Improving 3D Face Recognition Model Generation and Biometrics

3D face shape biometrics, with greater pose and lighting condition data invariancethan 2D (photometric), has the potential to yield superior performance to 2D data forsome applications. However, many of the capture limitations of 3D scanners are the sameas those of 2D biometric capture devices with respect to lighting, environmental, andsubject configurations. Many of the claimed advantages of 3D over 2D do not exist undercurrent capture configurations. In addition, 3D scanners themselves are more expensivethan 2D cameras, and 3D biometric data are unavailable on many of the subjects wewould like to identify. Further, the significant computational cost of 3D face recognitionhas made large scale deployment of 3D face recognition impractical. The focus of thisthesis is to address these issues to improve the feasability of 3D face recognition so that itis more applicable outside of a research environment. In particular, the focus is onimproving the methods and hardware needed to produce a 3D model of a face, improvingbiometric recognition and verification performance, and decreasing the computationalcost to allow for larger scale applications. In this thesis, I propose a new structure frommotion approach, a new fast 3D face biometric, and examine the impact of movement onexisting structure from light devices.