Abstract
The Rician distribution is widely used in SAR imagery to model the backscattering of a strong target inside a resolution cell. Nevertheless, the computation of the parameters of the Rice distribution remains a difficult task. In this letter, a new parameterization to model the Rice distribution is introduced. Thanks to the introduction of a new variable defined by the ratio of the target contribution to the speckle, the relationship between the coefficient of variation and this new parameter can be derived. An efficient numerical method is proposed to evaluate it from the coefficient of variation and a discussion on the variance of this estimator is led. A comparison with other methods of estimation showed that the proposed approach is a good compromise between the variance of the estimate and the computation time. At last, a link between the permanent scatterers and the Rice distributed targets is proposed through this new parameterization.
| Original language | English |
|---|---|
| Article number | 8933060 |
| Pages (from-to) | 2011-2015 |
| Number of pages | 5 |
| Journal | IEEE Geoscience and Remote Sensing Letters |
| Volume | 17 |
| Issue number | 11 |
| DOIs | |
| Publication status | Published - 1 Nov 2020 |
Keywords
- SAR data
- statistical modeling
- the Rician distribution
Fingerprint
Dive into the research topics of 'A New Parameterization for the Rician Distribution'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver