Handheld Person Verification System Using Face Information

摘要

This paper presents our newly developed wireless system for person verification based on the face verification technology. In our system, a handheld device cooperating with a client-server wireless network is used. Due to the limitation of the handheld device's computing power, most computations will be distributed to a remote control server. There are two main problems for implementing the system. One is the variations in lighting and background conditions. The other is that the camera on the handheld device can not be calibrated with the system in advance. For calibration on line, we use a three-point localization scheme for extracting appropriate face region according to the information of eyes and mouth. Furthermore, statistic based illumination normalization is used in preprocessing to decrease illumination influence under variant lighting conditions. Experiment shows the proposed system provides users a more flexible and feasible way to interact with the verification system through handheld device.