Feature extraction for different distances of visible reflection iris using multiscale sparse representation of local Radon transform

作者：

Highlights：

• Visible iris image is prune to investigate iris textures at different distances.

• Multiscale and sparse representation of local radon transform is proposed.

• Multiscale is used to reduce noise during down sampling of normalized iris.

• Sparse representation is used to generate compact information of iris features.

• The proposed combination is able to increase accuracy of iris recognition.

摘要

•Visible iris image is prune to investigate iris textures at different distances.•Multiscale and sparse representation of local radon transform is proposed.•Multiscale is used to reduce noise during down sampling of normalized iris.•Sparse representation is used to generate compact information of iris features.•The proposed combination is able to increase accuracy of iris recognition.

论文关键词：Local radon transform,Iris recognition,Feature extraction,Non-cooperative environment,Visible iris

论文评审过程：Received 15 December 2012, Revised 5 March 2013, Accepted 11 March 2013, Available online 22 March 2013.

论文官网地址：https://doi.org/10.1016/j.patcog.2013.03.009