A Novel Image Segmentation Approach Based on Particle Swarm Optimization

Image segmentation denotes a process by which an image is partitioned into non-intersecting regions and each region is homogeneous. Utilizing histogram information to aim at segmenting an image is a commonly used method for many applications. In this paper, we view the image segmentation as an optimization problem. We find a curve which gives the best fit to the given image histogram, and the parameters in the curve are determined by using the particle swarm optimization algorithm. The experimental results to confirm the proposed approach are also included.