Pupil Boundary Detection Based on Genetic Algorithm

Subject Areas : Information Technology in Engineering Design (ITED) Journal

Ali mousavi ¹ , hamid reza pourreza ² , mohammad etezad razavi ³

1 - Young Researchers and Elite Club, Mashhad Branch, Islamic Azad University, Mashhad, Iran
2 - Department of computer Engineering ferdowsi university mashhad, Mashhad, Iran
3 - Eye Research Center, Khatam Medical Eye Hospital, Medical Science University of Mashhad, Mashhad, Iran

Received: 2017-06-20 Accepted : 2017-06-20 Published : 2014-08-23

Keywords:

Abstract :

Droopy eyelid surgery which is one of the most widespread cosmetic surgeries removes extra skinaround the eyelid and improves it. Therefore, determination of eye parameters has significant effectson success of the surgery. To obtain the exact eye parameters, pupil boundary should be segmented. Inthis paper a new method based on genetic algorithm has been proposed to segment pupil in eye imageswhich is a dark area with lowest entropy and have the maximum pixel intensity compared with itsneighborhood iris. Performance evaluation measures are computed and compared with other methods.The average accuracy and sensitivity calculated for proposed method are 99.80 and 95.17 respectively.