Three-Dimensional Modeling and Crack Detection for High Pier Bridges Based on UAV Route Optimization Algorithm

This study proposes a path optimization algorithm specifically designed for high-pier bridges in mountainous areas. By integrating drone oblique photogrammetry with an improved dung beetle optimization algorithm, the research achieves high-resolution image acquisition and rapid modeling of refined 3D models. The study employs an improved dung beetle optimization algorithm, incorporating chaotic mapping, adaptive weights, and spiral search strategies to optimize flight path planning and enhance the accuracy of the bridge’s 3D model. Subsequently, the algorithm is utilized to automatically detect and quantitatively analyze surface cracks on bridge piers. Finally, a specific high-pier bridge is selected as the research object to validate the effectiveness of the path planning and the accuracy of the bridge surface crack identification method. The results indicate: (1) the algorithm can be effectively applied to practical engineering projects, with superior modeling outcomes; (2) significant improvements are observed in planar and elevation accuracy, demonstrating the effectiveness of optimized path planning in enhancing modeling quality under complex terrain conditions; and (3) the crack identification accuracy exceeds 90%, with a width calculation error of less than 10%, and the algorithm can accurately locate the cracks.