Exploration of Multispectral Image Data

Multi- and Hyperspectral imaging allows the acquisition of image series of a specified wavelength range. Those image series can be regarded as high-dimensional image cubes with large density of spectral information. Due to the high-dimensionality of multispectral image data and the associated complexity, the interpretation of these data is a major challenge for humans and time-consuming for computers. Moreover, recent developments in sensor technologies will lead to increasing spatial and spectral resolutions. Accordingly, there is the need of a system that allows a user to explore the data in an efficient and intuitive way to facilitate the interpretation and consequently to improve the understanding.

Beside the challenge of high-dimensonality, these imaging techniques have been applied in a growing number of application areas and are getting increasingly popular in further application domains. The increasing popularity requires tools for analyzing and processing of multi- and hyperspectral data in a generic way to ensure that important informations can be extracted also for new application domains.

In the context of this research direction, we focuses on the development of efficient visual analysis techniques for multi- and hyperpsectral data. Here, one of the major goals is the determination of the constituent spectra and the exploration of mixed spectra to gain comprehensive insights to spectral data.

This research project is funded by the German Research Foundation (DFG) as part of the research training group GRK 1564 'Imaging New Modalities'.