Detecting Hostellers Using Face Recognition

For education, a large number of people are migrating to different cities. The majority of them live on campus in a dorm. The majority of hostel students leave the campus without authorization from the tutors and wardens. It is difficult for universities to identify a hosteller who has fled the campus when hundreds of students depart at the same time. Institution lacks a proper system for identifying them. Hostellers who leave the campus without authorization tend to engage in unethical activities, resulting in a negative reputation for the institution and the hostel. The overall goal of this paper is to design and implement a system to detect hostel students who are leaving campus. Face recognition technology will be used to detect hostellers, and if they are found, the warden of the hostel and the tutor of the particular student will be notified through SMS. To keep track of the records, a CSV file is also created. This project uses Histogram of Oriented Gradients (or HOG) method at backend to detect faces. Dlib library is used to find facial recognition and to wrap the image. Convolutional Neural Network (CNN) is used to create 128 measurements which is later used to find the identification of the person. OpenCV is used to access. This method requires no human involvement, making it simple for institutions to implement. This technology can be utilised to avoid problems and incidents caused by unauthorised hostellers leaving campus.