DATA ACQUISITION POSSIBILITIES FOR FACE RECONSTRUCTION PURPOSE
P.Schrott, Gy. Szabo, K.Fekete
Budapest University of Technology and Economics, Department of Photogrammetry and Geoinformatics
H-1111, Budapest, Muegyetem rkp. 3. Hungary
- schrott.peter@fmt.bme.hu, gyszabo@eik.bme.hu, feketekaroly@mail.bme.hu
Commission V, WG V/6
KEY WORDS: Close Range, Three-dimensional, Biometrics, Scanner, Computer Tomography, Face reconstruction
ABSTRACT:
A new multidisciplinary project extending over a number of years was initiated in Hungary to combine knowledgebases of different
disciplines like anthropology, medical, mechanical, archaeological sciences etc. to computerize the face reconstruction.. A research
group (BME Cooperation Research Center for Biomechanics) was formed representing several organizations that are cooperateing
during the project period. In this paper we will show the first results of our work: the examination of the possible data gathering
methods from special aspects. First the data collecting method has to be able to produce geometric 3D data of the cranium of
damaged/mummified subjects. Second, the software development requires huge dataset of 3D face and skull models, which could
be produced from living persons, so the method have to be capable to mass data collection. Since any modification of the process
during the gathering period can result inhomogeneous database, the accuracy and the feasibility of the measuring methods is highly
important.
1. INTRODUCTION
Facial reconstruction is the process of reproducing the geometry
of faces of unidentified persons from skeletal remains. Recently
the most widespread face reconstruction methods are used by
highly qualified and experienced anthropologists and based on
artistic tools supported by scientific methods. (Kustar, 2004)
The first step is anthropological investigation, by which the
anthropologist estimates the age and sex of the person, observes
individual characteristics, possible illnesses and injuries, and
carry out morphometric measurements on the cranium.
The reconstruction usually starts with making a gypsum copy of
the skull. The thickness of soft tissues at the most important
anthropological landmarks is estimated based on the roughness
of the bone surface and/or on statistical data. The mimic
muscles are made of plastilin or clay, the eyes are of marbles
and the nose is formed from paraffin or wax. (Figure. 1.)
In the last decades more and more researchers turned to
computer technology to make the reconstruction process easier
and more reliable. Virtual 3D models of a face can be modified,
measured or compared easily, but the available face
reconstruction softwares are still in their infancy. These
softwares are modifications of common 3D modelling softwares
but there are very limited or no facial data and anthropological
relationship implemented in them - they are “virtual clay sets”,
and for correct operation they still need the specialists’
knowledgebase and experience. The BME Cooperation
Research Centre for Biomechanics aims to create a face
reconstruction software based on statistical samples (3D face
and skull models of 4000 living person) and guided by defined
mathematical correlations between the skull and the face
geometry.
2. ANALYSIS OF DATA GATHERING TECHNIQUES
The possible data collection technologies for face
reconstruction (and for the software development for the same
purpose) are: Computer Tomography, X-ray, different types of
3D scanners and photogrammetry. Combinations of the
mentioned technologies are able to produce 3D models from
dry and ‘living’ skulls, cadavers and living heads, mummies etc.
2.1 X-ray
Traditional X-ray is a possible technology for obtaining data for
face reconstruction. To evaluate X-ray images reference points
in the object space with known position are needed, and such
points are mapped on the images in an identifiable and
measurable manner. We designed an X-ray test field containing
X-ray opaque point-type markers embedded in a material
invisible for the X-ray.
Figure 1. . Basic steps of face reconstruction (Kustâr, 2004)