基于Gaussian-Hermite矩的SAR图像分割

SAR image segmentation is one of the most important tasks on SAR image processing. But the multiplication nature of the speckle noise in SAR images has been a troublesome problem in SAR images. In the paper, a SAR image segmentation approach based on Gaussian-Hermite moments is proposed. Because Gaussian-Hermite moments can better separate image features and enhance the object based on different modes. Different orders of Gaussian-Hermite moments are made use of as image features to extract objects from background scene. The proposed algorithm is compared with other five algorithms on real SAR images and simulation images. Experimental results show that the proposed method is effective to extract objects from background scene in SAR image. Simultaneously, the speckle noise is restrained.