Face Recognition using 2D Fast Fourier Transform
This article explains how to implement a simple face recognition system based on analysis throught their Fourier spectrum. Recognition is done by finding the closest match between feature vectors containing the Fourier coefficients at selected frequencies. The introduced method well compares to other competing approaches.
Despite the argument is certainly complex, the approach used and the tools already implemented make the process easy to implement by all users. The call is therefore not to be frightened by appearances.
The face is one of several features witch can be used to uniquely identify a person. It's the charateristic that we most commonly use to recognize others. Not two human faces are identical which makes them well suited for use in identification.
Besides being a challenging problem in itself the importance of face recognition systems lies in their potential applications such as access control, passport, etc...
The obviuous advantage of a face recognition system compared to competing methods is its low level of intrusion. It only requires looking into camera.
Automated face recognition systems generally evolved along two main routes, either the analysis of grey level information ( often called template based ) or the extraction of mainly geometrical features such as shape, profile or hair colour.
The work presented here comprises a novel template based approach that considering it's simple algorithm compares very wel to other more complex methods that are used commonly such Hidden Markov Models or back propagation Neural Network.
According to humans are thought to view faces primarly in a holistic manner and experiments suggest that holistic approaches are superior to geometrical recognition systems.
The tecnique presented is based on the Fourier spectrum of facial images, thus it relies on a global transformation , every pixel in the image contributes to each value in the spectrum.
The Fourier spectrum is a plot of the energy against spacial frequencies , where spatial frequencies rerlate to the spatial relations of intensities in the image . In our case this translate to distances between areas of particular brightness such as the overall area of the head or the distance of the eyes.
Higer frequencies describe finer details and contrary to what you might think we found them less useful for identification, just as humans can recognise a face from a brief look without focusing on small details.
The recognition of faces is done by finding the closet match ( the difference or distance ) between the newly presented face and all those faces known to the system. The distances are calculated between the feature vectors with entries that are the Fourier transform values at specially chosen frequencies. As few as 30 frequencies yield excellent results.