New facial recognition system developed
13 November, 2009
category: Biometrics
An engineer from the University of Miami named Mohamed Abdel-Mottaleb has developed a new biometric system that utilizes human face and ear photographs to authenticate identity within 95 to 100% accuracy, according to a New Kerala article. The technology can authenticate via either of two ways wherein it can utilize three-dimensional facial images or two-dimensional facial images with a three-dimensional image of the ear constructed from various still frames of video footage.
The 3-D facial recognition aspects of the system operate by weeding out the less discriminating regions of the face and focusing on the nose, eye brows, mouth and chin. The second mode that utilizes the same facial functions before also utilizes ear structure, however, the ear-mapping is tougher to obtain due to specific lighting requirements. However, when possible, researchers believe the accuracy rate of both together to be near 100%.
Read the full story here.