Abstract:
In FIRA MIROSOT robot-soccer game, the system of vision is the unique way throught which the whole system obtains the global information. The speed and precision of the vision system has important effect on the result of the game. According to the disadvantage of the traditional vision system, this paper puts forward a new design of color sign, and improves a method on color space conversion, color verdict and color image segmentation, so it dramatically reduces the work of calculation. Experimental results show the design has much supperiority on speed and precision.
Key words:
soccer robot,
image processing,
image segmentation
摘要: FIRA MIROSOT机器人足球比赛中,视觉系统是比赛系统中获得环境信息的唯一途径。视觉系统的识别速度和精度对比赛的胜负有很大的影响。针对传统的视觉缺陷,在色标设计方案、彩色空间的转换、色彩的阈值判别和图像分割几个方面进行改进,明显地减少了计算量,提高了实时性。实验结果表明,比赛中该方案在运算速度和精确度2方面都具有很大的优越性。
关键词:
足球机器人,
图像处理,
图像分割
CLC Number:
LIU Zuo-shi; HU Fa-huan. Fast Identification in Vision System for Soccer Robot[J]. Computer Engineering, 2007, 33(17): 234-236.
刘祚时;胡发焕. 足球机器人视觉图像的快速识别[J]. 计算机工程, 2007, 33(17): 234-236.