Facial Features based Recognition System using LBPH and AdaBoost

Face recognition stands as a pivotal domain within computer vision, boasting myriad applications in security, identification, and biometrics. This research study proposes a face recognition system using image processing techniques with the combination of two algorithms, Local Binary Pattern Histogram (LBPH) and AdaBoost. The LBPH algorithm was used to extract features from the images and create a feature vector for each face. The AdaBoost algorithm was then used to train a classifier on the feature vectors and classify new faces. After testing on the Face Recognition Database, the results reveal that the fusion of LBPH and AdaBoost algorithms achieves an impressive accuracy rate of 96.15% which is better as compared to other similar algorithms. The proposed algorithm is implemented in Python to recognize faces in a dataset of images.