Instinctive person verification system still faces various challenges in desirable performance due to dependent and independent noise. Most of the physiological biometric modalities are 2-D images, which may have high probability to get affected from noise. This work proposes a comprehensive analysis of robustness of various unimodal and multimodal biometric systems in clean and noisy conditions. On each stage of biometric system we emphasize, feature extraction, level of fusion and suitable normalization schemes. For feature extraction, methods we have employed subspace, kernel and texture based methods and we have subjected the data on all four levels of fusion schemes- sensor, feature, match score and decision level. The objective of this paper is to analyze the robustness of unimodal systems with distinct modalities and evaluate the robustness of a multimodal system with combinations of two, three and four modalities at different levels. All the experiments were evaluated for both clean and noisy data with virtually generated noises of Gaussian and Salt & Pepper methods, and were applied on all biometrics modalities considered for experimentation. The synthetic multimodal database was prepared from standard database of Face, Palmprint, Finger knuckleprint and Handvein. The obtained results and observations in terms of GAR (Genuine Acceptance Rate) show that palmprint with LPQ features are most effective in unimodal systems. In case of multimodal systems, combination of Face (KICA) and Palmprint (LPQ) are most beneficial. This work also suggests some important guidelines on selection of suitable biometric modality, feature extraction algorithms and fusion scheme.

Robustness, Noise, Subspace, Multimodal, Biometric

[1] M. Imran, S. Noushath, A. Abdesselam, K. Jetly and Karthikeyan, "Efficient multi-algorithmic approaches for face recognition using subspace methods," 2013 1st International Conference on Communications, Signal Processing, and their Applications (ICCSPA), Sharjah, 2013, pp. 1-6.
[2] Gomai, A. El-Zaart and H. Mathkour, ‘‘A new approach for pupil detection in iris recognition system,’’ in Proc. 2nd Int. Conf. Comput. Eng. Technol. (ICCET), Apr. 2010, pp. V4-415–V4-419.
[3] X. Wu and Q. Zhao, ‘‘Deformed palmprint matching based on stable regions,’’ IEEE Trans. Image Process., vol. 24, no. 12, pp. 4978–4989, Dec. 2015.
[4] Imran M., Rao A., Noushath S., Hemantha Kumar G. (2014) Some Issues on Choices of Modalities for Multimodal Biometric Systems. In: Babu B. et al. (eds) Proceedings of the Second International Conference on Soft Computing for Problem Solving (SocProS 2012), December 28-30, 2012. Advances in Intelligent Systems and Computing, vol 236. Springer, New Delhi.
[5] H. Al-Ghaib and R. Adhami, "On the digital image additive white Gaussian noise estimation," 2014 International Conference on Industrial Automation, Information and Communications Technology, Bali, 2014, pp. 90-96.
[6] C. Liu, R. Szeliski, S. Bing Kang, C. L. Zitnick and W. T. Freeman, "Automatic Estimation and Removal of Noise from a Single Image," in IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 30, no. 2, pp. 299-314, Feb. 2008.
[7] A. De Stefano, P. White, and W. Collis, “Training methods for image noise level estimation on wavelet components,” EURASIP J. Appl. Signal Process., vol. 2004, pp. 2400–2407, Jan. 2004
[8] T. Ojala, M. Pietikäinen, T. Mäenpää, Multiresolution gray-scale and rotation invariant texture classification with local binary patterns, IEEE Trans. Pattern Anal. Mach. Intell. 24 (7) (2002) 971–987
[9] D. Gabor, “Theory of Communication,” Journal of the Institution of Electrical Engineers, Vol. 93(26), 429-441, 1946
[10] S. Pyatykh, J. Hesser and L. Zheng, "Image Noise Level Estimation by Principal Component Analysis," in IEEE Transactions on Image Processing, vol. 22, no. 2, pp. 687-699, Feb. 2013.
[11] P. Kartik, R. V. S. S. Vara Prasad and S. R. Mahadeva Prasanna, "Noise robust multimodal biometric person authentication system using face, speech and signature features," 2008 Annual IEEE India Conference, Kanpur, 2008, pp. 23-27.
[12] José A. Sáez, Julián Luengo, and Francisco Herrera. Evaluating the classifier behavior with noisy data considering performance and robustness. Neurocomput. 176, C (February 2016), 26-35.
[13] Zhu, X. Wu, X. "Class Noise vs. Attribute Noise: A Quantitative Study", Artificial Intelligence Review (2004) 22: 177.
[14] Nettleton, D.F., Orriols-Puig, A. Fornells, "A study of the effect of different types of noise on the precision of supervised learning techniques" A. Artif Intell Rev (2010) 33:275
[15] M. S. Nair, K. Revathy and R. Tatavarti, "An Improved Decision-Based Algorithm for Impulse Noise Removal," 2008 Congress on Image and Signal Processing, Sanya, Hainan, 2008, pp. 426-431.
[16] S.Kother Mohideen, S. Arumuga Perumal, M.Mohamed Sathik, "Image De-noising using Discrete Wavelet transform ", IJCSNS International Journal of Computer Science and Network Security, VOL.8 No.1, January 2008.
[17] Ashok Rao, S. Noushath, Subspace methods for face recognition, Computer Science Review, Volume 4, Issue 1, February 2010, Pages 1-17, ISSN 1574-0137
[18] Hiew Moi Sim, Hishammuddin Asmuni, Rohayanti Hassan, Razib M. Othman, Multimodal biometrics: Weighted score level fusion based on non-ideal iris and face images, Expert Systems with Applications, Volume 41, Issue 11, 1 September 2014, Pages 5390-5404, ISSN 0957-4174.