Multilevel thresholding based image segmentation using Masi entropy and moth-flame optimization algorithm

This paper presents a multilevel thresholding based image segmentation technique using the generalized Masi entropy as the criteria to select the optimal thresholds. Moth-flame optimization (MFO) algorithm is utilized to efficiently search the multiple thresholds. Multilevel thresholding based image segmentation techniques require optimization of the search space. In this regard, the MFO algorithm is investigated in this work for its effectiveness in searching the multiple thresholds. The proposed technique is also compared with the particle swarm optimization algorithm based multilevel thresholding technique based on the Karur’s entropy function. The comparison is performed using standard benchmark image databases. Mean structural similarity (SSIM) index, feature similarity (FSIM) index, and peak signal to noise ratio (PSNR) are used to compare the quality of the segmented images. The experimental results suggest that the proposed MFO algorithm based multilevel thresholding technique performs better than the compared techniques. From the results it can be concluded that the proposed technique can be effectively used for multilevel thresholding based image segmentation applications.