http://www.bytefish.de

blog:fisherfaces_in_opencv

Philipp Wagner — Fri, 03 Feb 2012 02:20:16 +0100

fisherfaces in opencv

I've found some minutes to implement the Fisherfaces method with OpenCV2's C++ API. The code has been pushed into my github repository at:

https://github.com/bytefish/opencv/tree/master/lda

As some kind of documentation, I am pasting the README.markdown coming with the project.

If you still have any troubles using it, just comment below or drop me a mail.

github.com/bytefish/opencv/lda

This project implements the Fisherfaces method as described in: P. Belhumeur, J. Hespanha, and D. Kriegman, “Eigenfaces vs. Fisherfaces: Recognition Using Class Specific Linear Projection”, IEEE Transactions on Pattern Analysis and Machine Intelligence, 19(7):711–720, 1997.

Building the Project

This project now comes with the Eigen3 headers, so compiling the project is as easy as writing (assuming you are in this folder):

philipp@mango:~/some/dir/lda$ mkdir build
philipp@mango:~/some/dir/lda$ cd build
philipp@mango:~/some/dir/lda/build$ cmake ..
philipp@mango:~/some/dir/lda/build$ make
philipp@mango:~/some/dir/lda/build$ ./lda filename.ext

And if you are in Windows using MinGW it may look like this:

C:\some\dir\lda> mkdir build
C:\some\dir\lda> cd build
C:\some\dir\lda\build> cmake -G " MinGW Makefiles " ..
C:\some\dir\lda\build> mingw32-make
C:\some\dir\lda\build> lda.exe filename.ext

You probably have to set the OpenCV_DIR variable if it wasn't added by your installation, see Line 5 in the CMakeLists.txt how to do this. If you have problems working with CMake or installing OpenCV, you probably want to read my guide on Face Recognition with OpenCV2.

Using the Project

The project comes with an example, please have a look at the main.cpp on how to use the classes. You need some data to make the examples work, sorry but I really can't include those face databases in my repository. I have thoroughly commented the code and reworked it lately, to make its usage simpler. So if anything regarding the classes is unclear, please read the comments.

In the example I use a CSV file to read in the data, it's the easiest solution I can think of right now. However, if you know a simpler solution please ping me about it. Basically all the CSV file needs to contain are lines composed of a filename followed by a ; followed by the label (as integer number), making up a line like this: /path/to/image.ext;0.

Think of the _label_ as the subject (the person) this image belongs to, so same subjects (persons) should have the same label. An example CSV file for the AT&T Facedatabase is given here, which looks like this (assuming I've extracted the database to /home/philipp/facerec/data/at, file is without … of course):

/home/philipp/facerec/data/at/s1/1.pgm ;0
/home/philipp/facerec/data/at/s1/2.pgm;0
...
/home/philipp/facerec/data/at/s2/1.pgm;1
/home/philipp/facerec/data/at/s2/2.pgm;1
...
/home/philipp/facerec/data/at/s40/1.pgm;39
/home/philipp/facerec/data/at/s40/2.pgm;39

Once you have a CSV file with valid filenames and labels, you can run the demo by simply starting the demo with the path to the CSV file as parameter:

./lda /path/to/your/csvfile.ext

Or if you are in Windows:

lda.exe /path/to/your/csvfile.ext

License

All code is put under a BSD license, so feel free to use it for your projects.

opencv, cpp

blog:pca_in_opencv

Philipp Wagner — Fri, 03 Feb 2012 01:57:44 +0100

pca in opencv

There are hardly any useful examples on how to do a Principal Component Analysis (PCA) with the OpenCV2 C++ API. Maybe because it's simple for everyone else, but it took me some time to figure it out. So here is an example!

Most people will search for the Eigenfaces algorithm, so the example also shows how to display the average image and the first three principal components. Principal components are the eigenvectors sorted descending by their eigenvalue. The Eigenvectors have the same length as your images and resemble (ghostly) faces, hence the name Eigenfaces in the Eigenfaces algorithm (surprise!).

I provide an Eigenfaces implementation for OpenCV2 C++ at:

https://github.com/bytefish/opencv/tree/master/eigenfaces

The following is the README.markdown of the project, which should answer most questions. I am also providing you a simple example how to use the PCA, if you are not interested in the Eigenfaces.

German readers are also welcomed to have a look at a document Filip and I wrote on a similar topic: pdf/eigenfaces.pdf.

github.com/bytefish/opencv/eigenfaces

This project implements the Eigenfaces method as described in: Turk and Pentland, “Eigenfaces for recognition.”, Journal of Cognitive Neuroscience 3 (1991), 71–86.

Building the Project

This project comes as a CMake project, so compiling the project is as easy as writing (assuming you are in this folder):

philipp@mango:~/some/dir/eigenfaces$ mkdir build
philipp@mango:~/some/dir/eigenfaces$ cd build
philipp@mango:~/some/dir/eigenfaces/build$ cmake ..
philipp@mango:~/some/dir/eigenfaces/build$ make
philipp@mango:~/some/dir/eigenfaces/build$ ./eigenfaces filename.ext

And if you are in Windows using MinGW it may look like this:

C:\some\dir\eigenfaces> mkdir build
C:\some\dir\eigenfaces> cd build
C:\some\dir\eigenfaces\build> cmake -G " MinGW Makefiles " ..
C:\some\dir\eigenfaces\build> mingw32-make
C:\some\dir\eigenfaces\build> eigenfaces.exe filename.ext

Using the Project

/home/philipp/facerec/data/at/s1/1.pgm ;0
/home/philipp/facerec/data/at/s1/2.pgm;0
...
/home/philipp/facerec/data/at/s2/1.pgm;1
/home/philipp/facerec/data/at/s2/2.pgm;1
...
/home/philipp/facerec/data/at/s40/1.pgm;39
/home/philipp/facerec/data/at/s40/2.pgm;39

Once you have a CSV file with valid filenames and labels, you can run the demo by simply starting the demo with the path to the CSV file as parameter:

./eigenfaces /path/to/your/csvfile.ext

Or if you are in Windows:

eigenfaces.exe /path/to/your/csvfile.ext

License

All code is put under a BSD license, so feel free to use it for your projects.

Simple Example

#include <iostream>
#include "cv.h"
#include "highgui.h"
 
using namespace std;
using namespace cv;
 
Mat normalize(const Mat& src) {
    Mat srcnorm;
    normalize(src, srcnorm, 0, 255, NORM_MINMAX, CV_8UC1);
    return srcnorm;
}
 
int main(int argc, char *argv[]) {
    vector<Mat> db;
 
    // load greyscale images (these are from http://www.cl.cam.ac.uk/research/dtg/attarchive/facedatabase.html)
    db.push_back(imread("s1/1.pgm",0));
    db.push_back(imread("s1/2.pgm",0));
    db.push_back(imread("s1/3.pgm",0));
 
    db.push_back(imread("s2/1.pgm",0));
    db.push_back(imread("s2/2.pgm",0));
    db.push_back(imread("s2/3.pgm",0));
 
    db.push_back(imread("s3/1.pgm",0));
    db.push_back(imread("s3/2.pgm",0));
    db.push_back(imread("s3/3.pgm",0));
 
    db.push_back(imread("s4/1.pgm",0));
    db.push_back(imread("s4/2.pgm",0));
    db.push_back(imread("s4/3.pgm",0));
 
    int total = db[0].rows * db[0].cols;
 
    // build matrix (column)
    Mat mat(total, db.size(), CV_32FC1);
    for(int i = 0; i < db.size(); i++) {
        Mat X = mat.col(i);
        db[i].reshape(1, total).col(0).convertTo(X, CV_32FC1, 1/255.);
    }
 
    // change to the number of principal components you want
    int numPrincipalComponents = 12;
 
    // do pca
    PCA pca(mat, Mat(), CV_PCA_DATA_AS_COL, numPrincipalComponents);
 
    namedWindow("avg", 1);
    namedWindow("pc1", 1);
    namedWindow("pc2", 1);
    namedWindow("pc3", 1);
 
    // mean face
    imshow("avg", pca.mean.reshape(1, db[0].rows));
 
    // first three eigenfaces
    imshow("pc1", normalize(pca.eigenvectors.row(0)).reshape(1, db[0].rows));
    imshow("pc2", normalize(pca.eigenvectors.row(1)).reshape(1, db[0].rows));
    imshow("pc3", normalize(pca.eigenvectors.row(2)).reshape(1, db[0].rows));
 
    waitKey(0);
}

opencv

about

Philipp Wagner — Wed, 01 Feb 2012 20:34:51 +0100

… stalking me.

Personal Links

Contact Details

Email: bytefish AT gmx DOT de
Jabber: bytefish AT gmx DOT de
Address: Check the WHOIS record.
Public Key

blog:face_recognition_with_opencv2

Philipp Wagner — Wed, 01 Feb 2012 19:27:38 +0100

face recognition with opencv2

I've recently uploaded some code to perform face recognition with OpenCV2 to my github repository. Now if you research on face recognition you'll soon notice there's a gigantic number of publications, but source code is very sparse. Since I've got some positive feedback on Machine Learning with OpenCV2, I thought I write a document on:

Face Recognition with OpenCV2 (pdf)

It's the kind of guide I've wished for, when I was working myself into face recognition. It helps you with installing OpenCV2 on your machine and explains you how to build a project on Windows and Linux. Then the Eigenfaces and Fisherfaces method are explained and implemented with GNU Octave/MATLAB and OpenCV2's C++ API. A basic knowledge of C++ is assumed and administrative rights for your machine are a plus. I've decided to use MinGW (the GCC port for Windows) as C/C++ compiler for Windows, because it works great with OpenCV2 and comes under terms of a public license (please see mingw.org/license for details). If someone writes a the install guide for Visual Studio 2008/2010, I would be happy to add it to the document. I guess you only have to generate an Visual Studio project with CMake, open and compile it; but I don't have Visual Studio right now to check it.

The code and document is released under a BSD license, so feel free to use it for your commercial and academic projects. Note: the latest version of the document and code can be obtained from the projects github repository:

http://www.github.com/bytefish/facerecognition_guide

opencv, octave, cpp

documents

Philipp Wagner — Tue, 31 Jan 2012 23:11:39 +0100

… will be translated one day.

Here are some documents we wrote. Some of them are in german, but they'll get translated as soon as possible. Non-german readers are highly encouraged to look at all the beautiful pictures inside! If you have a question, just mail one of us.

Title	Author(s)	Document	Notes
Face Recognition with OpenCV2	Philipp Wagner	PDF (English)	The kind of guide I've wished for, when I was working myself into face recognition. Helps you with installing OpenCV2 and building projects on Windows and Linux. The Eigenfaces and Fisherfaces method are explained and implemented with GNU Octave/MATLAB and OpenCV2's C++ API. Sourcecode and latest revision available at github.com/bytefish/facerecognition_guide
Statistical Machine Learning with OpenCV	Philipp Wagner (but thanks to Filip for his amazing inkscape skills!)	PDF (English)	Is an introduction into statistical machine learning with OpenCV. It contains a guide for setting up OpenCV on Windows and Linux. A brief introduction and code for Support Vector Machines and Multilayer Perceptrons is given. Good to get started with OpenCV and its Machine Learning library!
Einfuehrung in MapReduce	Filip Martinovský, Philipp Wagner, Stefan Heinze	PDF (German)	Is an introduction to MapReduce and gives some ideas on how to implement it in Erlang.
Distributed Hash Tables	Filip Martinovský, Philipp Wagner	PDF (German)	Reads like a tutorial on how to implement distributed hash tables. Chord, a protocol for distributed hash tables, is described and implemented in Erlang.
Gesichtserkennung mit Eigenfaces	Filip Martinovský, Philipp Wagner	PDF (German)	Explains face recognition with Eigenfaces. Shows the performance of a prototype on two datasets (Yale Facedatabase A, AT&T Dataset).
XML Processing in Scheme	Filip Martinovský, Philipp Wagner	PDF (English)	Explains Xml processing in Scheme: SXML. Rather old.
XML Verarbeitung in Scheme	Filip Martinovský, Philipp Wagner	PDF (German)	Explains Xml processing in Scheme: SXML. Rather old.

erlang, distributed hashtable, mapreduce, eigenfaces, xml, scheme