PhD. of Computer Eng., University of Nottingham, England (UK)
Ass. Prof. of Computer Engineering, Dept. of Computer Eng.,
College of Engineering & Tech., AAST
Email: sherin@aast.edu, smy@syoussef.com,
1. Smart Video Surveillance and Monitoring Systems 
Sherin Youssef, Mohamed Aslan: smy@syoussef.com
Abstract: This project proposes a method to detect and track a single moving object automatically. It also provides a feasible way to classify the moving object using artificial neural network and determine whether it is a human intruder or not. The proposed smart surveillance system passes with the following phases: video capturing, video sampling, image pre-processing, moving object detection, trajectory tracking, rotation and velocity tracking, feature extraction and classification of moving object. Many experiments have been conducted to determine the efficiency and scalability of the proposed model. These experiments emphasize that proposed model can efficiently detect any intrusion accurately. There are a variety of enhancements that could be made to this system to achieve greater detection accuracy and increased robustness: Object tracking with moving camera, Multiple object tracking.
2. Image Compression and Information Hiding
Abstract: Information hiding is a new technology which integrates with theories and technologies of many academic and technical subjects. For information hiding, digital media are used as the carrier of the information to be hidden. The carrier conceals secret messages by covering the form of their existence. A method for hiding message information into media information in frequency space. The data hiding method has high resistance to removal or change of message information embedded into media information and effectively maintain hidden message information even when signal processing is performed by employing a frequency filter. More specifically, in order to hide message information (m) into media information (M), the frequency transform of the message information (m) and the media information (M) are performed, and frequency spectra f1 and f2 are obtained. Next, from the frequency spectrum f2 of the message information (m), a region containing feature frequency components representative of the features of the message information (m) in real space is extracted as the base region B. Then, n copies of the base region B are generated, and in frequency space, the n copies are dispersedly arranged.
The application and research trends for information hiding system are concerned. The information hiding technology based on digital image processing is closely related to human vision system. When the messages are having been concealed, the human eyes are due to verify the existence of hiding messages. That is, the status of information coverage depends on the human vision system. It is obvious that the characteristics of human vision system is to be taken advantage. The added secrete information in the digital image should have no any effect onto human eyes. In our research work, an implementation of information hiding technology system which is based on digital image encoding is proposed. First by analyzing knowledge of digital image processing and the model of human vision system, we discussed the algorithm of time domain appending method and the algorithm of substitution of lease significant bit. Secondly, we analyzed theory and algorithms of 2-D discrete wavelet transform and frequency domain algorithm based on discrete wavelet transformation hiding based on digital image.
3. Automated Retina Medical Diagnosis System
Sherin Youssef, Ahmed A. Aziz: smy@syoussef.com
Abstract: In this project, an Automated Medical Diagnosis System (AMDS) has been designed using a new architecture based on the integration between the wavelet packet transform and the neural network classifier model to effectively extract the feature from pre-processing medical signals for the purpose of automatic disease recognition among varying patients. An efficient feature extraction approach, based on wavelet packet entropy, is propsed and used for the design of intelligent medical diagnosis & disease discovery system. A multi-layer perception network is used for classification purpose. The project passes through the following phases. (1) Signal acquisition, (2) Signal processing and noise reduction, (3) Wavelet based feature extraction and selection, (4) Classifier design, (5) Diagnosis Decision (DD). Enormous experiments have been conducted to corroborate the efficiency of the proposed architecture. The proposed feature extraction method is proved to be robust against noise in the multimedia signal. Recognition rate was very high and showed the robustness of the proposed model. This system offer advantages in commercial and medical applications.
Abstract: In this project, an Automated Medical Diagnosis System (AMDS) has been designed using a new architecture based on the integration between the wavelet packet transform and the neural network classifier model to effectively extract the feature from pre-processing medical signals for the purpose of automatic disease recognition among varying patients. An efficient feature extraction approach, based on wavelet packet entropy, is propsed and used for the design of intelligent medical diagnosis & disease discovery system. A multi-layer perception network is used for classification purpose. The project passes through the following phases. (1) Signal acquisition, (2) Signal processing and noise reduction, (3) Wavelet based feature extraction and selection, (4) Classifier design, (5) Diagnosis Decision (DD). Enormous experiments have been conducted to corroborate the efficiency of the proposed architecture. The proposed feature extraction method is proved to be robust against noise in the multimedia signal. Recognition rate was very high and showed the robustness of the proposed model. This system offer advantages in commercial and medical applications.
4. Intelligent Access Security System Based on Iris recognition
Sherin Youssef, Yasmin El-Kazakz: smy@syoussef.com
Abstract: A wide variety of systems requires reliable personal recognition schemes to either confirm or determine the identity of an individual requesting their services. The purpose of such schemes is to ensure that only a legitimate user and not anyone else access the rendered services. Examples of such applications include secure access to buildings, computer systems, laptops, cellular phones, and ATMs. A biometric system provides automatic identification of an individual based on a unique feature or characteristic possessed by the individual. Iris recognition, a relatively new biometric technology, has great advantages, such as variability, stability and security, thus it is the most promising for high security environments. This project describes an alternate means to identify the iris recognition process by using preprocessed one-dimensional histograms. The methodology in forming these histograms, how they are used in identification. Experimental results show that the algorithm is efficient and adaptive to environment, e.g. it works well even for blurred iris image, and interference of eyelids and eyelashes.
6. Query Optimization Over Wireless Sensor Networks
Sherin Youssef, Salma Fayed: smy@syoussef.com
Abstract: Wireless sensor networks (WSNs) are one of the first real-world examples of pervasive computing, the notion that small, smart, and cheap sensing and computing devices will eventually permeate the environment. WSNs are often deployed for passive data gathering or monitoring in geographical region. Redundancy within a sensor network can be exploited to reduce the communication cost incurred in execution of such queries. Any reduction in communication cost would result in an efficient use of battery energy, which is very limited in sensors and hence maximizing its lifetime. Our objective is to design efficient mechanisms for QoS query execution in WSN. The aim is maximizing the network's lifetime by finding a near-optimal sensor query cover that is sufficient to answer a given query accurately. Moreover, it is desirable to select a near-optimal set of sensors that satisfy the conditions of coverage, connectivity, as well as minimum energy consumption and minimum communication overhead. This can be achieved by developing new competitive approaches for query coverage over WSN to solve the problem of self-organization of a sensor network and resolve many of the limitations of previous techniques. The designed techniques address the issue of optimizing the energy usage during query execution by reducing the communication cost, exploiting the redundancy especially when there are many more sensors in the network than are necessary to process a given query, and providing fault-tolerant energy conservative techniques. We introduce a new distributed approach that produces a near-Optimal Sensor Cover with minimum consumed energy and outperforms the performance of centralized approaches. Furthermore, we introduce a novel evolutionary-based approach that integrates an evolutionary-based mechanism with a decentralized query coverage algorithm for optimal query execution in self-organized WSN where more fitted connected sensor covers will be constructed subject to query imposed on the network while satisfying the coverage constraints.
7. Biometrics-based Face Recognition
Sherin Youssef, Areeg Fayed: smy@syoussef.com
Abstract: The project proposes new methods for extracting feature points from faces automatically. It provides a feasible way to locate the positions of near and far corners of eyes, midpoint of nostrils, projection of nose-area in both directions, and midpoint of mouth from face image. This approach would help to extract useful features on human face automatically and improve the accuracy of face recognition. Further mechanisms based on Wavelet Analyses are developed. The project passes through the following phases: preparing data set, image preprocessing, face localization, segmentation, feature vector extraction method, training and learning phase, and recognition phase. Experiments have been conducted to collaborate the efficiency of the proposed method. The experiments show that the method presented in this project could locate feature points from faces exactly and quickly.
8. Biometrics-based Finger Print Recognition
Sherin Youssef, Sara Mostafa: smy@syoussef.com
9. Automated Vehicle License Plate Allocation and Recognition System
Sherin Youssef, Shaza Basiony: smy@syoussef.com
Abstract: Nowadays license plate recognition has now become a key technique to many automated systems such as road traffic monitoring, automated payment of tolls on high ways or bridges, security access, and parking lots access control. Most of the previous License Plate Locating (LPL) approaches are not robust in case of low-quality images. Some difficulties result from illumination variance, noise, complex and dirty backgrounds. This report presents a real time and robust method for license plate location and recognition. Edge features of the car image are very important, and edge density and background color can be used to successfully detect a number plate location according to the characteristics of the number plate. The proposed algorithm can efficiently determine and adjust the plate rotation in skewed images. LP quantization and equalization have been applied as important steps for successful decryption of the LP. The program finds the optimal adaptive threshold corresponding to the intensity image obtained after adjusting the image intensity values. An efficient character segmentation algorithm is used in order to segment the characters in the binary license plate image. An Optical Character Recognition (OCR) Engine has then been proposed. The OCR Engine includes digit dilation, contours adjustment and resizing. Each digit is resized to standard dimensions according to a neural network dataset. The Back-Propagation Neural Network (BPNN) is selected as a powerful tool to perform the recognition process. Experiments have been conducted to corroborate the efficiency of the proposed method. Experimental results showed that the proposed method has excellent performance even in case of low-quality images or images exhibiting illumination effects and noise. Experimental results illustrate the great robustness and efficiency of this method.
10. Automated Speech Recognition
Sherin Youssef, Sohaila Kamal: smy@syoussef.com
Automated Speech Recognition System (ASRS) have been designed using new architectures based on the integration between the wavelet packet transform and the neural network classifier model to effectively extract the feature from pre processing real speech/voice signals for the purpose of automatic speech recognition among varying noisy. An efficient feature extraction approach, based on wavelet packet entropy, is proposed and used for the design of intelligent speech system. A multi-layer perception network is used for classification purpose. The project passes through the following phases: first; Signal acquisition, second; Signal processing, third; Wavelet based feature extraction and selection, fourth; Classifier design, and finally; Decision with different percentage of noise for signal. Enormous experiments have been conducted to corroborate the efficiency of the proposed architecture. The proposed feature extraction method is proved to be robust against noise in the multimedia signal. Recognition rate was very high and showed the robustness of the proposed model; that was about 80%. This system offer advantages in commercial and life applications.
11. A Spy Navigating Smart Robot
Wireless enabled Spy Robot with full audio and video streaming capabilities, accessible from anywhere in the world you have a Web connection and a Web enabled device like a PC, or cell phone. The robot is used in conjunction with software that makes designing navigation way points easy.
