×

You are using an outdated browser that does not fully support the intranda viewer.
As a result, some pages may not be displayed correctly.

We recommend you use one of the following browsers:

Full text

Title
New perspectives to save cultural heritage
Author
Altan, M. Orhan

VOLUMETRIC RECONSTRUCTION OF CULTURAL HERITAGE ARTIFACTS
Y. Kuzu, O. Sinram
Photogrammetry and Cartography
Technical University of Berlin
Str. des 17 Juni 135, EB 9
D-10623 Berlin, Germany
{kuzu,sinram}@fpk.tu-berlin.de
Commission V, WG V/2
KEY WORDS: Calibration, Close-Range, Correlation, Cultural Heritage, Photo-realism, Reconstruction, Visualization, Volume
Model.
ABSTRACT:
In this paper we present a voxel-based object reconstruction technique to compute photo realistic volume models of cultural heritage
artifacts from multiple color images. We use a CCD camera to acquire images of historical artifacts from several viewpoints. After
calibrating the camera, the orientation parameters of the acquired images are calculated with a bundle block adjustment. Since it is
impossible to mark control points on historical artifacts, natural texture was used. Our algorithm begins with initializing a volume
that encloses the 3D artifact to be reconstructed. A first approximation of the model is acquired by shape from silhouette which gives
the objects visual hull. Unfortunately, this method does not recover concavities on the object. In order to refine the representation,
several tools will be described. An important factor is the visibility information of a voxel in a specific image. It can be either on the
backside of the object, or occluded by other voxels in between. We present a fast method to recover this information with a line
tracing algorithm. Furthermore, a quality measure of the visibility is introduced, by using the surface normal vector of a voxel in
combination with the viewing direction of the image. These tools will help to generate the upgraded model more accurately.
KURZFASSUNG:
In dieser Arbeit stellen wir eine Technik zur Voxel-basierten Objektrekonstruktion vor, um fotorealistische Volumenmodelle von
Objekten aus mehreren Farbbildern zu erstellen. Wir verwenden eine CCD-Kamera um von verschiedenen Standpunkten Bilder
dieser Objekte aufzunehmen. Nach der Kalibrierung der Kamera werden die Orientierungsparameter der Bilder mit einer
Bündelblockausgleichung berechnet. Da es nicht möglich ist, Passpunkte auf dem Objekt anzubringen, wird natürliche Textur
verwendet. Am Anfang unseres Algorithmus wird ein Volumen initialisiert, welches das dreidimensionale Artefakt vollständig
umschließt. Eine erste Näherung des Modells wird mittels „Shape from Silhouette“ berechnet, welches die umgebende Hülle bildet.
Leider kann diese Methode keine konkaven Regionen des Objektes erfassen. Verschiedene Methoden werden beschrieben, die diese
Näherung verfeinern. Ein wichtiger Faktor ist die Sichtbarkeit eines Voxeis in einem bestimmten Bild. Ein Voxel kann sich auf der
Rückseite eines Objekts befinden, oder durch andere Voxel innerhalb der Sichtrichtung verdeckt sein. Wir stellen eine Methode vor,
die mittels eines Linienverfolgungs-Algorithmus sehr schnelle Information liefert. Darüber hinaus verwenden wir die
Oberflächennormale des Voxels als ein Qualitätsmerkmal für die Sichtbarkeit bezüglich der Blickrichtung eines bestimmten Bildes.
Diese Werkzeuge bilden die Grundlage, das Modell zu verfeinern.
1. INTRODUCTION
Reconstruction of the shape of 3D objects from a series of
images is a challenging problem both in the disciplines
photogrammetry and computer vision.
When generating 3D models with traditional CAD techniques,
the details of 3D objects should be entered manually using
graphical interfaces. Furthermore, with traditional CAD
techniques objects are modeled with polygon patches and
traditional material description which does not give high degree
of realism. Since it is a labor intensive and complex process, the
efforts concentrated in automatic modeling of 3D objects either
with active methods by scanning the objects or with passive
methods by using their images taken by a CCD camera. In this
paper, we describe an efficient image-based approach to
compute volume models of cultural heritage artifacts from their
color images which is a low cost therefore an attractive method.
Cultural heritage preservation is one of the application areas of
3D model reconstruction. A full three-dimensional
reconstruction serves as a permanent record of the heritage
artifacts in their original position. Such a reconstruction can be
used to detect the changes for conservation purposes. The
models may also serve as manufacturing blueprint for machine
production of replicas for exhibitions. If the object is an
historical device for example the visitors might like to see it in
action and perform their experiments. High precise
reconstruction of the artifacts or replica can be made accessible
to scholars and visitors.
Another application area is virtual museums where 3-D
reconstruction, modeling and visualization of cultural heritage
artifacts can be done volumetrically. A virtual museum is a
computer generated environment where the artifacts and
information resources of the museum can be viewed locally, or
on the internet. The users can view these artifacts from different