After more than ten years of technological development and application market cultivation, today's biometric technology has been applied on a large scale in many application fields. As far as the Chinese market is concerned, the application of biometric technology was mainly based on fingerprint recognition technology, which is one of its sub-technologies. In the Chinese market, the application of fingerprint recognition occupies most of the market share (according to the 2007 Data, China’s biometric market, fingerprint recognition applications account for more than 90% of the market share). There is a big gap between this and the general trend of global biometric technology applications. However, we are also pleased to see that the application of some new technologies, such as face recognition technology, in the Chinese market has also developed rapidly in the past one or two years, and its market potential cannot be underestimated. From the perspective of technology and application, this article will discuss the two most significant branches of biometric access control: face recognition and multiple biometrics.

　　 face recognition technology

　　As far as the global market is concerned, face recognition is the second-largest biometric technology after fingerprint recognition. Generally, face recognition technology is mainly used in the following three areas. In these application areas, face recognition has unique advantages:

　　1) Large-scale file management systems, such as driving license and passport management systems;

　　2) According to the blacklist, automatically monitor the target person through the camera surveillance network;

　　3) Access control system

　　 In fact, face recognition has great potential in personal use, commercial and government applications. This technology is still in continuous progress and development, and its applications are becoming more and more popular. As far as the current technology and product-level are concerned, the face recognition system can be implemented and applied smoothly at a cost level acceptable to customers.

　　Principle of face recognition technology

Simply put, the face recognition technology requires the computer to record the features of the face (either on-site collection or from photos), and then when it is used (collecting face images on site again, or submitting photos), it can automatically and accurately identify whether the same person.

　　 Like other biometric technologies, the face recognition system also first generates the feature template of the face and stores it in the database. These templates will be used for one-to-one matching comparison with the templates submitted for comparison. Generally, when two templates are compared, if the similarity exceeds the threshold set by the system, the system considers the comparison to be successful. --These two templates are from the same person.

　　 There are also different types of face recognition technology principles, including recognition based on facial features, recognition based on skin texture analysis, and even recognition based on facial temperature patterns. The latest technology uses three-dimensional face image creation templates-there are technologies that automatically generate three-dimensional image modeling from two-dimensional images of human faces, and there are technologies that truly collect facial and head image modeling in a three-dimensional manner.

　　 Of course, what these technologies have in common is that they are all based on facial image information. Except for this common starting point, they are very different.

　　Because face recognition depends on the image information of the camera/photo, it is possible to understand how much the image quality affects the accuracy of the system. The quality of face images collected on-site is affected by many factors: ambient light, camera image resolution, lens focal length and depth of field parameters, the moving speed of the target face, and many other factors.

　　 In addition to image quality, the second major factor affecting the reliability and accuracy of the face recognition system is the technical level of the recognition algorithm itself.

"Cameras and scanners are like eyes, but recognition algorithms are the brains of face recognition systems," said a supplier of core algorithms for multiple biometric technologies. "Without high-quality algorithms, a face recognition system cannot be reliable. For biometric applications, system reliability is undoubtedly the most important consideration for customers."

　　 Human factors are also important factors that affect the accuracy of face recognition technology. For example, the use of the camera, the lighting conditions, and the changes in the shadow on the target face, etc., may greatly compromise the performance of the face recognition algorithm. Algorithm vendors have been thinking of ways to reduce the adverse effects of these human factors, and for this, they have also developed many innovative methods and technologies. How to ensure that the right camera and algorithm are selected in a particular application is also very particular. Different applications require different technologies. Many camera products and face recognition algorithms on the market do not all have the same quality.

Author: Written by Mr.Juan from S4A INDUSTRIAL CO.,LIMITED