Minimum cross entropy thresholding

摘要

The threshold selection problem is solved by minimizing the cross entropy between the image and its segmented version. The cross entropy is formulated in a pixel-to-pixel basis between the two images and a computationally attractive algorithm employing the histogram is developed. Without making a priori assumptions about the population distribution, this method provides an unbiased estimate of a binarized version of the image in an information theoretic sense.