On minimum cross-entropy thresholding

作者:

Highlights:

摘要

Over last few years several papers have been written on entropy-based thresholding. Some of these methods use the gray-level histogram, while others use entropy associated with the two-dimensional histogram or the co-occurrence matrix. Recently, Li and Lee proposed a thresholding scheme that is aimed to minimize the cross entropy. This note first points out a certain conceptual problem and then proposes two algorithms which are free from that problem. The proposed schemes are tested on several data sets. The results are encouraging.

论文关键词:Information,Cross entropy,Thresholding,Segmentation,Object extraction

论文评审过程:Received 11 April 1994, Revised 9 March 1995, Accepted 11 August 1995, Available online 7 June 2001.

论文官网地址:https://doi.org/10.1016/0031-3203(95)00111-5