Face Detector Using PCA

Objective

Face Recognition using Principal Component Analysis - PCA using 5 photos for training and 1 image for testing for 9 individual persons.

Process

1. Prepare a face training dataset
2. Compute the average face vector
3. Subtract the average face vector from original images
4. Calculate the covariance matrix
5. Calculate the eigenvalues
6. Select the top K of all
7. Create features weight for training
8. Read the testing face image
9. Calculate the feature vector of the testing face
10. Compute the Euclidean distance between the test feature vector and all the training feature vector
11. Find the face class with minimum distance

Results and Analysis

The image when averaged will look like this:

All the images were verified correctly. The system successfully recognized the human faces and worked better in different conditions of face orientation.