Face detection is an essential component of computer vision in artificial intelligence, and it finds widespread applications in both industrial manufacturing and daily life. This paper investigates a face detection system using the Haar Cascades algorithm

Abstract

Face detection is an essential component of computer vision in artificial intelligence, and it finds widespread applications in both industrial manufacturing and daily life. This paper investigates a face detection system using the Haar Cascades algorithm to recognize face objects in real-time frames and locate faces within the frames. The system comprises a face recognition training module with input training data and a feature extraction algorithm for faces. After obtaining real-time images from the camera, data preprocessing is conducted, and the data is fed into the trained face recognition module to output the position of the face. Once the face's position on the frame is determined, the face area is highlighted to display the face detection results on the frame. The research results indicate an accuracy rate of up to 96% under stable lighting conditions. However, under unstable lighting conditions, with the camera focused on distant objects beyond 500mm, the accuracy rate drops to 66%.