Development of co-boresighted Vis-NIR-SWIR hyperspectral imaging systems

Abstract

Hyperspectral Imaging is used in many applications to identify or analyze materials in a scene based on the materials’ spectral signatures. Unique features in the spectral signatures can span beyond the spectral range of the hyperspectral imager. Additionally, lighting conditions and other factors can adversely affect the quality of data. Expanding the spectral range of hyperspectral imaging systems can therefore improve the accuracy of object/material recognition/analysis by allowing the system to “see” more of the spectral signatures as well as expand the number of objects/materials in a scene that can be identified/analyzed. This is particularly important in applications where erroneous identification or analysis can result in substantial risk or cost. More and more users are using two (or more) hyperspectral imagers to obtain different spectral ranges for their applications. Very few are effectively combining the data from the different hyperspectral imagers because it would require the hyperspectral imagers to be operated under tightly controlled conditions and the process of pixel coregistration is a very tedious and problematic post-processing step. In addition, this post-processing step prevents the use of the combined data in real-time applications. This paper describes a co-boresighted Vis-NIR and SWIR hyperspectral imaging system which Headwall Photonics is currently developing. It integrates two hyperspectral imagers, each optimized for its respective spectral range, into a single system with real-time pixel co-registration resulting in a system capable of producing wide-spectrum hyperspectral images with high spectral resolution. Aside from enabling real-time wide spectrum applications, such a system significantly simplifies the data acquisition and analysis for the user.