基于模糊相似性的彩色形态学图像处理方法

为了将传统灰度图像数学形态学扩展到彩色图像，提出一种结合矢量空间模糊相似性的彩色形态学图像处理方法。首先，在RGB彩色空间中利用彩色矢量间的距离和角度定义模糊相似性测度，以刻画与人类视觉感知相一致的彩色相似程度；以上述相似性测度为准则定义彩色空间中任意一组彩色的上确界和下确界；利用中心像素及其结构单元内像素的上确界和下确界构建彩色形态学的基本操作，包括膨胀、腐蚀、开、闭等操作；进一步将提出的彩色形态学操作应用于高分辨率遥感图像，通过实验对比验证其对地物目标的形变和平滑能力，说明其实用性和有效性。

Color morphology image processing method based on fuzzy similarity

To extend the mathematical morphology for grayscale image to its alterative for color image, a color mathematical morphology based on fuzzy similarity defined on RGB color vector space is proposed in this paper. Firstly, in RGB color space, the fuzzy similarity measure, which is consistent with the human visual perception, is defined by using the distance and angle between two vectors to describe the similarity degree between the two colors. The upper and lower bound functions in a collection of colors from in RGB color space are defined by their similarity measures. The basic operations of the color mathematical morphology are formulated by using the upper and lower bounds for a any pixel and its neighbor pixels within the structural unit centered the pixel, including dilation, erosion, opening and closing operations. The proposed color morphological operations are applied to the high resolution remote sensing images, and the deformation and smoothing ability to objects are verified by comparing experiment, and its practicability and validity are explained.