Affiliations: [a] Information Engineering College, Zhengzhou Institute of Technology, Zhengzhou, China
| [b] Information Engineering College, Huanghe Science And Technology College, Zhengzhou, China
| [c] Department of Statistics, University of Oxford, Oxford OX1, England
Correspondence:
[*]
Corresponding author. Jingtian Liu, Information Engineering College, Zhengzhou Institute of Technology, Zhengzhou 450000, China. E-mail: qingcuen002@163.com.
Abstract: The traditional algorithms reduce the recognition accuracy because of the influence of the fluctuation of the camera position during the walking of the robot. For this reason, a new intelligent recognition algorithm for color vision image position of soccer robot is proposed. The structure of the soccer robot vision system is designed. The panoramic visual sensor VS-C450 N-RC and the image acquisition device based on the IEEE 1394 standard are used to obtain color visual images, and the acquired distorted images are processed. Comparing color patches, an effective color patches scheme is proposed based on practice. RGB space is converted into HIS space, color, saturation and brightness are used to represent colors. According to the principle of contour extraction, an effective color patch extraction and recognition algorithm is proposed to match the robots on the actual field so as to obtain information such as the position of the soccer robot. The pose information of the robot is represented by the pose information of the color patches, and the position of the color visual image of the soccer robot is determined. Experimental results show that the proposed algorithm has high recognition accuracy.
