lanhoter/java_face-detection-and-recognition

2015 Final Year Project in INTI International University Malaysia, built from scratch

Java_Face-Detection-and-Recognition

IMPORTANT: you will need openCV library in order to get it run and work.

Final Year Project in INTI International University Malaysia

Email Notification

Algorithm:

Face Detection: Haar Cascades
Feature Extraction: Principle Component Analysis
Face Recognition: Eigenfaces

Construct Eigenfaces

Steps
1.Grayscale Face Images
2.A “STOP” Exposure
3.Calculate Eigenfaces(Corresponding to largest Eigenvalue)
Dataset training

Face Recognition Steps:

1.Automatically Crop Face Images
2.Grayscale and Resize Face images to 220 * 220 pixel
3.Feature Extraction and Calculate distance
4.Distance below 0.2 can be identified as same person

Image Face Recognition

1.Steps of Face Recognition from image
2.Load Face Image
3.Grayscale, a “STOP” Exposure and Resize
4.Face Detection using Haar Cascades
5.Feature Extraction using PCA
6.Calculate Eigenvalues and Similarity
7.Retrieve results from Face Database

System Strength

Fast Face Detection and Recognition within seconds
Secured based Email Notification
Detection and Recognition under difficult Light Conditions from 50 lux to 8000 lux
Low Cost Implementation and no additional Cost