Evaluating multi-sensor data for the development of three-dimensional photo realistic model of archaeological site documentation

Abstract

Stimulated by the highly demand on three-dimensional (3D) photo-realistic model of archaeological remains, a method which can rapidly record the remains meticulously has become a crucial requirement. Conventional method seems to be impractical to fulfil the demand because the output is in two-dimensional, destructive method, require bigger storage and a lot of manpower. Hence, this study was carried out in order to develop method for integrating multi-sensors data with the aim of producing 3D photo-realistic model of archaeological remains. Besides that, it is also intended to determine the reliability of Leica ScanStation C10 Terrestrial Laser Scanner in generating accurate geometric model of the monument. Based on the results obtained, the integration of ScanStation C10 and images data is capable in providing a 3D photo-realistic model of Bukit Batu Pahat shrine where feature mapping accuracy shown that the model is 90.56 percent similar with the real object. Additionally, the geometrical accuracy of the model generated by ScanStation C10 data is very convincing which is ±4 millimetres. Besides that, NextEngine scanner had produced an outstanding photo-realistic model of the artefact where feature mapping analysis shown that the artefact models are 88.89 percent similar with the real artefact while ±0.02 millimetres for geometric mapping accuracy. Visualization of the models in 3D environment is effective and fascinating. Instead of conducting virtual measurement and texture analysis, archaeologists can use the visualization as a bridge to share archaeological remains with the public. In a nutshell, the goals of this research have successfully achieved where a 3D photo- realistic model of Bukit Batu Pahat archaeological remains were generated from multi-sensors data and possess good geometric accuracy