Skip to content
/ cv Public

Computer Vision: Facial Recognition & OCR Project

Notifications You must be signed in to change notification settings

mlubega/cv

Repository files navigation

Computer Vision Project Spring 2019

This project implement facials recognition models using both traditional and deep learning methods. It also has a naive algorithm for Optical Character Recognition (OCR).

1 - bagOfFeatures.py

This is a basic implementation of the Bag of Visual Features (BoVW) algorithm. It extracts SIFT, SURF, and ORB features and trains MLP and SVM classifiers.

2 - TrainFacialRecognition.ipynb

This is the same implemenation of the Bag of Visual Features algorithm adapted to run in a Colab Notebook. Also includes an implementation of a VG166-based CNN for facial recognition using Keras. Colaboratory gives access to NVIDIA GPU which was necessary to speed up training when using the full data set.

3 - orc.py

This is a naive algorithm for Optical Character Recognition (OCR). It uses morphological transformations (thresholding, erosion, dilation) and MSER regions for text detection, and Pytesseract for text recognition.

4 - DetectNum.ipynb

This is the same algorithm from orc.py adapted for a Colaboratory Notebook.

5 - cropFaces.py

This file processes image and video raw data to extract faces, fix rotation, and generate train/test splits.

About

Computer Vision: Facial Recognition & OCR Project

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages