Home > Visualization >
Computational Forensics Project Background
Scenario:
Decomposed human remains are found.
It is determined that a reconstruction of the face will be necessary for
identification of the deceased.
Local law enforcement personnel obtain a scan of the skull from a
medical imaging facility. The 3-D data set is then transmitted via
the internet to a site where the computational reconstruction tools
reside. The reconstruction is accomplished, and
a completed rendering of the facial image is
downloaded by the local police to a relatively inexpensive computer, using
widely-available internet software. The rendering can then be used
for presentation to potential witnesses or family
members, and this can be done simultaneously by law enforcement personnel
at any number of locations.
Project Goals
To develop an internet-based set of
computational tools and data to support the 3-dimensional
forensic reconstruction of faces from human skeletal remains.
The reconstruction tools will include methods for predicting the
surface of the face and selected features, and techniques for
graphically providing missing details,
such as eyes and natural looking skin texture. The resources
required at the local level, i.e. law enforcement department,
will be minimal in terms of computing equipment and personnel
skills and training.
Core Capabilities
- Whole-head MRI and CT data base
- Automated techniques for segmenting soft from bony tissue
- Methods for predicting facial surface and features, from skull
- Visualization techniques for skin texture mapping and feature addition
These capabilites have required the establishment of a
multi-disciplinary team with expertise in computer science/artificial
intelligence, visualization algorithms, forensic anthropology,
medical imaging, and perceptual psychology.
project background |
potential benefits
pilot summary | image gallery
|
