5 research outputs found
Unsupervised Data Fusion With Deeper Perspective: A Novel Multisensor Deep Clustering Algorithm
No full textThis work was supported by the Federal Ministry of Education and Research (BMBF), and in the Client II Program, within theMoCa project under Grant 033R189B. The work of Behnood Rasti was funded by Alexander von Humboldt foundation. (Corresponding author: Kasra Rafiezadeh Shahi.
Hyperspectral Unmixing Overview: Geometrical, Statistical, and Sparse Regression-Based Approaches
No full textImaging spectrometers measure electromagnetic energy scattered in their instantaneous field view in hundreds or thousands of spectral channels with higher spectral resolution than multispectral cameras. Imaging spectrometers are therefore often referred to as hyperspectral cameras (HSCs). Higher spectral resolution enables material identification via spectroscopic analysis, which facilitates countless applications that require identifying materials in scenarios unsuitable for classical spectroscopic analysis. Due to low spatial resolution of HSCs, microscopic material mixing, and multiple scattering, spectra measured by HSCs are mixtures of spectra of materials in a scene. Thus, accurate estimation requires unmixing. Pixels are assumed to be mixtures of a few materials, called endmembers. Unmixing involves estimating all or some of: the number of endmembers, their spectral signatures, and their abundances at each pixel. Unmixing is a challenging, ill-posed inverse problem because of model inaccuracies, observation noise, environmental conditions, endmember variability, and data set size. Researchers have devised and investigated many models searching for robust, stable, tractable, and accurate unmixing algorithms. This paper presents an overview of unmixing methods from the time of Keshava and Mustard's unmixing tutorial [1] to the present. Mixing models are first discussed. Signal-subspace, geometrical, statistical, sparsity-based, and spatial-contextual unmixing algorithms are described. Mathematical problems and potential solutions are described. Algorithm characteristics are illustrated experimentally
THE FORMER MEMBERS OF THE VISIELAB
No full textProefschrift voorgelegd tot het behalen van de graad van Doctor in de Wetenschappen aan de Universitaire Instelling Antwerpen te verdedigen door TSANG ING RE
Pattern Recognition and Complex Systems
No full textffl Cluster Size Diversity, Percolation and Phase Transition . I.R. Tsang and I.J. Tsang. Presented at Statistical Physics and Probabilistic Methods in Computer Science and NP-hardness and Phase Transitions, ICTP Trieste, 1999. ffl Entropy and Diversity on Percolation Systems. I.R. Tsang and I.J. Tsang. Presented at Gordon Research Conference on Modern Developments in Thermodynamics, IL Ciocco, 1999
Remote sensing of grassland with contaminated soil using the spectral red-edge
No full textIn most cases contaminants are concealed in soil and under vegetation and therefore cannot be measured directly by remote sensing. However, soil contaminants were detectedusing the spectral red-edge to indicate vegetation stress caused by the presence ofthe contaminants. An improved red-edge position (REP) was developed and gave aslight improvement in the predictive capability over existing indices and an effectiveadditional diagnostic indicator of soil contamination was found to be the spatial patternof the REP. Where an area had high levels of hydrocarbon in the soil it also had ahigh level of variation. The indication was that spatial variation of spectral indices(especially the REP) may be more useful than the spectral index value for the detectionand mapping of soil contamination.Field analysis and radiative transfer modelling (using a coupled leaf and canopy model,LIBSAIL) showed the influence of vertical layering in the grassland canopy. The influenceof a vegetated under-storey on the red-edge was found to be greatest whendifferent absorption spectra were present and high within-the-leaf scattering. The formerdefined wavelength positions of features while the later determined if they wereresolvable in a spectrum. This greater understanding of the grassland canopy identifiedthe importance of fully surveying vegetation canopy structure, especially in complex,multi-layered canopies such as those found with contamination. With this understandingof what the red-edge can reveal, remote sensing is an effective tool for the detectionof contamination
