Automating Adaptive Scan Planning for Static Laser Scanning in Complex 3D Environments

Laser scanning is increasingly important for applications in the built environment and requires efficient planning for effective data collection. The conventional process is a time-consuming and repetitive manual task, presenting a strong case for automation. This paper introduces a novel method for scan planning in complex 3D environments, overcoming the limitations of existing approaches. It accepts any 3D model or point cloud as input by processing them as triangulated meshes. The method involves automated steps of mesh processing, viewpoint candidate generation, and evaluations of visibility and coverage. It facilitates optimized planning for static laser scanning missions by selecting appropriate viewpoints while considering targetless registration needs. Our method can handle uniform coverage requirements and specific local requirements. It is rigorously tested through method comparisons and extensive parameter studies and applied to several scenes, including a comparison to scan strategies manually created by laser scanning experts, demonstrating its practical applicability.