The system measures the facial hole and generates a facial pattern, which is a unique numerical code for the face. After storing the facial pattern, the system will compare it with thousands or millions of facial pattern data stored in the database. Each facial pattern is stored as an 84 byte file. The system can match facial pattern data in memory at a speed of 60 million faces per minute, and for facial pattern data in the hard drive, it can match 15 million faces per minute. Operators can view two photos declared as matches to determine if the computer's operation is accurate.

In order to determine someone's identity, facial recognition software compares newly captured images with images stored in a database. The core of the FaceIt facial recognition system is the Local Feature Analysis (LFA) algorithm. This is the mathematical technique used by the system when encoding holes. The system measures the facial hole and generates a facial pattern, which is a unique numerical code for the face.