In this work, different methods for material classification by hyperspectral images in the near infrared range are investigated. Specifically, the design of problem-adapted camera systems and the choice of optimal optical filters are discussed. In addition, a method for the fusion of multiple camera signals by spectral unmixing is presented. |