Research On Edge Detection Based On Prior Shape

Image is one of the most important means for people to obtain information.With the development of computer and the wide application of digital image processing,the image is also developed.Edge detection is very important in image segmentation,image processing,pattern recognition and image understanding,and it is widely used in many areas such as image analysis and computer vision,object recognition,detection and tracking.The edges of different objects are basically different,so these edges can be regarded as a simple description of image content.Edge is one of the important features of original image.So how to extract the edge from the image has become an important research problem.The edge detection algorithm based on Mumford-Shah model plays an important role in image processing.In this paper,we first propose a new variational model based on Modified Mumford-Shah model and automatically select prior shape model to automatically and adaptively search the prior shape from the library to guide edge detection.In order to calculate the new model,we use the high efficient algorithm of fixed point iteration and the split-Bregman algorithm to solve the new model.Numerical experiments show that our model can automatically select the appropriate shape prior to completion of the target according to the object edge.The result show that the excess edge almost invisible and the target of the edge is very clear.Our model can weaken the non-target edge and enhancement the edge of the object.In the case of noise and blurred image,the model of this paper can still obtain a complete edge.In addition,we also give the proof of the convergence of the new model.