A hybrid approach using human posture and contour for gait recognition under body occlusion

Abstract

This paper presents a hybrid method using a Gaitset network for gait recognition. We firstly use Alphpose model to obtain key points of human posture to improve the extraction of human contour under body occlusion. The human contour map is obtained by pose2Seg model with the key points of human posture. Both human posture and body contour are used as input of the network. By using the hybrid method, it can effectively improve the accuracy and robustness of gait recognition with the upper body information under occlusion. In scenes of human walking under body occlusion, our experimental results show that the accuracy of the proposed method in self-built database can reach 84.6% accuracy.