Computer

Permanent URI for this collectionhttps://dspace.univ-boumerdes.dz/handle/123456789/3082

Browse

Now showing 1 - 20 of 180

3D statistical shape modeling
(2016) Omari, Sabrina; Soual, Imene; Cherifi, Dalila (supervisor)
Statistical shape models (SSMs) have been firmly established as a robust tool for segmentation of images. Widespread utilization of three-dimensional models appeared only in recent years, primarily made possible by breakthroughs in automatic detection of shape correspondences; while 2D models have been in use since the early 1990s. The objective of this project is to build a 3D statistical shape modeling for a given data; the implemented process goes through those basic steps, first collect the given data then apply the alignment algorithm based on the ICP (iterative closest point) method which in turn relies on Procrustes analysis result as a starting point, next we apply fitting algorithm which is also based on ICP. Finally we obtain the model using PCA (principle component analysis). To achieve this work, we have implemented the above process on two different shape models, one tested with the Basel Face Model (BSF) and the other is the femur model data samples from the SICAS (Swiss Institute for Computer Assissted Surgery) Medical Image Repository which is used by the Basel University (Switzerland) for both samples, where these models allow the generation and the exploration of the possible shape variation.
Action detection using deep learning shoplifting detection framework
(Université M’hamed Bougara de Boumerdes : Institut de Genie Electrique et Electronique, 2024) Bettahar, Mohammed Nadir; Touzout, Walid ( Supervisor)
This project delves into the application of deep learning for action detection, with a specifi cfocu so nidentifyin gshopliftin gbehavior si nretai lenvironments. Th egrowing need for automated surveillance systems that can efficient lya ndaccurate lydetect suspicious activities has motivated this work. Shoplifting action detection is the process of identifying and localizing shoplifting activities in a video by findin gbot hwhere and when an action occurs within a video clip and determining what action is being performed. A key challenge lies in preparing a dataset that reflect sth ecomplexity of real-world scenarios, which was addressed by employing semi-supervised learning techniques. The use of You Only Watch Once version 9 (YOLOv9) object detection model,its tracking function, was instrumental in the automation of labeling and tracking objects within the shoplifting video dataset, ensuring a reliable foundation for action detection. To evaluate the effectivenes so fth esystem ,th eYo uOnl yWatc hOnc eversio n2 (YOWOv2) model was used, conducting comprehensive training and testing across a variety of shoplifting situations. This allowed for a detailed assessment of the model’s ability to recognize and generalize diverse shoplifting actions, even in challenging environments. The results show that the models can detect suspicious behavior, offerin ga promising tool for improving retail security. This work contributes to the broader field of shoplifting detection by providing insights into how deep learning techniques can enhance real-time surveillance and reduce theft in retail settings, with potential applications in other domains of anomaly detection. The YOWOv2-Medium-16-frames model gave the best performance with 54.74% frame mean average precision and 42.67% in video mean average precision.
Algeria licence plate recognition system using faster-RCNN and YOLO models
(2020) Boudissa, Mehieddine; Kissoum, Malik; Khouas, Abdelhakim (supervisor)
In some institutions, of?ce buildings, or government facilities the ?ow of incoming and outgoing traf?c of people and cars needs to be monitored and recorded for security purposes as well as practicality and automation of entry pass for vehicles. Over the last years, many techniques have been proposed in an attempt to solve the Automatic License Plate Recognition System (ALPRS) problem. These techniques rely mainly on hand-crafted approaches and basic computer vision algorithms such as edge detection with Sobel ?lter. These approaches are not accurate enough for real-world applications, nor are they robust enough to changes in size, shape, and rotation of the license plates. Recently, deep learning techniques have been shown to be a strong tool for solving computer vision and object detection problems, such as ALPRS. In this project, we propose a solution based on convolutional neural networks (CNN). A data set containing 1000 car images has been collected, labeled, and then split into a training set and testing set. The size of this data set would allow for a transfer learning approach and ?ne-tuning of models. In the next step, various models belonging to the “You Only Look Once” (YOLO) CNN and “Faster Region-based CNN” (Faster RCNN) families are trained to perform plate detection task only. Once the models are trained and optimized, they are used to crop images of plates from the original car images. These cropped images are used to train models to perform the digit recognition task, similar to those trained for plate detection. The training process was repeated for different structures and parameters of the models to obtain the best performance possible. Evaluating these models relies on the use of the mean average precision (mAP) used in the original papers of YOLO and Faster-RCNN. The evaluation of the ?nal model (plate detection and digit recognition) relies on the accuracy of performing the identi?cation of the license plate numbers. The end result is an application that achieved an accuracy of 81.36% with real-time video processing capabilities and robust to changes in size, shape, color, and rotation of the license plates. This project provides users of the application with a reliable and practical security tool. It would also supply Algerian academics and software developers with a benchmark data set for further research on the topic and evaluation of future models.
Arabic handwriting recognition using Curvelet transform and SVM
(2018) MOHAMMED TSABET, Younes; BOUMAAD, Bilal; DAAMOUCHE, A. (Supervisor)
Arabic cursive language recognition is an ever challenging problem in OCR applications. Traditional approaches to tackle this problem fail to adapt to the vast variability imposed by handwritten Arabic language, this necessitate the devising of more holistic techniques. Recent approaches to solve this challenge are making use of multidimensional analysis like wavelet and curvelet for feature extraction and then apply machine learning techniques for recognition. In this project we investigate the use of one of this approaches for feature extraction by applying Curvelet Transform to profile curvatures present in words without character segmentation mimicking the human way of recognition. The IFN/ENIT database of Tunisian towns is used and we apply SVM multi-classification for training a modal to intelligently classify those words. Results showed an accuracy of 66% though this accuracy can be elevated by following a certain train/test separation scheme.
Arabic handwriting recognition using curvelet transform and SVM
(2018) Mohammed tsabet, Younes; Boumaad, Bila; Daamouche, A.
Arabic cursive language recognition is an ever challenging problem in OCR applications. Traditional approaches to tackle this problem fail to adapt to the vast variability imposed by handwritten Arabic language, this necessitate the devising of more holistic techniques. Recent approaches to solve this challenge are making use of multidimensional analysis like wavelet and curvelet for feature extraction and then apply machine learning techniques for recognition. In this project we investigate the use of one of this approaches for feature extraction by applying Curvelet Transform to profile curvatures present in words without character segmentation mimicking the human way of recognition. The
Arabic speech recognition using recurrent neural networks
(2018) RABIAI, Zakaria; DAHIMENE, A(supervisor)
The purpose of this project is to implement an end-to-end automatic speech recognition system using recurrent neural networks, the Arabic language which ranks as the fifth most spoken language in the world has been chosen as the main language of the system. The Arabic language has been alienated from such type of projects due to its complexity, uniqueness and lack of free appropriate corpuses, but with new emerging algorithms in the domain of speech recognition such as the connectionist temporal classification, it is becoming more accessible to use unsegmented corpuses in the aim of building performant automatic speech recognition systems. The development of the project includes basic digital signal processing, exploration of the phonetic properties of the Arabic language, an adaption of a general corpus to fit the purpose of the project, feature extraction and a brief study on recurrent neural networks their performance in such a system. The full system with its various parts is implemented in Python and TensorFlow, different models inspired from literature are trained and tested using the Arabic speech corpus, leading to a selection of a final model that shows the lowest word error rate of 35.23%. The results encourage to explore more in depth the implementation of a speaker independent robust Arabic speech recognition system.
Atrial fibrillation analysis by deep learning.
(Université M’hamed Bougara de Boumerdes : Institut de Genie Electrique et Electronique, 2024) Agli, Wafa; Daamouche, Abdelhamid (Supervisor)
Atrial fibrillation (AF), an increasingly prevalent cardia carrhythmia, is a major contributor to stroke, heart failure, and premature mortality. Traditional manual screening for AF using electrocardiography (ECG) is not only time-consuming but also susceptible to human error, underscoring the urgent need for automated diagnostic tools. This study addresses this challenge by developing advanced computer-aided diagnostic methods leveraging deep learning for the automatic detection of AF. We introduce innovative one-dimensional (1D) and two-dimensional (2D) convolutional neural network (CNN) models specifically designed for the precise classification of ECG signals into normal or atrial fibrillation categories. Our methodology includes a meticulous preprocessing phase where each ECG record is filtere dan dpeak sare accurately detected using the XQRS algorithm. The signals are then segmented into beats with an 80-sample window, which serve as critical features for subsequent classification. The extracted features are fed into our CNN architectures for classification. The models are trained and evaluated using the MIT-BIH Atrial Fibrillation Database, and their generalization capability is further validated with unseen data from the PhysioNet/Computing in Cardiology Challenge 2017 database, following an inter-subject approach. To enhance the robustness of our models, we employ data augmentation techniques. Our comprehensive evaluation demonstrates that the 1D-CNN model achieves a remarkable total accuracy of 95% and an F1 score of 96.81%, while the 2D-CNN model attains an exceptional accuracy and F1 score of 99.84%. These results underscore the efficacy of our approach in accurately classifying ECG signals and highlight the potential of our models for real-world clinical applications, offering a substantial improvement in AF screening processes.
Automaion and remote management of cathodic protection of pipelines system
(Université M’hamed Bougara de Boumerdes : Institut de Genie Electrique et Electronique, 2023) Ghormrani, Amani; Belaidi, Hadjira (Superviseur)
In the realm of corrosion prevention methods, two broad categories exist: electrical and non-electrical approaches. While non-electrical methods encompass coatings, sealers, and corrosion inhibiting admixtures, cathodic protection falls under the electrical approach um- brella. Its main goal is to shift the reinforcing steel into a protected state and, consequently, prevent corrosion. This project aims to develop an automated and remotely managed cathodic protection system for pipelines. The proposed system utilizes a master code to communicate with smart voltmeters and collect voltage readings and establish a web-based dashboard for displaying the collected data. In addition, the system includes a control module that allows for the adjustment of the cathodic protection parameters remotely (probe distances, description). The proposed system is expected to improve the efficien cya ndeffectiven es sofcatho dicpro- tection systems, while reducing maintenance costs and increasing the lifespan of pipelines.
Automation and supervision for desalination station of SONELGAZ Combined Cycle power plant(Ras-Djinet)
(Université M’hamed Bougara de Boumerdes : Institut de Genie Electrique et Electronique, 2024) Benyettou, Sarah; Benmoussa, Yahia (Supervisor)
This project aims to replace the current automated system at the Ras Djinet power plant's desalination station, which employs an ABB PAC8000 (8521-HC-MT) controller. The existing controller has several drawbacks, including high cost, complexity, and challenges with upgrades, these drawbacks necessitate the exploration of alternative solutions to optimize the performance and maintainability of the desalination plant's control infrastructure. To address these issues, we propose replacing the ABB controller with a Siemens S7-300 controller that offers enhanced reliability and user-friendliness, making it a more suitable solution for the desalination station's control system. The handling of alarms and the switching between manual and automatic modes have also been taken into account. In addition, we have used the WinCC program integrated into TIA PORTAL to build a supervision system for the desalination station.
Big data compression
(2022) Boulkhiout, Mouaad; Hafri, Adel; Sadouki, Leila (Supervisor)
In order to make data storage more effective and to use up less storage space, data can be compressed. Additionally, data compression helps speed up the transmission of data exchange. Currently, a variety of techniques can be employed to data compression Moreover, the outcomes and approaches of each treatment vary. The comparison of data compression will be covered in this essay. We present a detailed analysis of Five separate algorithms, Shannon-Fano, Run-Length Encoding, the Huffman Algorithm, the LZW Algorithm, and the DELTA Algorithm. To address these issues, there is a growing need for greater data compression and communication theory research. Such study addresses the needs of fast data transfer through networks. This study focuses on deep learning analysis of the most widely used picture compression methods.
Bionic Robotic Hand controlled by myo-electric signals
(Université M’Hamed BOUGARA de Boumerdes : Institut de génie electrique et electronique (IGEE), 2020) Touait, Ayoub; Douki, Thiziri; Metidji, B. (Supervisor)
Several designs and products are currently available online as complicated limbs and hands for amputees, however many of these can be costly and heavy. This project has taken inspiration from these designs with the goal of reducing cost and weight without sacriﬁcing functionality. One major advantage of our project is ﬂex- ibility, easy to use and to wear. In this project we have developed a prosthetic hand controlled by myoelectric signals captured from the forearm muscle. This signals were ﬁltered and ampliﬁed to be easier to read by a micro-controller. The micro-controller then generated signals to control the mechanical hand based on these myoelectric input signals. This is done by the micro-controller increasing the pulse width sent to several servos making up the hand. The larger the pulse width, the more ﬂexed the hand becomes until it reaches the maximum pulse programmed into it. In addition to that, we have developed a dynamic website for the sale of the prosthesis, where the client can ﬁnd all the information related to this product.
Brain tumor classificaion using deep learning.
(2022) Berrichi, Ryad; Namane, Rachid (Supervisor)
Brain tumors are a common type of cancer that affects brain tissue. They often cause symp- toms such as headaches or seizures. They are usually diagnosed through brain scans such as magnetic resonance imaging (MRI). In recent years, computer scientists have developed algorithms that have shown promising results in automatically classifying these images into various types using deep learning models, which is a type of machine learning that uses artificial neural networks to recognize patterns in data. Publicly available MRI scans (1500 cancerous and 1500 non-cancerous) are used to train deep learning models: VGG16, VGG19, ResNet50, and Xception. Each model is implemented using three approaches, namely: implementation from scratch, transfer learning, and fine- tuning. This comparative study aims to find the best approach for training models on small datasets. The obtained overall accuracies ranged from 88% to 99%.
Building a management system for IGEE's general mean store
(2021) Bendjedia, Djamel Eddine; Zitouni, A. (Supervisor)
The project proposes technical solutions to problems related to IGEE’s general means store management. The aim of this project is to optimize the management of the store and to facilitate the work through creating an interactive web application that runs on a web server in order to simplify the work and automate the management of the store. Analysis of the current management process and tasks of the store employee is necessary for enhancing the management process of the store. The web application helps in avoiding mistakes and reducing time and efforts. It will automate tasks of the storekeeper related to products, their categories and their transaction records. This project has been carried out using the design and development processes. The design had been handled using the Unified Modeling Language (UML). For the implementation of this web application, several web development tools and skills had been used: HTML, CSS, PHP, and MariaDB.
Building a micro-processor based mobile robot using computer vision
(2016) Mezheri, Adel; Yahoui, Sofiane; Belaidi, Hadjira (Supervisor)
This project presents the concepts and techniques used for the design and implementation of an autonomous mobile robot platform able to navigate, detect and avoid obstacles in an unknown environment using computer vision. The first part of this project proposes the strategy used for the detection of obstacles using only a camera as sensor, implemented on a Raspberry microprocessor. The second part consists of the avoidance of the obstacles found by dealing with the read of digital data from the camera and processes them in order to make the right decisions to permit the robot to reach its goal. The work will be achieved with an experimental platform realization of the robot.
Building a self driving car using Ros
(2019) Aougaci, Ali; Khachouche, Lokmane; Belaidi, H. (supervisor)
This project consists of designing and building an autonomous mobile vehicle platform able to drive itself without human intervention. The appropriate hardware equipments and sensors suitable for the desired tasks and for the vehicle navigation are selected. Moreover, the car architecture is designed. Then, the Robot Operating System (ROS) is used for the software implementation, for sensors and actuators interfaces and for the approach execution. The robot Platform is a microprocessor and microcontroller based system.
Building a self driving car using Ros
(2019) Aougaci, Ali; Khachouche, Lokmane; Belaidi, H. (supervisor)
This project consists of designing and building an autonomous mobile vehicle platform able to drive itself without human intervention. The appropriate hardware equipments and sensors suitable for the desired tasks and for the vehicle navigation are selected. Moreover, the car architecture is designed. Then, the Robot Operating System (ROS) is used for the software implementation, for sensors and actuators interfaces and for the approach execution. The robot Platform is a microprocessor and microcontroller based system.
Building detection in remote sensing images using gabor filter
(Université M’Hamed BOUGARA de Boumerdes : Institut de génie electrique et electronique (IGEE), 2015) Fares, Dalila; Mallem, Ibtissem; Zemmouri, Karima; Daamouche, Abdelhamid (superviser)
buildings from remote sensing images. In this report, we propose an unsupervised method to extract building by means of Gabor filters and morphological operators. Our proposed method starts from a set of Gabor filter parameters which is selected empirically to generate a two-dimensional impulse response. The convolution of the impulse response with the image of interest enhances the buildings and attenuates the background. Moreover, morphological operators are applied to further refine the resulting class of interest. The developed algorithm is applied on many remotely sensed images; the obtained results are highly satisfactory and promising.
Cardiovascular diseases detection from phonocardiograms using deep learning
(Université M’hamed Bougara de Boumerdes : Institut de Genie Electrique et Electronique, 2023) Rahmani, Aymen Abderraouf; Boutellaa, Elhocine
Cardiovascular diseases are a significant public health concern, responsible for a high number of global deaths. Manual diagnosis of CVDs using heart sound signals requires extensive clinical expertise. In recent years, researchers have explored signal processing and machine learning techniques to automate the early detection of cardiovascular diseases from Phonocardiograms. However, the majority of these approaches depend on traditional features and classifiers, which may experience difficulties capturing the complexity of heart sounds. This study aims to develop a deep learning model capable of accurately classifying heart sounds as normal or abnormal. Making use of the publicly available PhysioNet 2016 dataset to train a hybrid CNN-LSTM model, a comprehensive comparison between different sound segmentation (windowed segments and heart cycle segments) and feature extraction techniques (Mel Spectrograms and Mel Frequency Cepstrum Coefficients) are conducted. The goal of this comparative study is to identify the optimal combination of segmentation and feature extraction methods to effectively represent heart sounds for efficient training of the adopted deep neural network architecture. We achieved an overall final score of 93% and an accuracy of 92% using the heart cycle segments and spectrogram features setting. Performance comparisons with the existing literature indicate the efficiency of this approach. This research aims to contribute to the advancement of automated CVD detection from Phonocardiograms, potentially aiding in early diagnosis and intervention.
Ceramic tiles classification using OpenCV under Embedded Linux
(2021) Bouakaz, Djihad Nacereddine; Maache, Ahmed (Supervisor)
The proposed project consists in developing a real-time system for automatic inspection of moving ceramic tiles and detection and identification of their surface defects at high-speed using OpenCV and FPGA HPS. Based on vision techniques, the system will consist of a lighting device (LEDs) and a LDR to provide the suitable lighting environment, camera, and embedded Linux running on FPGA DE10 board to create an image processing hardware and software environment. It will allow the visualization of the surface and by means of a data base, it will detect and proceed to the identification of the defects of surface during the production cycle. This process classifies ceramic tiles automatically and allows for making the necessary corrections in due time.
Chest medical image classification using deep network.
(Université M’hamed Bougara de Boumerdes : Institut de Genie Electrique et Electronique, 2023) Tis, Mohammed Amine; Daamouche, Abdelhamid (Supervisor)
Human lung which is among the most important parts in human body is facing mortal diseases especially after the COVID-19 pandemic. The scientific world is rapidly developing the health-care field to face these disorders and save millions of lives all around the world. The primary objective was to find aprecise an defficient strate gyf ort heaccura tea ndear lydetecti onand classificatio no flun gdiseases .T oachiev ethi sgoal ,w euse dth epowe ro ftwo essential medical imaging techniques: computerized tomography (CT-scan) and X-ray imaging. Additionally, we employed three deep learning models: Inception-v3, ResNet, and DenseNet, coupled with two distinct classification; binary classificatio nan dmulti-clas sclassificatio n.O urresear chjourn eystarted with binary classification ,focusin go ndistinguishin gbetwee nCOVID-1 9and non COVID-19, using both CT-scan and X-ray datasets in total of 17,599, all three models delivered outstanding results, with the highest accuracy reaching an impressive accuracy of 96%, achieved by DenseNet using CT-scan images. These results underscore the potential of deep learning in helping healthcare professionals with highly accurate disease classification .Shiftin gt oth emulti- class classificatio ndictate db yth enee dfo r amor ecomprehensiv ean drealistic approach to diagnosing and identifying a wide range of medical conditions in clinical practice and research. The new class added to COVID-19, non COVID-19 is: Community-acquired pneumonia (CAP), in total of 17,104 CT- scan images,and using the same models we challenged the system using different splitting data ratios. Through a series of experiments and evaluations, our system achieves an overall accuracy of 98% in classifying chest images across multiple categories, using DenseNet model and the 80:10:10 splitting ratio. The results showcase the significan tpotentia lo fdee plearnin gi nassisting healthcare.