In this study, we develop a telephoto pan-tilt drone search and track system for multicopters flying in a wide area spanning hundreds of meters, situated at a distance of hundreds of meters. It can detect periodic brightness changes around the drone propellers, which rotate at a high speed, in a high-frame-rate (HFR) video. The temporal frequency responses of the brightness signals in the HFR video are computed by performing pixel-level short-time Fourier transforms (STFTs) of the signals. By detecting the peak frequencies, the drone propellers are localized as vibration sources, and their rotation speed is estimated to monitor the flight status of the drone. For real-time localization and flight monitoring, the proposed system can perform pixel-level STFTs in a 500 fps video of 720×540 pixels using video processing accelerated by graphic processing units. This allows a multicopter to be tracked in real time at the center of the camera view by a galvanomirror pan-tilt active vision system with visual feedback. We verified its effectiveness by examining HFR videos for flying multicopters of different appearance, and conducted tracking experiments in outdoor scenes involving multicopters flying at an altitude of 70 m and 200 m ahead in a mountainous background.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.