Face Recognition OpenCV / EmguCV

Question

I am trying to create an application that detects faces and recognizes them. I did face detection, but I want some ideas when recognition. I use a webcam for tracking and it can detect a face. Then I take only part of the face to a new gray image and compare it using EigenObjectRecognizer with the list of images in the database.

But this does not give good results. Several times he found that something was wrong, several times nothing. I want to ask what to compare photos, what additional methods should I implement? How is histogram alignment or resolution of alignment of faces?

Answer 1

I have an OpenCV Face Recognition (Haar Face Detection + Histogram Equalization + Eigenfaces) tutorial and free source code that you can try: http://www.shervinemami.info/faceRecognition.html

Answer 2

The definition of a person and the classification of persons are completely different problems. I can tell you from my experience and read several face classification works that a good way to start is to read about Basic Component Analysis (PCA, also known as Eigenfaces), Fisher linear discriminant analysis (LDA), and vector machine support (SVM). These are classification methods that are extremely useful for classifying faces, and it turns out that OpenCV already includes excellent implementations on PCA and SVM . An excellent resource for face recognition and classification code for OpenCV in C ++ is this site .

One website offering resources and document links for the most relevant face classification methods is this one .

A well-explained example of PCA Eigenfaces and LDA with sample code in Matlab, which was extremely useful for my first face classification program here .

Answer 3

Own functions will only attract you more to face recognition. There is a limit to the number of people you can train. You will need to learn new algorithms, check new documents in this area and implement them.

You can realize the effectiveness of Eigenfaces by making sure that all of your images have the same orientation and size. Orientation and attitude are very important. So make sure you apply the correct transforms before comparing

Answer 4

Check my answer at the link below. It has some links that will be useful to you, mainly the YouTube video that I mentioned.

face value recognition in bone