Multi-sensor Data Fusion for Multi- and Hyperspectral Resolution Enhancement Based on Sparse Representations

This thesis presents solutions to two demanding, severely ill-posed multi-sensor data fusion problems, i.e., pan-sharpening and hyperspectral-multispectral data fusion. Incorporation of physical aspects, knowledge about image patches featuring sparse representations of dictionaries, and mutual correlation of spectral channels reduce the number of degrees of freedom. Parallel software solutions are optimized for operation on the SuperMUC. The quality of the data fusion products is assessed and compared to the state-of-the-art, for a large variety of sensor combinations.