Volume 19, Issue 4
A Two-Stage Image Segmentation Model for Multi-Channel Images

Zhi Li & Tieyong Zeng

Commun. Comput. Phys., 19 (2016), pp. 904-926.

Published online: 2018-04

Preview Purchase PDF 113 5308
Export citation
  • Abstract

This paper introduces a two-stage model for multi-channel image segmentation, which is motivated by minimal surface theory. Indeed, in the first stage, we acquire a smooth solution u from a convex variational model related to minimal surface property and different data fidelity terms are considered. This minimization problem is solved efficiently by the classical primal-dual approach. In the second stage, we adopt thresholding to segment the smoothed image u. Here, instead of using K-means to determine the thresholds, we propose a more stable hill-climbing procedure to locate the peaks on the 3D histogram of u as thresholds, in the meantime, this algorithm can also detect the number of segments. Finally, numerical results demonstrate that the proposed method is very robust against noise and superior to other image segmentation approaches.

  • Keywords

  • AMS Subject Headings

  • Copyright

COPYRIGHT: © Global Science Press

  • Email address
  • BibTex
  • RIS
  • TXT
@Article{CiCP-19-904, author = {}, title = {A Two-Stage Image Segmentation Model for Multi-Channel Images}, journal = {Communications in Computational Physics}, year = {2018}, volume = {19}, number = {4}, pages = {904--926}, abstract = {

This paper introduces a two-stage model for multi-channel image segmentation, which is motivated by minimal surface theory. Indeed, in the first stage, we acquire a smooth solution u from a convex variational model related to minimal surface property and different data fidelity terms are considered. This minimization problem is solved efficiently by the classical primal-dual approach. In the second stage, we adopt thresholding to segment the smoothed image u. Here, instead of using K-means to determine the thresholds, we propose a more stable hill-climbing procedure to locate the peaks on the 3D histogram of u as thresholds, in the meantime, this algorithm can also detect the number of segments. Finally, numerical results demonstrate that the proposed method is very robust against noise and superior to other image segmentation approaches.

}, issn = {1991-7120}, doi = {https://doi.org/10.4208/cicp.260115.200715a}, url = {http://global-sci.org/intro/article_detail/cicp/11113.html} }
TY - JOUR T1 - A Two-Stage Image Segmentation Model for Multi-Channel Images JO - Communications in Computational Physics VL - 4 SP - 904 EP - 926 PY - 2018 DA - 2018/04 SN - 19 DO - http://doi.org/10.4208/cicp.260115.200715a UR - https://global-sci.org/intro/article_detail/cicp/11113.html KW - AB -

This paper introduces a two-stage model for multi-channel image segmentation, which is motivated by minimal surface theory. Indeed, in the first stage, we acquire a smooth solution u from a convex variational model related to minimal surface property and different data fidelity terms are considered. This minimization problem is solved efficiently by the classical primal-dual approach. In the second stage, we adopt thresholding to segment the smoothed image u. Here, instead of using K-means to determine the thresholds, we propose a more stable hill-climbing procedure to locate the peaks on the 3D histogram of u as thresholds, in the meantime, this algorithm can also detect the number of segments. Finally, numerical results demonstrate that the proposed method is very robust against noise and superior to other image segmentation approaches.

Zhi Li & Tieyong Zeng. (2020). A Two-Stage Image Segmentation Model for Multi-Channel Images. Communications in Computational Physics. 19 (4). 904-926. doi:10.4208/cicp.260115.200715a
Copy to clipboard
The citation has been copied to your clipboard