Toward Drone Privacy via Regulating Altitude and Payload
Zupei Li, Chao Gao, Qinggang Yue, Xinwen Fu · 2019 International Conference on Computing, Networking and Communications (ICNC) · 2019
The commercial drone market is booming. However ubiquitous drones have raised great concern about people’s privacy. Computer vision algorithms may disclose details of human activities on the ground from the videos taken by drones. We may mitigate the threats from adversarial computer vision by exploring the limitations of computer vision techniques. In this paper, we propose to mitigate the drone security and privacy threats through regulating the drones altitude and its on-board camera capability. We systematically study the factors that can affect the effectiveness of drone attacks given the camera specification, drone altitude and drone dynamics. We validate our theoretical analysis by evaluating two scenarios, password inference and racial recognition. It can be observed that regulating drone altitude and camera capability shall preserve human privacy on the ground.