摘要
为解决水中机器人比赛易受比赛场地光线不均匀的影响而导致的机器人识别错误或丢失问题,提出一种基于区域处理技术的水中机器人定位方法。该方法综合利用直方图统计和区域处理方法对第一幅图像进行图像预处理、去背景操作,对机器人进行粗略定位,在得到的机器人区域中将色标分割出来,得到包括鱼的位置坐标和方向在内的各项信息。而后续图像的处理则主要采用区域处理方式,通过不断地修正机器鱼的位置信息,利用上帧图像的参数对本幅图像进行简化处理。结果表明,该方法能将光线等因素的影响降到最低,在一定程度上弥补了现在比赛平台色标识别上的不足。
To solve the problem that the image processing system used in robot fish water polo game can be easily influenced by the asymmetry of the light in the competition area and then the robot could be recognized wrongly or even got lost, this passage gives out a algorithm for image process which is based on regional treatment and then get the location of the underwater robot. This algorithm mainly process on the first image, remove the background and the get the coarse coordinate of the robot using histogram analysis and regional treatment in machine vision, and then segment the color code out, get the direction and location of the robot. The following process uses regional treatment, simplifies through using the data produced in last image and correcting the information of the robot. The result shows that this algorithm can reduce the influence of the light and make up the shortage in the image processing system used in robot fish water polo game.
出处
《兵工自动化》
2010年第11期74-78,共5页
Ordnance Industry Automation
关键词
区域处理
颜色空间
水中机器人
regional treatment
color space
underwater robot
