DSpace Collection:https://dspace.univ-ouargla.dz/jspui/handle/123456789/2072026-03-12T16:36:56Z2026-03-12T16:36:56ZImage Representation Using The Intrinsic Texture PropertiesDjilani, Belilahttps://dspace.univ-ouargla.dz/jspui/handle/123456789/403102026-02-10T09:18:08Z2026-01-01T00:00:00ZTitre: Image Representation Using The Intrinsic Texture Properties
Auteur(s): Djilani, Belila
Résumé: Texture is fundamental to computer vision, yet analyzing its structure re
mains difficult due to variations in scale and pattern. By employing multi
scale analysis based on wavelet theory, this thesis bridges the gap between
classical signal analysis and modern deep learning to address these challenges
and conduct an in-depth analysis of intrinsic texture properties.
We introduce two complementary methods. The Wavelet Texture De
scriptor (WTD) combines fixed wavelet decomposition with rigorous feature
selection to maximize efficiency in limited data environments. The Data
Driven Wavelet Transform (DDWT) takes this further by embedding a train
able wavelet layer into a neural network, allowing the model to learn task
specific wavelet filters rather than relying on rigid, fixed ones.
Experimental evaluation confirms that WTD achieves state-of-the-art re
sults, while DDWT offers superior adaptability for complex, heterogeneous
textures with negligible additional parameters and minimal computational
cost. Ultimately, this work proves that blending wavelet theory with modern
learning creates robust, interpretable representations for visual recognition,
extending the value of wavelets into the deep learning era.
Description: Artificial Vision2026-01-01T00:00:00ZA Transfer Learning-Based Approach for Arabic Sentiment AnalysisMohammed Elsadiq, BARMATIhttps://dspace.univ-ouargla.dz/jspui/handle/123456789/398712026-01-11T09:38:35Z2026-01-01T00:00:00ZTitre: A Transfer Learning-Based Approach for Arabic Sentiment Analysis
Auteur(s): Mohammed Elsadiq, BARMATI
Description: Artificial Intelligence2026-01-01T00:00:00ZDeep versus handcrafted approaches for improving medical image security using watermarkingHEBBACHE, Khaledhttps://dspace.univ-ouargla.dz/jspui/handle/123456789/382612025-03-09T10:05:42Z2024-01-01T00:00:00ZTitre: Deep versus handcrafted approaches for improving medical image security using watermarking
Auteur(s): HEBBACHE, Khaled
Résumé: As advancements in computer vision continue to transform healthcare, the need to
protect medical images, particularly in telemedicine, has become increasingly critical. This
thesis explores the domain of medical image watermarking, offering a detailed comparison
between handcrafted methods and deep learning approaches, with a specific emphasis on
feature extraction techniques. A comprehensive state-of-the-art review sets the stage,
identifying the strengths and limitations of various watermarking strategies aimed at
safeguarding medical images.
The primary contributions of this research include three proposed watermarking
methods. First, a novel blind watermarking method based on Local Binary Patterns and
Discrete Wavelet Transform (LBP-DWT) is presented, specifically designed for telemedicine
applications. Second, a gradient-based feature extraction technique, termed "GradWater," is
developed to further strengthen watermark security through rich image-driven features. Third,
a deep learning-based zero watermarking technique is proposed, utilizing a pre-trained
VGG16 model. This method generates the watermark without embedding any alterations into
the original medical image, ensuring its quality remains intact while maintaining strong
security.
An extensive experimental evaluation compares the feature extraction capabilities of
handcrafted methods against those of deep learning approaches, focusing on their resistance to
various attacks, such as noise, compression, and geometric distortions, while preserving the
quality of the medical images. The findings also offer a comparative analysis of zero and non-
zero watermarking schemes, providing valuable insights into their respective advantages.; Avec les avancées de la vision par ordinateur qui continuent de transformer le
domaine de la santé, la nécessité de protéger les images médicales, en particulier dans la
télémédecine, devient de plus en plus cruciale. Cette thèse explore le domaine du tatouage
numérique des images médicales, en offrant une comparaison détaillée entre les méthodes
artisanales et les approches basées sur l'apprentissage profond, avec un accent particulier sur
les techniques d'extraction de caractéristiques. Une revue complète de l'état de l'art est
présentée, identifiant les forces et les limites des différentes stratégies de tatouage visant à
protéger les images médicales.
Les principales contributions de cette recherche comprennent trois méthodes de
tatouage numérique proposées. Premièrement, une nouvelle méthode de tatouage aveugle
basée sur les motifs binaires locaux (LBP) et la transformation en ondelettes discrète (DWT)
est présentée, spécifiquement conçue pour les applications de télémédecine. Deuxièmement,
une technique d'extraction de caractéristiques basée sur le gradient, appelée "GradWater", est
développée pour renforcer davantage la sécurité des tatouages à travers des caractéristiques
riches et tirées des images. Troisièmement, une technique de tatouage zéro basée sur
l'apprentissage profond est proposée, utilisant un modèle pré-entraîné VGG16. Cette méthode
génère le tatouage sans introduire de modifications dans l'image médicale originale, assurant
ainsi la préservation de sa qualité tout en maintenant une sécurité robuste.
Une évaluation expérimentale approfondie compare les capacités d'extraction de
caractéristiques des méthodes artisanales à celles des approches d'apprentissage profond, en se
concentrant sur leur résistance à diverses attaques, telles que le bruit, la compression et les
distorsions géométriques, tout en préservant la qualité des images médicales. Les résultats
offrent également une analyse comparative des schémas de tatouage zéro et non-zéro,
fournissant des informations précieuses sur leurs avantages respectifs.
Description: Computer Science2024-01-01T00:00:00ZManagement and AI: Towards An Intelligent Decision Support System For Strategic ManagementHAMROUNI, Basmahttps://dspace.univ-ouargla.dz/jspui/handle/123456789/359472024-05-26T09:32:27Z2024-01-01T00:00:00ZTitre: Management and AI: Towards An Intelligent Decision Support System For Strategic Management
Auteur(s): HAMROUNI, Basma
Résumé: Case-Based Reasoning (CBR) is a problem-solving paradigm that uses knowledge of relevant past experiences (cases) to interpret or solve new problems. CBR systems allow for easily generating explanations as they typically organize and represent knowledge in a way that makes it possible to reason about and thereby generate explanations. An improvement to this paradigm is ontology-based CBR, an approach that combines, in the form of formal ontologies, case-specific knowledge with domain-specific knowledge in order to improve the effectiveness and explanation capability of the system. Intelligent systems as an applied domain of artificial intelligence (AI) include the processes, methodologies, and techniques of employing AI for real-world problem-solving. In this way, intelligent systems enhance management processes at all levels, operational and strategic. The business model represents a concept that has recently entered scholarly discussions within the fields of strategic management, entrepreneurship, and innovation. It is considered a strategic tool and a cognitive framework that helps strategy practitioners and endows entrepreneurs with a template for integrating and organizing strategically relevant elements in order to successfully exploit a business opportunity and remain profitable in the long run.
With this vision, the purpose of this project is to propose a new generation of intelligent decision support systems for Business Models that have the ability to provide explanations to increase confidence in proposed solutions. The performance results obtained show the benefits of the proposed solution with different requirements of an explanatory decision support system. Consequently, applying this paradigm to software tools for business model development will have great promise for supporting business model design and innovation. We have enriched research in the branch of intelligent systems for BM by developing a conceptual design for the intelligent decision support system for strategic decision-making by proposing an intelligent technique. This study is part of a growing body of research on tools for Business Model design; this project will contribute to future research on similar topics.
Description: Computer Science2024-01-01T00:00:00Z