/Face-Recognition

Implementation and testing of computer vision face recognition algorithms

Primary LanguagePythonMIT LicenseMIT

Face-Recognition

  • Implementation and testing of computer vision face recognition algorithms

  • This project consists of :

    1-mainwindow.py : main UI window

    2-main.py : Pyqt GUI script

    3- EigenFaces.py : implementation of face Recognition

    4- FaceDetection.py : implementation of face Detection

Libraries Versions

  • cv2 : version 4.5.5.64
  • matplotlib : version 3.3.4
  • scipy : version 1.6.2
  • numpy : version 1.22.2

To open UI

  • First the UI consists of two tabs :

    • Face detection

    • Face Recognition:

  • The result of face detectiom after applying image from "./image" folder :

  • The result of face recognition after applying image from "./testset" folder :

  • Another result from same folder :

Main Contents

Requiered Part Title
#Part 1 Face Detection
#Part 2 Face Recognition
#Part 3 ROC

Part 1

Face Detection

We used Open-Cv to detect faces in images (grayscale or rgb),by unsing pre-trained cascaded classifier Results:

Part 2

Face Recognition

introduction

The objective of this project is to highlight the importance of linear algebra in the field of computer vision and face recognition. Eigenface is the name of a set of eigenvectors computed from an image dataset. Eigenvectors is a set of features which characterize the global variation among face images.The basis of the eigenfaces method is the Principal Component Analysis (PCA).PCA is a dimensionality-reduction method that is often used to reduce the dimensionality of large data sets, by transforming a large set of variables into a smaller one that still contains most of the information in the large set.

Algorithm

  • first if all we Obtain face image and represent every image in a n^2 x m matrix Image.
  • Compute the Mean face vector(m).
  • Then subtract each image with mean.
  • Compute the eigen vectors(v).
  • Select eigenvectors(k).
  • Now project new image into (k).
  • The new image will be represented using the eigenvectors(x).
  • Face Detection
    • Subtract x with m.
    • If the difference is lower than a chosen threshold, the new image face is detected.
  • Face Recognitio.
    • Each image is represented using the eigenvectors.
    • Each image is then subtracted with x.
    • If the difference is lower than a chosen threshold, the new image face is classified to a class.

Mean image.

Each image when subtracted with the mean image.

Each image represented with eigenvectors.

Part 3

ROC