3D face recognition using multiple features for local depth information

Abstract

Depth information is one of the most important factor for the recognition of a digital face image. Range images are ve y usefil, when comparing one face with other faces, because of implicating depth information. As the processing for the whole face produces a lot of calculations and data, face images can be represented in terms of a vector of feature descriptors for a local area. In this papel; depth areas of a 3 dimensional(3D) face image were exiracted by the contour linefi-om some depth value. These were resampled and stored in consecutive location in feature vector using multiple feature method. A comparison beiween two faces was made based on their distance in the feature space, using Euclidian distance. This paper reduced the number of index data in the database and used fewer feature vectors than other methods. Proposed algorithm can be highly recognized for using local depth information and less feature vectors on the face.

An ordered topological representation of 3D triangular mesh facial surface: concept and applications

“…Another category of approaches [18][19][20][21] extended the eigenfaces paradigm developed in 2D-image based recognition to the 3D context. This paradigm stipulates that a face image can be defined as linear combination of a finite number of a particular facial images.…”

Disparity-Based 3D Face Modeling using 3D Deformable Facial Mask for 3D Face Recognition

“…Using the vertices locations of the model mask and their corresponding disparities, from the disparity map, the mask is deformed to the individual's face by computing their 3D coordinates using triangulations. Unlike the existing 3D algorithms that rely on range data, which can be accurate but suffer from the difficulty of extracting salient facial features [11][12][13][14], ours extracts a total of 68 facial features using 2D techniques because it is easier to extract corners and edges from the intensity frontal images than from 3D range images. In addition, some of these algorithms either extract or deform a model to local surface curvatures in 3D world plane.…”

Gabor Filter Bank Representation for 3D Face Recognition

“…More recently methods such as [5,6] have been proposed for comparing faces using pure 3D data. Pan [5] uses registration error of low resolution Structured Light Scans (SLS) while Lee [6] uses contour lines.…”

“…Pan [5] uses registration error of low resolution Structured Light Scans (SLS) while Lee [6] uses contour lines. However the most promising of recent research has focussed on using a combination of both intensity and 3D information.…”