The Research And Application Of Image Segmentation Based On Chaos Ant Colony Algorithm

Image segmentation is the fundamental step to do research in the field of graph and image processing and the premise of image analysis and image understanding,so the result of image segmentation directly affects the target expression and parameter measurement.Defects of the existing image segmentation algorithms are as follows.The segmenting between foreground and background is not clear.There exist discontinuities of the image edge information.And the shadow expression is not obvious.As one of the bionic algorithms,the ant colony optimization(ACO)can effectively separate foreground and background and achieve continuous edges with its fast convergence rate and strong robustness,but the defects arising from the positive feedback of the ACO itself such as early maturity have not been well improved.The proposing of Chaos Theory broke the predictability of the dynamics system.The theory became a hot focus of scholars with its peculiarly random and chaotic motion,as with such characteristics it can effectively make up for the precocious defect of ACO caused by the dominant position of the positive feedback.In the paper,on the basis of chaos ant colony optimization(CACO)the author introduced feature statistics in the optimization algorithm and optimized the calculation of Euclidean distance in the HSV image segmentation,which enabled the improved algorithm to segment the image effectively.The experiment showed that the improved algorithm could effectively segment the the foreground and background of images,in which the shading levels expressed clearly and the efficiency of the algorithm relatively accelerated,so as to verify the feasibility and accuracy of the algorithm in gray-scale image processing as well as in the processing of color images.