Skip to content

leon-bungert/blind_remote_sensing

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Blind Image Fusion for Hyperspectral Imaging with the Directional Total Variation

This MATLAB code allows to reproduce the results of Blind Image Fusion for Hyperspectral Imaging with the Directional Total Variation [1].

[1] Bungert, L., Coomes, D. A., Ehrhardt, M. J., Rasch, J., Reisenhofer, R., & Schönlieb, C.-B. (2018). Blind Image Fusion for Hyperspectral Imaging with the Directional Total Variation. Inverse Problems, 34(4), 044003. https://doi.org/10.1088/1361-6420/aaaf63 http://arxiv.org/abs/1710.05705

The aim of [1] is to fuse a hyperspectral image of low spatial resolution with a photograph of higher spatial resolution. Three examples are shown below. The examples on the left and middle where acquired from a plane flying over Spain to study vegetation. The example on the right has been acquired from a satellite.

The higher spatial resolution photograph is very important to resolve fine details. The example below compares the proposed regularizer "directional total variation" (dTV) to a more standard regularizer "total variation" (TV).

The mathematical model usually assumes that the hyperspectral image and the high-resolution photograph are perfectly aligned. For real data this is rarely the case. The proposed model estimates and corrects for a possible mismatch during the reconstruction. The example below shows the impact of the proposed "blind" approach (the mismatch is unknown prior to reconstruction).

Getting started

There are a number of examples which also reproduce the results as presented in the paper. To execute them all, run matlab/example.m. It will run all examples in matlab/src/scripts. These include

Further Improvements

As suggested in [2], the code can be made more robust to large deformations in the side information by a different initialisation of the image to be reconstructed. See [2] for more details.

References

[1] Bungert, L., Coomes, D. A., Ehrhardt, M. J., Rasch, J., Reisenhofer, R., & Schönlieb, C.-B. (2018). Blind Image Fusion for Hyperspectral Imaging with the Directional Total Variation. Inverse Problems, 34(4), 044003. https://doi.org/10.1088/1361-6420/aaaf63 http://arxiv.org/abs/1710.05705

[2] Bungert, L., Ehrhardt, M. J., & Reisenhofer, R. (2018). Robust Blind Image Fusion for Misaligned Hyperspectral Imaging Data. In Proceedings in Applied Mathematics & Mechanics (Vol. 18, p. e201800033). https://doi.org/10.1002/pamm.201800033

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages