Assessing accuracy and reliability of modern 3D reconstruction and measurement techniques in the evaluation of Byzantine baths: a case study

Ίων-Αναστάσιος Κάρολος, Konstantinos Bellos,Stylianos Bitharis,Vassilios Tsioukas,Christos Pikridas

Research Square (Research Square)(2023)

引用 0|浏览2
暂无评分
摘要
Abstract In recent years, sophisticated 3D reconstruction and measuring techniques such as Terrestrial Laser Scanning (TLS), LiDAR applications, and Photogrammetry have gained popularity for surveying and mapping of physical environments. Using a case study of UNESCO-recognized Byzantine Bath monument, we examine the accuracy and reliability of different techniques. To assess the accuracy and dependability of modern 3D reconstruction and measurement systems, we created 3D models of the area of interest using a combination of Terrestrial Laser Scanning, LiDAR technology in conjunction with Simultaneous Localization and Mapping (SLAM) systems and algorithms. The geometric features of the 3D models were evaluated by comparing the data generated by these approaches to ground-truth measurements. With errors ranging from 0.01 to 3.2 cm, our results demonstrated that the majority of the available techniques produced highly accurate and trustworthy 3D models of the monument. The TLS and SLAM approaches performed admirably in terms of capturing fine features and the monument’s geometry. Also, the iPhone solution occasionally was more effective at capturing the surface's geometry, texture and colour. Our work demonstrates the efficacy of modern 3D reconstruction and measurement techniques for surveying and mapping environments with precision and dependability. To evaluate our results, we use open-source algorithms and tools in order to quantify the differences from point cloud to ground truth point cloud in many different scenarios (vertical/horizontal sections and floor analysis on the z-axis component).
更多
查看译文
关键词
byzantine baths,measurement techniques,modern 3d reconstruction
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要