Author – Kalyan Acharjya

What is Biometric Identification?

Now a day everywhere you have to prove your identity whether it may be college, office, club, ATM machine, email account, home etc.  The way of doing so included smart cards ID card, ATM card along with pin number, password, some form identification key and many more. As the forefront of converging the world towards the smart global village and people are compelled for automated identification system, since the traditional are not reliable, unsafe, hence time consuming and burdensome. Suppose you forget the PIN of your ATM card or lost the ID card or may be password stolen etc leads to inevitable complexities for the user; user are unable to access the service. Another one example of annoying problem people’s face is to remember the all passwords of various accounts (say different e-mail accounts, internet banking, official accounts, social networking sites etc).
Iris Recognition
The only way to wash out all these problems concerning identification could be based on biological parameters of human body. As nature creates every human being with some unique features e.g., face, speech, iris, fingerprint, palm, sclera and ear shape etc. All these features have suitable but detectable differences in all human being, for example your face cannot exactly same with your brother (except in twin case) or your DNA match with your parents but your iris pattern doesn’t match with them. Many times human naked eye can’t make distinction between different biometric patterns (as it has limited vision); however it will easily do by computer or machine vision. All biometric identifications are gift with the fact that they cannot be stolen, cannot be replicated easily and cannot be forged as they are integral parts of body and we need not to put extra effort to carry them. A Biometric identification is the automated identification method by computer or machine based on person unique biological characteristics. The coordination of user with biometric machine during identification time is depends on type of the identification machine.

Introduction to Iris Recognition Technology

The iris is a segment of human eye; it has function to change the size of pupil in order to control the light entering in human eye. The interesting part of iris is that it posses unique random pattern which varies from person to person (it differ even in case of twins also). The recognition of person based on his/her iris unique pattern is called iris recognition.Iris Recognition- 2

It has lots of advantages as compare with the other biometric identification methods. For example in case of face recognition twins may have same face pattern thus much similarity between two faces, also face color will slightly vary along with age thus problems of persistence, some sort of facial surgery will completely change the face pattern, fingerprint identification has problems in winter days also it is easy to replicate the artificial finger same as original one. So finger print identification is not a good idea to be used in the critical secure zone, although it has less cost. In case of sclera, palm pattern and ear shape are at their infant stage in the field of digital image processing.

The best alternative is to solve the all sorts of problems is identification based on iris which is complex uniquely varying pattern. It stays exactly same throughout the life and impossible to replicate it.

Procedure for Iris Recognition Technology

The procedure for iris recognition technology is depicted in the flow diagram. At first image is captured by special type of camera (infrared or near infrared illuminated) followed with preprocessing step, which involves filtering from noise of making image ready for segmentation. In this step, segmentation deals with localizing the iris in the eye by extracting the outer boundary of pupil (which is inner boundary of iris) and detecting the outer boundary of iris itself. In segmentation normally we use the canny edge detector and some sort of transform (Circular Hough Transform) to exactly locate it. In the next step to convert the circular iris stripe to rectangular one (technically circular co-ordinate convert to rectangular) and it is ready for encoding. Encoding is the process to store the particular rectangular stripe in some digital encoded format for storing by applying some sort of computer algorithm and mathematical operations.

In the current level of decision making, check the particular input iris image in the previously stored iris database (Iris images encoded database) and if input iris image matches with any encoded iris images within the database then it gives the positive results, other hand it follows no. Based on the input iris image and its pattern matching with iris database image, biometric machine will identify a person and the whole process is completed within a seconds.


The iris recognition technology has tremendous scope and has vast applications areas. Some of them are as follows:

  • Identification in high security zone.
  • Banking Sector
  • E-airport (Automatic to and fro)
  • Automatic Attendance system in offices, universities and colleges.
  • Citizenship-of-countries verification based on unique iris pattern databases.
  • Replacement of other biometric system.
  • Door Lock System (High Security)


Biometric identification has potential to replace existing conventional identification systems because of it is automatic, has higher accuracy and consumes less time. Iris recognition technology is scientific method to identifying any person by analyzing his/hiris-4er unique iris pattern. For example if you sit inside your car, mounted camera will scan your iris without any slightest effort and automatically car will start. If the person stands in front the door, door will automatically open after succession of identification of home members.

The most important advantage of iris biometric system is that it is almost impossible to replicate (it tends for higher security) it and the system cost could be also brought within affordable range. Iris biometrics system has some challenges but still it is dominates the all other method as far as accuracy is concern.