Design And Research Of SUSAN Edge Detecting Algorithm

The edge is an important feature that human vision recognizes image target. Edge is also one of the most fundamental character of a image. Edge detection is one of the most important studying projects in image processing and computer vision. It is a hot point for researchers in the field of image measurement technology. On the basis of analyzing the existing methods of edge detection, this paper proposes an Canny adaptive method according to the requirement of edge detection in image measurement, Its main content is as follows:1. Digital image processing and its applications are introduced, then, the background and the significance of the image edge detection technique are elaborated, next to this, some exist problems of the image edge detection are discussed; Classic image noise filtering method is introduced, in detail elaborated the basic principle of these methods and procedures to achieve process; some classic methods are introduces, The basic concepts of these methods are summarized. These image edge detection methods are analyzed experimentally. The advantages and disadvantages are also analyzed.2. SUSAN operator and SUSAN edge detection algorithm which is not involved in gradient operation at all, only based on compared neighboring pixel grayscale are analyzed in details. This paper introduces the principle of the operator and realizes this SUSAN edge detector algorithm by using MATLAB. Finally, compared with other traditional detection operators in the process of testing images, a result shows that the effect of SUSAN edge detector to image containing noise is excellent than traditional edge detection method, for example, effectively improving the accuracy and reducing the missing rate, and making the edge more detailed and continuous.3. In order to extend the gray SUSAN edge detection method to color image, A method of SUSAN edge detection method is presented to detect the edges of color image. This new method first convert the RGB image into CIELAB color space, and then use the color difference to calculate the size of Univalue Segment Assimilating Nucleus (USAN) area. The final edges are extracted by thresholding. This new method can effectively detect edges of color images compared with traditional methods, the detected edges more consistent with the human visual characteristics.