The present invention relates to a monitoring method of an intelligent surveillance system which uses an object recognition technology. When a moving object is recognized from image data which is photographed by one of multiple PTZ cameras which are installed in the outside fence of a key facility, the object recognition technology identifies an object which is included in the corresponding image data. Especially, when the identified object is a human, the intelligent surveillance system recognizes the human as an intruder and informs an integrated control unit of the intruder, further makes an adjacently installed camera additionally track and monitor the intruder, and, at the same time, makes a wide area surveillance camera temporarily monitor a blind spot which can be generated by the additional surveillance of the adjacent camera. Furthermore, intelligent surveillance system is able to simultaneously verify whether the intruder is a risky person or not effectively by detecting the face area of the intruder and drawing age and/or sex information primarily. The present invention minimizes the generation of a blind spot by mutually connecting the wide area camera and the multiple PTZ cameras and conducting monitoring operation, identifies an object which is photographed by the cameras in real time by using the object recognition technology, especially, is able to obtain various kinds of image information by continuously tracking a moving path when the identified object is a moving person, is able to remarkably reduce the installation costs and maintenance costs of the cameras when being compared with an existing surveillance system, and furthermore not only increases surveillance accuracy but also minimizes monitoring malfunctions by recognizing only real risk elements by analyzing images which are photographed based on a face recognition technology.
展开▼
