Automatic attribute threshold selection for morphological connected attribute filters
作者：
Highlights：
• Attribute filters enhance and extract features without distorting borders.
• Visual image inspection is labor intensive, can be inaccurate and irreproducible.
• Automatic shape quantification attempts to agree with human intuition.
• Methods for automatic computation of attribute thresholds can be simple and fast.
• The choice of a technique for classification appears to depend on the imaging mode.
摘要
Highlights•Attribute filters enhance and extract features without distorting borders.•Visual image inspection is labor intensive, can be inaccurate and irreproducible.•Automatic shape quantification attempts to agree with human intuition.•Methods for automatic computation of attribute thresholds can be simple and fast.•The choice of a technique for classification appears to depend on the imaging mode.
论文关键词：Medical image enhancement,Attribute filters,Connected operators,Thresholding,Segmentation
论文评审过程：Received 23 June 2014, Revised 7 October 2015, Accepted 14 November 2015, Available online 2 December 2015, Version of Record 8 February 2016.
论文官网地址：https://doi.org/10.1016/j.patcog.2015.11.012
