Facial-Recognition-Haar-LBPH

PROJECT DESCRIPTION

Basic facial recognition implemented using Haar Cascades and Local Binary Pattern Histograms

DEPENDANCIES

• Python 3
• OpenCV
• Numpy
• os (built-in)
• Copy (built-in)

FILE DESCRIPTION

face.py - Main facial recognition code

test.py - Tester file to train and test on images

init.py - real time face recognition

CREATING DATASETS

Create 2 directories - Train and Test

• In Test folder, have around 10 images of the images you want to recognize (eg. images of you) called positive set and 10 images to check for misclassification (eg. images of NOT you) called negative set
• In Train folder, create 2 directories - 0 and 1
  • In directory 0, keep all negative set images (NOT you) and in directory 1 keep all positive set images (YOU)

Make sure you have around a 100 or so images for good results

RUN INSTRUCTIONS

• Make sure all dependancies are met

• Comment/Uncomment as reqd

• Make sure you have created the dataset maintaining directory structure as reqd

• Run test.py to train your model on the images you used to create your train and test dataset.

• Run init.py after you save the model for real time face recognition