A generalized multiclass histogram thresholding approach based on mixture modelling

作者：

Highlights：

• Generalizing thresholding to multi-modal class segmentation.

• Classes are modeled using mixtures of Generalized Gaussian distributions.

• Formulation of thresholding based on maximum likelihood estimation.

• Application to image foreground segmentation.

摘要

Highlights•Generalizing thresholding to multi-modal class segmentation.•Classes are modeled using mixtures of Generalized Gaussian distributions.•Formulation of thresholding based on maximum likelihood estimation.•Application to image foreground segmentation.

论文关键词：Image segmentation,Multi-modal class thresholding,Mixture of generalized Gaussian distributions (MoGG)

论文评审过程：Received 4 February 2013, Revised 13 July 2013, Accepted 10 September 2013, Available online 19 September 2013.

论文官网地址：https://doi.org/10.1016/j.patcog.2013.09.004