3D face recognition can be also used in many application like secure access into systems or recognize him for smart TV and allow him online shopping (e.g. can be allowed only for parents not for kids, etc.).
The 3D face recognition task requires like 2D face recognition an input from a camera. For 3D face recognition 3D facial surface is needed and has to be captured. The main face recognition process consists of following sub processes like:
- 3D facial surface capturing - there are several different ways how to achieve this task, for example stereo cameras, laser or depth camera (e.g. of the Kinect sensor), etc.
- preprocessing - the captured data are subsequently preprocessed
- feature extraction - the purpose of feature extraction is to extract the compact information from the images that is relevant for distinguishing between the face images of different people and stable in terms of the photometric and geometric variations in the images
- measurement of the distance -the last step of the 3D face recognition is the measurement of the distance between 3D face of the test user and the 3D faces stored inside of the database. There are several techniques to measure the distance. The simplest method is measuring a local and global distances of two faces where it is needed correctly and very accurate to determine facial points (like eyes, nose, mouth, chin, ears, etc.) and measure their given distances by established metrics. The more sophisticated methods are nearest-neighbor classifier, techniques including support vector machine etc.
Fig. 3.8 – Example of GUI for 3D face recognition