A Pupil Location Method Used for Video-Nystagmograph
1 State Key Laboratory of Power Transmission Equipment & System Security and New Technology, Chongqing University, Chongqing 400044, China
2 Department of Electronic Engineering, Putian University, Putian 351100, China
Abstract：In order to analyze infrared videonystagmograph, a pupil location method was presented. The method was based on morphology and regional connectivity sign algorithm, which included following steps: (1) using left-up neighborhood method for image segmentation, (2) judging the regions and merged the regions and eventually got the pupil area, (3) calculating the coordinates of the pupil of each frame images to track the pupil movement to obtain clinical diagnostic information. Experimental results of eye movement showed that the pupil location algorithm was simple and less time-consuming, only one time of image scanning and labeling was required, and as fast as twice of Hough transform. The method could meet the accuracy and real-time requirements of the pupil location of the infrared video-nystagmograph system. For the same data, the relevancy between the location results of the proposed method and the results by a Hough based eye direction detection algorithm without on-site calibration was 0.975. The pupil tracking by proposed method well reflected the eye movement, and provided useful information for nystgamus analysis.
