Abstract
With the improvement of student education concepts in major sports colleges, new advanced algorithms need to be introduced for the design and modification of handball training for college students. The aim of this study is to design and develop a comprehensive evaluation system for the handball training level of college students, in order to improve training efficiency and athlete skill performance. Given the advancement of current sports education concepts and technological means, we have introduced advanced Azure Kinect dynamic body sensing devices and combined them with the AFM (Advanced Feature Matching) algorithm to construct a handball motion pose capture and evaluation model. Utilizing the high-precision depth camera and colour camera of Azure Kinect, real-time capture of three-dimensional motion data of handball players performing specific actions such as standing in a large Z-shape c defence, raising hands for defence, right-hand upward passing, and shooting. Using the AFM algorithm to process the bone node data obtained from Azure Kinect, constructing a motion posture model, and quantitatively evaluating the athlete's movements by comparing them with the standardized movements in the standard database. Through comparative analysis, it was found that college handball players are generally able to complete the prescribed movements, but there are significant differences in the details of the movements, such as the stability of body posture and the accuracy of angles, compared to professional athletes. The handball training level evaluation system designed in this study based on Azure Kinect and AFM algorithm provides accurate and efficient training feedback for college handball players, which helps to improve training quality and athlete skill level.
Keywords
Get full access to this article
View all access options for this article.