生物特徵辨識是指基於一個人的生理或者行為特徵作為身分辨識機處的一種技術，虹膜辨識是生物特徵辨識中一種精確度、普遍性、獨特性很高，且侵入性很低的辨識方式。在一個典型的虹膜辨識系統當中包含了以下幾個階段：1. 影像擷取、2. 虹膜切割、3. 虹膜遮罩產生、4. 特徵提取、5. 特徵比對，為了提高虹膜辨識的準確率，許多的研究裡都關注在如何正確切割虹膜、提取特徵以及特徵比對，然而虹膜遮罩的正確與否也是虹膜辨識準確性的重要因素之一。
在本篇論文中，我們嘗試了多種的神經網路架構來對虹膜遮罩進行預估，最後提出了兩種基於深度學習(Deep Learning) 的演算法來學習輸入虹膜影像的遮罩，我會將虹膜影像和其對應正確的虹膜遮罩做些許前處理後，輸入進我們建置好的深度學習網路學習其特徵，學習完特徵後的網路在輸入新的虹膜影像時也能順利的預測其對應虹膜影像遮罩，使產生虹膜遮罩的正確率相對於rule-based 或其他演算法產生的虹膜遮罩高，且能提升虹膜辨識最終的準確性，使用patch-based CNN 的虹膜遮罩正確率可以達到92.87%、EER 為0.147%，使用multi-channel FCN 的虹膜遮罩正確率可以達到95.56%、EER 為0.0851%。

Iris recognition has a lot of applications. A typical iris recognition system has several stages, including acquisition, segmentation, iris mask generation, feature extraction and matching. In order to increase the accuracy of iris recognition, many studies focus on iris segmentation, feature extraction and matching. However, iris masks can also have a great impact on the accuracy of recognition.
In this study, we propose two iris mask estimation algorithm based on deep learning. After pre-processing the iris images and the corresponding masks, we train these data in convolution neural networks (CNN), which help to achieve a higher accuracy in matching iris masks for different images than rule-based algorithms. The accuracy of matching by using patch-based CNN is 92.87%, with the 0.147% EER (Equal Error Rate) and the accuracy of applying multi-channel fully convolution networks is 95.56%, with an even lower EER equal to 0.0851%.

[1] Y. Lecun, L. Bottom, Y. Bengio, and P. Haffner, “Gradient-based learning
applied to document recognition,” in Proc. of the IEEE, Nov. 1998, vol. 86,
pp. 2278–2324.
[2] vdumoulin, “Convolution arithmetic,” https://github.com/vdumoulin/
conv_arithmetic, 2016, [Online] Available.
[3] J. Daugman, “How iris recognition works,” IEEE Transactions on Circuits
and Systems for Video Technology, vol. 14, no. 1, pp. 21–30, Jan. 2004.
[4] J. Daugman, “High confidence visual recognition of persons by a test of statistical
independence,” IEEE Transactions on Pattern Analysis and Machine
Intelligence, vol. 15, no. 11, pp. 1148–1161, Nov. 1993.
[5] J. Daugman, “New method in iris recgonition,” IEEE Transactions on Systems,
Man, and Cybernetics, Part B, vol. 37, no. 5, pp. 1167–1175, Oct. 2007.
[6] R. P. Wildes, “Iris recognition: an emerging biometric technology,” in Proc.
of the IEEE, Sep. 1997, vol. 85, pp. 1348–1363.
[7] L. Ma, T. Tan, Y. Wang, and D. Zhang, “Personal identification based on
iris texture analysis,” IEEE Transactions on Pattern analysis and Machine
Intelligence, vol. 25, no. 12, pp. 1519–1533, Dec. 2003.
[8] C. Tisse, L. Martin, L. Torres, and M. Robert, “Person identification technique
using human iris recognition,” in Proc. Vision Interface, 2002, pp.
294–299.
[9] D. Zhang, D. M. Monro, and S. Rakchit, “eyelash removal method for human
iris recognition,” in IEEE International Confernece on Image Processing,
Oct. 2006, pp. 285–288.
[10] Y. Li and M. savvides, “An automatic iris occlusion estimation method
based on high-dimensional density estimation,” IEEE Transactions on Pattern
analysis and Machine Intelligence, vol. 35, no. 4, pp. 784–796, APR.
2013.
[11] M. Haindl and M. Krupicka, “Accurate detection of non-iris occlusions,”
in 2014 Tenth International Conference on Signal-Image Technology and
Internet-Based Systems, Nov 2014, pp. 49–56.
[12] A. Gangwar, A. Joshi, A. Singh, F. Alonso-Fernandez, and J. Bigun, “Irisseg:
A fast and robust iris segmentation framework for non-ideal iris images,” in
2016 International Conference on Biometrics (ICB), June 2016, pp. 1–8.
[13] Maroti Deshmukh and Munaga V. N. K. Prasad, Partial Segmentation and
Matching Technique for Iris Recognition, pp. 77–86, Springer India, 2015.
[14] N. Liu, H. Li, M. Zhang, Jing Liu, Z. Sun, and T. Tan, “Accurate iris segmentation
in non-cooperative environments using fully convolutional networks,”
in 2016 International Conference on Biometrics (ICB), June 2016, pp. 1–8.
[15] K. Bae, S. Noh, and J. Kim, “Iris feature extraction using independent component
analysis,” in Proc. of the 2003 4th International Conference on Audio
and Video-Based Biometric Person Authentication, 2003, pp. 838–844.
[16] E. Krichen, S. Garcia-Salicetti, and B. Dorizzi, “A new probabilistic iris
quality measure for comprehensive noise detection,” in IEEE International
Confernece on Biometrics: Theory, Applications, and Systems, Sept. 2007,
pp. 1–6.
[17] J. Thornton, Matching deformed and occluded iris patterns: a probabilistic
model based on discriminative cues, PhD thesis, Carnegie Mellon University,
2007.
[18] S. Attarchi, K. Faez, and A. Asghari, “A fast and accurate iris recognition
method using the complex inversion map and 2dpca,” in Proc. of the IEEE
International Conference on 2008 Computer and Information Science, Apr.
2008, pp. 179–184.
[19] G. E. Hinton and R. Salakhutdinov, “Reducing the dimensionality of data
with neural networks,” Science, vol. 313, no. 5786, pp. 504–507, Jul. 2006.
[20] J. Long, E. Shelhamer, and T. Darrell, “Fully convolutional networks for
semantic segmentation,” in Conputer Vision and Pattern Recognition, 2015.
[21] L. Xu, C. Lu, Y. Xu, and J. Jia, “Image smoothing via l0 gradient minimization,”
ACM transactions on Graphics, vol. 30, no. 5, pp. 174:1–174:12, Dec.
2011.