| Start Date: | 5/20/2014 | Start Time: | 2:00 PM |
| End Date: | 5/20/2014 | End Time: | 3:30 PM |
|
Event Description
Title: 3D Surface Modeling and Registration with Applications in Computational Archaeology and Face Recognition
Advisor: Dr. Fernand Cohen
Date: Tuesday, May 20, 2014
Time: 2:00 p.m.
Location: Materials Conference Room 348, 3rd Floor, LeBow Engineering Center
Abstract
In this thesis we present several novel methods in 3D surface modeling and registration as well as their applications in archaeology and face recognition. In 3D modeling, a novel curvature adaptive B-spline fitting method is proposed, where salient points on curves and surfaces are preserved providing robust 3D features for registration. For registration, we propose a novel personalized morphable generic model.
The first application where these methods are applied is in archaeology, where the proposed methods assist the archaeologist in the tedious process of reconstructing ceramic vessels from excavated fragments. These methods take advantage of the surface geometry features coupled with a series of generic models constructed by the archaeologists to produce a virtual reconstruction of what the original vessel may have looked like. Generic models are generated based on the expert historical knowledge of the period, provenance of the artifact, and site location. The generic models need not to be identical to the original vessel, but must be within a geometric transformation of it in most of its parts. The methods are suited for ceramic vessels with some relief (i.e., surface with molding, carving, or stamping). When the fragments have no relief on them, they are virtually mended to abutting fragments using intrinsic differential anchor points computed on the fragments break curves and a set of absolute invariants. For axially symmetric objects, a global constraint induced by the surface of revolution is applied to guarantee global mending consistency.
In face recognition, a 3D face is constructed from a limited number of 2D images and a generic face model. Better face recognition rate is achieved by using a gender/race based generic 3D face model. The 3D face model allows for accurate pose estimation as well as the face identification in 3D. This renders a difficult 2D face recognition problem into a much simpler 3D surface matching problem. Experimental results are shown in controlled environment with high resolution images, ideal illumination condition, as well as for images taken in uncontrolled environment with low resolution, arbitrary illumination and promising results are obtained. |
|
Location:
Materials Conference Room 348, 3rd Floor, LeBow Engineering Center |
Audience:
Current StudentsFacultyStaffGraduate Students |
|
