https://dspace.univ-ouargla.dz/jspui/handle/123456789/205 Sat, 18 Jul 2026 00:24:28 GMT 2026-07-18T00:24:28Z https://dspace.univ-ouargla.dz/jspui/handle/123456789/41037 Titre: Study of Waveguides Based on Photonic Crystal Components Auteur(s): Hathat, Mohammed Elfateh Résumé: In this thesis, photonic crystal-based waveguide devices are designed, simulated, and analyzed including photonic beam splitters and oil condensate quality sensors, with the overall aim of improving transmission efficiency and providing an opportunity to monitor the quality of oil condensate in real time. Using a combination of air/silicon material with a hexagonal lattice configuration of 2D-PhC, we firstly provided a conception of several splitter devices, such as 1x2, 1x3, and 1x4 splitters using 2D-PhC designs by applying different advanced splitting techniques. In this basis a new 1x2 Y- splitter was designed using an arc-shaped slot (ASS) with a transmission efficiency of around 44%in each output port, moreover the radiation losses have been reduced to 11%. The 1x3 splitter that combines asymmetrical and symmetrical 1x2 motifs showed a transmission efficiency of nearly 91% at two output ports when operating at 1550 nm and 1573 nm by using a defect Holes (DH) in the splitting region. The 1x4 splitter was optimized to establish a close-to-equal power splitting were near-equal distribution of power, with each output contributing about 24.82% of the total emitted power an almost optimal result. Furthermore, advanced 1x2 and 1x4 splitters have been proposed and designed using coupling technique in which the Polymethyl methacrylate (PMMA) is embedded in the coupling region. By heating the PMMA material the efficiency transmission of the splitting was improved and measured to 44% and 24.85% in each outputs for 1x2 and 1x4 splitters respectively. Besides a photonic crystal sensor of the quality of oil condensate was suggested. The sensor measures the change of refractive index (RI) that indicates an oil, the detecton analysis of the porposed sensor has proved that the resonant wavelength and the intensity power varied when the RI of the oil condensate change. Therefore, the propoed sensor has a dual sensitivity in terms of wavelength and the intensity power is about 631.33 nm/RIU and 172.32. Taken together, these findings support the idea that photonic crystal structures have a high potential of achieving a compact and multi-functional optical constituents to be integrated into photonic systems and used in industrial diagnostic situations.; Dans cette thèse, des dispositifs à base de guides d’ondes en cristal photonique ont été conçus, simulés et analysés, notamment des séparateurs de faisceaux photoniques et des capteurs de qualité des condensats pétroliers, dans le but global d’améliorer l’efficacité de transmission et de permettre une surveillance en temps réel de la qualité du condensat pétrolier. En utilisant une combinaison de matériaux air/silicium avec une configura tion en réseau hexagonal d’un cristal photonique bidimensionnel (2D-PhC), plusieurs dispositifs diviseurs ont d’abord été conçus, tels que des séparateurs 1x2, 1x3 et 1x4, en appliquant différentes techniques de division avancées. Sur cette base, un nouveau séparateur en Y (1x2) a été conçu en utilisant une fente en arc (ASS), présentant une efficacité de transmission d’environ 44 % sur chaque port de sortie, tandis que les pertes par rayonnement ont été réduites à 11 %. Le séparateur 1x3, combinant des motifs 1x2 asymétriques et symétriques, a montré une efficacité de transmission proche de 91 % sur deux ports de sortie lorsqu’il fonctionne à 1550 nm et 1573 nm, grâce à l’utilisation de cavités de défaut (DH) dans la région de division. Le séparateur 1x4 a été optimisé afin d’obtenir une répartition quasi égale de la puissance, chaque sortie contribuant à envi ron 24,82 % de la puissance totale émise — un résultat presque optimal. De plus, des séparateurs 1x2 et 1x4 avancés ont été proposés et conçus en utilisant une technique de couplage dans laquelle le PMMA est intégré dans la région de couplage. En chauffant le matériau PMMA, l’efficacité de transmission du dispositif a été améliorée et mesurée à 44 % et 24,85 % pour les séparateurs 1x2 et 1x4 respectivement. Par ailleurs, un cap teur à cristal photonique destiné à l’analyse de la qualité des condensats pétroliers a été proposé. Ce capteur mesure la variation de l’indice de réfraction (RI) caractéristique de chaque type d’huile. L’analyse de détection du capteur proposé a montré que la longueur d’onde résonnante et la puissance d’intensité varient avec le changement de l’indice de réfraction du condensat pétrolier. Ainsi, le capteur proposé présente une double sensibil ité, en termes de longueur d’onde et de puissance d’intensité, estimée respectivement à 631,33 nm/RIU et 172,32.Dans l’ensemble, ces résultats confirment le fort potentiel des structures à cristal photonique pour la réalisation de composants optiques compacts et multifonctionnels, susceptibles d’être intégrés dans des systèmes photoniques et utilisés dans des applications industrielles de diagnostic Description: Telecommunication Networks Thu, 01 Jan 2026 00:00:00 GMT https://dspace.univ-ouargla.dz/jspui/handle/123456789/41037 2026-01-01T00:00:00Z https://dspace.univ-ouargla.dz/jspui/handle/123456789/41036 Titre: Face Image Analysis via Transformers : Algorithms and Applications Auteur(s): Chahrazad, RAHMANI Résumé: Drowsiness is a major cause of traffic accidents worldwide, impairing perception, slowing reactions, and reducing decision-making. Accurate and reliable detection is therefore essential for enhancing road safety and mitigating fatigue-related incidents. Face image analysis has emerged as a promising approach for non-intrusive drowsiness detection, yet existing methods often struggle with robustness, temporal modeling, and reliance on large labeled datasets. This thesis investigates advanced deep learning frameworks to enhance driver drowsiness detection through robust learning of fatigue-related facial cues and data-efficient strategies. First, a Transformer-based framework is proposed, leveraging hierarchical self-attention mechanisms for extracting discriminative fatigue-related facial features. Driver faces are detected and aligned using Multi-Task Cascaded Convolutional Neural Networks (MTCNN), followed by a Swin Transformer to capture spatial facial patterns. The method demonstrates competitive performance on the NTHU Drowsy Driver Detection (NTHU-DDD) dataset, highlighting the effectiveness of Swin transformers in face-based fatigue analysis. Second, a semi-supervised learning framework is developed to reduce reliance on large labeled datasets while maintaining real-time applicability. YOLOv8 is employed for fast face detection, and a Swin Transformer learns drowsiness-related representations from sequential frames. Pseudo-labeling enables progressive incorporation of unlabeled data, improving model generalization across diverse datasets, including NTHU-DDD, YawDD, and UTA-RLDD. Finally, a computationally efficient driver-monitoring system is introduced, combining a custom CNN enhanced with channel attention and semi-supervised training. This framework leverages both labeled and unlabeled data to improve generalization while reducing annotation requirements. Evaluated on the UTA-RLDD dataset, the system achieves high accuracy and low computational complexity, ensuring suitability for practical deployment in real-world driving scenarios. Description: Automatic and Systems Thu, 01 Jan 2026 00:00:00 GMT https://dspace.univ-ouargla.dz/jspui/handle/123456789/41036 2026-01-01T00:00:00Z https://dspace.univ-ouargla.dz/jspui/handle/123456789/40787 Titre: Performance Improvement and Complexity Reduction of Massive MIMO Communication System Auteur(s): Smail, LABED Description: Telecommunication Systems Thu, 01 Jan 2026 00:00:00 GMT https://dspace.univ-ouargla.dz/jspui/handle/123456789/40787 2026-01-01T00:00:00Z https://dspace.univ-ouargla.dz/jspui/handle/123456789/40310 Titre: 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 Vision Thu, 01 Jan 2026 00:00:00 GMT https://dspace.univ-ouargla.dz/jspui/handle/123456789/40310 2026-01-01T00:00:00Z