Error Motion Tracking Method for Athletes Based on Multi Eye Machine Vision

Traditional methods are unable to perform three-dimensional detection of erroneous movements, resulting in insufficient accuracy in tracking athlete erroneous movements. Therefore, the tracking method for athlete erroneous movements based on multi eye machine vision is highlighted. Using multi eye machine vision technology to construct an athlete error motion tracking framework and obtain athlete error motion image timing. From the perspective of regional consistency and similarity, segment machine vision images of athlete’s incorrect actions. Apply Canny operator to detect athlete’s incorrect actions, obtain pixel values of edge images, and remove false edges. The design is based on a multi eye machine vision athlete error action recognition process, obtaining unknown vectors. With the support of a multi eye machine vision detection system, the absolute value of brightness difference between two frames of images is calculated, and the Hom Schunck algorithm is combined to track the optical flow field to achieve athlete error action tracking. From the experimental verification results, it can be seen that the tracking curve of this method for three types of erroneous actions is consistent with the actual curve, and the maximum tracking accuracy is 93%, which can accurately track athlete’s erroneous actions.