Face Detection and Face Recognition
Face recognition and detection is like a very old topic for the CV world and I am totally a new person for this. So just follow the routine to get some result first. And next I will learn and use the current most popular way---- the deep learning algorithm to optimize the result later.
So I first apply the Haar classifier to get the face area from images and videos, then train the faces with Fisher algorithm.
I use a funny video for testing the result. Get the faces first with the Haar classifier.
So I first apply the Haar classifier to get the face area from images and videos, then train the faces with Fisher algorithm.
I use a funny video for testing the result. Get the faces first with the Haar classifier.
After running all the video long, I get quite a lot faces.
The target is the folder-3. A funny looking guy.
So I picked several images for each person for training and re-run the video for recognition.
The accuracy for detection and recognition are 95% and 85% for now. I think this is a good start. Next I will learn the deep learning, quite exciting!!!!