Research On Image Segmentation Method Based On Superpixel And MMTD

Image segmentation is to divide the image into a number of specific,unique regions and propose the technology and process of interest target.It is one of the most important research contents in computer vision and image processing.In recent years,the image segmentation method based on the graph cut theory represented by GrabCut has become a research hotspot.Graph cutting is an energy minimization method combining image edge information and texture information.GrabCut algorithm is an excellent interactive image segmentation algorithm based on graph theory.This algorithm is widely used because of its simple interactivity and good segmentation effect.However,when dealing with images with similar foreground and background,the GrabCut algorithm doesn't work very well,aiming at this problem,this article proposes an image segmentation method based on super pixels.In this aiticle,we first introduce the method of medium mathematics and the method of medium truth degree(MMTD),On this basis,the following superpixel extraction and image segmentation are performed.The main work is as follows:(1)First,the image enhancement algorithm for the image is processed.The image enhancement algorithm of this article is the histogram equalization algorithm.For the histogram equalization algorithm,the data is not selected,an improved histogram equalization algorithm is processed.Image enhancement is for the next work.(2)Secondly,extracting the image pixel block.For the extraction of traditional SLIC super-pixel block,the determination of weight value needs to be specified manually,and the same weight value for different images can lead to the problem of unsatisfactory segmentation effect.,this aiticle puts forward a super pixel extraction algorithm based on the mediation degree of truth value measurement method.(3)In view of the poor segmentation effectof GrabCut algorithm in processing foreground and background similar images,an image segmentation algorithm based on super pixels is proposed.Based on previous work,First,the image is enhanced to enhance the contrast of the image,then the image is extracted from the collection of pixels,and then the image is segmented using super pixel blocks instead of pixels.The experimental results show that the segmentation effect is significantly improved by using this method.