Palmprint is universal, unique, and acceptable biometric trait with higher reliability. Real world systems demand highly precision-wise identification systems; multispectral imaging provides dominant discriminative features and is robust against anti-spoofing. Each individual spectral band highlights different kinds of palm features. This paper proposes a CNN-based multispectral palmprint biometric system which is developed under a novel deep learning architecture that meets real-time application requirements. The adopted PolyU multispectral database consists of palmprint images collected under red, green, blue, and near-infrared (NIR) illuminations. We have investigated the performance of the system in individual spectral band and also analyzed the verification rate by combining images obtained from different spectral band. After analyzing the features from different spectral bands, we have developed biometric system under pre- and post-classification approaches to combine multispectral data.

In this paper, we propose different approaches for the segmentation of handwritten Devanagari word documents into constituent characters (or pseudo-characters). For accurate identification and segmentation of shiroreakha we exploited ShiroreakhaNet which is encoder-decoder based convolutional neural network. After, segmenting the shiroreakha structural patterns/properties are exploited for the segmentation of upper and lower modifiers. For the corroboration of the efficacy of the results, we collected dataset from different domains. Comparison is also performed with the state-of-the-art methods, and it was revealed that proposed approaches significantly perform better.

Post-traumatic spigelian hernia is a rare entity and requires a high index of suspicion for its detection. Mostly the cases are diagnosed at the time of first presentation when a swelling is seen at the site of injury. We report a case of missed post-traumatic abdominal wall hernia at the left spigelian point in a 3-year child missed during initial exploration for traumatic jejunal perforation which led to re-exploration for incarceration of hernia. The rarity of presentation as a post-operative complication has not been reported in literature.

Shirorekha identification and removal is an important and a challenging pre-processing stage in almost all machine interpretations for handwritten Devanagari documents. Within this area of investigation, all studies are designed based on traditional image processing techniques. Which are mainly based on hand-engineering and learn local transformations only. However, it can also be viewed as a supervised classification task in which each pixel, in a document, is examined/ queried so that those classified as shirorekha are removed. For this purpose, we extended this area of investigation by designing an encoder-decoder based convolutional neural network (EDCNN). Which have demonstrated, from various studies, that they learn image intricacies very well. The contribution of this work is three-fold, first, we created our own handwritten word dataset comprising of words with and without shirorekha, such that, effective training takes place. Next, we trained the proposed network with binary as well as in gray scale formats. Finally, we demonstrated that the proposed approach is accurate and generalizable.

In this paper, we are proposing deep biometric verification system based on multispectral palmprint images. We have built our own deep learning architecture, deployed on multispectral palmprint unimodal system, deep features and traditional texture namely log-gabor features are compared in this work. Multispectral palmprint biometric are captured under different illumination. Images from multispectral database provides images obtained from varied wavelength in four different instance colours Red, Green, Blue and Near InfraRed. Deep learning is gaining extraordinary leaps by consolidating most of the machine learning algorithms essence and proved tremendous in computer vision and image recognition. We have evaluated our system by performing feature level fusion on traditional log gabor method and from the novel deep learning CNN (Convolution Neural Network) architecture proposed in our work. We have compared the performance of the system under different spectral wavelengths. Experimental results demonstrate the effectiveness of our proposed system by outer performing the traditional method results and shown reliable.

The paper addresses the unimodal and multimodal (fusion prior to matching) biometric recognition system from the promising traits face and iris which uniquely identify humans. Performance measures such as precision, recall, and f-measure and also the training time in building up the compact model, prediction speed of the observations are tabulated which gives the comparison between unimodal and multimodal biometric recognition system. LPQ features are extracted for both the modalities and LDA is employed for dimensionality reduction, KNN (linear and weighted), and SVM (linear and nonlinear) classifiers are adopted for classification. Our empirical evaluation shows our proposed method is potential with 99.13% of recognition accuracy under feature level fusion and computationally efficient.

Single biometric modality may not be enough to achieve the best performance in real time applications. Multimodal biometric system provides high degree of recognition and more population coverage by combining different source of biometric modalities which in turn enhances the performance superiority and error rate shrinking properties. This work addresses the performance comparison in terms of verification rate of the unimodal and bimodal biometric systems holding face trait as the primary modality and Fingerprint, Iris, Palmprint and Handvein as secondary modality. The potential of the proposed verification system is judged by conducting empirical analysis on both the types of fusion strategies- pre and post classification. Overall gist of the paper summarizes few issues, (a) Which secondary physiological modality would be the optimal combination that gives the complimentary information along with the face trait, (b) impact and adoptability of fusion strategies at various levels of fusion to be adhered, (c) finally a brief overview that addresses face centric bimodal systems developed at all levels of fusion on both clean and noisy data.

In this paper, biometric verification system is built adopting five physiological traits. We have opted to choose score level fusion, due to its ease in accessing and combining scores generated by different matchers. In our work, we have explored simple transformation rules (min, max, sum), dynamic weighting (DLC, WF) and t-norms (Frank, hamacher and Sugeno weber) score level fusion schemes. All these score level fusion schemes are checked with different desirable parameters such as accuracy, scalability and robustness in declaring the most efficient method for both unimodal and multimodal biometric system.

In this paper, we have developed Biometric recognition system adopting hand based modality Handvein, which has the unique pattern for each individual and it is impossible to counterfeit and fabricate as it is an internal feature. We have opted in choosing feature extraction algorithms such as LBP-visual descriptor ,LPQ-blur insensitive texture operator, Log-Gabor-Texture descriptor. We have chosen well known classifiers such as KNN and SVM for classification. We have experimented and tabulated results of single algorithm recognition rate for Handvein under different distance measures and kernel options. The feature level fusion is carried out which increased the performance level.

In the proposed multimodal biometric verification system, the system is implemented on all levels of fusion strategies (i) fusion prior to matching (sensor level and feature level) and (ii) fusion post matching (score level and decision level), a binding required and a deserving step that outputs a reliable and robust biometric identification and verification systems. We have chosen benchmark databases for our experimentation and considered physiological modalities such as face, palmprint, finger knuckle print, handvein. The performance measures considered here are FAR (False Acceptance Rate) and FRR (False Rejection Rate). Extracting texture features from a well-known texture operator-LPQ (local phase quantization), we have performed sensor level fusion adopting HAAR wavelets, feature level fusion using Z-Score normalization, score level fusion employing simple sum rule and decision level fusion with AND rule. For the implemented biometric recognition system, score level fusion strategy outer performs than the other fusion techniques in terms of EER, yielding good verification rate on all benchmark threshold values (0.01%, 0.1%, 1%), with the GAR=100% at 1% FAR. The tabulated results of the experiments are visualized by BAR chart

In this paper, we propose Unimodal Biometric Identification System Based on Ensemble machine learning techniques such as Stacking, Boosting and Bagging. The article framework gives the comparative analysis between Ensemble techniques adopting different classifiers (Geometric based classifiers-KNN, SVM, NN and Decision tree basedRF, RPART) and the best feature extraction algorithms such as Texture based-LPQ, Appearance basedICA1. The experimentation is carried out on physiological biometric traits adopting the standard benchmark databases such as AR facial database and Poly-U fingerprint database. The objective of the paper is to understand whether concentrating on discriminating feature extraction algorithms or performing extensive computation on ensemble techniques with different classification models would contribute greatly in performance of the system.

Selecting the most accurate rules under the fusion strategies depends on many factors such as quantity of training data, quality of given training data, understanding reliability of ground truth for training data, modeling tools etc. In our work we have adopted score level fusion scheme as it contains rich source of input data and has proximity features, we have explored promising rules such as density based likelihood features-support vector machines((LF-SVM)), likelihood features random subspace of AdaBoost (LF(RS-ADA)) and classifier based (support vector machines(SVM), random subspace of AdaBoost (RS-ADA), dempster-shafer (DS).

In this work, we have proposed novel deep CNN framework architectures that effectively represent complex image characteristics which performs feature extraction in just two convolution layers and has successfully proved to be an reliable biometric verification system on employment of physiological traits face and iris for our system development. Extensive experiments in configuring the CNN hyper parameters such as number of convolution layers required, filters and its size in each layer, batch size, epochs, iterations and learning rate is a paramount, determining these factors truly depends on the nature of data and its size. Our work has relinquished our novel idea and has obtained 99% of GAR in unimodal biometric verification system itself and definitely the approach has rendered great results when compared with conventional feature extraction and classification techniques.

International Journal of Computer Sciences and Engineering (A UGC Approved and indexed with DOI, ICI and Approved, DPI Digital Library) is one of the leading and growing open access, peer-reviewed, monthly, and scientific research journal for scientists, engineers, research scholars, and academicians, which gains a foothold in Asia and opens to the world, aims to publish original, theoretical and practical advances in Computer Science,Information Technology, Engineering (Software, Mechanical, Civil, Electronics & Electrical), and all interdisciplinary streams of Computing Sciences. It intends to disseminate original, scientific, theoretical or applied research in the field of Computer Sciences and allied fields. It provides a platform for publishing results and research with a strong empirical component. It aims to bridge the significant gap between research and practice by promoting the publication of original, novel, industry-relevant research.

Noise is a random signal that usually arises during electronic transmission, preprocessing, acquisition, faulty storage that distorts the original information of the image. Image can be deteriorated by various kinds of noises. In our work, we are proposing both the unimodal and multimodal biometric verification system adopting hand-based traits and we are also analyzing the robustness of the deployed system by adding salt and pepper noise that takes either maximum or minimum grey values. The experimentation is performed on multimodal preclassification and post classification techniques and the verification results are tabulated for standard threshold values at 0.01%, 0.1 %, 1 %.

The impulse of ubiquitous biometrics may refer to the unique identification of an individual by analyzing psychological and behavioral traits. We have employed psychological biometric modalities such as AR database for face trait, poly-U database for palm trait and behavioral biometric modalities such as MCYT-100 database for signature trait, TIMIT database for speech trait. Desirable features extraction algorithms are employed, on both psychological and behavioral traits and fused at score level. The each of scores from the matcher from respective modalities fused into a single score by adopting sum rule. The proposed system evaluated using the performance measures of standard biometric verification GAR (Genuine Acceptance Rate), FAR (False Acceptance Rate), and the obtained results analyzed graphically using ROC (Receiver Operating Characteristics). We have motivated in conducting the experiments on fusion of both psychological and behavioral traits and observed how the verification rate increases with the increase in fusion of more than one modalities.

In this paper, we have proposed a multimodal biometric verification system adopting the physiological traits-face and fingerprint. The system has been evaluated for robustness analysis imposing the Additive White Gaussian noise (AWGN) on clean data of the employed databases-AR facial database, PolyU High-resolution fingerprint database. The unimodal and multimodal (pre and post matching fusion strategies) verification system is investigated extracting the log-gabor features. Considering the performance measures such as GAR (Genuine Acceptance Rate) and FAR (False Acceptance Rate) at the benchmark threshold values - 0.01%, 0.1%, 1%. The accuracy of the system for clean and noisy data can be visualized by ROC (Receiver Operating Characteristics) curve, area under it expresses the performance of the implemented system.

: Multi-algorithmic approach to enhancing the accuracy of iris recognition system is proposed and investigated. In this system, features are extracted from the iris using various feature extraction algorithms, namely LPQ, LBP, Gabor Filter, Haar, Db8 and Db16. Based on the experimental results, it is demonstrated that Mutli-algorithms Iris Recognition System is performing better than the unimodal system. The accuracy improvement offered by the proposed approach also showed that using more than two feature extraction algorithms in extracting the iris system might decrease the system performance. This is due to redundant features. The paper presents a detailed description of the experiments and provides an analysis of the performance of the proposed method.