基于PSO优化空间约束聚类的SAR图像分割

SAR Image Segmentation Based on Spatially Constrained FCM Optimized by Particle Swarm Optimization

相干斑噪声的存在使得合成孔径雷达(SAR)图像的分割问题变得非常复杂.本文提出一种具有鲁棒抗噪特性的SAR图像分割方法,该算法通过将多尺度条件下的边缘信息以及相邻像素的相对位置和强度信息所构成的空间信息融入模糊C-均值(FCM)聚类算法的相似性测度中,增强了分割方法的抗噪性.空间信息对FCM聚类算法的调控由粒子群优化(PSO)算法完成,优化的调控参数有助于获得良好的分割结果.该算法对初始分割不敏感,并具有鲁棒的抗噪性能.MSTAR数据的分割实验结果表明:该算法能够有效地分割SAR图像,与通过改进型FCM(IFCM)算法获得的分割结果比较,分割结果明显改善.

The segmentation of synthetic aperture radar （SAR） images is greatly compficated due to the presence of speckle. A SAR image segmentation method robust to speckle is proposed in this paper. Spatial information, containing the edge information and the relative location and intensity information of neighboring pixels across scales,is incorporated into the similarity measure of fuzzy c-means （FCM） algorithm, which enhances the robustness of the method to speckle. The FCM algorithm is spatially adjusted by the parfide swarm optimization （PSO） algorithm to obtain appropriate adjustment parameters that can provide better segmenta- tion results. The proposed method is not sensitive to initial seginentation result and is robust to speckle. Experimental results on the MSTAR dataset demonstrate that the proposed method is capable of effectively segmenting SAR images and achieving better results than the improved FCM （IFCM） algorithm.