Real Time-Based Face Recognition, Tracking, Counting, and Calculation of Spent Time of Person Using OpenCV and Centroid Tracker Algorithms

The study of human vision is one of the current controversial subjects in the computer vision consortium. Real-time tracking, face recognition, and counting of persons from video footage and CCTV or Webcam are almost new in flexible activities. Face recognition is often the first impression when used in video tracking, interfaces, and facial recognition. In this paper, we resolve to implement human face recognition, tracking, and counting, and spent time calculating in real-time using OpenCV and Python programming. The pilot method was implemented by using different types of OpenCV libraries like face recognition, Imutils, DateTime, and Centroid Tracker algorithm which we acquired the exact and absolute outcome for face recognition, tracking, counting, and spent time calculation depending on real-time. Face recognition algorithms is recognized faces in video files or webcams by showing people's identities. People are tracked and assigned an Individual ID using OpenCV's centroid tracking algorithm. Then the system calculates spent time for indicates the person was in the frame. After that, counting the number of people with the help of different personal IDs of other people which were provided by the tracker. The live frame captures the number of people (Live person count) and the total number of people (Total people count). Real-time and date are also displayed for security reasons. Through this research, we have been able to identify person through face detection, track and count all individuals with spent time calculating in real-time, which plays an important role in the security of important public and private institutions, especially banks, shopping malls, universities, etc.