An Algorithm of Eyelashes Detection for Iris Recognition

In order to reduce the eyelash influencing on iris recognition rate, an eyelashes detection algorithm is proposed based on adaptive threshold. Based on the iris localization, gray morphological operation is used to accurately localize eyelids. According to the gray distribution of the parabolic pixel point after the upper eyelid parabolic moving down, the adaptive threshold is used to segment the eyelashes. The experimental results show that compared with the traditional eyelash detection algorithm, it is not necessary for the algorithm to set up more parameters, which embodies the simplicity to maximize the results of eyelash segmentation.