The iris recognition technology developed by Mobbeel lets you identify the users of any system with a mobile device using only its camera. There is no need for other hardware or additional sensors, just the built-in camera.
Iris recognition is one of the most reliable biometric methods available. It is virtually impossible to find two identical irises, thereby offering an extremely high level of security.
As previously stated, iris recognition is the process of recognising a person by analysing the random pattern of the iris. Typically, this kind of recognition requires the use of infrared light to illuminate the iris and facilitate the acquisition of the iris pattern without any irregularity.
In general, this is how the recognition works. The first step is to capture the picture. Then the eye is isolated and the inner and outer edges of the iris are located using complex image processing algorithms. In addition, noise removal processes are triggered to allow for eyelids, eyelashes and reflection and remove them from the analysis. Once the iris is located and isolated it is encoded using mathematical algorithms resulting in a code that contains the unique characteristics of the iris. Although any two pictures at different times and in different conditions will not be exactly equal, the process is able to validate whether the iris belongs to that person, or not.
As shown in the diagram, iris recognition is classified among biometric technologies that make use of the physiological characteristics of an individual.
Mobbeel has concentrated its efforts in the development of this technology, specifically targeting current mobile devices and the wide range of opportunities their features offer.
With our technology, iris pattern acquisition does not need a special sensor or infrared light in the mobile device, it simply uses the built-in camera. Most current smart-phones have cameras with enough resolution and sufficient technical specifications to allow for this.
The technology has been specially designed to be able to perform the whole recognition process, from the initial capture of the eye, through the acquisition of the biometric iris pattern and final identification of the individual. The technology is capable of processing all the information, taking into account that the samples might be very different. The sophisticated algorithms allow the samples to be taken in a non-controlled environment by the mobile device's built in camera, where the environmental conditions (no special light, different places and distances to the eye, etc) may have had a major impact on the outcome.
The technology also works within client/server architecture, where the client (mobile device) is used just as a capture device and the server receives the sample taken by the client, through a secure channel, and performs the recognition process.
One of the most important things about this kind of technology, from the user's point of view, is privacy. In this system, biometric templates are built up and stored with user's biometric information, so they can be compared in the recognition process. It is essential that those templates are kept 100% safe, either in the mobile or in the server.
With any biometric data system one the the primary concerns from a user point of view will be privacy. Mobbeel never stores the biometric templates obtained from users so there is no risk of a user's credentials being compromised (the mobile phone is lost or stoled, a security failure on the server side). This is possible thanks to our biometric encryption techniques.
Biometric encryption can be defined, basically, as a process that uses a PIN, a password or a cryptographic key along with biometric data in a way that neither (biometric data or password/key) can be retrieved from a sample stored. This means that with some algorithms the original key can be regenerated from a new biometric template without having to store the initial biometric template of the user. The two biometric templates do not have to be exactly the same. Using these techniques, biometrics and their intrinsic variability can be effectively combined with cryptography and its absolute need for accuracy.
Mobbeel iris recognition technology can be easily integrated with others systems or third-party applications using its API (Application Programming Interface)
If you want to know more about our technologies, APIs or user case studies you can contact us here or by email to info (at) mobbeel (dot) com
