Matlab/Octave code implementation of block subspace filter (BSF) for removing radio interference artefacts in SAR images

Function description

function data_flt = BSF(data,block_len,Ns)     
% Input:     
%   data      -----  single look complex SAR image data matrix, complex double datatype.     
%   block_len -----  block length use to perform block-wise processing, e.g., block_len = [1000 1000].
%   Ns        -----  number of subspace dimension. Tuning this parameter to obtain satisfactory results.
% Output:
%   data_flt  -----  filtered single look complex SAR image data matrix

Usage example

load data                           % load SLC SAR data
block_len = [1000 1000];            % block lengths
Ns = 2;                             % emprical RFI subspace dimensions
data_flt = BSF(data,block_len,Ns);   

Example results

left: input image containing RFI; mid: filtered image; right: removed interference artefact

More descriptions can be found in the following paper:

Huizhang Yang et al., “BSF: Block Subspace Filter for Removing Narrowband and Wideband Radio Interference Artefacts in Single Look Complex SAR Images,” IEEE Trans. Geosci. Remote Sens., 2021. DOI:10.1109/TGRS.2021.3096538

If you find this repo useful, consider to cite this paper in your research. If you have any question, send an email to [email protected] or [email protected]