Efficient multilevel image segmentation through fuzzy entropy maximization and graph cut optimization

作者：

Highlights：

• Iterative scheme for accelerating fuzzy c-partition (c>2) entropy calculation.

• Evaluation of different optimization techniques for maximizing fuzzy entropy.

• Enforcing local spatial coherence in global threshold-based segmentation approaches.

• Outperform existing algorithms in terms of both result quality and processing speed.

摘要

•Iterative scheme for accelerating fuzzy c-partition (c>2) entropy calculation.•Evaluation of different optimization techniques for maximizing fuzzy entropy.•Enforcing local spatial coherence in global threshold-based segmentation approaches.•Outperform existing algorithms in terms of both result quality and processing speed.

论文关键词：Image segmentation,Multilevel thresholding,Fuzzy c-partition entropy,Artificial bee colony,Iterative scheme,Graph cut

论文评审过程：Received 19 April 2013, Revised 5 March 2014, Accepted 12 March 2014, Available online 24 March 2014.

论文官网地址：https://doi.org/10.1016/j.patcog.2014.03.009