Entropy-based circular histogram thresholding for color image segmentation

Abstract

Circular histogram thresholding on hue component is an important method in color image segmentation. However, existing circular histogram thresholding method based on Otsu criterion lacks the universality. To reduce the complexity and enhance the universality of thresholding on circular histogram, the cumulative distribution function is firstly introduced into circular histogram. Then, this paper expands circular histogram into the linearized one in anticlockwise direction or clockwise one by using optimal entropy of cumulative distribution function. In the end, fuzzy entropy thresholding method is utilized on linearized histogram to select optimal threshold for color image segmentation. Experimental results indicate that the proposed method has better performance and adaptability than the existing circular histogram thresholding method, which can increase pixel accuracy index by 30.12% and structure similarity index by 27.53%, respectively.