Nandita Bhattacharjee

Partial Fingerprint Matching through Region-Based Similarity

Despite advances in fingerprint matching, partial/incomplete/fragmentary fingerprint recognition remains a challenging task. While miniaturization of fingerprint scanners limits the capture of only part of the fingerprint, there is also special interest in processing latent fingerprints which are likely to be partial and of low quality. Partial fingerprints do not include all the structures available in a full fingerprint, hence a suitable matching technique which is independent of specific fingerprint features is required. Common fingerprint recognition methods are based on fingerprint minutiae which do not perform well when applied to low quality images and might not even be suitable for partial fingerprint recognition. To overcome this drawback, in this research, a region-based fingerprint recognition method is proposed in which the fingerprints are compared in a pixel- wise manner by computing their correlation coefficient. Therefore, all the attributes of the fingerprint contribute in the matching decision. Such a technique is promising to accurately recognise a partial fingerprint as well as a full fingerprint compared to the minutiae-based fingerprint recognition methods.The proposed method is based on simple but effective metrics that has been defined to compute local similarities which is then combined into a global score such that it is less affected by distribution skew of the local similarities. Extensive experiments over Fingerprint Verification Competition (FVC) data set proves the superiority of the proposed method compared to other techniques in literature.

Adaptive approach to fingerprint image enhancement

The purpose of fingerprint image enhancement is to improve the clarity and quality of its local features. The quality of fingerprint image determines the accuracy and reliability of minutia extraction which also determines the accuracy of an automatic fingerprint recognition system. In this paper, we propose an adaptive image pre-processing approach that can significantly improve poor quality images according to their noise level based on contrast stretching, power-law transformation and Gabor filter. The original image smoothing is performed initially by Gaussian filter, and then processed by the proposed adaptive algorithm, and finally the resultant image is filtered by Gabor filter to get the improved binarized image. Experimental results indicate that the proposed approach improves the quality of image and reduces the noise significantly as compared to other fingerprint image preprocessing approaches. Furthermore, the result of Goodness Index (GI) shows that our proposed approach improves the performance by 9% as compared to conventional Gabor filter based approach and is also better than other reported results. Especially, the performance of GI is improved by more than 40% in some poor quality images.

Security Analysis and Improvements on WLANs

IEEE 802.11i standard defines the security specifications of IEEE 802.11 series Wireless Local Area Network (WLAN). It is the replacement of the old security standard named Wired Equivalent Privacy (WEP), and it aims to eliminate all known attacks against WEP. It well defines the solutions to the confidentiality, mutual authentication and integrity aspects of the WLAN security except the availability aspect. Many researchers have shown that IEEE 802.11i standard cannot prevent various Denial of Service (DoS) attacks including de-authentication, disassociation and memory/CPU DoS attacks. Besides, IEEE 802.11i reserved the PSK mode of WEP for flexibility and backward compatibility. However, the PSK mode in IEEE 802.11i standard fails to provide sufficient security to prevent offline dictionary attacks and internal attacks . In this paper, we present our solutions which can effectively improve IEEE 802.11i. For memory/CPU DoS attack against 4-way Handshake protocol, we propose an alternative Enhanced 3-way Handshake protocol which can effectively prevent this attack and can save computation cost compared to the original one . For the vulnerability in PSK mode, we proposed a novel ECDH protocol to prevent the offline dictionary attacks and internal attacks. The formal proofs of above two proposed protocols are also provided using Protocol Composition Logic (PCL).

An Efficient and Accurate Iris Segmentation Technique

Accurate segmentation is a crucial phase in the implementation of an iris recognition system. In this paper we investigate a novel technique for iris segmentation. Morphological operations and area computation are applied together with other iris segmentation techniques in order to increase the speed and accuracy of the preprocessing phase. A rough approximation of the pupil’s location is first determined in the initial stage, followed by edge detection and circular Hough transform for accurate iris segmentation. The edge image used to localize the outer iris border is modified increasing the speed and accuracy of the process. Finally, we investigate the effect of eyelids detection using a parabolic curve fitting technique. Two data sets of eye images are used to evaluate the proposed techniques. Experimental results show that the proposed segmentation technique is efficient and performs well on both data sets of images.

Increasing error tolerance in biometric systems

Using biometrics for authentication or cryptosystems would require tolerance of errors of the order of 40%. Most of the proposed techniques either use the reliable bits of a biometric or good quality biometric which does not have such a large variation. However, biometric authentication can be viewed as a design of error correcting codes wherein the redundancy provides the tolerance to errors. This paper explores addition of redundancy to support variations as high as 40% in biometrics. We investigated the error characteristics of biometric templates and found that some blocks within the template tend to produce errors higher than the mean error. To handle this behavior we propose a two level correction. We extensively tested our proposed method using iris images from the commercial Bath dataset [13] and found that the proposed method can tolerate the errors with a success rate of 99.07%, while not accepting any imposters.

Fingerprint reference point detection based on local ridge orientation patterns of fingerprints

This paper introduces an approach to detect reference points for fingerprint images from their local ridge orientation maps. This approach classifies detected singular points into several core and delta point pairs. The validation procedure contains two stages in which the first stage is to remove false pairs of core and delta points and the second stage is to do some post processing of these unpaired singular points. Furthermore, this approach could be extended for reference point detection of arch and partial fingerprint images. The experiments of singular point detection part are conducted on FVC 2002 database 1 and 2. The result shows the average correct rate of singular point detection is 94.05% on the above databases.

On the commonality of iris biometrics

Biometric fuzziness occurs due to variations during data acquisitions and degrades the performance of the iris biometric system. Most of the proposed techniques overcome the fuzziness by removing some parts of the biometric image and/or using image processing methods which are usually computationally inefficient. However, we approach the problem from a post-processing perspective and we start our work with the raw binary iris codes. Our approach is based on the concept of commonality between two codes. The common substrings extract significant structural similarities and the misalignment distance accounts for rotational inconsistencies. This research explores how the length of common substrings can be used as a metric of similarity of iris biometric. We tested our metric using commercial Bath dataset and noticed that it achieves an EER of 1.3%, is robust at handling heterogonous biometric, and is more suitable for low FAR-based applications compared to the widely used Hamming Distance.

Rotating ultrasonic signal vectors with a word-parallel CORDIC processor

The CORDIC algorithm is an iterative method for the efficient computation of vector rotations and several other trigonometric and hyperbolic functions. We have developed a fast, redundant, constant-scale factor, word-parallel implementation of a CORDIC algorithm to rotate ultrasonic signal vectors. The implementation is an improvement of a similar 1996 algorithm know as the differential CORDIC. The CORDIC processor is a part of an ultrasonic imaging system under development and has been implemented using logic synthesis of VHDL descriptions on a Xilinx Virtex 800 FPGA. The algorithm has been simulated with MATLAB. The results of simulation and testing of the CORDIC rotator using a novel VHDL testbench have been presented. The error resulting from truncations is well within the expected limit.

Post-Quantum One-Time Linkable Ring Signature and Application to Ring Confidential Transactions in Blockchain (Lattice RingCT v1.0)

In this paper, we construct a Lattice-based one-time Linkable Ring Signature (L2RS) scheme, which enables the public to verify if two or more signatures were generated by same signatory, whilst still preserving the anonymity of the signatory. The L2RS provides unconditional anonymity and security guarantees under the Ring Short Integer Solution (Ring-SIS) lattice hardness assumption. The proposed L2RS scheme is extended to be applied in a protocol that we called Lattice Ring Confidential transaction (Lattice RingCT) v1.0, which forms the foundation of the privacy-preserving protocol in any post-quantum secure cryptocurrency such as Hcash.

A Region-based Alignment-free Partial Fingerprint Matching

Fingerprint is one of the most researched biometrics and applied to many applications like boarder security, access security, security of electronic devices to name a few. However, alignment and matching of partial fingerprints convincingly is still a challenging problem. This paper proposes a region based partial fingerprint matching approach that matches the fingerprints without aligning them. The approach focuses on extracting the matching regions in unaligned fingerprints. As extraction of regions happens without alignment, a new similarity metric is proposed that can compute similarity value between unaligned fingerprints. The results demonstrate that the proposed approach can eliminate the need for aligning the fingerprints without affecting the accuracy. The popular Fingerprint Verification Competition (FVC) 2002 dataset is used for conducting experiments. Moreover, the results of fingerprint recognition are compared with the best alignment based fingerprint matching techniques.