Research on an Encryption Processing Method for Aerial Triangulation of UAV

Aiming at the key technical problems such as insufficient automation and limited adaptability of complex terrain in aerial triangulation encryption in UAV low-altitude photogrammetry, this study proposes a new processing method for photogrammetry system. Through the fusion of image pyramid multi-resolution analysis architecture and GNSS / IMU integrated navigation and positioning technology, the bundle block adjustment model is constructed, and the adaptive terrain feature constraint mechanism is innovatively introduced, which significantly improves the matching reliability of homonymous points in weak texture regions. In the experiment, the 1 : 2000 scale UAV aerial photography image of a county in Anhui Province was taken as the research object. Under the condition of laying 20 uniformly distributed ground control points, after three-stage iterative adjustment, the average mean square error of plane position was ± 0.12 m ( X : ± 0.132 m, Y : ± 0.108 m ), the mean square error of elevation direction was ± 0.173 m, the maximum residual error was controlled within the range of 0.535 m, and the overall processing efficiency was improved by 42.7 % compared with the traditional method. The research shows that by establishing the automatic processing system of the whole process of aerial triangulation, this method effectively solves the engineering problems such as high dependence of artificial prick points and accuracy attenuation of terrain fluctuation areas, and provides high-precision and high-efficiency technical methods for UAV remote sensing surveying and mapping.