A multilevel color image thresholding scheme based on minimum cross entropy and alternating direction method of multipliers

In this paper, we focus on the problem of fast color image segmentation and propose a two-stage image segmentation method using multi-level thresholding algorithm and statistical region merging (SRM) technique. With the help of the alternating direction method of multipliers (ADMM), the optimal threshold values can be determined by the centroid of sub-modes of the image histogram that follow minimum cross entropy (MEC). In the second stage, a modified SRM is used to remove over-segmented regions of thresholded image to bring the final result. Experimental results depicts that the proposed approach selects threshold values more efficiently as compared to other MCE-based thresholding techniques and produces high quality of the segmented color images than the other methods like mean-shift, normalized cuts (Ncuts) and differential evolution (Q=7).