基于颜色空间映射的自适应足球机器人视觉系统

为了解决足球机器人比赛中视觉信息检测问题,给出基于嵌入式多处理器体系结构的视觉系统框架,并研究了一种自适应阈值目标分割方法。该方法将视觉系统采集的图像从YCbCr颜色空间转换为一维颜色映射图像,并提出基于竞争学习的空间映射方法。同时使用小波变换方法来平滑颜色映射空间直方图和降低阈值搜索维数。然后使用基于映射图像直方图动态搜索阈值的方法来分割图像。最后,通过实验表明研究的视觉系统的有效性,能够实时准确地完成视觉处理任务。

Self-adaptive vision system of robot soccer based on color space mapping

In order to solve the problem of vision information detection in robot soccer,the system structure of multi-processor embedded autonomous robot soccer and the vision system framework are introduced.Furthermore,a self-adaptive threshold segmentation method is proposed.The image with YCbCr color space is transformed into a one-dimension color mapping image.The color space mapping method based on competitive learning is presented.The wavelet transform method is used to smooth the histogram curves and reduce the dimensions of threshold searching.Then,using dynamic threshold method to segment image based on the color mapping image histogram.Finally,the experiment results showed the effectiveness of the proposed methods.And this method can accomplish the vision procession tasks accurately.