Development of 3D hyperspectral camera using compressive sensing

Abstract

Hyperspectral cameras are capable of collecting and processing imaging data across a wide range of electromagnetic spectrum. However, in comparison to general visible light cameras, developing high-performance hyperspectral cameras is difficult due to the complex and high-cost manufacturing process. Moreover, the conventional imaging system records the total intensity of different light rays arriving to each point on a sensor plane, resulting in the loss of the directional information of the light rays. In order to overcome the current limitation of hyperspectral cameras, in this paper, we report the development of a new generation of 3D hyperspectral imaging system based on integration of compressive sensing and light-field imaging. Here, we demonstrate the use of the hyperspectral imaging system in infrared regime. The system is different from traditional cameras, which does not require a large-scale sensor array. Instead, individual images can be obtained by individual photodetectors using compressive sensing. Then, the individual images can be used to compute a synthetic image, which is the computational result of light-field imaging. In particular, stereoscopic infrared imaging will be demonstrated in the paper. This represents a potential advancement of hyperspectral imaging. © 2012 IEEE.