Fast nonparametric active contour adapted to quadratic inhomogeneous intensity fluctuations
作者：
Highlights：
• A segmentation approach robust to inhomogeneous intensity fluctuations is developed.
• Spatial inhomogeneities of intensity are modeled as 2D quadratic variations.
• No a priori knowledge on the noise fluctuations is required.
• Approach based on the optimization of a MDL criterion without any tuning parameter.
• It results in a fast and unsupervised active contour segmentation technique.
摘要
Highlights•A segmentation approach robust to inhomogeneous intensity fluctuations is developed.•Spatial inhomogeneities of intensity are modeled as 2D quadratic variations.•No a priori knowledge on the noise fluctuations is required.•Approach based on the optimization of a MDL criterion without any tuning parameter.•It results in a fast and unsupervised active contour segmentation technique.
论文关键词：Segmentation,Nonparametric noise modeling,Inhomogeneous intensity fluctuations,Polygonal active contour,Minimum Description Length (MDL)
论文评审过程：Received 8 January 2014, Revised 29 April 2014, Accepted 18 May 2014, Available online 28 May 2014.
论文官网地址：https://doi.org/10.1016/j.patcog.2014.05.013
