Huang, Yong-ZhiYong-ZhiHuangTAI-TIEN WANGJeng, Fu-ShuFu-ShuJeng2026-04-142026-04-142026978981954262823662557https://www.scopus.com/record/display.uri?eid=2-s2.0-105031517773&origin=resultslisthttps://scholars.lib.ntu.edu.tw/handle/123456789/737158With the advancement of remote sensing technologies, engineers can employ various modeling methods to generate detailed three-dimensional point cloud models. These methods include unmanned aerial vehicle (UAV) modeling, close-range photogrammetry modeling, ground-based LiDAR, etc. Coupled with point cloud analysis, geometric parameters of discontinuity surfaces on outcrops can be obtained, leading to the characterization of engineering properties of rock masses. This technology enables large-scale and efficient quantitative analysis, and there have been numerous successful implementations globally. However, there is a lack of exploration into the impact of model resolution, with most studies merely mentioning resolution in the context of model generation without explicitly addressing its influence on interpretation results. In this study, three different and common modeling methods were employed to create outcrop point cloud models along Highway 20 in Taiwan. This research aims to illustrate and compare the differences in interpretation results among models produced by different methods. The results indicate that close-range photogrammetry produces the highest model resolution, followed by UAV modeling, and ground-based LiDAR with the lowest resolution. Regarding orientation interpretation results, the point cloud percentage distribution on various discontinuity sets is consistently ranked, with dip and dip-direction are mostly falling within the range of ±10°.false3D point clouds interpretationClose-range photogrammetryGround-based LiDARUAVconference paper10.1007/978-981-95-4263-5_142-s2.0-105031517773