Mansoura journal for computer and information sciences /
[9002998.] رقم البحث : 9002998 -
Video Analysis For Human Action Recognition Using Deep Convolutional Neural Networks /
تخصص البحث :
Mansoura journal for computer and information sciences / / Vol.14 - No.1
تاريخ تقديم البحث 07/07/2019
تاريخ قبول البحث 07/07/2019
عدد صفحات البحث 7
Nehal N.Mostafa( nihalnabil1990@ gmail.com - ) - مؤلف رئيسي
Mohammed F. Alrahmawy( mrahmawy@mans.edu.eg - )
Omaima Nomair( omnomir@yahoo.com - )
Video analysis, Human action recognition, Deep learning, machine learning, video analysis, Image segmentation, feature extraction, Kalman filter, human action, Convolutional Neural Network.
In the last few years, human action recognition potential applications have been studied in many fields such as robotics, human computer interaction, and video surveillance systems and it has been evaluated as an active research area. This paper presents a recognition system using deep learning to recognize and identify human actions from video input. The proposed system has been fine-tuned by partial training and dropout of the classification layer of Alexnet and replacing it by another one that use SVM. The performance of the network is boosted by using key frames that were extracted via applying Kalman filter during dataset augmentation. The proposed system resulted in oromising performance compared to the state of the art approaches. The classification accuracy reached 92.35%.
[9002999.] رقم البحث : 9002999 -
Arabic characters descriptors for lexicon reduction in Arabic handwriting /
تخصص البحث :
Mansoura journal for computer and information sciences / / Vol.14 - No.1
تاريخ تقديم البحث 07/07/2019
تاريخ قبول البحث 07/07/2019
عدد صفحات البحث 7
Nada Essa( nadaessa2012@gmail.com - ) - مؤلف رئيسي
This paper introduces an advanced Arabic handwriting recognition technique using lexicon reduction. The lexicon reduction technique stands on extracting the Arabic character shape descriptors. The technique implementation consists of two major stages. The first stage presents a method for extracting the shape descriptor of each character. The second stage suggests Aho-Corasik string searching algorithm for Arabic character recognition. Various stages have been evaluated on the IFN/ENIT database. The results demonstrate the efficiency of the suggested technique.
[9003000.] رقم البحث : 9003000 -
Swarm Intelligence based Fault-Tolerant Real-Time Cloud Scheduler /
تخصص البحث :
Mansoura journal for computer and information sciences / / Vol.14 - No.1
تاريخ تقديم البحث 07/07/2019
تاريخ قبول البحث 07/07/2019
عدد صفحات البحث 10
A. S. Abohamama( abohamama@mans.edu.eg - ) - مؤلف رئيسي
M. F. Alrahmawy( mrahmawy@mans.edu.eg - )
Mohamed A. Elsoud( moh_soud@mans.edu.sa - )
Taher T. Hamza( Taher_hamza@yahoo.com - )
Real-Time Applications, Ant Colony Optimization, Fault Tolerance, Cloud Scheduling
Cloud computing is a distributed computing paradigm that is deployed in many real-life applications. Many of these applications are real-time such as scientific computing, financial transactions, etc. Therefore, improving the dependability of cloud environments is extremely important to fulfill the reliability and availability requirements of different applications, especially real-time applications. Fault tolerance is the most common approach for improving the system’s dependability. In addition to traditional fault tolerance techniques such as replication, job migration, software rejuvenation, etc, fault-tolerant scheduling algorithms can play a great role toward more dependable systems. In this paper, an ACO based fault-tolerant soft real-time cloud scheduler is developed to minimize deadlines missing rate, makespan, and the imbalance in distributing the workload among the different machines. The performance of proposed scheduler has been assessed under different scenarios. Also, it has been compared to other well-known scheduling algorithms and the experimental results have shown the superiority of the proposed algorithm
[9003001.] رقم البحث : 9003001 -
ETL Semantic Model for Big Data Aggregation, Integration, and Representation /
تخصص البحث :
Mansoura journal for computer and information sciences / / Vol.14 - No.1
تاريخ تقديم البحث 07/07/2019
تاريخ قبول البحث 07/07/2019
عدد صفحات البحث 10
Abeer Saber( abeer_saber@fci.kfs.edu.eg - ) - مؤلف رئيسي
Semantic web introduces new benefits for many research topics on big-data. It semantically maintains a large amount of data and provides meaningful meaning of unstructured data contents. Big data refers to large scale. It is used to describe a massive collection of datasets in different formats. The semantic and structural heterogeneity are the biggest problems that still face the aggregating, integrating, and storing big data. In this paper, we solved both of the problems of columns redundancy that are produced from the semantic heterogeneity and the problem of structural heterogeneity through developing and implementing a new ETL model based on semantic and ontology technologies. Geospatial data is used as a case study because its integration is complex and usually suffers from the variety of resources and the representation of the produced big data. The results of using this model showed that it solves the problem of heterogeneity in several data sources and it improves the data integration and representation.
[9003002.] رقم البحث : 9003002 -
Representation Learning Framework of Object Recognition via Feature Construction /
تخصص البحث :
Mansoura journal for computer and information sciences / / Vol.14 - No.1
تاريخ تقديم البحث 07/07/2019
تاريخ قبول البحث 07/07/2019
عدد صفحات البحث 6
Muhammad H. Zayyan( mhaggag@mans.edu.eg - ) - مؤلف رئيسي
Samir Elmougy( mougy@mans.edu.eg - )
Mohammed F. AlRahmawy( mrahmawy@mans.edu.eg - )
Object recognition, ECO features, Adaboost, Random Forest, Pooling, Genetic Algorithm.
In this paper, we recognize objects within images by collecting information from a large number of random-size patches of the image. The different backgrounds accompany the foreground object demand to have a learning system to identify each patch as belonging to the object category or to the background category. We strengthen a recent method called Evolution-COnstructed (ECO), which is based on the ensemble learning approach which combines several weak classifier. The improvement is relying on decreasing the overfitting problem. Two different improving ideas are proposed: 1) Pooling operation, which is applied to the weak classifiers data, 2) Random Forest algorithm, which combines the weak classifiers outcomes. Experimental results are reported for classification of 9 categories of Caltech-101 data sets and proved that our modifications boost the performance over the base method and other existing methods.
[9003009.] رقم البحث : 9003009 -
Wavelet-based Video Enhancement Using Contrast Limited Adaptive Histogram Equalization /
تخصص البحث :
Mansoura journal for computer and information sciences / / Vol.14 - No.1
تاريخ تقديم البحث 10/07/2019
تاريخ قبول البحث 10/07/2019
عدد صفحات البحث 10
Fatma Faek( tomafayk@gmail.com - ) - مؤلف رئيسي
Omaima Nomir( omnomir@yahoo.com - )
Ehab Essa( ehab.essa@gmail.com - )
Ebrahim El-henawy( henawy2000@yahoo.com - )
Video enhancement, Wavelet transforms, Interpolation, histogram equalization, CLAHE.
A combination of wavelet transforms and contrast limited adaptive histogram equalization (CLAHE) techniques are used to efficiently enhance videos. The proposed technique handles the noises within video frames and enhances the resolution of the video. Lifting wavelet transform (LWT) and stationary wavelet transform (SWT) are applied to separate original frame into the low-frequency sub-bands, and the high-frequency sub-bands.Thereafter, we applied the interpolation to correct the coefficients of the high-frequency and the original frame separately. Next, Inverse Lifting wavelet transform (ILWT) is utilized for the integration of each all these enhanced sub-band. Finally, the CLAHEalgorithm is applied to make the details of the frame more visible, and meaningful to generally improve the resolution of the video. The output video shows that the proposed technique enhances the quality of resolution videos under various environmental conditions, alleviates noises and avoids the over-enhancement problems.
[9003010.] رقم البحث : 9003010 -
A Multimodal Wireless System for Instant Quizzing and Feedback /
تخصص البحث :
Mansoura journal for computer and information sciences / / Vol.14 - No.1
تاريخ تقديم البحث 10/07/2019
تاريخ قبول البحث 10/07/2019
عدد صفحات البحث 7
Khaled Mohammed( khaled.alburaihi@gmail.com - ) - مؤلف رئيسي
A. S. Tolba( ast@astolba.com - )
Mohammed Elmogy( melmogy@mans.edu.eg - )
Educational Platform, Student Attendance Management, Quiz Management System, Radio Frequency Identification (RFID), Face Verification, Students Alert System
This paper presents a wireless system for instant quizzing in the classroom and collecting students’ feedback on teachers performance. This system is integrated with a student attendance management system to facilitate management of quizzing and quiz marking in addition to questionnaires about Quizzes. Such a system is very essential for following attendance and student learning progress in addition to formative assessment. The system uses two communication technologies: Wi-Fi, and Radio Frequency Identification (RFID). Such a low-cost system assures attendance follow up to assure abiding by the university bylaws, avoid spoofing and cheating, and enhance both teaching and learning. A student recommendation system is also implemented to increase student retention and enhance students success rate
جميع البيانات التي يتم تسجيلها على النظام هي ملك للجهة صاحبة هذه البيانات
لا يتم الاطلاع على هذه البيانات من قبل مهندسي المركز إلا للضرورة القصوى ولغرض الدعم الفني وبالاتفاق مع الجهة صاحبة البيانات شفوياً أو كتابياً
لا يتم إعطاء أي بيان من بيانات النظام لأي فرد او جهة فرعية تنتمي للجهة صاحبة البيانات او لأي جهة أخرى إلا بموافقة كتابية من الجهة صاحبة البيانات
يمكن استخدام البيانات المسجلة على النظام بهدف عمل إحصائيات يستفاد منها فى قياس اداء النظام وكفاءة استخدامه من قبل مستخدميه
يتم تأمين جميع البيانات الخاصة بالنظام من خلال تأمين الحاسب الخادم الذي يستضيف البيانات ومن خلال تأمين انتقال البيانات بين جهاز المستخدم والحاسب الخادم وكذلك تأمين واجهة النظام التي يستخدمها المستخدم
كل مستخدم للنظام مسئول مسئولية كاملة عن كلمة المرور الخاصة به وعليه اتخاذ كافة الاحتياطات الممكنة حتى لا يستطيع أى شخص آخر الحصول عليها
تسجيل دخول 
