A study on face and face mask detection using HAAR cascade

Abstract

Computer technology used in various programs to recognize people's faces in digital images is apprehended as facial recognition. There is a wide range of applications in the fields of content-based content retrieval, video encoding, video conferencing, crowd viewing, and smart computer interaction. Object detection is a computer vision approach that allows us to recognize and locate objects in an image or video. The main objective of this research was to develop a security system that could detect a face mask, a person's face, and the number of people standing in front of the camera in real-time using realtime video capture. OpenCV is a cross-platform library of over 2500 algorithms that have been optimized and were chosen for our project due to its benefits. One of the most basic machine learning-based methods is the Haar cascade classifier. Haar features are particularly useful for mask detection because they are excellent at detecting edges and lines, simultaneously, cascade classifiers are one of the few real-time algorithms available. The accuracy level was very poor when a built-in Haar cascade was used for the project, to solve this issue, a new Haar cascade was trained to improve accuracy which was proved to be effective. Three separate features, such as face mask, face, and human detection, have been brought under one project. A GUI was created using the Tkinter module to achieve that. Face mask detection, face detection, and human detection are the three key features of this project to ensure social distancing digitally. This designed system can be used effectively in applications like, keeping any location protected from intruders, ensuring face-mask wearing, and preserving social distance