A NEW APPROACH FOR MULTILEVEL THRESHOLD SELECTION

This paper describes a new method for multilevel threshold selection o f gray level images. The proposed method includes three main stages. F irst, a hill-clustering technique is applied to the image histogram in order to approximately determine the peak locations of the histogram. Then, the histogram segments between the peaks are approximated by ra tional functions using a linear minimax approximation algorithm. Final ly, the application of the one-dimensional Golden search minimization algorithm gives the global minimum of each rational function, which co rresponds to a multilevel threshold value. Experimental results for hi stograms with two or more peaks are presented. (C) 1994 Academic Press , Inc.