Corner Detection based Target Tracking and Recognition for UAV-based Patrolling System

The growing interest in the application of unmanned aerial vehicles (UAV) in civil applications has been observed in recent decades. The UAV can be considered as a hardware platform for patrolling the critical infrastructures by carrying camera devices. Such UAV based patrolling system can significantly improve the efficiency of system monitoring and fault detections, e.g. prognosis and diagnosis services in large-scale photovoltaic (PV) farms. However, the motion of the camera carried by the UAV can make it difficult to track and recognize the target, and the available classical target tracking algorithms cannot be directly applied to such scenario. To address such technical challenge, this paper presented a real-time algorithmic solution based on the corner detection method for accurately detecting and tracking the objects. The corners can effectively represent the target, which increase the discrimination of the target and background, and target model based on these corners can reduce the impact of background information. The proposed solution is evaluated through a set of experiments and the results confirm that the fixed target monitored by an UAV can be detected and tracked effectively.