Abstract: Aiming at the large deviation problem of traditional measurement methods in motion parameters estimation, an adaptive motion parameters recognition method based on deformable shape was presented to improve the measurement accuracy of the motion parameters. Firstly, because athlete motion is a random process, the motion state shape matrix can be acquired by computing athletes centroid coordinates. Then the variability of shape number is calculated using rank of measurement matrix. Finally, the optimal calculation results are obtained under noise interference. The simulation results showed that this method could more accurately measure the movement parameters and improve the measurement accuracy of parameters.