IMAGE SEGMENTATION USING PARTICLE SWARM OPTIMIZATION AND HARMONY SEARCH ALGORITHM

Abstract

Segmentation holds a very important place in the field of Image Processing. It is a technique used to divide an image into segments which correlate with the real world. The regions having common characteristics are grouped together and are thus differentiated from other regions of the image making the image more clear and distinguishable. It also helps to differentiate the foreground and background of the image thus helping to analyze the image better. Segmentation can be performed by variety of evolutionary algorithms. In the present work Particle Swarm Optimization and Harmony Search Algorithm have been applied. PSO is an evolutionary computational technique which imitates the social behavior of bird flock. It is use to optimize variety of computational problems. Harmony Search algorithm is a population based algorithm which imitates the music improvisation process. It consists of Harmony memory which is initialized with initial population which are further updated and improvised until a termination criteria is met. The combination of the above two algorithms increases the convergence rate and accuracy of the result of the segmentation with low computational overhead. Finally the Berkley data set is used to to display the result of the algorithm.