Suneeta Agarwal

Texture and Wavelet-Based Spoof Fingerprint Detection for Fingerprint Biometric Systems

This paper describes an image-based system to detect spoof fingerprint attacks in fingerprint biometric systems. It is based on the observation that, real and spoof fingerprints exhibit different textural characteristics. These are based on structural, orientation, roughness, smoothness and regularity differences of diverse regions in a fingerprint image. Local binary pattern (LBP) histograms are used to capture these textural details. Wavelet energy features characterizing ridge frequency and orientation information are also used for improving the efficiency of the proposed method. Dimensionality of the integrated feature set is reduced by running Pudilpsilas Sequential Forward Floating Selection (SFFS) algorithm. We propose to use a hybrid classifier, formed by fusing three classifiers: neural network, support vector machine and k-nearest neighbor using the ldquoProduct Rulerdquo. Classification rates achieved with these classifiers, including a hybrid classifier are in the range ~94% to ~97%. Experimental results indicate that, the new liveness detection approach is a very promising technique, as it needs only one fingerprint and no extra hardware to detect vitality.

Curvelet-based fingerprint anti-spoofing

Existing perspiration-based liveness detection algorithms need two successive images (captured in certain time interval), hence they are slow and not useful for real-time applications. Liveness detection methods using extra hardware increase the cost of the system. To alleviate these problems, we propose new curvelet-based method which needs only one fingerprint to detect liveness. Wavelets are very effective in representing objects with isolated point singularities, but fail to represent line and curve singularities. Curvelet transform allows representing singularities along curves in a more efficient way than the wavelets. Ridges oriented in different directions in a fingerprint image are curved; hence curvelets are very significant to characterize fingerprint texture. Textural measures based on curvelet energy and co-occurrence signatures are used to characterize fingerprint image. Dimensionalities of feature sets are reduced by running Pudil’s sequential forward floating selection (SFFS) algorithm. Curvelet energy and co-occurrence signatures are independently tested on three different classifiers: AdaBoost.M1, support vector machine and alternating decision tree. Finally, all the aforementioned classifiers are fused using the “Majority Voting Rule” to form an ensemble classifier. A fingerprint database consisting of 185 real, 90 Fun-Doh and 150 Gummy fingerprints is created by using varieties of artificial materials for casts and moulds of spoof fingerprints. Performance of the new liveness detection approach is found very promising, as it needs only one fingerprint and no extra hardware to detect vitality.

Wavelet energy signature and GLCM features-based fingerprint anti-spoofing

This paper proposes a texture-based method to spoof-proof a fingerprint biometric system. The fundamental basis of this anti-spoofing method is that, real fingerprint exhibits different textural characteristics from a spoof one. Textural measures based on wavelet energy signatures and gray level co-occurrence matrix (GLCM) features are used to characterize fingerprint texture. Dimensionalities of the feature sets are reduced by running Pudilpsilas sequential forward floating selection (SFFS) algorithm. We test two feature sets independently on various classifiers like: AdaBoost.M1, support vector machine and OneR. Then, we fuse all the mentioned classifiers using the ldquoproduct rulerdquo to form a hybrid classifier. Classification rates achieved for wavelet energy signatures range from ~94.35% to ~96.71%. Likewise, classification rates for GLCM features range from ~94.82% to ~97.65%. Thus, the performance of a proposed method is very promising and it can be efficiently used to spoof-proof a real-time fingerprint biometric system.

Normalized Cross-Correlation Based Fingerprint Matching

To perform fingerprint matching based on the number of corresponding minutia pairings, has been in use for quite sometime. But this technique is not very efficient for recognizing the low quality fingerprints. To overcome this problem, some researchers suggest the correlation technique which provides better result. Use of correlation-based methods is increasing day-by-day in the field of biometrics as it provides better results. In this paper, we propose normalized cross-correlation technique for fingerprint matching to minimize error rate as well as reduce the computational effort than the minutiae matching method. The EER (equal error rate) obtained from result till now with minutiae matching method is 3%, while that obtained for the method proposed in this paper is approx 2% for all types of fingerprints in combined form.

Offline signature verification using grid based feature extraction

Signature verification is one of the most widely used biometrics for authentication. This paper presents a novel approach for offline signature verification. The proposed technique is based on the grid features extraction. For verification, the extracted features of test signature are compared with the already trained features of the reference signature. This technique is suitable for various applications such as bank transactions, passports etc. The threshold used in the proposed technique can be dynamically changed according to the target application. Basically, the threshold here is the security level which the user can input as per his requirement. The proposed technique deals with skilled forgeries and has been tested on two databases: Database A and a standard Database B (Set 1 and Set 2). The proposed technique gives FAR of 9.7% and FRR of 17.9% for Database A, FAR of 12.6% and FRR of 10.2% for Database B (Set 1) and FAR of 13.5% and FRR of 10.8% for Database B (Set 2) which is better than many existing verification techniques.

A hybrid DCT-SVD based robust watermarking scheme for copyright protection

A hybrid watermarking scheme exploiting the properties of the Discrete Cosine Transform (DCT) and Singular Value Decomposition(SVD) has been proposed here. A reference image is being formed from the cover image and then its singular values are modified to hide the secret information in an imperceptible way. The security is further enhanced by the zig zag scrambling of the cover image and gray scale watermarks. The robustness of the methodology against the various image processing attacks has been validated with high Normalized Cross Correlation (NCC) values. Also, the imperceptibility of the watermarked image with the original cover image comes out to be high as indicated by high achievable Peak Signal to Noise Ratio (PSNR) values.

An efficient fragile watermarking scheme with multilevel tamper detection and recovery based on dynamic domain selection

Self-recoverable fragile watermarking is meant for accurate tamper localization as well as image recovery with superior visual quality. However, most of the existing state of art approaches perform authentication and recovery on block basis owing to which the entire block is categorized as tampered in case of alteration of one or more pixels of it. This, results in staircase formation of tamper detected regions, hence lacking in accuracy. Furthermore, the visual quality of the recovered image also deteriorates as an approximate value is assigned to all the block pixels corresponding to the altered region. The proposed watermarking scheme performs both authentication and recovery pixelwise. The authentication of each pixel is done via multi level tamper detection(MLTD) through three authentication bits based on value, location and neighbourhood information. The domain for image recovery is chosen dynamically based on the content of the block, may it be in spatial domain for smooth blocks or frequency domain for the rough ones. This provides high accuracy in recovery. As the embedding of recovery information is done in the frequency domain, the imperceptibility of the watermarked image scheme remains high. Also, embedding of authentication information in the spatial domain maintains its fragile nature. Even for higher tampering ratios, the lost content is rebuilt with high peak signal to noise ratio(PSNR) of the recovered image. The probabilities of false rejection and false acceptance head towards the ideal value for most of the empirical analysis. Comparative study via metric evaluation of probability of false rejection (PFR), probability of false acceptance (PFA) and PSNR of recovered image for different standard test cover images demonstrate the efficacy of the proposed scheme over other existing state of art approaches. Further, the security of the proposed scheme remains high due to usage of multi-layered secret keys and chaos based random mapping handling worst tamper scenarios.

A novel approach of data compression for dynamic data

This paper presents a compression algorithm for dynamic data, the size of which keeps on increasing frequently. It is an efficient data compression technique comprising of a block approach that keeps the data in compressed form as long as possible and enables the data to be appended to the already compressed text. The approach requires only a minimal decompression for supporting update of data. The algorithm reduces the unnecessary time spent in compression-decompression approaches for dynamic documents deigned till now to a minimum. Further, the text document can be modified as required without decompressing and again compressing the whole document.

DCT Based Approach for Tampered Image Detection and Recovery Using Block Wise Fragile Watermarking Scheme

This paper proposes a self- embedding block wise fragile watermarking scheme with tamper detection and content recovery capability. The proposed scheme embeds the shuffled extensive ten bit Recovery data and two bit Authentication data of the image block into the least significant bits (LSB) of its corresponding mapping block.The integrity of a test block is decided by comparing 2×2 non overlapping block of the test block with its corresponding mapping block. Experimental results show that the suggested scheme outperforms conventional self-recovery fragile watermarking algorithms in alteration detection as well as in tamper recovery of the image.

A new parameterized string matching algorithm by combining bit-parallelism and suffix automata

In the parameterized string matching, a given pattern P is said to match with a substring t of the text T, if there exist a bijection from the symbols of P to the symbols of t. This problem has an important application in software maintenance, where we wish to find the equivalency between two sections of codes. Two sections of codes are said to be equivalent, if one can be transformed into the other by renaming identifiers and variables only. Crochemore et al., 1994, has developed an algorithm (BDM) for exact string matching problem using suffix automata. Kimmo Fredriksson et al., 2006, has developed parameterized bit-parallel algorithm (parameterized shift-or) and parameterized BDM (PBDM). Parameterized shift-or (PSO) simulates finite automata in their nondeterministic form. The main drawback of PSO is: it is unable to skip text characters while matching forward. In this paper, we develop a new algorithm for parameterized string matching problem. This algorithm is based upon both suffix automata and bit parallelism concepts. This algorithm is faster than PBDM, since it processes the suffix automata in their non-deterministic form.