摘要
Hough变换是形状识别的有效方法,在Robocup小型组(SSL)比赛的视觉处理系统中,可利用Hough变换对场地边界及机器人色标进行直线和圆检测,通过几何特征的约束,及适应场地情况的算法改进,可减少Hough变换的处理时间,以解决比赛场地标定繁复及在不同光照条件下机器人识别困难的问题。上述方法与颜色识别方法相结合,提高了比赛系统的稳定性及准确性。
Hough transform is an effective method of form identification. In the vision system of Robocup Small Size Leage（SSL）, it can be used to recognize the borderline of competition field, and identify the patch circles lay on the top of robots.With the effort of geometric characteristic restraint, and some algorithm improvements which are adapted to the competition conditions,the calc ulating time of the Hough transform can be reduced to solve the problems that the field＇s calibration work is complicated and it will be difficult to recognize the robots under varying illumination conditions.Cooperating with color recognition, this method can improve the stability and accuracy of the vision system.
出处
《机械设计与制造》
北大核心
2008年第1期167-169,共3页
Machinery Design & Manufacture
