VSCL paper “Combining Visible and Infrared Spectrum Imagery using Machine Learning for Small Unmanned Aerial System Detection”, by Vinicius G. Goecks, Grayson Woods, and John Valasek, has been selected as the winner of the 2020 SPIE Automatic Target Recognition Best Student Paper Award. This paper presented a novel approach to combine data from RGB and long-wave infrared (LWIR) cameras to detect drones through previously difficult environments such as flying above and below the treeline/horizon, in the presence of birds, and glare from the sun.
“The main insight of our approach is that it enables detection and tracking of vehicles at any time of the day, around-the-clock, and in real-time.” says Dr. Goecks. “It also can be built on top of any existing camera system with minimal computation overhead.”
A summary video of the system can be found here, along with videos with all predictions for the single-vehicle case and multiple-vehicle case.
The paper was presented at the 2020 SPIE Defense + Commercial Sensing Conference Digital Forum held 27 April – 8 May 2020. The paper is available at the SPIE Digital Library and the preprint version is available at arXiv.

Long-wave Infrared and Visible Spectrum sensors