Face recognition technology is based on human facial features, it is the input face image or Video streaming. First to judge whether there is a face, if there is a face, it will show the location and size of each face and the location information for main facial organs. And according to this information, we further extract the identity features contained in each face and compare them with stored face to verify the identity of each face
How Does Face Recognition Work?
Face recognition technology contains three parts:
1) Face Detection
Face detect refers to judging if there is a face in dynamic situation and complex background and separating it. There are several methods as follow：
A. Reference Template
It means establish a standard face template first, then analysis the similarity degree of captured sample and standard face template
B. Establish Face Rule
As human face has some structure features, then we can capture these structure features and create corresponding rules to judge whether there is face.
C. Learning Samples
Learning samples is neural network method in pattern recognition, means generate classifier by learning face samples and non-face samples.
D. Skin Color Model
This method is based on the create classifier consists of defining the bounding limits of the regions corresponding to a color belonging to the skin, by means of some numerical (and often empirical) rules.
Eigenface is a set of eigenvectors when they are used in the computer vision problem of human face recognition. In the recognition process, an
eigenface is formed for the given face image, and the Euclidian distances between this eigenface and the previously stored
eigenfaces are calculated.
2) Face Tracking
Face tracking refers to tracking the detected face as per dynamic target. A model-based method or a bombinated method of motion and model are adopted. In addition, skin color model tracking is a simple and effective method
3) Face Comparison
Face comparison is to identify the detected face image or search the target in face database. In fact, it means that the captured face image will be compared with the stored face image in turn, and the best matching will be found
So, the specific method and reliability of face recognition depends on facial description. Mainly use two kinds of description method:
A. Eigenvector B. Face Texture Template
The core technology of face recognition is “Analysis of Local Human Body Feature” and “Graphic/Neural Recognition Algorithm.” This algorithm is based on human facial organs and feature parts. For example, compare, judge and confirm the identification parameters of correspondence geometrical relationship with all the original parameters in database. Generally, the judge time should be less than 1s
Advantages of CAMABIO Face Recognition Technology
CAMABIO face recognition technology is base CAMABIO patented face recognition algorithm and 10 years biometric hardware manufacturing experience. Now CAMABIO can provide full range of face recognition product like face recognition algorithm SDK, face recognition algorithm chip, face recognition module,face recognition access control and time attendance solution, citizen identity card and face recognition system,face recognition cloud service,etc.
CAMABIO face recognition have following advantages
1. Nerural network algorith with deep learning features
2. 3D live face detection which can reject picture and 3D printing model effectively
3. With a strong environment capability.
4. With the strong recognition ability which far more than eye recognition