A method for evaluating the hovering accuracy of drone based on visual guidance total station

This study evaluates the positioning accuracy of Unmanned Aerial Vehicles (UAVs) using a visual guidance system integrated with a total station. Traditional total station measurements face challenges in image readability and limited zoom range, prompting the adoption of visual target detection and automated search techniques. The proposed method employs a two-axis turntable-mounted zoom camera to capture initial angular and vertical adjustments, enabling approximate coordinate calculation of a reflector plate on the UAV. Through coordinate transformation using extrinsic parameters, these values are converted into the total station’s framework, where precise distance measurements (via Distance Measurement Engine) and angle decomposition refine spatial positioning. Experimental results demonstrate a hovering accuracy of approximately 35.2 mm (1σ) at 30 m altitude, with 3D coordinate data showing standard deviations of 28.2 mm (X), 33.4 mm (Y), and 35.1 mm (Z). These findings validate the system’s capability to achieve reliable, high-precision UAV localization under visual guidance, supporting applications in surveying, construction, and automated machinery alignment.