基于数学形态学的彩色图像边缘检测

Color image edge detection based on mathematical morphology

图像边缘检测大都是基于灰度图像的,其不能充分利用彩色图像的全部色彩信息。本文利用数学形态学在描述灰度图像方面的优势,将灰度形态学边缘检测法运用到彩色图像中。首先在RGB空间对原彩色图像的R、G、B三副伪灰度图像分别进行形态学边缘检测,然后对三副图像的边缘强度进行合并,并采用阈值分割法提取出图像的轮廓边缘。实验结果表明:该算法计算量小,并且充分利用了图像的色彩信息,能有效地检测出彩色图像边缘。

Image edge detection methods are mostly based on gray image,so they cannot make most use of all color information of the color image. This paper makes use of the advantages of mathematical morphology which describes gray image, to implement the gray morphological edge detection in the color images. Firsdy, the 3 pseudo - gray images R.G. B of the original image are processed by morphological edge detection. Then, the 3 edge intensities are combined, and the edge figure of the image is extracted by thresholding algorithm. The experiment addressed that to use this kind of algorithm costs less computation, and to mostly use the color information of the image, furthermore, to detect the edge of the color image efficiently.