OTSU’S MULTI LEVEL THRESHOLDING USING MODIFIED FIREFLY ALGORITHM

Abstract

Segmentation plays a crucial role in most tasks requiring image analysis. Basically segmentation is the process of partitioning an image into multiple segments. Segmentation changes the representation of image into a more meaningful and easier to analyze one. Pixels with the same characteristics are grouped together. Image segmentation can be done using several techniques. Here we are using the threshold selection method. Threshold selection is a significant technique for image segmentation and is broadly applied in many fields like computer vision, character recognition, analysis of medical images etc. Here, we are using the modified firefly algorithm for threshold selection embedded with the otsu’s method. The intra cluster variance is minimized for threshold selection so that pixels with more similarity and hence less intra cluster variance are grouped together. The firefly algorithm is a meta heuristic algorithm which is inspired by the flashing behavior of fireflies. It has been used here as firefly has a high convergence rate.The minimization of intra cluster variance is the fitness function taken in firefly.The results have also been compared to other evolutionary algorithms like particle swarm optimization and ant colony optimization which shows better performance of modified firefly algorithm.