s1-noisefloor 0.1.1

Algorithms to remove noise floor intensity patterns that are prevelant in Sentinel-1 cross-polarized images in EW and IW modes

About

This library provides algorithms to remove noise floor intensity patterns that are prevelant in Sentinel-1 cross-polarized images in EW mode, and to a lesser degree, IW mode.

The algorithms in this library make use of convex optimization to construct noise floors that can be subtracted to correct the image. Two main algorithms are implemented that make use of the information provided in the Sentinel-1 xml files.

1. A linear noise floor removal method that rescales the default noise floor that is provided in each Sentinel-1 product. This is the application of the method in P. Q. Lee, L. Xu, D. A. Clausi. 2020. Sentinel-1 additive noise removal from cross-polarization extra-wide TOPSAR with dynamic least-squares. Remote Sensing of Environment. 248. https://doi.org/10.1016/j.rse.2020.111982 . Currently only available in GRD EW mode images.

2. A non-linear noise floor removal method that computes the noise floor as a power function of the antenna pattern. Parameters are estimated with linear programming. This is the application of the method in http://www.eng.uwaterloo.ca/~pqjlee/lp_paper_aug14_2020.pdf . Can be applied to both EW and IW GRD mode images.

For more info, see https://github.com/PeterQLee/sentinel1_denoise_rs .