Informatics Publishing Limited

Yahia Belayadi ¹, Abdallah Khababa ¹, Abdelouahab Attia ², Zahid Akhtar ³

Affiliations
1 Department of Computer Science, Ferhat Abbas University, DZ
2 Department of Computer Science, Mohamed El-Bachir Ibrahimi University of Bordj Bou Arreridj, DZ
3 State University of New York Polytechnic Institute, US

Abstract

In recent years, the field of the Internet of Things (IoT), including smart and wearable devices, has witnessed a tremendous advancement leading to the collection of a wide variety of information not only about users but also their activities via various systems such as social networks, apps and so on. Thus, the collection of this large amount of data allows social systems to reach a wide variety of targets and gives more visibility about users and their profiles. It can also help to improve the services and functionalities of the users. Besides, the analysis and prediction of user’s activities in location-based social networks (LBSNs) have received much attention both from industries and research communities, especially in smart city developments, which give much importance to the automation of the LBSNs. In this paper, we present a new method based on association rules for user activity analysis in LBSNs. In particular, the Apriori algorithm has been applied to extract the consequential and advantageous rules to categorize users’ profiles. Empirical evaluations on a publicly available large-scale real-world dataset, named Gowalla, demonstrate the effectiveness of the presented association rules-based system in analyzing users’ activities via LBSNs.

Keywords

Complex System, Social Networks, Association Rules, Apriori Algorithm, Gowalla Dataset.

Full Text

     

Abdelouahab Attia ¹, Abdelouahab Moussaoui ², Mourad Chaa ³, Youssef Chahir ⁴

Affiliations
1 Department of Computer Science, Mohamed El Bachir El Ibrahimi University of Bordj Bou, Arreridj, DZ
2 Department of Computer Science, Ferhat Abbas University, DZ
3 Department of New Technologies of Information and Communication, Ouargla University, DZ
4 Department of Computer Science, University of Caen Lower, FR

Abstract

In this paper, a new method based on Log Gabor- TPLBP (LGTPLBP) has been proposed. However the Three Patch Local Binary Patterns (TPLBP) technique used in face recognition has been applied in Finger-Knuckle-Print (FKP) recognition. The 1D- Log Gabor filter has been used to extract the real and the imaginary images from each of the Region of Interest (ROI) of FKP images. Then the TPLBP descriptor on both images has been applied to extract the feature vectors of the real image and the imaginary image respectively. These feature vectors have been jointed to form a large feature vector for each image FKP. After that, the obtained feature vectors of all images are processed directly with a dimensionality reduction algorithm, using linear discriminant analysis (LDA). Finally, the cosine Mahalanobis distance (MAH) has been used for matching stage. To evaluate the effectiveness of the proposed system several experiments have been carried out. The Hong Kong Polytechnic University (PolyU) FKP database has been used during all of the tests. Experimental results show that the introduced system achieves better results than other state-of-the-art systems for both verification and identification.

Keywords

Biometric Systems, Three Patch Local Binary Patterns, 1D Log Gabor Filter, Finger Knuckle Print.

Full Text

     

Rabah Hammouche ¹, Abdelouahab Attia ², Samir Akrouf ³

Affiliations
1 Department of Computer Science, Mohamed El Bachir El Ibrahimi University, DZ
2 LMSE Laboratory, Mohamed El Bachir El Ibrahimi University, DZ
3 Department of Computer Science, Mohamed Boudiaf University, DZ

Abstract

The authentication of individuals based on Finger Knuckle print (FKP) is a very interesting system in the biometric community. In this paper, we introduce a biometric authentication system based on the FKP trait which consists of four stages. The first one is the extraction of the Region of Interest (ROI). The Phase Congruency method with Gabor filters bank descriptors has been used in the feature extraction stage. Then to enhance the performance of the proposed scheme the Principle Component Analysis (PCA) + Linear Discriminant Analysis (LDA) method has been used in the dimensionality reduction stage. Finally, cosine Mahalanobis distance has been used in the matching stage. Experiments were conducted on the FKP PolyU Database which are publicly available. The reported results with comparison to previous methods prove the effectiveness of the proposed scheme, as well as the given system can achieve very high performance in both the identification and verification modes.

Keywords

Finger Knuckle Print, Phase Congruency, Gabor Filters Bank, Score-Level-Fusion.

Full Text

     

Nour Elhouda Chalabi ¹, Abdelouahab Attia ², Abderraouf Bouziane ¹

Affiliations
1 Department of Computer Science, Mohamed El Bachir El Ibrahimi University of Bordj Bou Arreridj, DZ
2 LMSE Laboratory, Mohamed El Bachir El Ibrahimi University, DZ

Abstract

Hand-based recognition systems with different traits are widely used techniques and are trustworthy ones. We can find it in different real life fields such as banks and industries due to its stability, reliability, acceptability, and the wide range features. In this paper, we present a finger dorsal knuckle print multimodal recognition system, where we use PCAnet (principal component analysis network) deep learning to extract the features from both Major and Minor finger dorsal knuckles to allow a deeper insight into the exploited trait. Then SVM is used for the matching stage of the two modalities, followed by a score level fusion method to combine the scores using different rules. In order to establish the effectiveness of the proposed system, several experiments in the course of this work have been done on the finger knuckle images of the publicly available database PolyUKV1. The results show that the proposed method has better results in comparison with a unimodal system.

Keywords

Finger Knuckle Print, Major, Minor, PCAnet, Score Level Fusion, SVM.

Full Text

     

Soumia Djaghbellou ¹, Zahid Akhtar ², Abderraouf Bouziane ³, Abdelouahab Attia ¹

Affiliations
1 Department of Computer Science, Mohamed El Bachir El Ibrahimi University of Bordj Bou Arreridj, DZ
2 Department of Computer Science, University of Memphis, US
3 LMSE Laboratory, Mohamed El Bachir El Ibrahimi University Bordj Bou Arreridj, DZ

Abstract

Optical character recognition systems for handwritten Arabic language still face challenges, owing to high level of ambiguity, complexity and tremendous variations in human writing styles. In this paper, we propose a new and effective Arabic handwritten characters recognition framework using multi-scale histogram oriented gradient (HOG) features and the deep rule-based classifier (DRB). In the feature extraction stage, the proposed framework combines multi-scale HOG features, and then the DRB is applied on comprehensive HOG features to obtain the final classification label/class. This study involves experimental analyses that were conducted on the publicly available cursive Arabic Handwritten Characters Database (AHCD) containing 16800 characters. Experimental results demonstrate the efficacy of the proposed recognition system compared to the existing state-of-the-art-systems.

Keywords

Arabic Character Recognition, Writing, DRB Classifier, HOG, AHCD.

Full Text

     

Abdelouahab Attia ¹, Abdelouahab Moussaoui ², Youssef Chahir ³, Mourad Chaa ⁴

Affiliations
1 Computer Science Department, Mohamed El Bachir El Ibrahimi University of Bordj Bou Arreridj, DZ
2 Computer Science Department, Ferhat Abbas University, DZ
3 Department of Computer Science, University of Caen, FR
4 Ouargla University, DZ

Abstract

3D Palmprint recognition has become a promising alternative tool for resolving problems compared to the robustness of 2D palmprint recognition. Regarding robustness, biometric systems that use 2D Palmprint suffer from being attacked by using a fake Palmprint identical. Given this, the current paper introduces a new 3D Palmprint recognition approach. Firstly, a set of preprocessing techniques has been applied on 3D depth image such as Tan and Triggs method which can effectively and efficiently eliminate the effect of the low-frequency component with keeping the local statistical properties of the processed image. Then, Gabor wavelets have been employed to extract features. After that, the extracted features have been used as an input in the collaborative representation based classification with regularized least squares (CRC_RLS) to classify the 3D Palmprint images. To evaluate its performance, the proposed algorithm has been applied on the PolyU 3D Palmprint database which contains 8.000 samples. The experimental results successfully and greatly improve the recognition results, especially when, we use Tan and Triggs method for preprocessing and Gabor for feature extraction with CRC_RLS for presentation and classification. We achieve a significant recognition rate of 100 % in lowest Runtime which reflects the robustness of the proposed recognition system.

Keywords

Three-Dimensional Palmprint, Biometric, Gaussian Difference Filtering, Gradient Palms, Weberpalms, Gabor Features, Self-Quotient Image Algorithm.

Full Text

     

Abdelouahab Attia ¹, Zahid Akhtar ², Samir Akhrouf ³, Sofiane Maza ⁴

Affiliations
1 LMSE Laboratory, Mohamed El Bachir El Ibrahimi University of Bordj Bou Arreridj, DZ
2 Department of Network and Computer Security, State University of New York Polytechnic Institute, US
3 Department of Computer Science, Mohamed Boudiaf University, DZ
4 Department of Computer Science, Mohamed El Bachir El Ibrahimi University of Bordj Bou Arreridj, DZ

Abstract

Diabetic Retinopathy (DR) is one of the mainly causes of visual loss worldwide. In fact, DR is leading source of impaired vision in people between 25 and 74 years old. DR exists in wide ranged and its detection is a challenging problem. The gradual deterioration of retina leads to DR with several types of lesions, including hemorrhages, exudates, micro aneurysms, etc. Early detection and diagnosis can prevent and save the vision of diabetic patients or at least the progression of DR can be slowed down. The manual diagnosis and analysis of fundus images to substantiate morphological changes in micro aneurysms, exudates, blood vessels, hemorrhages, and macula are usually time-consuming and monotonous task. It can be made easy and fast with the help of computer-aided system based on advanced machine learning techniques that can greatly help doctors and medical practitioners. Thus, the main focus of this paper is to provide a summary of the numerous methods designed for discovering hemorrhages, microaneurysms and exudates are discussed for eventual recognition of non-proliferative diabetic retinopathy. This survey will help the budding researchers, scientists, and practitioners in the field.

Keywords

Diabetic Retinopathy, Deep Learning, Machine Learning, Computer-Aided Diagnosis.

Full Text

     

Nour Elhouda Chalabi ¹, Abdelouahab Attia ², Abderraouf Bouziane ¹

Affiliations
1 Department of Computer Science, Mohamed El Bachir El Ibrahimi University of Bordj Bou Arreridj, DZ
2 LMSE Laboratory, Mohamed El Bachir El Ibrahimi University of Bordj Bou Arreridj, DZ

Abstract

Faced by problems such as lack of robustness from 2D palmprint recognition system which can result to be attacked using a fake palmprint or having the same palmprint as another individual, 3D can present an alternative solution to deal with this problem, hence in this paper we are going to introduce a novel approach based on 3D palmprint recognition system named TT-P-BSIF: first, a preprocessing technique based on Tan and Triggs method was applied on a 3D depth image in order to effectively and efficiently eliminate the effect of low frequency component and at the same time keeping the local statistical properties of the treated image. Then the processed image is divided into a regular number of blocks using two parameters (a and b), after that the Binarized Statistical local features (BSIF) has been applied on each block in order to extract the features vector. These vectors are all combined to produce one larger vector for each processed image. Afterwards nearest neighbor classifier is used to classifier the 3D palmprint images. To examine the proposed method, this latter has been evaluated on a 3D palmprint database that contains 8,000 samples, the obtained results were consistent and promising which proves that the introduced method can massively and effectively improve the recognition results. Therefore, this proposed work using Tan and Triggs method for preprocessing and BSIF for feature extraction was able to generate a recognition rate up to 99.63% and verification rate at 1% up to 100% with EER equals to 0.12%.

Keywords

3D Palmprint, Tan and Triggs, BSIF, Nearest Neighbor Classifier.

Full Text