A color edge extraction method on color image

In the field of medical image recognition, industrial defect detection and other fields, because it is impossible to extract the edge of the color image directly, the recognition effect is poor. In order to extract the edge of the color image directly and retain the color information of the edge. This paper proposes a color image edge extraction method called function series. In this paper, the template threshold denoising method is used to denoise the input image. Then the key features of the input image are enhanced by using the image key feature enhancement function proposed in this paper. Then use the function series to extract the color features of the H, S, I space of the input image. The extracted color features are decomposed by eigenvalue decomposition, and the color information with larger eigenvalue is retained to enhance the color information of the edge. Finally, the edge of the target image with color information is obtained by the inverse transformation of the function series. The experimental results show that the correct edge recognition rate of this method is 92.09%. Moreover, this method can accurately extract the contour of the color image, and the color information is preserved completely. Compared with other methods, the correct recognition rate of this method is higher than that of the comparison method and can retain more abundant color information. The method proposed in this paper provides a new extraction method for color image edge extraction, which has certain practical value.