Surveillance of Gathering by Using Computer Vision

Abstract

The importance of facial recognition systems has accelerated in recent decades. A facial recognition system is part of the processing of biometric information. The field of facial recognition is rapidly growing into a vital part of our everyday lives. The use of facial recognition systems has been extended primarily from security purposes to Social networking sites, managing fraud, and improved user experience. The purpose of this study is to develop a system which will help event organizer or law enforcement to monitor and recognize crowd in a gathering. Traditional observational surveillance methods are typically limited by the stationary nature of the camera, which is usually handled manually or fixed upon a tripod or other structure therefore in this project drone will be used as a mean for surveillance. Real-time Person tracking in required to find the authentication of individual in gathering, the first stage is to detect the person's face using a drone. To detect face, it is required to preprocess images. In this project image processing is done by converting RGB images into grayscale, and contrast stretching is used to enhance image. Pre-process images to remove noise and blurriness in images and system detect the person face in an image. After face detection, facial features are extracted to train model and stored in the database.Person tracking is done by face recognition, to detect face in a frame the propose system uses Haar cascade classifier. Person tracking is done throguh face area and center point of face and will show us the current location of a person. This survillance system is very usefull for keeping safety and security of a person in crowd gathering.