Research on Gesture Recognition Based on Improved Template Matching Algorithm
Wanying Zhang, Hao Liu, Hao Ma, Wenjie Zhao · 2021 IEEE 2nd International Conference on Information Technology, Big Data and Artificial Intelligence (ICIBA) · 2021
Aiming at the problem of low recognition rate when recognizing the left and right hands with traditional gesture features, a feature$H$(Hand-to-hand angle) based on the angle between the hands is designed to describe the positional relationship of the left and right hands. At the same time, a new gesture recognition algorithm is proposed to solve the interference problem of undefined gestures in the recognition process. Based on the template matching method, similarity judgment is introduced to filter out interfering gestures. Finally, real-time classification and recognition of static gestures are realized. Experiments show that this gesture recognition algorithm can effectively filter out interfering gestures, and the recognition rate of static gestures is up to 96%.