Boca Raton, Florida – A new algorithm could improve the speed an accuracy of facial recognition, and even see through a disguise.
Face recognition software has never really taken off, because although systems are accurate they require a lot of computer power. Approaches have been based on neural networks, dynamic link architectures (DLA), fisher linear discriminant model (FLD), hidden Markov models and Gabor wavelets.
Now, scientists at Florida Atlantic University claim to have drastically reduced the amount of processing power required for accurate facial recognition. By applying a one-dimensional filter to the two-dimensional data from conventional analyses, such as the Gabor method, they say, they have done this without compromising accuracy.
The team tested the performance of their new algorithm on a standard database of 400 images of 40 subjects. Images are grey scale and just 92 x 112 pixels in size. They found that their technique was faster and more effective with low resolution images, such as those produced by standard CCTV cameras.
The team claims it also solves the variation problems caused by different light levels and shadows, viewing direction, pose, and facial expressions. It can even see through certain types of disguises such as facial hair and glasses.
The details are published in the International Journal of Intelligent Systems Technologies and Applications.