DROACOR^®-THERMAL: AUTOMATED TEMPERATURE / EMISSIVITY RETRIEVAL FOR DRONE BASED HYPERSPECTRAL IMAGING DATA

Abstract

Thermal remote sensing from unmanned aerial vehicles is a slowly but steadily growing field of application. New hyperspectral systems operating in the thermal infrared are deployable on such systems and are also usable for ground based monitoring, such as in mining applications. Temperature/emissivity retrieval methods have to be adapted for these new situations. This contribution presents an extension of the Drone Atmospheric Correction method (DROACOR®) for thermal infrared imaging spectroscopy. The method includes an implementation of the semi-automatic normalized emissivity mapping (NEM) method for temperature/ emissivity separation. Furthermore, an extension of the method for correction of low emissivity targets, appearing as cold targets in the temperature mapping is introduced. Two examples of DROACOR-thermal processing are presented for a nadir looking drone based and a horizontal ground based data acquisitions are shown. The resulting spectral emissivitiy distributions and temperature mappings are plausible. They are well comparable to spectral library references and allow for the detection of materials only visible in the thermal infrared range.