Method for Capturing Tennis Training Actions under Machine Vision Images

In recent years, tennis has gradually emerged, attracting a large number of participants, and computer technology has also been widely applied in this field. However, there are still some shortcomings in the professional training of tennis in motion capture. This article proposes a tennis training motion capture method based on machine vision images (MVI), which utilizes MVI technology to collect and capture tennis training movements. The collected tennis training motion images are used for motion object detection, motion feature collection and extraction, motion feature analysis and tracking. This article uses machine vision deep learning algorithms, combined with OpenPose pose estimation algorithm and YOLO algorithm, to reconstruct, segment, extract, and classify image features that capture actions. This article utilizes MVI enhancement technology to improve the resolution and adaptability of tennis training motion capture. The experiment shows that the method proposed in this article achieves optimization of tennis training motion capture, improving the accuracy and reliability of tennis training motion in MVI capture technology.