The rule is simple more the images in training more the accuracy in the testing. In the training process, we give more than 50 images to the recognizer so it can learn the different facets of the same person. For face detection, we use haarcascade_frontal_face.xml file to train haar cascade classifier.
More than 50 images of each person.
- LBPH Face Recognizer
- Haar cascade classifier
Face Detection on pictures with different emotion, gesture & posture using LBPH face recognizer. Then extracted unique characteristics of face through Data Gathering & trained the model which leads to Data Comparision when provided with test face. Face Recognition will determine who the person is.
Sharad Kumar Tiwari