IRIS Attack Detection Using Multiscale BSIF and Person Identification

I.INTRODUCTION

In imaging science, image processing is processing of images using mathematical operations by using any form of signal processing for which the input is an image, such a photograph or video frame, the output of image processing may be either an image or a set of characteristics or parameters related to the image. Most image processing techniques involve testing the image as two dimensional signals and applying standard signal processing technique. Image processing usually refers to digital image processing but optical and analog image processing is also possible. The acquisition of image is referred to as imaging. Digital image processing is the use of computer algorithms to perform image processing on digital images. As a subcategory or field of digital signal processing, digital image processing has many advantages over analog image processing. It allows a much wider range of algorithms to be applied to the input data and can avoid problems such as the build-up of noise and the signal distortion during processing. Iris recognition is an automated method of biometric identification that uses mathematical pattern-recognition techniques on video images of one or both of the irises of an individual's eyes, whose complex random patterns are unique, stable, and can be seen from some distance.

Retinal scanning is a different, ocular-based biometric technology that uses the unique patterns on a person's retina blood vessels and is often confused with iris recognition. Iris recognition uses video camera technology with subtle near infrared illumination to acquire images of the detail-rich, intricate structures of the iris which are visible externally. Digital templates encoded from these patterns by mathematical and statistical algorithms allow the identification of an individual or someone pretending to be that individual. Databases of enrolled templates are searched by matcher engines at speeds measured in the millions of templates per second per (single-core) CPU, and with remarkably low false match rates.

Several hundred millions of persons in several countries around the world have been enrolled in iris recognition systems for convenience purposes such as passport-free automated border-crossings, and some national ID programs. A key advantage of iris recognition, besides its speed of matching and its extreme resistance to false matches is the stability of the iris as an internal and protected, yet externally visible organ of the eye.

An iris-recognition algorithm can identify up to 200 identification points including rings, furrows and freckles within the iris. First the system has to localize the inner and outer boundaries of the iris (pupil and limbus) in an image of an eye. Further subroutines detect and exclude eyelids, eyelashes, and seculars reflections that often occlude parts of the iris. The set of pixels containing only the iris, normalized by a rubber-sheet model to compensate for pupil dilation or constriction, is then analyzed to extract a bit pattern encoding the information needed to compare two iris images.

The remainder of the paper is organized as follows. In section 2 the review of the related papers for the field of this work. Next in section 3, the overview of PAD scheme. In section 4, the performance of detecting the presentation attack will show in Mat lab. In section 5, the conclusion of this paper is discussed.

II.RELATED WORKS

The fake detection method that can be used in multiple biometric systems to detect different types of fraudulent access attempts. So the new approach has been introduced for the biometric system. The objective of the proposed system is to enhance the security of biometric recognition frameworks, by adding liveness assessment in a fast, user-friendly, and non-intrusive manner, through the use of image quality assessment. It presents a very low degree of complexity, which makes it suitable for real-time applications, using 25 general image quality features extracted from one image (i.e., the same acquired for authentication purposes) to distinguish between legitimate and impostor samples[5]. The results, obtained on publicly available data sets of fingerprint, iris, and 2D face, show that the proposed method is highly competitive compared with other state-of-the-art approaches and that the analysis of the general image quality of real biometric samples reveals highly valuable information that may be very efficiently used to discriminate them from fake traits.

The new fake iris detection methods based on wavelet packet transform. First, wavelet packet decomposition is used to extract the feature values which provide unique information for discriminating fake irises from real ones. Second, to enhance the detecting accuracy of fake iris, Support vector machine (SVM) is used to characterize the distribution boundary based on extracted wavelet packet features, for it has good classification performance in high dimensional space and it is originally developed for two-class problems. The results indicate the proposed method is to be a very promising technique for making iris recognition systems more robust against fake iris spoofing attempts [8].

Iris recognition as a reliable method for personal identification has been well-studied with the objective to assign the class label of each iris image to a unique subject[4]. In contrast, iris image classification aims to classify an iris image to an application specific category, e.g., iris livens detection (classification of genuine and fake iris images), race classification (e.g., classification of iris images of Asian and non-Asian subjects), coarse-to-fine iris identification (classification of all iris images in the central database into multiple categories). This paper proposes a general framework for iris image classification based on texture analysis[6]. A novel texture pattern representation method called Hierarchical Visual Codebook (HVC) is proposed to encode the texture primitives of iris images. The proposed HVC method is an integration of two existing Bag-of-Words models, namely Vocabulary Tree, and Locality-constrained Linear Coding. The HVC adopts a coarse-to-fine visual coding strategy and takes advantages of both VT and LLC for accurate and sparse representation of iris texture. The proposed iris image classification method achieves state-of-the-art performance for iris liveness detection, race classification, and coarse-to-fine iris identification. A comprehensive fake iris image database simulating four types of iris spoof attacks is developed as the benchmark for research of iris liveness detection [13].

Liveness detection (often referred to as presentation attack detection) is the ability to detect artificial objects presented to a biometric device with an intention to subvert the recognition system[14]. This paper presents the database of iris printout images with a controlled quality, and its fundamental application, namely development of liveness detection method for iris recognition. The database gathers images of only those printouts that were accepted by an example commercial camera, i.e. the iris template calculated for an artefact was matched to the corresponding iris reference of the living eye. This means that the quality of the employed imitations is not accidental and precisely controlled. The database consists of 729 printout images for 243 different eyes, and 1274 images of the authentic eyes, corresponding to imitations. It may thus serve as a good benchmark for at least two challenges: a) assessment of the liveness detection algorithms, and b) assessment of the eagerness of matching real and fake samples by iris recognition methods. To our best knowledge, the iris printout database of such properties is the first worldwide published as of today [2].

In its second part, the paper presents an example application of this database, i.e. the development of liveness detection method based on iris image frequency analysis. We discuss how to select frequency windows and regions of interest to make the method sensitive to “alien frequencies” resulting from the printing process. The proposed method shows very promising results; since it may be configured to achieve no false alarms when the rate of accepting the iris printouts is approximately 5% (i.e. 95% of presentation attack trials are correctly identified). This favorable compares to the results of commercial equipment used in the database development, as this device accepted all the printouts used. The method employs the same image as used in iris recognition process, hence no investments into- the capture devices is required, and may be applied also to other carriers for printed iris patterns, e.g. contact lens.

III. OVERVIEW OF PROPOSED SCHEME

Overview of the proposed PAD algorithm that explores both periocular (eye region) and iris region to accurately identify the presentation attacks on the iris recognition system. The proposed scheme can be structured in the following two important components namely:

1) Multi-Scale Binarized Statistical Image Feature

Extraction (M-BSIF)

Unsupervised filter learning is widely used to learn a new filter by exploring the statistics from the natural images. These methods have emerged as a feasible alternative to the manually design filters, for instance like Local Binary Patterns (LBP). The most popular techniques of the unsupervised learning includes: Restricted Boltzmann Machines (RBMs) Auto-encoders sparse coding and Independent Component Analysis (ICA). Among these schemes, these of ICA has proved to be a more appealing choice as it overcomes the tuning of large sets of hyper-parameters and can also provide a statistically independent basis that in turn can be utilized as the filter to extract the features from the given image.

The objective of the BSIF is to learn a set of filters from natural images using a ICA based unsupervised scheme. These learned filters can be used to represent each pixel of the given image as a binary string by simply computing its response to the learned filters. The binary code corresponding to the pixel can be considered as a local descriptor of the image intensity pattern in the neighborhood of pixel. Finally, the histogram of the pixels code values allow one to characterize the texture properties within the image sub-regions. Thus, the applicability of the BSIF especially for the visible iris presentation attack detection appears to be an elegant choice as it effectively captures the micro-texture information that can be used to detect the artefact.

In this work, we have employed the open-source filters that are trained using 50000 image patches randomly sampled from 13 different natural scenic images.

The training process to construct BSIF filters involves three main steps,

a) Mean subtraction of each patch

b) Dimensionality reduction using Principle Component Analysis (PCA)

c) Estimation of statistically independent filters (or basis) using Independent Component Analysis (ICA).

2) Decision Level Fusion

In this work employed 8independent linear Support Vector Machine (SVM) classifiers corresponding to both iris and periocular biometrics whose decisions are combined using weighted majority voting. Out of 8 different linear SVM classifiers, the first four are applied on periocular and the remaining four on the iris modality. Among four different linear SVM classifiers that are used on the periocular modality are distributed such that, one each is used on the three independent M-BSIF features and one on the feature level fusion of these M-BSIF features..

Fig. 1. Overview of the Proposed PAD Algorithm

Similarly, out of four linear SVM classifiers that were used on the iris modality, three classifiers are used on three independent M-BSIF features and one on the feature level fusion of M-BSIF features. Each of these linear SVM classifiers is first trained using a set of positive (either with normal (or real) iris or periocular samples) and negative (or artefact or spoof) samples according to the standard protocol described for each of the database used in this work

In this work, weights for the individual classifiers (or experts) are computed according to their performance such that larger weights are assigned to the expert with high accuracy advice-versa as mentioned in. The weight assignment scheme will consider the individual performance of the expert and depending upon performance the method will compute the weights such that wr= 1. These weights are assigned on the development dataset from VSIA database and kept constant throughout experiments.

IV. SIMULATION AND RESULTS

Iris biometric systems are highly vulnerable to presentation attacks that can be carried out using both a photo print and electronic screen display. In this work, we explored the vulnerability of iris recognition systems to various presentation attacks. Further, we also proposed a novel algorithm to accurately detect and mitigate the presentation attacks on the iris recognition system. To this extent, introduce a new relatively large scale visible iris artefact database that comprised of 550 normal and 2750 iris artefact samples. This is one of the important contributions of this work as the VSIA database will be made available to the research community. We then proposed a novel scheme based on M-BSIF and the linear SVM technique that emerged as the best iris presentation attack detection algorithm on both visible and NIR iris recognition system.

Fig.2.Periocular image

First of all the normal iris image is taken as the input image. If the image is color means that are converted into gray scale image. Here the hue and saturation value is eliminated. Then the grayscale image is converted to binary image. Here the nonzero imaginary parts are ignored. The high value is marked as 1(white) and the low value is marked as 0(black). The iris region is presented in the periocular region. The rectangular box is fixed at the iris region for crop the iris region. Then the cropped image and the original image are multiplied then truncation and round off is done. Then the size of the image is finding out and again crops the image.

Fig.3.Normalization

Then the iris image is normalized by the image registration technique. Linear and non-linear methods are used to shape the image. Take the ratio of the inner and outer boundaries of the iris region. Normalization changes the range of the pixel value. It is used to improve the iris recognition. Image registration is the process of transforming different set of data into one co-ordinate system for aligns the image.

Fig.4.Scaling image

The above figure is the scaling output the different value of the iris image is obtained. Scaling is a linear transformation that enlarge object by a scalar factor that is the same in all direction. The result of uniform scaling is similar to the original. More general is scaling with a separate scale factor for each axis direction. It also includes the case in which one or more scale factors are equal to zero. Then the feature selection and feature classification are done. By using the SVM classifier it will detect the image is fake or not.

Fig.5.Normal output

This is the final output. If there is any attack is not presented in the iris means it displayed as normal. The mean value and the dimension values are summed and given the output.

Fig.6.Attack output

This is the attack output. There is any attack is presented that is color, size, intensity means that will display the output as the attack by using the principle component analysis. The database images are already labeled with 1 or -1. The given image is compared to the database. If the most value is indicated1 means it will display the positive output otherwise the attack is presented in the iris region.

V.CONCLUSION

Iris biometric system is highly vulnerable to presentation attacks that can be carried out using either a photo print or electronic screen display. This work, explored the vulnerability of iris recognition system to various presentation attacks. Further, we also proposed a novel algorithm to accurately detect and mitigate the presentation attacks on the iris recognition system. To this extent, introduce a new relatively large-scale visible iris artefact database that comprised of 550 normal and 2750 iris artefact samples. This is one of the important contributions of this work as the VSIA database will be made available to the research community. The proposed a novel scheme based on M-BSIF and the linear SVM technique that emerged as the best iris presentation attack detection algorithm on both visible and NIR iris recognition system.

ACKNOWLEDGEMENT

This work was carried out under the guidance and support of Mrs. Subah Martina. A, M.E., [Assistant professor / ECE], DMI college of Engineering.

REFERENCES

[1] J Unique Identification Authority of India. [Online]. Available: http://uidai.gov.in//, accessed Jun. 2014.

[2] A. Czajka, “Database of iris printouts and its application: Development of liveness detection method for iris recognition,” in Proc. 18thInt. Conf. Methods Models Autom. Robot. (MMAR), Aug. 2013, pp. 28–33.

[3] J. Daugman, “Iris recognition and anti-spoofing countermeasures,” in Proc. 7th Int. Biometrics Conf., 2004.

[4] J. Galbally, S. Marcel, and J. Fierrez, “Image quality assessment for fake biometric detection: Application to iris, fingerprint, and face recognition,” IEEE Trans. Image Process., vol. 23, no. 2, pp. 710–724, Feb. 2014.

[5] J. Galbally, J. Ortiz-Lopez, J. Fierrez, and J. Ortega-Garcia, “Iris liveness detection based on quality related features,” in Proc. 5th IAPR Int. Conf.Biometrics (ICB), Mar./Apr. 2012, pp. 271–276.

[6] X. He, S. An, and P. Shi, “Statistical texture analysis-based approach for fake iris detection using support vector machines,” in Advances in Biometrics (Lecture Notes in Computer Science), vol. 4642, S.-W. Lee and S. Z. Li, Eds. Berlin, Germany: Springer-Verlag, 2007, pp. 540–546.

[7] K. Hughes and K. W. Bowyer, “Detection of contact-lens-based iris biometric spoofs using stereo imaging,” in Proc. 46th Hawaii Int. Conf.Syst. Sci., Jan. 2013, pp. 1763–1772.

[8] X. He, Y. Lu, and P. Shi, “A new fake iris detection method,” in Advances in Biometrics, vol. 5558, M. Tistarelli and M. S. Nixon, Eds. Berlin, Germany: Springer-Verlag, 2009, pp. 1132–1139.

[9] X. He, Y. Lu, and P. Shi, “A fake iris detection method based on FFT and quality assessment,” in Proc. Chin. Conf. Pattern Recognition, Oct. 2008, pp. 1–4.

[10] Z. He, Z. Sun, T. Tan, and Z. Wei, “Efficient iris spoof detection via boosted local binary patterns,” in Advances in Biometrics. Berlin, Germany: Springer-Verlag, 2009, pp. 1080–1090.

[11] M. Kanematsu, H. Takano, and K. Nakamura, “Highly reliable liveness detection method for iris recognition,” in Proc. Annu. Conf. SICE, Sep. 2007, pp. 361–364.

[12] E. Lee, K. Park, and J. Kim, “Fake iris detection by using purkinje image,” in Advances in Biometrics (Lecture Notes in Computer Science), vol. 3832, D. Zhang and A. K. Jain, Eds. Berlin, Germany: Springer-Verlag, 2005, pp. 397–403.

[13] S. J. Lee, K. R. Park, and J. Kim, “Robust fake iris detection based on variation of the reflectance ratio between the iris and the sclera,” in Proc. Biometric Consortium Conf., Sep./Aug. 2006, pp. 1–6.

[14] A. F. Sequeira, J. Murari, and J. S. Cardoso, “Iris liveness detection methods in mobile applications,” in Proc. 9th Int. Conf. Computer. Vis.Theory Appl., 2013, pp. 1–5.