Title :
A new method for multilevel thresholding using symmetry and duality of the histogram
Author :
Yin, Peng-Yeng ; Ling-Hwei Chen
Author_Institution :
Dept. of Comput. & Inf. Sci., Nat. Chiao Tung Univ., Hsinchu, Taiwan
Abstract :
A new approach for multilevel thresholding based on two interesting properties of the histogram is proposed. First, a peak-finding method is presented based on the symmetry of histogram that the hillsides of each hill are symmetric about the central curve. Next, to identify the valleys of the histogram, the duality that the peaks and the valleys are opposite is introduced. Based on this property, a valley-finding method is proposed to construct the hierarchical order of the various valleys. The proposed method is computationally faster than the traditional ones (the variance-based and the entropy-based)
Keywords :
computational complexity; image segmentation; duality; hierarchical order; hillsides; histogram; multilevel thresholding; peak-finding method; symmetry; valley-finding method; Analysis of variance; Computer networks; Digital images; Entropy; Histograms; Image processing; Image segmentation; Information science; Neural networks; Speech processing;
Conference_Titel :
Speech, Image Processing and Neural Networks, 1994. Proceedings, ISSIPNN '94., 1994 International Symposium on
Print_ISBN :
0-7803-1865-X
DOI :
10.1109/SIPNN.1994.344969