Topography-Based Detection of the Iris Centre Using Multiple-Resolution Images

Abstract

Low cost iris tracking is one of the most challenging research topics for the eye-tracking community. To this end, accurate detection of the iris centre in images captured by a web cam is a major goal. We propose a novel method for iris detection that is based on image topography using multi-resolution to detect the most stable "valley" over different resolutions, which is assumed to be the iris centre. Our algorithm was tested using the BioID database obtaining the best average behavior. Our algorithm functions in real time and does not require complex post processing stages.