The aim of the research project which resulted in this work is to achieve a cost‐effective approach for instantaneous hyperspectral imaging.
This paper presents a simulation study and an experimental evaluation of a novel imaging spectroscopy technique, where multi‐channel image data are acquired instantaneously and transformed into spectra by using a statistical modelling approach. A digital colour camera equipped with an additional colour filter array was used to acquire an instantaneous single image that was demosaicked to generate a multi‐channel image. A statistical transformation approach was employed to convert this image into a hyperspectral one.
The feasibility of this method was investigated through extensive simulation and experimental tasks where promising results were obtained.
The small size of the initially acquired single instantaneous image makes this approach useful for applications where video‐rate hyperspectral imaging is required.
For the first time, a simplified prototype of this novel imaging spectroscopy technique was built and evaluated experimentally. And the results were compared with those of a more ideal simulation study. Recommendations for how to improve the prototype were also suggested as a result of the comparison between the simulation and the prototype evaluation results.
