Hyperspektralt kamera og metode for å ta opp hyperspektrale data

摘要

A hyperspectral camera includes light mixing chambers projected onto an imaging sensor. The projection of each chamber is slightly larger than a sensor pixel. The chambers are placed as a linear array in a slit plane or as a two dimensional matrix in front of the imaging sensor. The mixed light from each chamber is depicted by several sensor pixels. The sensor outputs information used to form an overdetermined equation set. The set is solved and optimized for the solution giving the lowest overall error or the best fit. The solution of the equation set combined with the optimization is the intensity values of the chambers constituting imaginary pixels being calculated. These imaginary pixels form the output of an improved hyperspectral camera system, which has significantly lower optical errors like keystone and point spread function variation for different wavelengths.