Development of a Low-Cost Method for Evaluating the Accuracy of 3D Scanned Meshes Using 3D Printing and Hausdorff Distance

抄録

This study proposes a low-cost experimental method for evaluating 3D scanning meshes, using LiDAR and photogrammetry as case studies. The proposed method includes the design of an evaluation function to assess the fidelity of digitized objects through multiple stages, such as 3D printing, reprinting, and rescanning. To enable accurate comparisons at each stage, both manual and ICP (Iterative Closest Point) alignment techniques are used, and the Hausdorff distance is used as a key metric for quantifying accuracy. Using a Barong sculpture from Indonesia as a case study, the study evaluates the performance of LiDAR and photogrammetry in terms of both accuracy and visual consistency. The designed evaluation function yielded logically valid results for both LiDAR and photogrammetry. However, the findings also indicate that visual appearance does not always correlate with scan accuracy, and the limited number of evaluation samples remains a challenge. This study highlights the potential of consumer-grade technologies in supporting cultural heritage preservation and presents a practical method for assessing scanning workflows, offering valuable insights for future applications in digital archiving and artifact documentation.