Face detection method based on OpenCV and Dlib

• Main Authors: LYU DONGYE, LI MENGTONG, ZHANG ZIJUN, MA JUN, CHENG LIN, LIU RUIYUAN, SUN WENCAI
• Format: Patent
• Language: Chinese; English
• Published: 05.07.2024
• Subjects: CALCULATING; COMPUTING; COUNTING; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; PHYSICS

The invention discloses a face detection method based on OpenCV and Dlib, and belongs to the field of computer vision. The face comprises a human face and human eyes. Firstly, face video images are collected in real time through a camera, graying processing is carried out on the images, and bilateral filtering and CLAHE image processing methods are adopted. And then, a Haar feature classifier of an OpenCV library is used to realize face detection, a face ROI is intercepted, detection parameters of the Haar classifier are optimized through a grid search method, and a CSRT target tracker is introduced to track the detected face. And finally, converting the detected human face ROI into a Dlib image, and detecting human eyes by using human face key points in a Dlib library. The fast and simple characteristics of the Haar classifier are utilized, the face ROI is intercepted, the CSRT target tracker is introduced to track the face, and the face key points of the Dlib library are combined, so that the fast and accur