University of Bordj Bou Arreridj

Welcome to the Digital Repository of University Mohamed El Bachir El Ibrahimi - Bordj Bou Arreridj










 

Communities in University of Bordj Bou Arreridj Repository

Select a community to browse its collections.

Now showing 1 - 8 of 8

Recent Submissions

Thumbnail Image
Item
Une solution IA pour médias et institutions face aux fake news SUR LES RÉSEAUX SOCIAUX
(University of Mohamed El Bachir El Ibrahimi - Bordj Bou Arréridj, 2026) Sebaihi riadh
This work aims to develop an intelligent system for the automatic detection of fake news amid the widespread circulation of digital content. The system relies on natural language processing techniques and the analysis of textual content in news articles. Machine learning and deep learning algorithms were employed to classify news as true or false. Experiments were conducted on real datasets, and performance was evaluated using metrics such as accuracy and F1-score. The results demonstrate the effectiveness of intelligent models in reducing the spread of misinformation
Thumbnail Image
Item
Fingerprint Classification for AFIS Search-Space Reduction Using Deep Learning
(University of Mohamed El Bachir El Ibrahimi - Bordj Bou Arréridj, 2026) Mechri abdelhafid; Tebbi mohammed amine
Fingerprint-based Automatic Fingerprint Identification Systems (AFIS) perform 1:N matching by comparing each query against every enrolled template in the gallery, a procedure whose computational cost scales linearly with gallery size and becomes prohibitive at operational scale. This thesis addresses that bottleneck by developing a deep learning-based Henry-class pre-classification stage that partitions the enrolled gallery into five class-labelled bins before matching, restricting each identification query to the relevant subset and reducing the active comparison set without discarding correct candidates. The classification system was developed and evaluated on NIST Special Database 4 (NIST SD4), a 2 000-image benchmark comprising matched F/S impression pairs across all five Henry classes. Three fingerprint input representations were investigated — raw CLAHE-enhanced images, Orientation Field Enhancement Maps (OFEM), and HSV-encoded orientation fields — and five ImageNet-pretrained convolutional backbone architectures were evaluated: EfficientNet-B3, ResNet50, MobileNetV3-Small, VGG16-BN, and ConvNeXt-Tiny. Orientation-field encoding consistently outperformed raw and OFEM representations across all architectures, establishing input representation quality as the dominant performance factor. ConvNeXt-Tiny, trained on a dual raw-and-orientation-field input, achieved 92.50% standalone accuracy. A binary VGG16-BN specialist, trained with Focal Loss and controlled oversampling to address the chronic underperformance on the Tented Arch minority class, raised the Tented Arch F1-score from 0.606 to 0.808. Integration of the specialist through a confidence-gated veto mechanism produced the final dual-architecture system, reaching 93.25% overall classification accuracy. The operational contribution of the classification system was validated by integrating it with the SourceAFIS 3.18.1 minutiae-based matching engine within a complete 1:N identification pipeline. Three search-space scenarios were evaluated: full-gallery baseline (Scenario A, 2 000 comparisons), single-bin classification-guided search (Scenario B, ≈400 comparisons), and dual-bin search (Scenario B+, ≈800 comparisons). The B+ strategy preserved near-complete identification accuracy — 87.50% Rank-1 — while reducing active comparisons by 2.61× relative to the full-gallery baseline. A Gabor score-fusion stage, applying Gabor-enhanced template matching in parallel with raw matching and combining scores via a conservative max-fusion rule, lowered the Equal Error Rate from 13.36% to 8.58% by recovering 153 identification failures attributable to poor-quality probe impressions, while breaking only 14 previously passing matches. The principal finding of this thesis is the empirical demonstration that a deep learning classification front-end, integrated with a traditional minutiae matcher, delivers measurable and reproducible search-space reduction without sacrificing identification accuracy. The dual-architecture design, the Gabor fusion strategy, and the honest characterisation of the conditions under which the approach succeeds and fails together constitute a reproducible and extensible foundation for future work in adaptive AFIS search-space reduction.
Thumbnail Image
Item
Détection de la somnolence du conducteur à l’aide de techniques d’intelligence artificielle
(University of Mohamed El Bachir El Ibrahimi - Bordj Bou Arréridj, 2026) SALI Oussama; Benameur Zahra
This thesis focuses on the design and implementation of a driver drowsiness detection system based on the analysis of eye closure from video images. The objective is to automatically identify signs of drowsiness in real time by monitoring the driver's eyelid state. When the system detects that both eyes remain closed for a critical period, an audible alarm is immediately triggered to alert the driver and help prevent potential accidents. The proposed system relies on image processing and artificial intelligence techniques. The video is first analyzed to detect the driver's face using a face detector based on the Single Shot Detector (SSD) architecture, implemented through OpenCV's Deep Neural Network (DNN) module. The eyes are then localized using Haar Cascade classifiers. The extracted eye regions are analyzed by a Convolutional Neural Network (CNN) specifically trained to distinguish between open and closed eyes. The driver's level of alertness is continuously monitored using a scoring system that differentiates normal blinking from actual drowsiness episodes. When the score exceeds a predefined threshold, an audible alarm is activated to warn the driver.
Thumbnail Image
Item
E-Health Digital Records Platform
(University of Mohamed El Bachir El Ibrahimi - Bordj Bou Arréridj, 2026) Zouaoui Abdelghafour; Bouhali Abdelghafour; Salhi Amine; Haddad Abdelouadoud
This thesis presents the design and implementation of a multimodal biometric recognition system integrated into a full-stack E-Health web and mobile platform. Three physiological biometric modalities (iris, face, and fingerprint) are individually trained using ResNet-50 deep neural networks with ArcFace angular margin loss, then fused at score level using several strategies, including a Verification-Based Learned Fusion approach. The resulting system achieves a Rank-1 accuracy of 98.98% and an Equal Error Rate (EER) of 0.84% on a 64- subject test cohort of 3,444 probe samples. The biometric engine is deployed as a Python FastAPI microservice consumed by a Next.js 16 web application with role-based access control, while a Flutter mobile application extends access to Android patients. The platform follows Privacy-by-Design principles by storing only 512- dimensional embedding vectors
Thumbnail Image
Item
Étude des Techniques d’Interprétabilité Appliquées aux Modèles CNN pour la Détection Précoce de la Maladie d’Alzheimer à Partir d’Images IRM.
(university of bordj bou arreridj, 2026) Chihab Imane Khadidja; Zouache Hanane
This thesis aimed to develop an intelligent and innovative system to facilitate the early diagnosis of Alzheimer’s disease, using magnetic resonance imaging (MRI) and deep learning techniques. To achieve this, several convolutional neural network (CNN) models, such as Xception, MobileNet, and EfficientNet, were used to extract rich and subtle features from the images. This data was then integrated with classical machine learning algorithms, such as SVM, KNN, and Random Forest, to classify the stages of disease progression. The experimental results demonstrated that combining convolutional neural networks and machine learning algorithms provides optimal accuracy in disease classification. The hybrid model (Xception +SVM) achieved the greatest generalizability and the best diagnostic accuracy. To ensure the reliability of medical results and foster trust with clinicians, "Explainable AI methods, such as Grad-CAMandSHAP, have been integrated to provide a visual and systematic analysis that clarifies the model’s decision-making process. Furthermore, the research has led to the design of an integrated medical platform using the Flask framework. This platform allows users to intelligently manage patient data, upload MRI images, run prediction algorithms, view classification results, and interact with the model’s visual interpretations.