Abstract—This paper studied a real-time visual tracking and remote controlling application for an aerial surveillance robot. An automatic aerial surveillance has an intrinsic problem because it is hard to follow a designated target without its relative position information. In addition, the problem becomes worse in indoor situation, because the aerial robot cannot use the GPS information. However, the visual tracking technique can be a good alternative to keep the targets under the indoor environment or not. The visual tracking algorithm utilized by us uses correlation filters which can track the designated target in the complicated background. In our application, the aerial robot is quad-copter which can fly in 6 DOF (Degree Of Freedom). The aerial robot is navigated by method of the PNG (Proportional Navigation Guidance). The proportional navigation is generally used for targeting objects in aerial moving system. The image sequences captured from the aerial robot are transferred to a server and the server processes the visual tracking algorithm. Then the server sends the control command to the aerial robot in real-time. The aerial robot can route an aerial surveillance scenario. For that surveillance, we implemented a visual image processing and a control method using the AR-Drone 2.0 SDK. The application controls the quad-copter in workspace remotely through Wi-Fi. We generate ground-truths by manual surveillance. The result shows remarkable performance in real time, and the aerial robot achieves to fly for aerial surveillance scenario.

Index Terms—real-time application, surveillance, aerial robot, visual tracking, remote control, proportional navigation

Cite: Yeungkwon Woo, Seyoung Kim, and Eunjin Koh, "Real-Time Visual Tracking and Remote Control Application for an Aerial Surveillance Robot," Jounal of Automation and Control Engineering, Vol. 3, No. 6, pp. 487-491, December, 2015. doi: 10.12720/joace.3.6.487-491