Live human-face recognition with OpenCV using Haar Cascade Classifiers; frontal_face dataset and LBPH Algorithm, works with small computing devices.
- Run
face_add.py
to genrate face samples, add id - Train the sample images using
face_train.py
- Run facial recogniton script with
facial_recognition.py
- Control confidence/success level in
face_recognition
when face is detected using LBPH - Edit/add id names in
usr_id_label
in facial_recognition.py
- Face images are stored inside images
- Trained data is located in
train>data.yml
- OpenCV, install using
pip install opencv-contrib-python
-
LBPH docs compressive guide. Sample code is in C++, you might be interested.
-
Haarcascades pre-trained dataset useful for various subjects.
-
Marcelo for introducing confidences level.
-
Aswinth similar guide using Raspberry Pi.