Building a UAV-Hyperspectral System I

Abstract

Hyperspectral imaging offers spectral resolution unavailable with traditional remote sensing platforms, yet integrating a sensor onto an unmanned aerial vehicle (UAV) platform is a nontrivial task. Although commercial off-the-shelf systems are available at attractive prices, there are few comparative performance metrics, and the community has yet to establish a modus operandi for scientific-grade hyperspectral UAV platforms. Researchers investing in hyperspectral UAV assets must have a firm grasp of their operational needs in conjunction with an understanding of hardware costs versus performance trade-offs. Hyperspectral sensors, being the core of the UAV platform, varies considerably in their design and performance, with optical physics imposing a hard limit on their capabilities. This chapter will review the main differences between the types of pushbroom hyperspectral sensors, and the key performance indicators that contribute to a scientific-grade instrument based on our experience. In particular, how sensor characteristics such as dynamic range, full well capacity, and bit depth translate into sensor performance in real-world applications is discussed. Understanding sensor limitations helps manage researcher expectations, especially since low-quality equipment can lead to imaging artifacts, that is, the vignetting effect, which may require correction as part of the image acquisition workflow.