Research on denoising acceleration signal based athlete motion state detection

Due to the non-standard and inaccurate detection of the movement state of athletes, the subsequent adjustment and positioning error increases when facing a variety of motion states. Therefore, this paper proposes an athlete motion state detection method based on denoising acceleration signal. First of all, in order to overcome the influence of noise signal on the measurement accuracy of normal motion signal in motion state detection, the acceleration signal of athletes' inertial measurement is processed by empirical mode decomposition denoising method based on continuous mean square error criterion, and the motion acceleration signal after denoising is integrated twice in succession, thus the spatial trajectory of athletes is obtained. Secondly, the Generalized Likelihood Ratio Test is used to detect and identify different motion states of athletes, and adaptively match the corresponding detection modes to achieve accurate state detection in multiple motion states.