Color Image Edge Detection Based On Canny Theory

Edge detection is the base of the image measuring, image segmentation, image compressing and pattern recognition and so on. Edge detection is one important subject in the field of computer vision and image measuring. How to detect the color edge exactly and quickly is one research hotspot in the area of digital image processing.Based on Canny algorithm, an adaptive Canny color edge-detection method using improved grad magnitude and direction algorithm is proposed. Canny color edge-detection not only keeps the traditional Canny's excellent performance in good localization, only one response to single edge and good detection, but also improves the performance in detail of color edge-detection and in restraining false edge. Adaptive Canny color edge-detection divides image into sub-image, and detects them with adaptive threshold value according to the whole image's edge information and sub-images', that improves the automation of color edge-detection. This method achieves good performance in practice. With introducing the basic thesis and concrete method, through experiment and test result analysis, we testify the adaptive Canny color edge-detection is an efficient improving approach on color edge-detection.