A Direct Georeferencing Imaging Technique to Identify Earth Surface Temperatures Using Oblique Angle Airborne Measurements

Abstract

This thesis describes a novel, open-source image processing method that directly georeferences oblique angle thermal images and calculates Earth Surface Temperatures (STs) at a high spatiotemporal resolution. Images were collected from a thermal camera mounted on a tethered balloon. Median STs are represented spatially in six four-hour time interval plots to display diurnal ST variation. The technique is applied to two datasets collected during two separate field campaigns, one from a northern Canadian mining facility and one from the University of Guelph, Guelph, Ontario, Canada. A comparison between STs for images recorded from the mining facility and a satellite image is completed with a resulting median absolute error of 0.64 K, bias of 0.5 K, and root mean square error of 5.45 K. Further applications of this direct georeferencing workflow are numerous and can be evaluated with other cameras such as red green blue, multispectral, and hyperspectral imaging systems.