Non-environment-sensitive facial recognition system using two CCTV cameras

Abstract

With the rapid growth of the high image quality CCTV monitoring market, research on facial recognition from long as well as short distances is needed. We proposed a non-environment-sensitive hybrid facial recognition system using fixed-format and pan-tilt-zoom cameras. The proposed system enables efficient system load distribution and facial recognition in the transferred images from multiple sites and channels. Furthermore, this system enables the multiple face detection and image processing of multiple CCTV channels. This system can be used for security monitoring in the public areas, and can create a new intellectual video monitoring market. © 2016 SERSC.