The existed mathematical morphology for edge detection algorithm can' t detect the complete and sharp information of noise color image. This paper proposed a new approach to edge detection of color image, this approach was based on HIS color space to multi-scale and multi-structural elements of mathematical morphology. Using two unit elements: the scale and structure, the method expanded them from the horizontal and vertical two directions, and detected the color noise image through the form of face. Based on this method, separately detected the edge of the two components H and S which were carrying color information. Finally, weighted-synthesized them to get the color edge. The experimental results show that the algorithm has obvious denoising effect,it can obtain clearly color edge and abundant details, and the algorithm is feasible and effective for the color edge extraction.%针对已有的数学形态学边缘检测算法对彩色噪声图像检测到的彩色边缘信息不够完整、清晰,提出了一种基于HSI色彩空间的多尺度多结构元的数学形态学边缘检测算法,采用以尺度和结构两个单位元素进行横向和纵向的拓展,以面的形式对彩色噪声图像进行全面的边缘检测.基于这种理念分别对H和S两个携带颜色信息的分量进行边缘检测,最后将两分量的边缘信息通过加权合成得到彩色图像的彩色边缘.实验证明,该算法的去噪效果明显,得到的彩色边缘轮廓清晰、细节丰富,对彩色边缘的提取具有可行性和有效性.
展开▼
