Srovnání kvality objemových výpočtů realizovaných fotogrammetrickým zpracováním snímků z dronů různých výrobců

Abstract

This bachelor’s thesis compares the accuracy of volume calculations from photogrammetrically processed 3D models derived from images captured by Parrot Anafi Ai, Autel EVO II Pro V3, and DJI Mavic 3 Enterprise drones at a Stora Enso timber yard. The theoretical section summarizes photogrammetric image processing using Structure from Motion techniques, cybersecurity, and the current market situation for unmanned systems applicable to photogrammetry. The practical section covers data collection and volumetric analysis using Agisoft Metashape software. The calculated volumes were compared with Stora Enso’s reference data, which are based on LiDAR measurements of raw materials at the production input. The Parrot Anafi Ai drone achieved the highest accuracy, with an average relative error of 7.4%, followed by the Autel (8.1%) and DJI (10.6%) platforms. These findings confirm that the tested alternatives are viable substitutes for the dominant manufacturer’s machines for industrial inventory needs. The study also identifies a technological gap in the market and challenges for European manufacturers in the field of industrial solutions.