图像分割是图像处理的重要步骤, 由于彩色图像含有的信息比灰度图像还多, 因而对彩色图像分割的研

究越来越受到人们的关注. 提出一种新的基于RGB 空间颜色相似性的彩色图像分割方法. 首先比较各种颜色模

型的优势与不足, 然后根据RGB 颜色空间的颜色信息和亮度信息提出一种计算在RGB 空间下颜色相似性的方

法, 再结合提出的图像颜色分量计算方法, 从而形成颜色分类地图, 最后根据颜色分类图进行像素划分, 得到分

割结果. 实验在Matlab 平台上进行, 结果表明: 对于颜色分明的图像, 该算法准确性高, 有较好的健壮性和较低

的计算复杂度, 能很好应用在图像前景与背景分割应用上.-Image segmentation is an important step in image processing, since the information contained in the color image more than the grayscale image, so the color image segmentation more and more people' s attention. A novel based on the RGB color space color similarity image segmentation. first comparative advantages and disadvantages of various color models, and propose a method to calculate the color similarity in the RGB color space based on information and brightness information RGB color space, combined with the proposed image color component calculation method, thereby forming a color-coded map of the last pixel is divided according to the color classification map, segmentation results obtained in experiments on Matlab platform, the results show that: for distinct color image, high accuracy of the algorithm has better robustness and low computational complexity, can be well applied in image segmentation foreground and background applications.