Research of face recognition system based on visual intelligent monitoring

The key technology of intelligent face recognition based on the visual monitoring was researched. Firstly, obtained the real-time scene by video camera, and the background difference method was used to judge whether there is a character goal in monitoring region, in view of the existing target, the color space of YCgCr and YCgCb were chosen to detect skin color, the geometric constraints that the aspect ratio of facial contour is 0.8 to 1.85 was used to realize the positioning of face region. Then the improved 2DPCA algorithm was used for face recognition, and data matching with the established database of face to determine the validity character identity. Finally the identification results were sent to the mobile terminal through the control system. System test results show that the processing speed of face recognition is 7 s/times, and the accuracy of face recognition results was 85%, which can provide technical support for the identity intelligent recognition of visual monitoring. © 2016 SERSC.