Image Segmentation Using Meta-heuristic Algorithms

In this paper, inter-class variance of the two classes of image pixels formed on either side of the threshold value is maximized to perform image segmentation. This process has been automated by the use of evolutionary meta-heuristic algorithms such as Bat Algorithm (BA), Cuckoo Search Algorithm (CSA) and Interior Search Algorithm (ISA). The performance of these algorithms was compared on the basis of various parameters such as accuracy of convergence to optimal threshold values under noisy conditions and the speed of convergence.