The biometrics system is one of the fastest growing technologies around the world because of its unique nature. It uses physiological or behavioral traits of an individual to identify a person. It is really hard to fraud here, and it is completely safe and secure. Many organizations are adopting various biometric techniques. Let us discuss iris recognition which is one of the topmost techniques in biometrics.



Iris is nothing but a colored ring around the pupil of the eye. The color of the iris is determined by the density of stromal pigmentation. The function of the iris is to control the amount of light entering through the pupil.

12 mm is the average diameter of the iris.

Fun Fact- The blockbuster movie Vikram consists of a scene where Agent Tina’s iris scan was used as a biometric to access a secret chamber. Refer to this to know more.


Iris recognition is considered to be the most accurate biometric identification technique that is available today. The concept of iris recognition was first proposed in the year 1939 by Dr. Frank Burch.

Iris scan is adopted because it is a highly protected internal organ (externally visible), it is a reliable method of identification and it cannot be copied since it has unique patterns. It also works on blind persons. Iris scan uses pattern recognition techniques based on high-resolution images of the iris of an individual.

Let us see case study regarding iris scan.


This case study deals with the input images in Iris Recognition Systems. It is essential to develop robust algorithms that work in noncollaborative environments to overcome the constraints of the iris recognition system. Two different types of potential input images are analyzed here.

  • Noisy and Artificial Iris images

To determine the distance between the noisy and artificial samples OSIRIS [Open Source for Iris] system has been used here. This system is developed in the framework of the Biosecure Network of Excellence inspired by Daugman’s works. It consists of two steps: Segmentation and Classification.

  • To detect the contours of the iris, the Segmentation part uses the circular Hough transform and an active contour approach.
  • The Classification part is based on Gabor phase demodulation and Hamming distance classification. In this case, Hamming distance is chosen to carry out the analysis.

A total number of 40 subjects between the age of 16 and 70 is considered here and the total number of images in the database is 1600. The images have been taken in different scenarios under several different lighting conditions.

The aim of the paper is to analyze the influence of different input iris images that is captured in non collaborative environments


Here the input samples are iris images. Any robust/strong iris recognition system should identify the user correctly. It is not common that all iris systems can work properly in noncollaborative environments in which noisy samples are more frequent.

Noise sources are nothing but it modifies the input images in such a way that the conditions for recognition are not optimal. Five different groups of noisy images are analyzed here.

  1. Gaze deviation
  2. Eyelid obstruction
  3. Mydriasis (Excessive dilution of the pupil)
  4. Glasses
  5. Contact lenses

Glasses and conventional contact lenses are not that problematic but most genuine samples are falsely rejected in cases of gaze deviation and eyelid obstruction. Mydriasis is somewhat not very common and it gets affected in its final stage. So, to make the systems more robust, extra algorithms must be considered.

Two different approaches have to be considered for artificial images.

  • A person who uses artificial iris for health/cosmetic reason
  • A person who intent to cheat the recognition system with such artificial iris.

Here six different groups of artificial iris images have been analyzed.

  1. Color Cosmetic lenses
  2. Fantasy Cosmetic lenses
  3. Nonintegrated prosthetic eyes
  4. Integrated prosthetic eyes
  5. Prosthetic lenses
  6. Printed photographs

Cosmetic or prosthetic lenses made identifications impossible since they are opaque. So alternative recognition methods have to be used in such cases. It is concluded that it is difficult to distinguish an artificial and natural iris from an imposter by only taking into account the Hamming distance. Extra Alternative algorithms have to be implemented to avoid this problem.

The research works have proved that there are many possibilities to spoof/imitate automatic iris recognition systems by using certain lenses or printed iris photographs. It is essential to analyze iris forgeries to avoid spoofing(copying/imitating) in this kind of system. Robust algorithms that implement anti spoofing countermeasures are better.

Happy Reading!!

Click to read about IoT (Internet of things)

Click here to learn about the technology that agent Tina used in the movie Vikram