In a recent work [1] Lebrun et al. proposed an algorithm able to adapt itself to the noise characteristics by learning them locally. The main objective of this project is to study the proposed approach and adapt it to the processing of SAR images. Among the necessary adaptations, the format of the image should be studied (SAR images have a very high dynamic so the coding on a reduced size may have a big impact). In theory this blind algorithm should be able to adapt itself to the specific distributions of the SAR data.
A comparison with a state of the art algorithm dedicated to SAR images [2] will be led on synthetic and real images.
Programming
Some code is available via IPOL. which could be used for the project.
Supervision
F. Tupin, A. Almansa
References
[1] Marc Lebrun, Miguel Colom, and Jean-Michel Morel, The Noise Clinic: a Blind Image Denoising Algorithm, Image Processing On Line, (2015), pp. 1–54. http://dx.doi.org/10.5201/ipol.2015.125
[2] Charles-Alban Deledalle, Loïc Denis, Florence Tupin, Andreas Reigber and Marc Jäger NL-SAR: a unified Non-Local framework for resolution-preserving (Pol)(In)SAR denoising, Technical report HAL, hal-00844118 (HAL) to appear in IEEE Transactions on Geoscience and Remote Sensing TGRS 2015