Artificial intelligence based tracking model for functional sports training goals in competitive sports

Abstract

As an advanced training concept, functional physical training is gradually recognized by top athletes for its high training effect and low sports injury. Functional physical training should gradually develop from elite athletes to grassroots athletes, so as to lay a solid foundation for the development of competitive sports. Because particle filtering is susceptible to external factors in moving target tracking, this paper designs a method for sparse coding using local image blocks of the target, establishes a static “impression” and dynamic model for the appearance of the target. The tracking problem is regarded as a binary classification problem between the foreground target and the background image. During the tracking process, the dual particle filter is implemented to alleviate the tracking drift, so that the algorithm can adaptively capture the changes in the target appearance At the same time, it can reduce the update caused by wrong positioning. The subjects’ FMS test and Y balance test have improved in varying degrees; the pressure distribution of the forefoot, arch, and heel tends to be rationalized, and the ratio of internal and external splayed feet has decreased. Experiments show that this particle filter moving target tracking scheme can adapt to changes in the environment and overcome the inflexibility of the global template when dealing with local changes in the target.