作、有效发力且保持最优动作达到设定时间后,对该动作进行计分,以分解动作的总得分评判健美操整体分数.首先说明了系统的目标图形监测流程及硬件设置,阐述了朴素贝叶斯最优分类器在动作识别中的应用,最后针对系统进行了实验.结果表明:该自动评分系统能够有效地对目标健美操动作进行评分,方案完善,识别率较高,具有较强的应用价值.

In view of the excessive influence of the judges' subjective evaluation criteria in the traditional aerobics scoring process, an automatic scoring system for aerobics strength and action is designed by using the target graph monitoring and naive Bayesian optimal classifier theory. When the target is in the optimal movement,effective exertion and maintaining the optimal movement for the set time, the action is scored, and the total score of the movement is used to judge the overall score of aerobics. First, the system's target graphics monitoring process and hardware settings are explained, and the application of naive Bayesian optimal classifier in motion recognition is expounded. Finally, experiments are carried out on the system. The results show that the automatic scoring system can score the target calisthenics effectively, the scheme is perfect, the recognition rate is high, and it has a strong application value.