Iris recognition is one of the most accurate biometric technologies. The key to iris recognition is the failure of a test of statistical independence. The accuracy is largely due to many degrees of freedom exhaustively match with called as “Iris Code” which 2048 bits (256 bytes) to record each individual unique Iris pattern. To avoid obscured by eyelids, eyelash, specula reflections, and boundary artifacts of hard contact lenses, a major improvement is to computer an equal number of masking bits. Each iris record has composed of an Iris code and a Mask code.
Matching speed is the most concerned especially for a large population. The Iris system must reliably match the new template with one enrolled into the database. Iris recognition technologies were originally developed for user with central processing unit (CPU)-based system which is known as sequentially processing device. On a single 300 MHz CPU, such exhaustive searches are performed at a rate of about 100,000 irises per second. On a single 3.0 GHz server, one million Iris codes can be compared in about 1 second2. Iris recognition performance is dependent of the CPU power
One of fastest growing application such technology is to deploy handheld or mobile device to a remote battle field or road patrolling. The requirements for such device include weight and size and more important is the battery usage time and thermal management. Since those devices could be deployed in any environment, additional to requirements of ragged, waterproof, a cooling fan is not an option. The thermal management is critical for device successful. The high temperature is harmful to CPU performance and also harmful to the battery life. Even with 1.0 GHz CPU, thermal issue can determine the success or failure for the device.
Our patent technology is to introduce a rapid pattern matching engine without faster CPU. It can be used as either an external search engine or an embedded hardware.
