Sub-pixel Edge Detection Of Color Images Based On HSV Space

The Paper Supported by Natural Science Basic Research Plan in Shaanxi Province of China (Program No.2013JM8043).In recent years, computer measurement technology gets rapid development, it’s been a lot of new detection technologies, so, online measurement may achieve high accuracy and high speed standard. An important aspect of image measurement technology is the image edge detection, edge detection is extraction accurately useful information of image, it occupies an important position in digital image processing and computer vision, it has become the main technology to solve the problem of high precision in the current measurement technologies. Now there are many edge detection algorithms, but the lack of calculation speed and the algorithm itself leading to many limitations in practical edge detection process.The results of color image edge detection depend on the choice of color space largely, this paper first introduces the color space to solve the space expression problem of color image, and choose the HSV space which based on human visual perception to detect the edge. Then introduce some traditional dege detection algorithms, and classify some color image dege detection method, provide a theoretical basis for the algorithm in this paper. Analysis that the key of color image edge detection is the color distance problem, the definition of color distance is introduced, propose the HSVSin-cos color distance model combining the characteristics of HSV color space and the Euclidean distance calculation principle. Improve the SUSAN operator which has the superiority of detecting the image edge angle points to make it can be fully applied to the HSV space. Use the two menthods and the wavelet multi-scale which is researched hot recent years to detect the pixel-level edge. Further study the sub-pixel interpolation algorithm, analysis other sub-pixel methods, focus on the sub-pixel processing based on sub-pixel interpolation algorithm for pixel-level edge, thus we can get the sub-pixel edge of the color image.This algorithm makes some reasonable improvements on the color distance algorithm and SUSAN algorithm, combines sub-pixel interpolation algorithm, make up some shortcomings of traditionalcolor image edge detection algorithm to a certain degree. The experimental results show that this algorithm is an effective color image sub-pixel edge detection method in HSV space, it has some advantages that the traditional algorithms haven’t in the correct rate of edge detection and sub-pixel quality.