An Improved Fuzzy Clustering Segmentation Algorithm Based on Animal Behavior Global Optimization

Abstract

The bio-inspired optimization algorithms play vital role in many research domains and this work analyzes animal behavior optimization algorithm. Medical image segmentation helps the physicians for disease diagnosis and treatment planning. This work incorporates ABO algorithm for cluster centroid selection in Fuzzy C-means clustering segmentation algorithm. The Animal Behavior Optimization (ABO) algorithm was developed based on the group behavior and was validated on 13 benchmark functions. The dominant nature of an animal species decides the fitness function value and each solution in problem space depicts the animal position. The ABO algorithm was coupled with the classical FCM for the analysis of region of interest in abdomen CT and brain MR datasets. The results were found to be efficient when compared with the FCM coupled with artificial bee colony (ABC), firefly, and cuckoo optimization algorithms. The promising results generated by ABC makes it an efficient one for real-world problems.