Title: Soccer player leg motion feature recognition method based on decision tree algorithm

Authors: Huimin Du

Addresses: Chengdu Sport University, Chengdu, Sichuan 610041, China

Abstract: In order to improve the accuracy of feature classification and reduce the identification time, this study designed a feature recognition method of soccer players' leg movements based on decision tree algorithm. Firstly, on the basis of collecting the leg movement images of soccer players, the image quality is improved by preprocessing. Then, the leg movements of soccer players are classified based on decision tree algorithm, and the leg movements feature sequence potential function is established according to the classification results. Finally, AdaBoost algorithm is used to identify the leg movement characteristics of soccer players. The experimental results show that the maximum accuracy of feature classification by this method can reach 97.0%, and the maximum time of feature recognition is only 35.47 s, indicating that this method can quickly and accurately identify the leg movement features of soccer players.

Keywords: soccer; leg action; image preprocessing; decision tree algorithm; feature recognition.

DOI: 10.1504/IJRIS.2022.10048821

International Journal of Reasoning-based Intelligent Systems, 2022 Vol.14 No.2/3, pp.91 - 96

Received: 14 Mar 2022
Accepted: 27 Apr 2022

Published online: 09 Sep 2022 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article