Title: Research on human behaviour recognition method of sports images based on machine learning

Authors: Baojun Liu; Zimin Wang

Addresses: Department of Physical Education, Guilin University of Electronic Technology, Guilin, 541004, China ' School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, 541004, China

Abstract: In order to improve the efficiency of human behaviour recognition in sports images, a human behaviour recognition method based on machine learning is proposed. The adaptive mechanism is used to improve the ViBe algorithm to eliminate shadows, the dense structure is utilised to improve OpenPose to extract key bone points of the human body, and the support vector machine (SVM) classifier is adopted to perform human behaviour recognition, which realises the human behaviour recognition of sports images. The simulation results show that the proposed method can accurately recognise human behaviour in sports images, and the average recognition accuracy reaches 87.5%.

Keywords: sports; elimination of image shadows; extraction of key bone points; machine learning; support vector machine; SVM.

DOI: 10.1504/IJBIC.2024.136728

International Journal of Bio-Inspired Computation, 2024 Vol.23 No.2, pp.99 - 110

Received: 14 Dec 2022
Accepted: 10 Aug 2023
Published online: 19 Feb 2024 *

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