Color Image Edge Detection Arithmetic Based on Color Space

A method to determine whether a pixel is the edge point of a color image is presented. First, according to the characteristic of RGB color space, a color triangle of a pixel is constructed with its color coordinates RGB, and then the perimeter and the internal angle of the color triangle which can be seen as the information value of a pixel is calculated. Last, comparing the value with that in its neighborhood, a pixel is in the edge point or not is determined. This method took the correlation of every color component into account reasonably, and converted the computation of a vector space to scalar quantity in a natural way. Comparing it with traditional method, experimental result shows that more legible edges can be detected with this method and can be realized easily.

I. Introduction

Edge is the place which has more or less mutations in image gray level or structure information, is the end of one area and another area to start [1]. Not only edge can convey a large number of information of the image edge and draw the basic contour of the object, so edge detection is very important to computer vision system, and has very important role in image segmentation, pattern recognition, computer vision and so on. Edge detection is an important part in the image processing, and is first-term work of all based on boundary segmentation method, such as 3-D reconstruction and image segmentation and so on. The field of image processing active research problem is edge detection, especially color image edge detection is a problem of deserve to research.