Validation of the Positioning Accuracy Parameters of the UAV Control System in Controlled-Environment Agriculture

The paper demonstrates that positioning an unmanned aerial vehicle (UAV) in controlled agricultural environments is possible without reliance on satellite navigation. Modifications to the standard flight controller software, specifically, the integration of modules for processing and decoding data from a video stream sensor and a laser rangefinder, ensures high accuracy in determining coordinates both in height and in plan. ( Research purpose ) The study aims to determine the positioning accuracy parameters of a UAV when using a video stream sensor and a laser rangefinder as primary instruments for coordinate calculation. ( Materials and methods ) Data from the Optical Flow & LIDAR Sensor 3901-L0X, transmitted to the UAV flight controller via the debugger port, were analyzed. The Canny detector and Gaussian filter were applied to extract precise contours of high-contrast objects on a horizontal plane and to compute the coordinates of multiple points in the processed video stream. The scaling factor of these coordinates was determined based on laser rangefinder measurements. Methods of mathematical statistics were used to process the research data and calculate errors in determining positioning coordinates. ( Results and discussion ) The findings indicate that combining data from the video stream sensor with height measurements from the laser rangefinder yields high accuracy and enables aerial imaging of agricultural biological objects in greenhouse environments. ( Conclusions ) The study determined that the software for processing video stream and laser rangefinder data enables aerial imaging in greenhouse environments, achieving UAV spatial coordinate calculation accuracy exceeding 95 percent.