Fast and accurate iris localization based on improved Chan-Vese active contour model

Abstract

Iris segmentation is the main process of the iris recognition system. In this process, the iris part of the eye image is segmented from the other parts such as the sclera, pupil, eyebrows, eyelashes and eyelids. This process is important because the accuracy of the iris recognition system can be negatively affected by poor and false segmentation. The speed of iris segmentation must be fast enough to deal with millions of irises. The speed is also an important factor in real-time implementation of iris recognition systems. This research is important since poor localization and segmentation because of reflections, eyelids and eyelashes interferences are the main factors that can reduce the accuracy of the system. Other than that, the modified methods of Integro-differential operator and Hough transform still consume a great deal of time and complexity. Many localization and segmentation methods still assume an iris as a circle which is the source of the reduction in accuracy. In this paper, the Chan-Vese active contour model is improved to achieve fast and accurate iris localization in the iris recognition system. Iris localization is a part in iris segmentation which to localize and detect the iris boundaries. The proposed method is also compared with other state-of-the-art methods in terms of execution time, memory usage and accuracy. According to the results, the proposed method achieves faster execution time, low memory usage and high accuracy compared to other methods.