Hyperspectral Image Unmixing With Endmember Bundles and Group Sparsity Inducing Mixed Norms
L. Drumetz, T. R. Meyer, J. Chanussot, A. L. Bertozzi and C. Jutten
IEEE Transactions on Image Processing
doi: 10.1109/TIP.2019.2897254
Year: 2019 , Volume: 28 , Issue: 7
Pages: 3435 – 3450
Description :
In blind spectral unmixing, endmember variability can be handled using a dictionary of candidate endmembers, with several instances corresponding to the same material. However,he usual abundance estimation techniques do not take advantage of the particular structure of these bundles, organized into groups of spectra. We use group sparsity by introducing mixed norms in the abundance estimation optimization problem. In particular, we propose a new penalty which simultaneously enforces group and within group sparsity, using a mixed fractional norm. Unmixing using group lasso or elitist lasso (favoring inter or within group sparsity alone, respectively) can also be used. The present code come with a demo allowing to reproduce some of the results of the paper, from the bundle extraction to the abundance estimation.