摘要
针对TerraSAR-X图像中的斑点噪声抑制问题,分析了噪声的基本数学模型,并考察了经典处理方法,进而结合Kuan滤波的基本思想,给出了一种改进型梯度倒数加权(GIW)滤波算法,利用邻域像素和中心像素的灰度梯度倒数及其邻域统计信息构造滤波加权模板,然后通过卷积操作实现斑点噪声的抑制。对TerraSAR-X图像处理的结果表明该算法原理简单,实现方便,可有效抑制图像中的斑点噪声,同时具有优良的边缘细节保持能力,非常适合边缘提取、目标检测等后续处理。
As for the reduction of speckle noise in TerraSAR-X image, several classical algorithms are investigated based on the mathematical model analysis of the noise. A modified Gradient Inverse Weighted （GIW） filter algorithm is presented, which used the idea of the Kuan algorithm. In the algorithm, the gray gradient reciprocal of the neighborhood pixels and the center pixel, and the local statistics are used to construct the weighted filter mask, followed by the convolution operation to restrain the speckle noise. The experiment results with the TerraSAR-X images indicates that the algorithm, which is simple in principle and easy to realize, can suppress the speckle noise effectively while keeping fine edge details, thus it is suitable for the follow-up processes as edge extraction and target detection.
出处
《电光与控制》
北大核心
2010年第4期17-21,共5页
Electronics Optics & Control
