期刊名称：International Journal of Advanced Research in Computer Engineering & Technology (IJARCET)
印刷版ISSN：2278-1323
出版年度：2014
卷号：3
期号：2
页码：396-398
出版社：Shri Pannalal Research Institute of Technolgy
摘要：This paper present a rapid algorithm for synthetic aperture radar (SAR) image segmentation based on the G0-Statistical-Model. The proposed approach considers the segmentation of SAR images as an intensity minimization problem in a variation framework. The energy (intensity) functional is formulated based on the statistical characteristic of SAR images. The total change regularization is used to impose the smoothness restraint of the segmentation result. To solve the problem efficiently the energy functional is firstly modified to be convex and differentiable by using convex relaxing and variable dividing techniques, and then the constrained optimization problem is change to an unconstrained one by using theG0 statistical model. Finally the energy is minimized with loss of minimization algorithm. The effectiveness of the proposed algorithm is validated by experiments on both synthetic and real SAR images.
