1007 
3D DIGITAL INVESTIGATION ON DISEASES OF TIBET MURALS 
Shaoxing Hu a ’ , Feng Gao b , Aiwu Zhang c , Yulin Chen c 
Guangjie Cai c , Wenguo Duan c , Xiao Zhou b 
a School of Mechanical Engineering & Automation, Beijing University of Aeronautics and Astronautics, Beijing 
100083, - husx@buaa.edu.cn 
b Chinese Academy of Cultural Heritage, Beijing, 100029, - gaofenga@sina.com 
c Key Laboratory of 3D Information Acquisition and Application of Ministry of Education, Capital Normal University, 
Beijing 100037, - zhangawl63@163.com 
Interactive Session of Special Sessions (SS) 
KEY WORDS: 3D laser scanning, Range images, Photogrammetry, Multi view geometry, Orthophoto, Mural Survey 
ABSTRACT: 
Murals are the important components of culture and arts of Tibet. Unhappily, all of them have been ruined or are being ruined by 
such diseases as cracks, hollowing, salting, shedding, bleaching, discolouration, erosion and so on. The original data collection and 
disease inspection of murals is an important and complicated task for mural preservation. We describe a hardware and software 
system for 3D digital investigation on diseases of Tibet murals under non-laboratory conditions. Our system employs laser 
triangulation rangefinders, laser time-of-flight rangefinders, digital cameras, and a suite of software for acquiring, computer-aided 
modeling, disease marking and disease reporting. We digitized the whole prayer-wheel gallery with about 700m 2 of Jokhang Temple 
including 6 seriously damaged murals. We discussed the challenges we faced in 3D technology applying to investigation on diseases 
of murals. We focus in particular on processing huge laser scanning data and building a metrological 3D orthophoto model. We 
proposed a systemtic method of investigation on diseases of murals by integrating of 3D laser scanning and digital photogrammetry 
based on multiview geometry, not only the geometric and texture information of the murals are completely fused, but also the 
location, length and area of mural diseases are measured. 
1. INTRODUCTION 
Jokhang Temple was built 1,000 years ago, and it conserves 
Tibet's most ancient murals. But due to many reasons, all of 
them have been ruined or are being ruined by such diseases as 
cracks, hollowing, salting, shedding, bleaching, discoloration, 
erosion and so on, the preservation of murals is a key question. 
The traditional means, such as drawings, photographs and 
written records, can not capture as detailed as disease and shape 
information of murals. 
3D laser scanning technology in recent years developed a high 
technology, by high-speed scanning, allow us to accurately 
digitize the shape and surface characteristics of many objects. 
As an application of this technology, June 2007, a team from 
Capital Normal University and Chinese Academy of Cultural 
Heritage collected the whole prayer-wheel gallery with about 
700m 2 of Jokhang Temple including 6 seriously damaged 
murals. 
In fact, there are many successful stories in 3D laser scanning 
technology applying to heritage preservation. A team of 30 
faculty, staff, and students from Stanford University and the 
University of Washington spent the 1998-99 academic year in 
Italy digitized sculptures and architecture by Michelangelo, and 
then they developed a series of 3D laser data-processing 
algorithms [Cignoni P., 2004, Curless B., 1996, Davis, S.R., 
2002, Garland M.,2002, Garland M.,2002, Rusinkiewicz S., 
2000, Shaffer E., 2005]. The National Research Council of 
Canada (NRC) has done a lot of research work in heritage 
preservation [NRC-IIT, 2007], published more than 50 
academic papers since 1988. 
However, there are few reports about the application of 3D laser 
scanning technology in investigation of mural diseases. The 
technical goal of this paper is to propose a systemtic method of 
investigation on mural diseases using 3D laser scanning 
technology and determine mural disease location, size, length, 
etc, in the true 3-D environment. 
Figure 1. The pipeline of investigation on mural diseases 
In the remaining sections, we describe range data capturing 
(section 2), our post-processing pipeline (section 3). In section 4, 
Corresponding author.