Salient object detection via local saliency estimation and global homogeneity refinement
作者：
Highlights：
• We use local saliency and global homogeneity to detect saliency.
• We model it as a competition of the strengths of foreground and background seeds.
• Our approach achieves remarkable precision–recall performance.
• The performance is not sensitive to the population of the selected seeds.
• Our approach outperforms many existing methods in quantitative way.
摘要
Highlights•We use local saliency and global homogeneity to detect saliency.•We model it as a competition of the strengths of foreground and background seeds.•Our approach achieves remarkable precision–recall performance.•The performance is not sensitive to the population of the selected seeds.•Our approach outperforms many existing methods in quantitative way.
论文关键词：Salient object detection,Local contrast,Global homogeneity
论文评审过程：Received 19 February 2013, Revised 29 October 2013, Accepted 13 November 2013, Available online 22 November 2013.
论文官网地址：https://doi.org/10.1016/j.patcog.2013.11.015
