Research On 3D Face Recognition Based On Statistical Analysis

Face recognition has been one of the most popular topics in the field of pattern recognition and image processing.This thesis mainly focuses on 3D face recognition problem. Beginning with a survey of existing methods applied to two-dimensional (2D) and three-dimensional (3D) face data, we focus on subspace techniques, registration is done to transform face models into the same pose and position. Finally precise and effective recognition algorithms are used to achieve high accuracy. The main contributions of the work are as follows.1,The grid control points is used to simulate the point-cloud data which is created by B-spline surface fitting. We standardize the point-cloud data to reduce the quantity of point-cloud data to raise efficiency of our algorithm.2,The partial-ICP method is used to align all the 3d face model, which could implicitly and dynamically extract the rigid parts of facial surface by selecting a part of nearest points pairs to calculate dissimilarity measure during registration of facial surfaces. The method is expected to be able to get much better performance than other methods in 3D face recognition under expression and pose variation.3,The thesis investigates principle component analysis (PCA) approach and the methods based on fisher discriminant analysis deeply, and then the choice of feature vector and distance measure criterion is discussed.4,Investigating the use of image pre-processing applied as a preliminary step in order to reduce error rates, we implement the eigenface and Fisherface methods of face recognition with both 2D and 3D face, finally we combine them, this research leads to an innovative multi-subspace face recognition method capable of combining 2D and 3D data, yet multi-subspace combination increase the accury rate to 98.3%.