Identification Of the Main 3D Scanning Techniques Suitable for Use in Cultural Heritage Objects Tobias Reich i3mainz, Institute for Spatial Information and Surveying Technology University of Applied Sciences Mainz, Germany
Agenda Basics Terrestrial Laserscanning Structure from Motion Application of the techniques in an excavation project The collapsed wood accumulation in the well-room of the Royal Palace of Qatna (Syria) High-Precision Measuring Techniques (3D acquisition of surfaces) Basics Different projects 1-16
Basics of Laserscanning Technology Measures distances and angles to object Generates 3D points similiar to a Total Station Difference: TLS creates homogeneous cloud with thousands of 3D points in short time Is established product for visualizing, reconstruction, deformation analysis, documentation and other applications The quality of TLS and its different influence factors are correlated i.e.: surface area, angle of incidence, recording distance 2-16
Basics of Laserscanning Technology (TLS) http://www.faro.com gib.uni-bonn.de Distance One sigma standard deviation 1 25 m 5 mm 50 m 9 mm Both information with a mean albedo and 127000 points (pts)/sec scan rate Point accuracy Leica GeoSystems HDS6000 out of the user s manual 3-16
Basics of structure from motion Technology Technology originates in Robot Vision Technology tracking of objects, collision tests, motion control, etc. Used in different software packages to generate 3D point clouds / models from a group of pictures automatically examples: VisualSFM (Open Source); AgiSoft Photoscan (commercially) Workflow similar in different software packages http://www.imageclef.org Unsorted pictures Orientation & bundle adjustment 3D reconstruction Extraction of features, contours and deep edges interior and relative orientation by assignment of homologous image areas Optionally with meshing and texturing Figures: http://www.carlos-hernandez.org/research.html 4-16
Basics of structure from motion Technology Workflow on a terracotta figure with around 50 images. Images were generated with a Nikon D300 - typical Referencing capturing with 3D points scenario on the object for an Isolate Object Comparison with highly precise 3D model (created with fringe projector type GOM Atos III) Nikon D300 GOM Atos III terracotta figure 5-16
Basics of structure from motion Technology - Terracotta figure Workflow of used software - Collecting image set (import) - Getting the tie points (SIFT operator) - Matching homoglues image points - Adjust camera & orientation parameters - Undistored the images - Generate a 3D point cloud, mesh (tex.) 6-16
Basics of structure from motion Technology - Terracotta figure - Collecting image set (import) - Getting the tie points (SIFT operator) - Matching homoglues image points - Adjust camera & orientation parameters - Undistored the images - Generate a 3D point cloud, mesh (tex.)
Basics of structure from motion Technology - Terracotta figure - Collecting image set (import) - Getting the tie points (SIFT operator) - Matching homoglues image points - Adjust camera & orientation parameters - Undistored the images - Generate a 3D point cloud, mesh (tex.)
Basics of structure from motion Technology - Terracotta figure - Collecting image set (import) - Getting the tie points (SIFT operator) - Matching homoglues image points - Adjust camera & orientation parameters - Undistored the images - Generate a 3D point cloud, mesh (tex.)
Basics of structure from motion Technology - Terracotta figure - Collecting image set (import) - Getting the tie points (SIFT operator) - Matching homoglues image points - Adjust camera & orientation parameters - Undistored the images - Generate a 3D point cloud, mesh (tex.)
The collapsed wood accumulation in the wellroom of the Royal Palace of Qatna - A 3D Recostruction 7-16
12 m Qatna The Well-Room U 9 m 8-16
The waterlogged timber accumulation 9-16
The 3D dokumentation with TLS in 6 Field Scan Sessions Layer 2008 Layer 1 (top Layer) Layer 2 Layer 3 HDS3000 Layer 4 Layer 5 Layer 6 (lowest Layer) Layer 2009 10-16
Detailed 3D documentation of individual pieces of wood - structure from motion 3d model of a piece of wood 3D Camera positions and orientations 11-16
Combination of the 3D result from TLS and SFM - 3D dokumentation of the collapsed wood accumulation 12-16
collapsed wood accumulation - generated from all measured layers 13-16
HIGH-PRECISION MEASURING TECHNIQUES - 3D acquisition of surfaces Fringe Projection Technology: photogrammtric solution equipped with an active light projection system is characterized by a stable and invariant mounting of two cameras and a pattern projector with high luminosity cameras are arranged with converging optical axes, so they see the same object part projector illuminates the object with different patterns This allows to solve correspondence problems and to calculate the spatial position of each object pixel GOM ATOS III Point spacing: 0,03 mm (measurement volume 65mm³) 3D point accuracy: 1/5 of its size 0,75 mm (measurement volume 1500mm³) 14-16
HIGH-PRECISION MEASURING TECHNIQUES - Examples of projects (Fringe projection System) Documentation and analysis of stone-carved Buddhist texts of the 8 th - 9 th Century in Sichuan, China The 3D data sets were the basis for the development of a analytical tool for a Photography Web-Atlas of a wall cutout of Scanned with stone-carved Sutra Inscriptions (cooperation 3D with model Academy of a wall of Sciences cutout with Heidelberg) stone-carved Buddhist texts Buddhist texts 15-16
HIGH-PRECISION MEASURING TECHNIQUES - Examples of projects (Fringe projection System) High-precision measuring techniques for the monitoring of surfaces from heritage objects (cooperation with the Institut für Steinkonservierung e.v.) Comparison between different measurement campaigns in a false color scheme 3D model 2008 Comparison: 2008-2009 Comparison: 2008-2010 Comparison: 2008-2012 16-16
Thank you!