Lockheed Martin, our capstone sponsor, focuses on unmanned vehicle systems and develops VCSi, a ground control station (GCS) used for commercial and military use. The GCS allows an operator to control both the vehicle and payload (imaging) systems onboard. Currently, most payload systems (camera and gimbal) can track objects and people of interest, independent of the flight system. If the object moves out of the camera and gimbal’s field of view, the drone itself does not adjust its position, causing the object to be lost. For example, if a drone is deployed to follow a person, the gimbal can track them while they remain in frame, but once they move out of range, tracking is lost. This technology has many uses, including military surveillance, law enforcement, tracking animal migration, and supporting search and rescue. The objective of this capstone is to develop a drone system that is capable of autonomously adjusting its flight path to keep an object/person of interest in view without requiring operator or pilot intervention.
The objective of this project is to deliver a drone system capable of autonomously following objects or individuals of interest. The software solution will be portable, so that it can be easily integrated with other drone systems. To achieve this, we will use open source network protocols to be compatible with the majority of open source autopilots. Alongside the drone, a website serving as a control station will be delivered to allow for custom payload and imaging control.
Our approach will use computer vision to detect and track objects, and by pairing these detections with geodata, we will feed this information into our solution for responsive flight control.