A simple boundary extraction technique for irregular pupil localization with orthogonal polynomials

A simple boundary extraction technique for irregular pupil boundary localization is proposed in this paper by designing a new Where-to-go approach that extracts the irregular pupil boundary points from the edge map. Initially, the input eye image is subjected to the orthogonal polynomials model and then Hartley’s statistical hypothesis testing is employed to separate out the responses towards noise from those towards strong edges in the eye image. Then the proposed boundary extraction technique is employed to extract the pupil boundary points perfectly and it is independent of the specific eye image characteristics such as pupil deformation, poor contrast, poor brightness, etc. The proposed approach is adaptable to all monochrome databases and exhibits encouraging results when compared with the existing techniques.

► We model irregular pupil boundary localization with orthogonal polynomials (OP). ► We examine local regularities using the point spread operators in the OP model. ► Edge map separates significant edge from the noise using statistical test. ► Pupil boundary points are extracted with the help of weightage assignment patterns. ► It is free from pupil deformation, poor contrast, poor brightness, etc.