UAV Navigation has developed a Visual Navigation System (VNS) that massively reduces the accumulated positional error during dead-reckoning navigation. The VNS leverages visual odometry techniques to determine the position and orientation of the aircraft by analyzing and processing the images captured by a camera installed on its underside.
Initial testing in real-time flight conditions has been a success, reports UAV Navigation. The system integrates well with the company’s flight-control solution to improve navigation in GNSS-denied environments.
Vector autopilot. (Photo: UAV Navigation)
UAV Navigation’s sensors are tolerant toward GNSS failures (typically, in GNSS-denied scenarios) and can operate in dead-reckoning mode without compromising flight safety. However, a prolonged GNSS failure can lead to a significant navigation drift, and this is where the VNS comes in.
The VNS system includes a simple belly-mounted camera and image processing computer. Images from the camera are processed by a lightweight onboard computer, translating them into a relative change in the aircraft position. This information can be combined with the inertial sensors to reduce the overall drift to < 1% of the distance traveled, eliminating any drift associated with time.
Combined with the Vector autopilot, the VNS components provide a complete and robust autonomous flight control and navigation solution.