Ultra-sensitive hyperspectral camera using compressive sensing

Abstract

With the advancement of machine vision, image analysis utilizing high-dimensional data has become feasible, and its application has been vigorously studied. Hyperspectral cameras, which can capture wavelength information and spatial information at the same time, have been developed for practical use in various fields such as food, agriculture, chemistry, and medicine. However, existing hyperspectral cameras are limited by significantly low sensitivity and shutter speed associated with the spectrometry process, as well as by the increased size of the systems, hindering their widespread adoption in industrial applications. In this paper, we show an overview of an ultra-sensitive hyperspectral camera we developed by employing compressive sensing technology, its implementation into a digital camera, and its performances.