A Spatial Regularity Constrained Active Contour Model for PolSAR Image Segmentation
A variational active contour model based on the statistical model and local information of PolSAR images is proposed for PolSAR image segmentation. The energy functional of the proposed model consists of two parts: the likelihood data term, and the regularization term. We introduce a spatial regularization term which imposes a regional homogeneity effect on the segmentation results. The contours are evolved by minimizing the functional using fuzzy region competition method. With the above two modification, the proposed method can lead to more accurate and efficient PolSAR image segmentation algorithm than method based on standard level set method. Experimental results on both synthetic and real PolSAR images are shown. Performance evaluation and comparison with another method are also given.
KeywordsPolarimetric aperture radar Image segmentation Variational method Spatial regularization
- 12.Li C, Xu C, Gui C (2005) Level set evolution without re-initialization: a new variational formulation. In: Proc IEEE CVPR, San Diego, CA, 2005, pp 430–436Google Scholar