Image Segmentation Method And Application Based On Improved Coyote Optimization Algorithm

Image segmentation is a very classical problem of computer vision,which plays an important supporting role in image analysis and image understanding.The essence of image segmentation is to separate the target and background of an image,so as to extract the region of interest.Image segmentation technology has a wide range of applications,such as face recognition,driverless and security monitoring.Image segmentation also plays an important role in medical image assisted therapy.Medical image segmentation is one of the most important research directions in the field of image segmentation.In recent years,with the continuous development of artificial intelligence,swarm intelligence algorithm has been widely used in image segmentation algorithm,and good experimental results have been obtained.This paper selects the Coyote Optimization Algorithm in Swarm Intelligence Algorithm and improves it in a variety of schemes,then combines the improved Coyote Optimization Algorithm with image segmentation technology,analyzes the experimental results of a variety of improved Coyote algorithms in image segmentation,and finally summarizes the advantages and disadvantages of a variety of improved Coyote algorithms in image segmentation.The content of this paper is mainly divided into the following parts:(1)Firstly,the traditional Coyote Optimization Algorithm is introduced in detail,and the assumption of Otsu and Kapur entropy as the objective function is introduced,which are applied to the Coyote Optimization Algorithm respectively.(2)In order to avoid the shortcomings of slow convergence and premature,an improved fuzzy Coyote Optimization Algorithm based on differential dynamic mutation disturbance strategy is proposed,and the image segmentation experiment is carried out on the selected standard image.In this paper,PSNR(peak signal-to-noise ratio)and FSIM(feature similarity)are used to evaluate the effect difference of the algorithm in multi threshold image segmentation.(3)Taking Otsu as the objective function,the concept of chaos initialization is introduced,and a Coyote Optimization Algorithm based on chaos initialization is proposed.Then,on this basis,the objective function Otsu is replaced by fuzzy Kapur entropy,and then the reverse learning strategy is introduced to propose a Coyote Optimization Algorithm based on chaos initialization and reverse learning strategy,so as to obtain the optimal segmentation threshold vector.(4)The improved Coyote algorithm proposed above is applied to medical image segmentation,the experimental results of several improved Coyote algorithms in medical image segmentation are compared,and the advantages and disadvantages of their segmentation effects are analyzed.