Research On Image Edge Detection Algorithms Based On Artificial Immune Principle

Human beings see the world through images,graphics,video,and other forms.Edge detection can not only get the useful information about structure of the boundary,but also greatly reduce the data useless to be processed.Image edge detection has become one of the most important information tool in this rapidly developing society.Therefore,Research on image edge detection and extraction method is very important.The main characteristic of artificial immune system is that has a very complete dynamic adaptation,self learning,and,self feedback ability.Image edge detection is similar to artificial immune principle.Through simulating immune principle,the image edge is used to identify the "self" and "non-self" edge.In this research,we use artificial immune principle to detect edge.This paper puts forward the algorithms of image edge detection based on artificial immune principle.The algorithms,mainly simulate the positive selection mechanism of biological immune system,to construct detector set.We study the characteristic value matching rules,and integrate image gradient,the maximum inhibition,maximum TiDuCha characteristics to construct detector set.Two detectors are respectively used in binary detector and real value detector for image edge detection.We carried out some experiments on images with no noise and images with the salt and pepper noises,and performed a comparative analysis among the motheds presented in this paper,the classical canny algorithm and prewitt algorithm.