In today’s blog, we continue with the different types of Facial Recognition technologies:

The Various Facial Recognition Technologies

There are various technologies that are employed in facial recognition systems to capture the unique features of the face.  These technologies can be categorized as Eigenface; Neural Network; Automatic Face Processing; and Feature Analysis.

With Eigenface technology, the enrollment and verification templates are constructed via a database consisting of many 2-D, grayscale images of faces.  So for example, if you were to be enrolled or verified by a facial recognition system, the image of your face would be reconstructed using the various 2D, grayscale images (the Eigenfaces).  This reconstructed image would then serve as the appropriate template of your face.

With neural network technology, the system tries to “learn” which unique facial features will work best for verification and identification.  Various algorithms have been developed and are utilized in order to accomplish this task.  In order to help the system “learn” which facial features will work best, different weight factors are assigned to the unique features found between the matched and unmatched templates.

Automatic Face Processing is an older technology.  With this, the distances as well as the corresponding distance ratios are calculated between the unique features of the face.

Feature Analysis is the most commonly used technology.  With this technology, the unique features are captured from the different parts of the face, as well as the relative position of these features.  Also, this technology can take into consideration to a certain extent any changes in the appearance of the face.

In tomorrow’s blog, we look at some of the privacy issues of Facial Recognition.