Environmental Sensing Using Land-Based Spectrally-Selective Cameras and a Quadcopter

Fig. 1: Illustration of the experimental setup to sample the surface temperature of a patch of land. In this work, we address rapid sampling of environmental fields using static and mobile imaging sensors. Specifically, we demonstrate effective reconstruction of the surface temperature field of a patch of vegetation. Accurate monitoring of surface temperature is desirable for various applications, e.g., atmospheric boundary layer studies over complex terrain [1]. We describe an experiment where the surface temperature of a region is observed with a land-based thermal imaging camera, augmented with truecolor images from a parallel-mounted commercial digital camera, and a downwardfacing non-contact infrared temperature sensor mounted on a quadcopter. The land-based thermal imaging camera is mounted at an elevation along with the truecolor camera, providing a synoptic snapshot of the temperature field and a high resolution true color image respectively. The quadcopter serves as a fast aerial observation platform, allowing rapid sampling of surface temperature using its downward-facing temperature sensor. This provides both speed and flexibility compared to land-based observation platforms such as robotic rovers. This work has three goals: a) to compare the temperature data acquired by the quadcopter with the denser synoptic thermal image captured from the thermal camera, b)