Procedures for processing thermal images using low-cost microbolometer cameras for small unmanned aerial systems

Remote sensing data using thermal-infrared (TIR) cameras can be very helpful for many applications including agriculture and thermal refugia for fish habitat. Small unmanned aerial vehicles (UAS) can be efficient platforms for providing high-resolution thermal imagery at a low-cost; however most thermal cameras used for remote sensing are cooled systems and can be very large, expensive, and consume large amounts of power. Because of these constraints they are not easily integrated with small UAS. For surveillance and military applications, microbolometer thermal cameras are widely used on UAS because they are smaller, less expensive and consume less power than cooled thermal cameras. However, mi-crobolometer thermal cameras are not as sensitive and accurate as cooled systems. Also, many are not calibrated and only measure relative temperature. This presents a challenge when microbolometer thermal cameras are used for scientific and ecological applications: absolute surface temperature is necessary for these applications. This paper presents procedures that can be followed to convert the raw images from a microbolometer camera to accurately represent surface temperature (without compensating for emissivity). The steps needed for this include preparing the images for orthoretification, geometric calibration, orthorectification, and correction for external disturbances.