Heft 81/2006

des Instituts für Geodäsie

Heft 81/2006


Digital 3D Documentation of Cultural Heritage Sites Based on Terrestrial Laser Scanning

139 S.

Auflage:  200

ISSN:  0173-1009





Vollständiger Abdruck der von der Fakultät für Bauingenieur- und Vermessungswesen der Universität der Bundeswehr München zur Erlangung des akademischen Grades eines Doktor-Ingenieurwissenschaften (Dr.-Ing.) genehmigten Dissertation.

Vorsitzender: Univ.-Prof. Dr.-Ing. K.-H. Thiemann
1. Berichterstatter: Univ.-Prof. Dr.-Ing. O. Heunecke
2. Berichterstatter: Univ.-Prof. Dr.-Ing. habil. W. Niemeier
3. Berichterstatter: Univ.-Prof. Dr.-Ing. habil. L. Gründig

Diese Dissertation wurde am 8. Februar 2006 bei der Universität der Bundeswehr München eingereicht.

Tag der mündlichen Prüfung:  29. Mai 2006

(in Englisch)

Chapter 1  -  Introduction
Chapter 2  -  Cultural Heritage Documentation 10
2.1 General Explanation 10
2.2 Digital 3D Documentation 14
2.3 Case Scenarios 14
Chapter 3  -  Data Acquisition 17
3.1 Geodetic Measurements 17
3.1.1 Traditional Surveying 17
3.1.2 Image Processing and Close-Range Photogrammetry 19
3.1.3 Terrestrial Laser Scanning  
3.2 Non-Geodetic Measurements 23
3.3 Combining Laser Scanning Data with Surveying and
Close Range Photogrammetry Data

3.4 Thermography Data Processing 26
Chapter 4  -  From Point Cloud to 3D Model 28
4.1 Point Clouds Registration 28
4.1.1 Point Clouds Registration Using Targets 29
4.1.2 3D Similarity Registration 30
4.1.3 Feature Registration 31
4.2 Point Cloud Treatment 32
4.2.1 Remove Unwanted Points 32
4.2.2 Detect and Remove Outliers 33
4.2.3 Detect and Remove Disconnected Components 34
4.2.4 Noise Reduction 34
4.2.5 Redundancy Reduction, Point Clouds Sampling and Decimation
4.2.6 Adding Points 37
4.3 Create 3D Geometry from Point Clouds 39
4.4 Create and Fit Basic Geometric Primitives 39
4.4.1 Fitting a Planar Patch to 3D Point Sets 41
4.4.2 Fit a Sphere to a 3D Point Sets 45
4.5 Create and Fit Basic Shapes 49
4.6 Create 3D Objects 53
4.7 Create Triangular Meshes 58
4.8 Clean and Edit Triangular (Polygonal) Meshes 59
4.9 NURBS (Non-Uniform Rational B-Spline) Curves and Surfaces

Chapter 5  -  Virtual Reconstruction 70
5.1 Introduction 70
5.2 The Virtual Reconstruction Process 73
5.2.1 Data Collection 73
5.2.2 Process Design 74
5.2.3 Virtual Reconstruction Aspects
Chapter 6  -  Visualization
Chapter 7  -  Case Scenarios 87
7.1 Project Maximilian Straße 87
7.2 Schönberg (Dealu Frumos) 90
7.3 St. Michael Church 93
7.4 Test Objects 98
7.5 Hohenburg Ruins 101
7.6 Aspects of a 3D Cultural Heritage Documentation System
Chapter 8  -  Conclusions 110
Acknowledgments 112
Biography 113
Annex I Basic 3D Object Creation 114
Annex II Software 115
Annex III Video Format Test 116
Annex IV Data flow - case scenarios 117
Annex V Project Schönberg (Dealu Frumos), Romania - Virtual Reconstruction at different building stages
Annex VI Data Index to Historical Buildings and Monuments of the Architectural Heritage
VI.I Maximilian Straße 121
VI.II Schönberg (Dealu Frumos) 124
VI.III Hohenburg 127
Annex VII Collection of Common Shaders and Standard Mapping Techniques Used by Common CG Software (3ds Max, Alias Maya, etc.)

Annex VIII Case Scenarios - Enclosed MultimediaCD Content

(in Englisch)

The classical measurement techniques used to acquire spatial data needed in documentations for areas like architecture, industry or cultural heritage are replaced, for the moment just at the beginning, by the newly emerged method of laser scanning. Nowadays, it seems that the technical community got through the initial enthusiasm of the first years from its development and the users are not exclusively interested in how fast and how many points this tool can acquired, rather more in how it can be optimally included in the current workflows. For now, the involvement level of the laser scanning technique in spatial data acquisition is given more by its disadvantages rather than by its abilities. The users are facing the option of including a tool and corresponding software rather expensive with the involvement of the established methods like photogrammetry or classical surveying. Due to this aspect, until now the laser scanning technique is used mainly in projects where the time factor is crucial. Moreover, the users are learning that it is more likely to try to combine the technique with previous tools than to replace entirely the current workflows.

Currently, starting with a more project-related market campaign of the hardware and software developing companies, the users are reorienting to the laser scanning tools that best serves their quality and economic needs. The economic benefits are coming mainly in areas like direct surveying costs, construction, asset management and safety dividend [SparView, 2005]. According to the same source the employment of 3D scanning tools generate savings in the order of 15-20% of the survey budget, or 5-10% on construction total project cost.

Starting with these pre-requisites and considering the laser scanning technology’s growing potential, this thesis is concentrated mainly on spatial data acquisition with 3D scanning instruments and processes that generate 3D geometry. A data flow of the processes from the recording phase to the presentation stages is described in detail and analyzed. The work also considers the processes included in currently available commercial software and describes a possible workflow for specific projects. The thesis also refers to the operations performed in order to combine laser scanning data with surveying and photogrammetry data, or data specific to other types of measurements. A concept of 3D digital environment for cultural heritage documentation is proposed and exemplified with different case scenarios.

One of the advantages of the laser scanning technology is the ability to provide initial 3D representations (point clouds) for the documentation of the recorded item right after the acquiring processes. A better representation is realized by cleaning the recorded point clouds from all the unwanted or erroneous data. From this point further it might not be akways necessary to create a 3D solid model due to the presence and development of plug-ins that allow the import of the point clouds in the major CAD software or modeling application. This implies the fact that not all users need the entire capabilities of a specific 3D scanning software in their workflows and some of the purchase costs are not justified. This asks for a module-differentiation in the software licensing method that allows the customer to acquire only the needed modules. Such a differentiation in modules is present in some of the available software (e.g. Cyclone 5.4) only on larger categories. However, a more sub-level orientation is needed. This would also help to a better management of the workflow where sub-level processes are distributed to separated users within the framework of the same company. In this way a company would not face higher costs when acquiring software and would be more encouraged to employ in it’s workflow the needed tool. In the actual globalization conjuncture, it is rather unlikely that all the processes of an entire workflow to be realized involving the software products of the same developer, thus a universal data file format for 3D models is more than needed in order to remove the frustrating and time consuming methods used nowadays when changing the application platform.

It is expected that the current 3D scanning systems will evolve in instruments capable to acquire data with quality similar to the current classical measurement instrumentation. Thus, a more comprehend and generally accepted quality control system has to be developed, not only by identifying possible problems but also by correcting these errors. Other hardware categories are given by the machines that perform the algorithms needed to visualize and generate data specific to the scanning technology and 3D model creation. The development of the computer systems includes the enlargement of disk storage, disk access [ACM Queue, 2003], memory optimization and other components that are contributing to optimize the software usage.

Future developments in this research area might include aspects related to the expansion of the automatization level on the entire workflow, advance in the concept of virtual large sites (cultural, architectural, industrial, entire cities etc.), a new restoration concept that combines real data (captured with laser scanning instrumentation) and virtual data, augmented reality (e.g. museum with virtual exhibits), etc.

Another aspect is referring to the personal qualification regarding these new data acquisition, visualization and documentation techniques. Currently, personal training is realized mainly on company level or by the hardware and software providers. A larger immersion in these technologies at different study levels in universities or other form of studies it is expected and wanted.

The present circumstances encourage a social and economical globalization, which makes sustained campaign for cultural heritage promotion necessary. Mainly, due to the following reasons: firstly, to increase the cultural heritage’s level of awareness between different nations and civilizations; secondly, to enhance the individual relationship to the cultural heritage. This could be realized also by overcoming the current phase of singular projects to large scale 3D digital cultural heritage documentation.

The 3D digital documentation of objects and sites having cultural heritage importance becomes more and more significant. Based on innovative measuring technique like terrestrial laser scanning – in addition with conventional procedures like tacheometry and GPS – and the possibilities of nowadays computer graphic software this can be performed in a way never known so far. These new possibilities are explored within this thesis in few case studies. In practice the next step must take full advantages of these new chances.


zurück zum Heft-Verzeichnis





(12,1 MB)