Research on Martial Arts Action Recognition and Matching Based on Kinect Data

With the continuous development of sports technology and computer vision, action recognition technology based on Kinect data has found wide applications in fields such as sports training, health monitoring, and intelligent entertainment. As a complex sport, martial arts require accurate recognition and matching of actions, which is of significant research value. This paper proposes a martial arts action recognition and matching method based on Kinect sensor data. By analyzing the skeletal data captured by Kinect, key features of martial arts actions are extracted, and action recognition is performed using machine learning algorithms. First, this paper preprocesses the action data collected by Kinect, including denoising, normalization, and feature extraction. Then, a hybrid action recognition model based on Support Vector Machine (SVM) and Convolutional Neural Network (CNN) is proposed, and it is compared with the traditional K-Nearest Neighbors (KNN) algorithm. Finally, an efficient martial arts action matching method is designed, where similarity measurement algorithms such as Dynamic Time Warping (DTW) and Euclidean distance are used for action matching. Experimental results show that the proposed method has significant advantages in action recognition accuracy and matching precision, enabling effective automated recognition and matching of martial arts actions, with high practical application value.