Open-source Imaging Spectroscopy for the Remote Sensing Practitioner

Abstract

Imaging spectroscopy, or hyperspectral imaging, is a powerful tool for acquiring detailed information about the physical and chemical properties of the Earth's surface. One of the main advantages of hyperspectral remote sensing is its ability to identify and classify materials based on their spectral signatures. For example, it can be used to detect and map mineral deposits, monitor vegetation health, and identify different types of land cover. It can also be used to monitor changes in the environment over time, such as the impact of climate change on ecosystems.

This thesis assembles all the ingredients needed for a remote sensing practitioner to build, calibrate, and operate an imaging spectrometer in the field. Importantly, all the tools are open-source and documented to facilitate broad use with minimal barriers of entry. This includes hardware designs and software tools. With a custom imaging spectrometer in hand, I demonstrate the calibration procedure and validate the performance with a drone trial. I discuss how the effects of the grating second order can be corrected. Following this, I provide a comprehensive overview of the data acquisition system and the challenges resolved. With the navigation and orientation data collected, I explain how hyperspectral data from a pushbroom can be directly georeferenced to a coordinate grid. In light of the new platform and form factor, I also include a new application for hyperspectral imagers that was not possible before to image the diffuse radiance of the sky, a key component to accurately model the Earth's atmosphere.