Current methods for facial reconstruction are tedious and time-consuming, and require forensic
artists with years of practical experience. Furthermore, the complexity of the reconstruction
problem greatly increases when time-related factors come into play, such as those that
occur in missing children scenarios. This thesis describes a software system for simulating the
growth of the craniofacial skeleton. It is a first step towards our goal of a complete software
package for three-dimensional craniofacial reconstruction. There is a tremendous amount of
data on craniofacial growth in the form of studies that collect frontal and lateral cephalograms,
which can be used to generate three-dimensional coordinates of landmarks on the
craniofacial skeleton at various ages. We define a simplified model of bone growth that uses
these landmarks to drive the growth of the rest of the craniofacial skeleton. The inputs to our
growth model include a triangular mesh acquired from the bone to be grown (e.g. skull, mandible),
a set of vertices on the mesh identified as landmarks, the coordinates of these landmarks
through time, and vertex weights which are a measure of the influence exerted by
landmarks on the rest of the vertices. The output is a triangular mesh, "grown" either forwards
or backwards in time to a specified age. An expert in craniofacial growth assigns these vertex
weights by using a specialized tool called Krayola. We also provide a tool for automatically
generating a first approximation for the vertex weights of a new mesh given the weights previously
assigned to a mesh of similar bone type (e.g. skull, mandible). Validation of our
growth model is an outstanding issue; we lack three-dimensional data (e.g. from CT scans)
for an individual through time, with which we would compare the output of our software. For
now, we must be content with the expert opinion of our colleagues in the Department of Dentistry's
craniofacial reconstruction group, who are quite pleased with our results so far. / Science, Faculty of / Computer Science, Department of / Graduate
| Identifer | oai:union.ndltd.org:UBC/oai:circle.library.ubc.ca:2429/5919 |
| Date | 05 1900 |
| Creators | Coughlan, Kevin Michael |
| Source Sets | University of British Columbia |
| Language | English |
| Detected Language | English |
| Type | Text, Thesis/Dissertation |
| Format | 6828987 bytes, application/pdf |
| Rights | For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use. |
Page generated in 0.0017 seconds