T. Arnold, M. De Biasio, G. Lodron, and R. Leitner (Austria)
spectral imaging, imaging spectroscopy, tunable filter.
Classical color imaging and spectroscopy are both separately invaluable for a wide range of applications like material inspection or cancer diagnosis. However, classical
color images contain little spectral information while spectroscopy is limited to single point measurements. Spectral
imaging (SI) combines both approaches, allowing spatial
and spectral features to be analyzed together. The combination of imaging acousto-optical tunable filters (AOTF),
highly sensitive cameras and fast application-optimized
statistical data evaluation algorithms allows the analysis of
spectral and spatial information in real-time. This technology, using available off-the-shelf components, makes it
possible to acquire hyper-spectral images with two spatial
dimensions and one spectral dimension sequentially and visualize it as 3D hyper-spectral cubes. However, moving
objects cause motion artefact due to the sequential acquisition that have to be corrected through a registration process. The imaging system proposed in this paper is capable
of displaying several registered hyper-spectral data cubes
per second. Future miniaturization of tunable filters will
make it possible to build a hand-held device suitable for
applications ranging from food-safety to life-sciences.