Face Recognition in Biometric Systems Using Haar Cascade Classification
L. Kh. Safiullina, A. Sh. Gabdullin, Igor Vyacheslavovich Anikin · 2021
the paper considers the problem of face recognition in biometric systems using Haar cascade classifiers. We evaluated four types of Haar cascade classifiers from the OpenCV library by their performance and accuracy: haarcascade_frontalface_ default; haarcascade_frontalface _alt; haarcascade _frontalface _ alt2 and haarcascade _frontalface _alt _ tree. The framework was developed for providing experiments. Training and recognition time was considered as well as FAR and FRR rates for these classifiers. Evaluation has been done based on 1,120 images collected from 150 users. As a result, haarcascade _frontalface _alt and haarcascade _frontalface _alt2 classifiers were recommended for solving the face recognition problem in biometric systems.