Face Recognition for Perimeter Security

Abstract

Human face recognition and real time recognition from video sequences to application tests is a challenging task due to variations in the background, the look of the face and the light. At present, there is very little information and use of the facial recognition system in India. The project proposes a system that will use facial Recognition to track or search for a target person from real-time video feeds, such as video feeds from a surveillance system. First, the program is provided with live video of the area to be tested. Thereafter it is provided in the input data set of the target image for example, known person, etc. As soon as the input is provided, the program will extract the specified set of facial expressions from the Dataset Input and create a training module that will enable them to search for the person from the original video footage. When a match is discovered, the system will identify and tag the person and if the match is unavailable, the system will identify it as unknown and notify the manager of the same. And one of the main goal of this project is to develop the aforementioned program in line with the existing Surveillance program i.e. make it compatible with pre-installed surveillance cameras to keep costs down and hassle of using them to a minimum. The implementation of the proposed system would be for Government agencies such as Police, Military, Municipal Companies, Major Companies, etc.