Influence of Nonstationarity on SAR Imagery Models and De-Speckling Filters
-
摘要: 从随机过程理论角度按平稳程度把合成孔径雷达(SAR)图像分为分辨单元内平稳、分辨单元内不平稳、处理窗内局部平稳和处理窗内局部不平稳4种情况,研究了非平稳性对图像的斑点乘性噪声模型、场景-斑点乘积模型和结构邻域模型的影响,得出各自的平稳性适用条件,并在此基础上评价了各种滤波降噪方法的性能,用实际数据进行了验证.Abstract: From the stochastic process point of view, Synthetic Aperture Radar (SAR) images are divided into four cases as stationary within resolution cell, nonstationary within resolution cell, stationary within local processing window and nonstationary within local processing window according to their stationarities. Then the influence of nonstationarity on multiplicative noise model of speckle, scene-speckle product model and structured neighborhood model are investigated, stationarity satisfactory conditions of these models are derived respectively. The performances of various kinds of de-speckling filters are re-assessed based on the above analyses, and filtering results of real imagery are given.
-
Key words:
- synthetic aperture radar /
- stationary /
- filter /
- simulated annealing
-
[1] Jahangir M, Oliver C J. Estimating parameters for a statistical model describing weak scattering. DERA Radar 1997, 1997. 129~134. [2] Kuan D T, Sawchuk A A. Adaptive noise smoothing filter for images with signal-dependent noise[J]. IEEE Trans Pattern Anal Mach Intell,1985,7(2):165~177. [3] Lewinski D J. Nonstationary probalilistic target and clutter scattering models[J]. IEEE Trans Ant Propagation,1983,31(3):490~498. [4] Oliver C, Quegan S. Understanding synthetic aperture radar images[M]. London:Artech House, 1998. [5] Touzi R. A protocol for speckle filtering of SAR images. http://www.estec.esa.nl/ceos99/papers/p119.pdf,1999. [6] McConnell I, White R, Oliver C, et al. The DRA despeckling algorithm. IEE Colloquium. 1995. 9/1~9/6. -
点击查看大图
计量
- 文章访问数: 7948
- HTML全文浏览量: 179
- PDF下载量: 1711
- 被引次数: 0
下载:
