The Application Of The Iris Recognition Technology In Eye Movement Rotation

With the rapid development of science and technology, the research on eyes is more and more popular. So, the iris recognition technology has a very good development prospect, and it applies mostly in identification. And the research on eye-movement is reflected in the manufacture on Eye Tracker. Based on estimating the eye movement parameters of pilots, the thesis uses the Iris Recognition Technology to describe the parameters of eye movements. The main algorithms include iris image pre-processing, pupil center positioning, iris boundary localization and iris rotation angle estimation. The task of the thesis is to apply these algorithms in a video image, to get the eye-movement component parameter results, to draw some corresponding waveform graphs for describing the eye-movements.The details of the thesis are as follows:1) During the pre-processing of iris image, some classical pre-processing algorithms are studied, and their merits and demerits are analyzed. In this thesis, the morphological filter method is used to reduce the noises in the image and to make the boundary of the pupil more smooth. The experiment proves that this method raises the accuracy of pupil center locating.2) During the location of pupil center, many pupil center positioning methods are analysed, and their merits and demerits are compared. In this thesis, the pupil center locating algorithm based on dot-Hough transform is proposed. The major advantage of proposed algorithm is that some point-group but not all points of the edge are chosen. This algorithm can also be applied in the images, in which the pupil are covered by eyelid or eyelash.3) During the external boundary location of iris, the combination of the Sobel edge detection operator and Hough transform is improved. This method is easy and efficient. This step is the basement process of estimating the iris rotation.4) After the iris boundary the localization, the target area becomes the iris area that is closer to a circle. Some features'coordinates of iris can be described using the iris feature extracting algorithm. With some mathematic principles, the procedure program is written to estimate the angle of irises.In order to prove the feasibility of the algorithm, some video images are tested on the platform of VC6.0 & OpenCV1.0 in the computer of Core2 & WinXP. The coordinates of pupil center is obtained by the pupil center locating algorithm. According to two pupil centers in two adjacent frames, the horizontal and vertical components of eye-movement are described by the waveforms. However, the rotation component of eye-movement is described by the change angle of iris features. Finally, the experimental results show that the iris recognition method applied in eye-rotation has good performance and is convincing.