Iris Localization Technology And Its Applications Under Various Light Conditions

Iris localization is a key step in the process of judging the human eye movement state. In recent years, with the continual development of computer image processing technology and pattern recognition technology, the software and hardware platform of iris localization technology gradually matured. The application fields of iris localization technology expanses rapidly, which has been widely used in biometric identification, eye tracking technology, neural system disease detection and treatment and other fields, which made it gradually recognized and accepted by the public.The purpose of the iris localization is to fix the inner and outer boundary of the iris and extract the iris part accurately from the eye image. In this paper, the physiological structure of the iris and the common iris databases at home and abroad is introduced first, and then an iris localization algorithm is proposed which can be used in a variety of light conditions. Finally, the algorithm is used in iris recognition and eye tracking field which made good localization results. The main contents of the paper are as follows:Firstly, based on collecting and collating the iris databases which are commonly used at home and abroad, video images of the eye are collected under two different light conditions. The paper analyzes the various factors that affect the iris localization results. And on this basis, the principles and characteristics of the common iris localization algorithms are compared in this paper.Because of the variety methods and standards in collecting of the iris image library images, which bring difficulties to the comparison between different iris location algorithms, the performance of algorithms between different iris databases also have significant difference. Only through establishing the iris image databases that contain more contents and covers a wider range, can detect the performance of the algorithms fairly.Secondly, propose an improved fast iris localization algorithm. On the basis of remove the eyelid, eyelash interference and influence of the corneal spots, a discrete gray gradient differential algorithm is used to threshold the eye image and position the pupil center coordinates roughly. Then an improved starburst algorithm is used to detect the feature points of the iris. Finally use the least squares method to calculate the profile parameters of the iris, which find the exact position of the iris.Finally, by introducing the eye tracking technology, the proposed algorithm is applied in the field of eye tracking. Through image preprocessing, corneal spot processing, feature points of iris outer edge detection, an improved PERANSAC algorithm is used to detect the iris. Saccade and blink of eye in tracking process are also analyzed in this paper.