"The development of a robust and integrated multi-camera surveillance device is an important requirement to ensure public safety and security. Being able to re-identify and track one or more targets in different scenes with surveillance cameras. That remains an important and difficult problem due to clogging, significant change of views, and lighting across cameras. In this paper, traditional surveillance systems developed and supported by intelligent techniques. That system have ability to performance the parallel processing of all cameras to track peoples in the different scenes (places). In addition, show information about authorized people appearing on surveillance cameras and issue a warning whistle to alert security men if any unauthorized person appears. We used Viola and Jones approach to detected face, and then classifying the target face as one of the authorized faces or not by using Local Binary Patterns (LBP)."