Alignment-Free Cancelable Fingerprint Templates Based on Local Minutiae Information

Invasion of user's privacy due to compromised biometric template draws great public concern recently. Cancelable biometric is proposed as a promising resolution by converting the original biometric template into a distorted version via a transformation function. However, certain schemes such as Densely Infinite-to-one mapping (DITOM) are merely applicable to fixed-length and ordered biometric features. This limits their application capacity on unordered and variable size features such as fingerprint minutia or descriptors. In this paper, we proposed an extended form of infinite-to-one transform that is capable to operate the two-dimensional feature vectors to generate a cancelable fingerprint template. The proposed method was evaluated using four public databases, i.e. FVC2002 DB1, DB2, DB3 and DB4. Experimental results demonstrate satisfactory accuracy performance. Besides, the changeability and non-invertibility are also justified.

The current cryptographic algorithms are suffering from the drawback of maintaining the secrecy of private keys. These private keys must be taken care of so that the cost of privacy can be avoided. Stored private keys are generally secured by the user tailored passwords, which can be hacked by brute-force attacks. Also, it is quite a horrid issue for the user to play with those large keys. So, the major issue is to protect and manage such a long private key. The generation of cryptographic key from individual user's biometric feature is a viable option. In this way it is too difficult for the attacker to guess the cryptographic key without having access of the user's biometrics. But the problem is that once it gets compromised it goes to waste. To solve the aforesaid issues cancelable biometrics approach has been proposed which is a wonderful solution for cancelling and reissuing the biometrics template. In this paper, a way to generate cryptographic key from the cancelable fingerprint template of the user has been suggested. Shuffling based transformation method is applied on the minutia points extracted from the user's fingerprint to generate cancelable template. Then a pair of private and public key is generated for RSA cryptosystems from cancelable template. This cancelable template maintains the secrecy of fingerprint template and produce revocable key for asymmetric cryptography.

Despite their usability advantages over traditional authentication systems, biometrics-based authentication systems suffer from inherent privacy violation and non-revocability issues. In order to address these issues, the concept of cancelable biometrics was introduced as a means of generating multiple, revocable, and noninvertible identities from true biometric templates. Apart from BioHashing, which is a two-factor cancelable biometrics technique based on mixing a set of tokenized user-specific random numbers with biometric features, cancelable biometrics techniques usually cannot preserve the recognition accuracy achieved using the unprotected biometric systems. However, as the employed token can be lost, shared, or stolen, BioHashing suffers from the same issues associated with token-based authentication systems. In this paper, a reliable tokenless cancelable biometrics scheme, referred to as BioEncoding, for protecting IrisCodes is presented. Unlike BioHashing, BioEncoding can be used as a one-factor authentication scheme that relies only on sole IrisCodes. A unique noninvertible compact bit-string, referred to as BioCode, is randomly derived from a true IrisCode. Rather than the true IrisCode, the derived BioCode can be used efficiently to verify the user identity without degrading the recognition accuracy obtained using original IrisCodes. Additionally, BioEncoding satisfies all the requirements of the cancelable biometrics construct. The performance of BioEncoding is compared with the performance of BioHashing in the stolen-token scenario and the experimental results show the superiority of the proposed method over BioHashing-based techniques.

Cancelable biometric techniques play a vital role in the privacy and security of remote surveillance systems to keep the genuine users’ confidential data safe and away from intruders. This research presents an efficient cancelable biometric framework by exploiting an irreversible hybrid encryption technique that incorporates Deoxyribonucleic, Ribonucleic acid sequences (DNA and RNA), and the evolutionary optimization technique of the Genetic Algorithm (GA). So, these techniques are employed to create completely deformed templates from their original ones. Thus, our main contribution is offering a novel biometric security framework that achieves unique randomness characteristics using RNA and DNA sequences and the evolutionary GA-based encryption scheme. The proposed framework produces an entirely unrecognized biometric template by ciphering the main discriminative features of the biometric traits of the authorized clients. The proposed algorithm is first initialized by creating several encrypted biometric images of the original users by employing the logistic map function. Secondly, the initialized encrypted images are transformed into one vector of a binary array. Then, they are converted to their corresponding introns, exons, and consequently, their relevant codons are stored in the cloud database. Thirdly, these relevant codons are replaced by new ones after generating encrypted RNA lists. The utilized encryption key is extracted from the original biometric image after utilizing excessive permutations between its pixels. Then, the GA-based optimization technique is applied to select the most convenient biometric features. Finally, after employing the GA-based cross-over and mutation operations, the chosen features are used to generate the cancelable biometric traits. To validate the security efficacy of the proposed framework, different six biometric databases are used, which are Olivetti Research Laboratory (ORL) Faces (gray), CASIA v.5 Faces (color), UPOL Iris (gray), Indian Institute of Technology Delhi (IIT Delhi) Ear (color and gray), Fingerprint, and CASIA Palmprint (color and gray). The security performance of the proposed framework is compared to those of recent studies in this field, such as OSH (Optical Scanning Holography) and DRPE (Double Random Phase Encoding). The simulation results prove the superior performance of the proposed framework in terms of all tested evaluation metrics. The proposed framework provides high AROC (Area under the Receiver Operating Characteristic) that reaches 0.9990, low FAR (False Acceptance Rate) of 0.0015, more uniform histograms, high correlation values for genuine users, and complete hidden of biometric features. In addition, from a security perspective, the proposed framework accomplishes good entropy, UACI (unified average changing intensity), and NPCR (number of pixels change rate) values that reach 7.9960, 33.55%, and 99.65%, respectively.

In this paper, we propose a salting based two-factor cancelable biometrics construct, dubbed Random Permutation Maxout (RPM) transform for facial template protection. The RPM transform is inspired from a member of rank-based Locality Sensitive Hashing (LSH), namely Winner Takes All hashing, which was devised for data retrieval. With externally generated user-specific parameters, RPM converts a continuous facial feature vector into a max ranked indices vector as cancellable template. Since the features magnitude of facial features have been transformed to the discrete index form, the resulting template is robust against noises and it is strongly concealed from the adversary learning on the original facial features. This lays a strong promise on non-invertibility requirement The LSH theory compliance RPM is shown minimal performance deterioration after transform. The experimental results render reasonable accuracy performance on benchmark AR and FERET datasets. We also perform several rigorous security, privacy, revocability and unlinkability analyses, which are required for cancellable biometrics techniques.

Biometric based authentication can provide strong security guarantee about the identity of users. However, security of biometric data is particularly important as compromise of the data will be permanent. Cancelable biometrics store a non-invertible transformed version of the biometric data. Thus, even if the storage is compromised, the biometric data remains safe. Cancelable biometrics also provide a higher level of privacy by allowing many templates for the same biometric data and hence non-linkability of user's data stored in different databases. In this paper, we propose an approach for cancelable fingerprint templates by using the idea of fuzzy vault. By integrating the fuzzy fingerprint vault with the existing password-based authentication system, we can use a different "long and random" password for each application, and the fuzzy fingerprint vault can be changed by simply changing the password.

Different fingerprint recognition systems store minutiae-based fingerprint templates differently. Some store them inside a small token; some can be found in a server database. As the minutiae template is very compact, many take it for granted that the template does not contain sufficient information for reconstructing the original fingerprint. This paper proposes a scheme to reconstruct a full fingerprint image from the minutiae points based on the amplitude and frequency modulated (AM-FM) fingerprint model. The scheme starts with generating a binary ridge pattern which has a similar ridge flow to that of the original fingerprint. The continuous phase is intuitively reconstructed by removing the spirals in the phase image estimated from the ridge pattern. To reduce the artifacts due to the discontinuity in the continuous phase, a refinement process is introduced for the reconstructed phase image, which is the combination of the continuous phase and the spiral phase (corresponding to the minutiae). Finally, the refined phase image is used to produce a thinned version of the fingerprint, from which a real-look alike gray-scale fingerprint image is reconstructed. The experimental results show that our proposed scheme performs better than the-state-of-the-art technique.

Cancellable fingerprint template has increasingly received interest in research thanks to not only the security for the user's original features but also the stable performance for the system. In this paper, we propose a new method to design cancellable fingerprint template with local structure by clustering the minutiae using the k Nearest Neighbor (kNN) algorithm. In other words, $k$ minutiae in a fingerprint that are closest to a reference minutia form a local structure. Pairwise features from the reference minutia and each of the member in the cluster are extracted and used for local structure matching. The partial Discrete Fourier Transformation was applied as the non-invertible transformation. This method has been evaluated with four public databases FVC2002 DB1-DB3, and FVC2004 DB2. The Equal Error Rate achieved for each database is 0.2%, 0.04%, 4.78%, and 7.64%, respectively.

AFIS (Automated Fingerprint Identification System) is the most widely used identification system. But the most critical part in fingerprint identification is reliable extraction and matching of fingerprint features which depends on quality of input fingerprint image. But it is quite rare that the input fingerprint image will be of very good quality. So there is necessity of some fingerprint enhancement technique to enhance the features of fingerprints before extraction and matching. In this experiment, variable traditional gabor filter (TGF) and modified gabor filter (MGF) are compared in terms of performance, particularly for average and poor quality fingerprint images. Simulation results indicate that, as the quality of fingerprint image degrades, MGF performs more better than TGF in terms of reducing spurious minutiae.

We present a deep artificial neural network (DANN) model that learns latent fingerprint image patches using a stack of restricted Boltzmann machines (RBMs), and uses it to perform segmentation of latent fingerprint images. Artificial neural networks (ANN) are biologically inspired architectures that produce hierarchies of maps through learned weights or filters. Latent fingerprints are fingerprint impressions unintentionally left on surfaces at a crime scene. To make identifications or exclusions of suspects, latent fingerprint examiners analyze and compare latent fingerprints to known fingerprints of individuals. Due to the poor quality and often complex image background and overlapping patterns characteristic of latent fingerprint images, separating the fingerprint region of interest from complex image background and overlapping patterns is very challenging. Our proposed DANN model based on RBMs learns fingerprint image patches in two phases. The first phase (unsupervised pre-training) involves learning an identity mapping of the input image patches. In the second phase, fine-tuning and gradient updates are performed to minimize the cost function on the training dataset. The resulting trained model is used to classify the image patches into fingerprint and non-fingerprint classes. We use the fingerprint patches to reconstruct the latent fingerprint image and discard the non-fingerprint patches which contain the structured noise in the original latent fingerprint. The proposed model is evaluated by comparing the results from the state-of-the-art latent fingerprint segmentation models. The results of our evaluation show the superior performance of the proposed method.

This paper proposes an internet-based personal identity verification system using lossless data hiding and fingerprint recognition technologies. At the client side, the SHA-256 hash of the original fingerprint image and sensitive personal information are encrypted and embedded into the fingerprint image using an advanced lossless data hiding scheme. At the service provider side, after the hidden data are extracted out, the fingerprint image can be recovered without any distortion due to the usage of the lossless data hiding scheme. Hence, the originality of the fingerprint image can be ensured via hash check. The extracted personal information can be used to obtain the correct fingerprint feature from the database. The fingerprint matching can finally verify the client's identity. The experimental results demonstrate that our proposed system is effective. It can find wide applications in e-banking and e-government systems to name a few.

Biometric template protection approaches have been developed to secure the biometric templates against image reconstruction on the stored templates. Two cancellable fingerprint template protection approaches namely minutiae-based bit-string cancellable fingerprint template and modified minutiae-based bit-string cancellable fingerprint template, are selected to be evaluated. Both approaches include the geometric information of the fingerprint into the extracted minutiae. Six modified fingerprint data sets are derived from the original fingerprint images in FVC2002DB1_B and FVC2002DB2_B by conducting the rotation and changing the quality of original fingerprint images according to the environment conditions during an emergency situation such as wet or dry fingers and disoriented angle of fingerprint images. The experimental results show that the modified minutiae-based bit-string cancellable fingerprint template performs well on all conditions during an emergency situation by achieving the matching accuracy between 83% and 100% on FVC2002DB1_B data set and between 99% and 100% on FVC2002DB2_B data set.

An Outsized keys of various sizes such as 256,192 or 128 binary digits are being used for the existing symmetric cryptographic algorithms. These long keys are very tough to recollect. Consequently, key is deposited in a plastic card with a built-in microprocessor, tokens which cannot be tampered. Also an authentication based on password can be used to restrict the access of key related to cryptography. But there is every chance that these user nominated passwords can be lost, taken or predicted by dictionary attacks. Moreover, in symmetric cryptography there is another probable risk of safety that is copy of same secret key exists with receiver and sender and this must be kept secret. Probable solution to above deliberated key management problem is by using biological data of both receiver and sender in all sessions of communication. One of the foremost problem with biometrics is that once it gets negotiated it can't be reused. For cancelling and reissuing biometrics template, cancelable biometrics has been anticipated so that the above mentioned problem may be overcome. A method to generate 128 bit symmetric key using cancelable fingerprint template of receiver and sender is being proposed in this current work. This present approach confirms the privacy of fingerprints by one way transformation of original template into cancelable one as well as resolves the difficulty of key storage and key distribution. The reason is that in this approach the key is neither deposited in any location nor it is sent to the receiver.

This paper proposed a biometric key generation method for fingerprint and finger vein fusion. A matrix extraction algorithm is proposed, which can fuse the aligned fingerprint and finger vein images to extract the biological key. Furthermore, blind alignment of fingerprint and finger vein biometric images are used to improve the accuracy and stability of subsequent biometric key extraction. Compared with the original direct eigenvector extraction method, the proposed bio-key extraction method is more stable and accurate.

In this paper, we propose a histogram of oriented gradient inspired cancelable biometrics - Random Augmented Histogram of Gradients (R∙HoG) for iris template protection. The proposed R∙HoG is built upon on two main components: 1) column vector random augmentation and 2) gradient orientation grouping mechanisms to transform the unaligned irisCode feature into the alignment-robust cancelable template. The alignment-robust property of the proposed R∙HoG enables the fast template comparison which is crucial for an efficient authentication process. Experiments were performed on CASIA-IrisV3-Internal and CASIA-IrisV4-Thousand datasets. The results demonstrate the proposed R∙HoG could achieve acceptable verification performance in both datasets. Other than that, the irreversibility and security properties are studied based on major security and privacy attacks in biometric system. Lastly, results from the benchmarking evaluation framework show the proposed method is satisfying the unlinkability property.