Study of Iris Pattern Matching and Detection of the Persons having Squint

摘要

Objectives: The present paper has attempted a quantitative study to measure the amount of off centricity of the iris from its normal position. Methods/Statistical Analysis: The eye image having squint has been taken from real life image (CASIA DATABASE). First of all the central position of the eye opening between the eyelids are detected by Matlab code. Then the distance between the locations of the squint eye are taken by using 3X3 spatial filter with a local searching method. Findings: Squint is the off centre location of iris and pupils and may occur either in one eye or both of the eyes and both male and female persons may have squint. The local searching method terminates when the centre of the filter coincides with the centre of the pupil of the squint eye. The filters are moved from the centre of the eye by coinciding the centre of the spatial filter with the centre of the eye in vertical, horizontal and angular directions. Application/Improvement: By this process noises will be reduced which are prevalent in squint eye. The number of pixels responsible for the exact locations may be compensated by making the successive position of the filters bit overlapping which will in turn reduce the possibility of having checker board appearance in the output of the image.