| One of the most basic features of the image is edge which reflects most information of it and is an important basis of the image segmentation, pattern recognition, intensification and the analysis and understanding of the image. Edge detection has been a hot focus of research which plays an important role in image processing. For a long time, scholars are mainly devoted to the research of the edge detection for grayscale images, but neglect the research of color image. As is known to all that most of the images in life are colored which can provide more information than the gray image. Therefore, color image processing is increasingly concerned.This article summarized the background and significance of edge detection as well as the history and development of it, introduced several classical edge detection operators, at the same time elaborated the basic theory of mathematical morphology and its application in image edge detection. And after analyzing and researching on the HIS color model, the innovation work is conducted as following:First, an H*I*S model is obtained by improving the original HIS color model. And then to detect HIS color space and improved HIS color space by means of the combination of the Vector gradient operator with Roberts operator, Sobel operator and Prewitt operator edge detection. And the result shows that the improved HIS model has a better image edge detection effect.Second, A multi-scale multiple structure and color image edge detection method is proposed by the combination with mathematical morphology which has a dominant position in image processing to solve the problems in the edge detection of the HIS model. Such as to cause breakpoint easily and the edge lost which lead to a discontinuity about some edges. The experimental results show that, the proposed algorithm can make full use of the hue, intensity and saturation information to eliminate the noise effectively and extract the complete edge information adaptively. |
PDF Full Text Request