Research On The Techniques Of Image Edge Detection

Image edge detection has become one of the most important parts of image processing. It is also the basis of image analysis.In the thesis, first of all, the background, the significance and also the development status of the image edge detection technique are introduced. In addition, the process and direction of image processing research is also described. As the basic concept of image processing is discussed, we know the most importance of the image edge detection.Secondly, some classical edge detection algorithms such as Roberts, Sobel, Prewitt and Laplacian are discussed. Theory analysis and experiments are carried out to compare their advantages and disadvantages. Also some improved methods for the classical operators are simply introduced after each one in this chapter.Then, the traditional Canny edge detection algorithm is lack of the automation process capabilities and limited orientations from calculating gradient in the edge detection. On the basis of the traditional Canny edge detection algorithm, a new method of scale-adaptive Gaussian filter with multi-directional gradient together is proposed as the improved Canny algorithm. The new method smooths the original image with scale-adaptive Gaussian filter, which retains the edge information and removes the impact of noise. In addition, according to gradient direction on each point, the new method can reduce noise and improve the egde location precision.Finally, with theoretical and experimental results, it is demonstrated that the improved Canny algorithm is simple and easy to realize, fulfilling the optimization criteria of the original algorithm. Moreover, compared with the traditional algorithm, the new method provides the automation process capabilities and shows the better accuracy and location precision in the edge detection.