Iris automatic segmentation method and system based on multi-model voting mechanism

Abstract

The invention discloses an iris automatic segmentation method and system based on a multi-model voting mechanism. The method comprises the following steps: obtaining a data set and performing preprocessing operation to obtain a preprocessed data set; inputting the obtained preprocessed data set into a trained first target detection convolutional neural network to obtain a rectangular frame of an iris inner circle and an iris outer circle in each human eye iris image; obtaining the boundary contour of the outer circle and the boundary mask of the inner circle of the iris through fitting of the obtained rectangular frame, cutting the whole iris off according to the rectangular frame, then amplifying and sending the cut-off iris image to another two trained segmented deep convolutional neural networks, and obtaining the masks of the inner circle of the iris respectively; and finally voting through the masks of the two segmentation models and the inner circle boundary mask obtained by the target detection model to obtain a final iris inner circle result. Fitting and positioning of the inner and outer boundaries of the iris region are facilitated, and accurate input parameters are provided for subsequent iris normalization.

Description

technical field [0001] The invention belongs to the technical field of image segmentation, and more specifically relates to an automatic iris segmentation method and system based on a multi-model voting mechanism. Background technique [0002] Iris recognition is widely used in various fields because of its precise recognition characteristics. However, in the actual application process, due to the influence of the surrounding environment of iris image acquisition, such as illumination and distance, the effect of iris image acquisition is different, and the effect of some images is poor. In subsequent processing, it is difficult to segment the boundary of the iris area , which has a greater impact on iris recognition. [0003] Currently used methods for segmenting iris regions in iris images include traditional image processing methods and deep learning-based segmentation methods. However, both methods have some defects in use. The segmentation method based on deep learning...

AI Technical Summary

Problems solved by technology

However, both methods have some defects in use. The segmentation method based on deep learning is prone to errors; the traditional method is susceptible to interference, which affects the effect of iris image segmentation.

Images