An Automated Image Thresholding Scheme for Highly Contrast-Degraded Images Based on a-Order Fuzzy Entropy

Abstract

This paper presents an automated thresholding algorithm for highly contrast-degraded images based on the minimization of the a-order fuzzy entropy of an image. The advantage of the proposed method is that it is based on a flexible parametric criterion function that can be automatically tuned according to the histogram statistics, in order for the thresholded image to preserve as much of the object properties of the initial image as possible, despite the contrast degradation. The effectiveness of the new algorithm is demonstrated by applying our method to different types of contrast-degraded images. Performance assessment is based on comparison of the results derived using the proposed method with the results obtained from various existed image thresholding algorithms using objective empirical discrepancy measures.