Biometrics
While biometrics industry activity has shown much promise and growth, the reliability of technology has been the limiting factor for wide-scale adoption. The technology acceptance curve has been slowed by issues of limited accuracy and a high level of failure to enroll subjects. The importance of these performance metrics is critical when considering the cost of a security system failure in high level security locations.
The primary sources of biometric information for security systems include fingerprint, hand, voice, iris/retna, and facial data. Face recognition is currently, and projected to be, the fastest growing mode of biometric technology. For greater adoption, accuracy levels must be improved by overcoming lighting, angle, and color variables.
The largest developers of facial recognition technology all use two-dimensional (2D) data for image capture and matching. In a well-controlled environment facial recognition technology can perform to a moderate level of accuracy. However, in less controlled environments where lighting, angle, impediments such as hats, facial hair, glasses, facial expressions, or even skin coloring, can greatly throw off the accuracy levels, the error rate dramatically rises.
The emergence of three dimensional (3D) systems offer the promise of overcoming the accuracy limitations of current 2D systems. While current 3D systems are able to create an extremely good representation of the face, the subsequent verification stage requires reconstruction algorithms that must perform numerous extremely complex calculations, making real-time matching difficult.
Pixel Velocity, has collaborated with biometric technology industry leaders in an effort to develop and bring to market the most accurate biometric verification system in the world. This team has leveraged a optical 3D facial recognition technology to engineer an extremely accurate, non-intrusive, multimodal biometric verification system. Although still in a development stage, the technology shows promise in expanding the current state of the art, providing a passive access control module for real-time identify verification.