Fast processing for spectral discrimination in hyperspectral imaging

Hyperspectral imaging systems based on interferometers are intrinsically faster than the systems based on dispersive means and on optical filters. This advantage could be hampered by the big computing power necessary to elaborate the big amount of data to extract the spectral content of the image. In applications where it is not important to obtain the full hypercube of the image but it is enough to discriminate spectral contents, the processing time and the computing resources could be considerably reduced. We present two techniques to perform fast processing to be implemented in our hyperspectral imaging device.