Human-Machine Interaction Gesture Feature Recognition and Edge Repair Algorithm Research

Abstract

How to accurately detect and segment gesture parts from the facial area is a major research focus in the field of human-machine interaction. The similarity in texture between the hand and the face poses significant challenges to the stability of gesture recognition. This paper proposes a gesture recognition algorithm based on edge repair using a computer-based method to handle the issue of overlapping between hands and faces. Based on the Chamfer distance matching method, an edge repair algorithm is introduced to handle blurred edge areas at the intersec-tion of hands and faces. Experimental results show that the accuracy of the edge repair algorithm reaches 94.6%, significantly outperforming other methods with an accuracy of only 8.7%. This algorithm is particularly effective in scenarios involving frequent gesture changes and facial movement.