Active contour model for inhomogenous image segmentation based on Jeffreys divergence
作者：
Highlights：
• Differing from the local data fitting term of RSFACM, a novel local data fitting term based on Jeffreys divergence is constructed instead of Euclidean distance, which achieves relatively better segmentation.
• Similarly, a novel global data fitting term based on Jeffreys divergence is also constructed to improve the versatility of the model.
• An adaptive weight of local and global data fitting terms is introduced to automatically adjust the role of these two terms, which can increase the robustness to the initial curve.
• The proposed ACM outperforms some of the state-of-the-art ACMs in segmentation accuracy and efficiency and it is not strictly dependent on curve initialization.
摘要
•Differing from the local data fitting term of RSFACM, a novel local data fitting term based on Jeffreys divergence is constructed instead of Euclidean distance, which achieves relatively better segmentation.•Similarly, a novel global data fitting term based on Jeffreys divergence is also constructed to improve the versatility of the model.•An adaptive weight of local and global data fitting terms is introduced to automatically adjust the role of these two terms, which can increase the robustness to the initial curve.•The proposed ACM outperforms some of the state-of-the-art ACMs in segmentation accuracy and efficiency and it is not strictly dependent on curve initialization.
论文关键词：Active contour model,Inhomogenous image segmentation,Local and global data fitting energies,Jeffreys divergence,Adaptive weight
论文评审过程：Received 28 August 2019, Revised 14 December 2019, Accepted 23 June 2020, Available online 25 June 2020, Version of Record 1 July 2020.
论文官网地址：https://doi.org/10.1016/j.patcog.2020.107520
