Iris Liveness Research Articles

Entropy with Local Binary Patterns for Efficient Iris Liveness Detection

Iris anti-spoofing is one of the most important topics, in which the development is increasing rapidly. This paper introduces an efficient system for detecting iris attacks. The system avoids the segmentation and the normalization stages employed traditionally in fake detection systems. Wavelet packets (WPs) are used to decompose the original image into wavelet approximation and detail channels. Entropy values are extracted from the wavelet channels, and also from the local binary pattern (LBP) images of the channels. These features are used for discriminating between real and fake iris images. Support vector machines are used for the classification purpose. The aim is to contribute for improved classification accuracy with less computational complexity and reduced processing time. Entropy of the WP channels gives 99.9237% classification accuracy, and the entropy of the LBP images yields 99.781%, using ATVS-FIr-DB. Fusion of these features yields 100% classification accuracy. Entropy of the wavelet channels is sufficient to obtain 100% accuracy using CASIA-Iris-Syn database, without fusion. All images in both databases are used, without the need to discard images with unsuccessful segmentation. Segmented images from both databases are used for comparison. Results show that more discriminative features can be obtained using the proposed algorithm. System complexity and processing time are reduced noticeably, and the system is robust to different types of fakes.

Iris liveness detection for next generation smartphones

This paper presents a novel liveness detection method that exploits the acquisition workflow for iris biometrics on smartphones using a hybrid visible (RGB)/near infra-red (NIR) sensor. These devices are able to capture both RGB and NIR images of the eye and iris region in synchronization. This multi-spectral information is mapped into a discrete feature space. An intermediate classifier which uses a distance metric close to Jenson-Shannon divergence is employed to classify the incoming image. Further, a fast, multi-frame pupil localization technique using one-dimensional processing of the eye region is proposed and evaluated. This is used to analyze the pupil characteristics of the images classified as 'live' in the previous stage. It is shown that such an analysis could detect presentation attacks, even with a 3-D face model made of materials that has properties similar to human skin and the ocular region1.

Iris liveness detection using regional features

In this paper, we exploit regional features for iris liveness detection. Regional features are designed based on the relationship of the features in neighbouring regions. They essentially capture the feature distribution among neighbouring regions. We construct the regional features via two models: spatial pyramid and relational measure which seek the feature distributions in regions with varying size and shape respectively. The spatial pyramid model extracts features from coarse to fine grid regions, and, it models a local to global feature distribution. The local distribution captures the local feature variations while the global distribution includes the information that is more robust to translational transform. The relational measure is based on a feature-level convolution operation defined in this paper. By varying the shape of the convolution kernel, we are able to obtain the feature distribution in regions with different shapes. To combine the feature distribution information in regions with varying size and shape, we fuse the results based on the two models at the score level. Experimental results on benchmark datasets demonstrate that the proposed method achieves an improved performance compared to state-of-the-art features.

Iris liveness detection for mobile devices based on local descriptors

Iris recognition is well suited to authentication on mobile devices, due to its intrinsic security and non-intrusiveness. However, authentication systems can be easily tricked by attacks based on high-quality printing. A liveness detection module is therefore necessary. Here, we propose a fast and accurate technique to detect printed-iris attacks based on the local binary pattern (LBP) descriptor. In order to improve the discrimination ability of LBP and better explore the image statistics, LBP is performed on a high-pass version of the image with 3 × 3 integer kernel. In addition a simplified interpolation-free descriptor is considered and finally a linear SVM classification scheme is used. The detection performance, measured on standard databases, is extremely promising, despite the resulting very low complexity, which makes possible the implementation for the relatively small CPU processing power of a mobile device.

Iris Image Classification Based on Hierarchical Visual Codebook

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. In contrast, iris image classification aims to classify an iris image to an application specific category, e.g., iris liveness 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. 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 (VT), and Locality-constrained Linear Coding (LLC). 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. Extensive experimental results demonstrate that 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.

Interview

Dr Luisa Verdoliva from Università di Napoli Federico II, Italy, talks about the research behind the Letter ‘Wavelet-Markov local descriptor for detecting fake fingerprints’ on page 439. Dr Luisa Verdoliva My research field is image processing and computer vision. Although images convey a huge amount of information, human beings are able to process them in real time and effectively extract useful information. I have always been fascinated by this ability which is what led me to work in this area, in particular on image compression, segmentation and denoising. A few years ago, Carlo Sansone, who is an expert in pattern recognition, presented me with the problem of fingerprint liveness detection. We started to collaborate, along with the other co-authors, and began to obtain encouraging results. Fingerprint-based biometric systems have become increasingly popular and are commonly used for authentication in various security applications (recently also to access smartphones, e.g. iPhone 5s). Relying on features that are unique to each individual, they avoid the problems typical of systems based on the use of passwords. Unfortunately, it is possible to spoof a variety of fingerprint-based systems by reproducing fake fingers on simple moulds made of silicone, play-dough, clay or gelatin. It is important to improve the security of these systems using software-based approaches which are economic, non-invasive and flexible. Telling apart live from fake fingerprints is a binary classification problem. The main ingredient for success is to extract some discriminative (and compact) features from the image, on which a classifier can base its decision. For example, if live fingerprints were crisp and fake fingerprints were blurred, a reliable measure of blurriness would make a very good feature. However, current fakes are very realistic and we must look for much finer differences. The features we propose are based on the joint statistical analysis of wavelet-domain coefficients. They try to capture the behaviour observed locally in very small areas of the image by means of histograms collected over all of these areas. Extensive tests on the databases used in this field show that the local descriptor considered in our work guarantees a very good detection performance, in most cases better than competing techniques typically based on global features. Even if they try to exploit the specific characteristics of the fingerprint pattern (the ridge-valley structure), they are based on macroscopic features that have been shown to be less discriminative than microscopic ones. This work provides some new insights into which features can best capture the differences between live and fake fingerprints. Some existing authentication devices are already equipped with a liveness detection system. Given the simplicity and low complexity of our proposed approach, it can be readily implemented in real-world systems, improving their performance. Using such technologies in scenarios like frontiers and airports is challenging. The number of subjects requiring identification would be so large that even a very low error rate, like that reported in this work, would lead to a huge number of controls. Local descriptors can be used for many important tasks. Besides fingerprint liveness detection, we are using them for iris liveness detection and image forgery detection. We have also recently worked on sensor noise-based forgery localisation. Moreover, I am continuing to work on more traditional image processing problems, such as the compression, denoising and segmentation of remote sensing images, both optical and SAR. Like many problems involving security, this is a two-party game with attacker and defender evolving rapidly; innovation is key to keep the lead. Today's state-of-the-art techniques could not be forecast just ten years ago and we have little idea of what will be the major ideas ten years from now. Fake fingerprints may become much more difficult to detect. The use of fingerprints for authentication, though well established and accepted, will probably lose ground to new and more secure approaches, maybe based on a combination of biometrics.

Liveness detection for iris recognition using multispectral images

Liveness detection is a necessary step towards higher reliability of iris recognition. In this research, we propose a novel iris liveness detection method based on multi-features extracted from multispectral images. First, we analyze the specific multispectral characteristics of conjunctival vessels and iris textures. To ensure the effective utilization of these characteristics, iris images are simultaneously captured at near-infrared (860nm) and blue (480nm) wavelengths. Then we respectively define and measure relative number of conjunctival vessels (RNCV) and entropy ratio of iris textures (ERIT) using 860-nm and 480-nm images. Finally, the feature values of RNCV and ERIT are arranged to form a robust 2-D feature vector. The trained Support Vector Machine (SVM) is used to classify the feature vectors extracted from live and fake irises. Experimental results demonstrate that the proposed method can discriminate between live irises and various types of fake irises with high classification accuracy and low computational cost.

Automated personal identification system based on human iris analysis

In general, a typical iris recognition system includes iris imaging, iris liveness detection, iris image quality assessment, and iris recognition. This paper presents an algorithm focusing on the last two steps. The novelty of this algorithm includes improving the speed and accuracy of the iris segmentation process, assessing the iris image quality such that only the clear images are accepted so as to reduce the recognition error, and producing a feature vector with discriminating texture features and a proper dimensionality so as to improve the recognition accuracy and computational efficiency. The Hough transform, polynomial fitting technique, and some morphological operations are used for the segmentation process. The phase data from 1D Log-Gabor filter is extracted and encoded efficiently to produce a proper feature vector. Experimental tests were performed using CASIA iris database (756 samples). These tests prove that the proposed algorithm has an encouraging performance.

Personal identification based on iris texture analysis

With an increasing emphasis on security, automated personal identification based on biometrics has been receiving extensive attention over the past decade. Iris recognition, as an emerging biometric recognition approach, is becoming a very active topic in both research and practical applications. In general, a typical iris recognition system includes iris imaging, iris liveness detection, and recognition. This paper focuses on the last issue and describes a new scheme for iris recognition from an image sequence. We first assess the quality of each image in the input sequence and select a clear iris image from such a sequence for subsequent recognition. A bank of spatial filters, whose kernels are suitable for iris recognition, is then used to capture local characteristics of the iris so as to produce discriminating texture features. Experimental results show that the proposed method has an encouraging performance. In particular, a comparative study of existing methods for iris recognition is conducted on an iris image database including 2,255 sequences from 213 subjects. Conclusions based on such a comparison using a nonparametric statistical method (the bootstrap) provide useful information for further research.