Based on double threshold segmentation gait recognition method to design of motion control system for middle size league soccer robots
Zhenli Lu1, 2, 4, Kai Tian1, 3, Limin Mao1, Bin Li2
COMPUTER MODELLING & NEW TECHNOLOGIES 2013 17(5B) 7-13
1 School of Electrical Engineering and Automation, Changshu Institute of Technology, 215500, Changshu, China
2 State key Laboratory of Robotics, Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang, 110014, China
3 China University of Mining and Technology, 221116, Xuzhou, China
4 Institute of Electronics and Telematics Engineering of Aveiro, University of Aveiro, 3810-193, Aveiro, Portugal
To achieve a soccer match where robots against human, the key issue for robots is to effectively identify human action. Through analyzing the characteristics of the human gait, and combing with the motion character of the meddle size league soccer robot, based on the tracking information of human skeleton captured by Kinect, a gait recognition method of double threshold segmentation is proposed in this paper. The gait recognition field is demarcated by acquiring the depth information of the human skeleton. The “go-forward” and “go-backward” and “piaffe” gait rules are developed, which can also generate motion commands to control the middle size league soccer robot to represent the gait. The experimental results show that the proposed method can realize motion recognition in real time, which has strong practical performance and can provide technical reserves for related research.