Faculty Of Computer Science Graduation Project 2019 - 2020

Permanent URI for this collectionhttp://185.252.233.37:4000/handle/123456789/3761

Browse

Now showing 1 - 20 of 63

3D Reconstruction of Multi-View Flat Sketches
(MSA university Faculty of Computer Science, 2020) Hussein, Ahmed Fayez
We propose and implement a system for 3D reconstruction of flat sketches. Our system takes two sketches of two views of the shape as an input and outputs 12 corresponding multi view images with normal and depths maps. Architecture of the system is composed of two components which are Generator and Discriminator. Generator is responsible for generating the 12 images from the sketches while discriminator’s objective is to fake images generated from generator and real images which are the targets. Generator succeed when it manages to fool discriminator into classifying generated images as real ones. Generating images is accomplished through two steps: (i) encoder which convert input sketches into a feature map representation, and (ii) decoder which map this feature representation into images. Discriminator is a linear classification convolutional network which classifies images as real or fake. Our system produced nearly perfect multi-view images from detailed and undetailed synthetic sketches, and from hand-drawn sketches.
3D shoes model angle tracing using deep learning
(MSA university Faculty of Computer Science, 2020) emad abbas taghayan, khaled
In this days online shopping is playing a big role in our world as it is more easy to buy things online with using any device the connects to the internet and in any place without spending a lot of time in moving from one place to another also you can buy thing online from any place around the world but one major problem in the online shopping is testing the clothes before you buy how can this be possible with using only your phone , that’s what our project was aiming for to help the process of testing the cloth which in our case we tried testing shoes in your place only by using a camera we were able to develop a two deep learning model the first one is able to detect if there was a shoes in an image or not and the other model is able to trace the shoes angle from an image the study included a data of 300 images was collecting by taking photos of different shoes types from different angles and different leg angles also in order to use It in a deep learning we had to do some augmentation on the data to make it larger so we reached the number of 1500 image then the data was divided to train and test , the CNN model was trained to learn the different angles of the shoes image to have the best accuracy , the trained model was evaluated by getting the angles that the model outputs and put it on a unity program to that tests the angle on a 3D shoes model to see if it rotates the same angle as the image or not , the results we got for the final model showed that the model can trace the shoes angle from an image with accuracy equals to 75.3% which is a good result considering the number od data that we got the model can help in the future to be connected to a AR application to trace the angles automatically
An app to aid blind people in Kitchen
(MSA university Faculty of Computer Science, 2020) Zaher, Abdallah Mohamed
Recognition is one among the most areas in computer-vision, it yields high level understanding by computers, one among the foremost necessary areas in recognition is seeing that is that the method of finding a particular object in a picture or video sequence. During this paper we tend to gift Associate in application that employs a neighborhood of seeing, this application is in a position completely between different varieties of fruits/vegetables. This application relies on color and size through examination image histograms to search out the most effective matching image, Associate as a result the applying shows an accuracy of 95+% of distinguishing fruits and vegetables.
Augmented reality for shooting clay target
(October University for Modern Sciences and Arts, 2020) Mahmoud, Abdelrahman Mohamed Mohamed
The main idea of this project is to help the beginners and intermediate players to improve levels of shooting by detecting and tracking target and nozzle of shotgun then by detecting target and nozzle of shotgun i can identify the location of target and shotgun then calculating distance between nozzle of shotgun and target to take a suitable decision to launch the bullet and break the target at a real-time and reduce ﬁnancial losses to shooting clubs because it reduce wasting bullets and targets.
Augmented Reality For The Future Of Dentistry
(MSA university Faculty of Computer Science, 2020) Ayman Gerges, Alber
Most people find worn, broken or discolored teeth unattractive, and many of those affected opt for cosmetic dental reconstruction using crowns or ceramic veneers, for example. However, any change to the front teeth can also have a major effect on a person’s facial expressions. To allow patients to see how their new smile will look, it is common practice nowadays to first make a plaster cast of the patient’s dentition. The dental technician then uses wax to insert any missing teeth and creates a 3D model which the patient can try out a complicated and time-consuming procedure. So, some applications using augmented reality to help dentists with their hard work and motivates their patients to see their new smile in live video by detecting the patient’s mouth and replace the nature teeth with the augmented teeth.
Auto analysis for sport trends in social media
(MSA university Faculty of Computer Science, 2020) Khaled Megahed Akl, Ahmed
An auto analysis of sport trends in social media is the goal of this project. We know that social media has many debates and discussions which some peoples do not able to read all of them, many companies want to know the reaction of people and there is opinion in their products. Many of companies want to know it about their services. So our system make it easy to them. Social media platform which we could contact them and get permission to be able to get data from it is twitter. We will illustrate our project and how can we solve it and problems which we faced. The end of all them we could achieve our goal and get data from twitter to make analysis of them and know what percentage of agreement from people on specified trend is. So let’s go and begin from chapter 1.
Blood Donation Management mobile Application
(MSA university Faculty of Computer Science, 2020) Bassem, Ahmed
This program seeks to create a management framework for blood banks, which tries to build a strong link among the two groups of people who donate blood and blood type, and insurance without a disease, and the accepter, who need blood to use the blood. The purpose of the application is to create a blood bank network. And to provide a good analysis of accurate statistics for blood transformation in Egypt to increase number of donors to meet the need of the state of blood. This system helps to manage records that are saved in the stored database of the donor requests for the application, provides a good system of Blood Bank management to enable quick access to the collected data or information, enables good contacts by location identification and telephone numbers.
Building a web-based chatbot
(MSA university Faculty of Computer Science, 2020) Mahmoud, Menna talla el-dessouky
The main idea of this project is to study the requirements needed to design and implement an efficient self-service chatbot. The development process will be including AI technology, machine learning and Natural language processing. The main objective of this project is to enhance the communication methods between users of the chatbot and the business adapting the chatbot technology as an advanced and efficient method of communication. Moreover, the project aims to enhance the accuracy and the time spent completing tasks thus improving efficiency and productivity of workflow.
Building-Block using Genetic Algorithm
(October University for Modern Sciences and Arts, 2020) Ezzat, Mustafa Ihab
3D modeling using blocks have a lot of potentials as an interaction method that can enrich the interaction between designers or young people and modeling. This system uses one of the world’s most widespread toys around the world to make it easier and fun for modeling in 3D structures and widen the imagination and creativity of users since the lack of easy 3D modeling software. The system create a knowledge to be able to develop a creative and complex well-structured models represented using genetic algorithm that uses randomizing techniques similar to the biological process to enhance the results.
Car Parts Image Classification and search engine for CDCM Company
(MSA university Faculty of Computer Science, 2020) Botros, Beman Emad Ramzy
This article is talking about the problem of spare parts in Peugeot, Citroën, and DS. This article solves this problem by merge spare parts of the three cars like make the turbo of Peugeot combine it in Citroën and DS and so on. The system made that by detecting the old spare part using an android mobile camera and get the new one from another car. So the system used image classification and machine learning by python, open CV, and tensor flow by making a new dataset for the different spare parts to classify it with high accuracy and the system to get the nearest location for the shops of spare parts. And the system adds a pdf for each spare part to help the user to know how to tie and untie each spare part.
Car plates detection and recognition
(MSA university Faculty of Computer Science, 2020) Magdy Mohamed, Mohamed
License plate recognition (LPR) has been A dilemma as it causes traffic jam and hard for employees to recognize due to high car numbers. The idea of this model is to notice high violation without affecting traffic or be a dilemma on employees to find the car as the project recognizes violated car’s model-color and name, which compared it to the governmental information which will be a revolution in maintain traffic in the world.
Chest Diseases Detection Using Convolutional Neural Networks
(MSA university Faculty of Computer Science, 2020) Amr Attia Ewais, Abdelrahman
X-ray pictures are utilized in medication to photo the inner structure of the human body such as bones, skulls, teeth, and chest …etc. [1]. Today, we are facing huge chest diseases that may affect the body and lead to the death of humans. Chest diseases are a very popular and dangerous problem. And it is a very earnest health problem so we must be stricter about diseases of that type because it may lead to death [12]. So it very important to save time in chest cases because time is very important do chest patient may feel he/she cannot take a breath so doctors need to diagnose the patient’s case to control the disease as fast as can. Nowadays a lot of researches has been working how to detect chest diseases by AI methodologies that is help in a lot of different fields [1]. As you can see below that using also X-ray images to applying CNN algorithms that make to detect diseases so x-ray has a very important role in detect diseases generally. Now can detect diseases by the power of computation resources. And now they reach software that takes a photo to x-ray and then applying an algorithm on this picture to detect disease
Cityfy: Big Data Solutions for a Smarter Cairo
(MSA university Faculty of Computer Science, 2020) Ahmed Abdel Khalik, Amr
In response to the ever changing and growing technological aspects of the world. I present my project offer of a software-intensive Smart Cit. The target being Egypt and its citizens. Cityfy is a Big data solution center that contains components that would improve the wellbeing of the users, and provide insight that could lead to newer stages of a greener city.
Deep-learning based trucks violation detection
(October University for Modern Sciences and Arts, 2020) Kiwan, Mohamed Gamal el din mohamed
This project builds a truck detecting model for automatically supporting the traffic department Decide if the truck is overloaded or Normal-loaded to help the traffic department in controlling of high-way roads. We build the model based on convolutional neural network model. The dataset of the truck is constructed and hyper parameters modified of the convolutional neural network. A basic network model has successfully been trained by KERAS library. The model by KERAS library achieves 91.67% on overload / normal-overload truck classification, which isn't a bad result. we dived deeply in the model by changing the model to work by TENSORFLOW library. we optimized the model by TENSORFLOW library, the optimized TENSORFLOW model achieved 96% on the test set, that's better than that 91.67% of the KERAS model.
dental cart application
(MSA university Faculty of Computer Science, 2020) Nasef, Yousef
Dental tools are a vital part of dentistry. To clean, remove, repair and eradicate carriages in teeth, dental practitioners including dentists, hygienists, and dental assistants use different kinds of instruments. Many dental instruments are referred to by their use and others by their technical name. For general dentistry, there are three main classes of dental instruments used: general instruments used in a number of procedures, extraction instruments, and instruments used for teeth restoration.
Detecting And Classifying Diabetic Retinopathy Using Deep Learning
(2020) Waheed Mohamed Sabry, Islam
Diabetic Retinopathy (DR) is human eye disease among people with diabetics which causes damage to retina of eye and may eventually lead to complete blindness. Detection of diabetic retinopathy in early stage is essential to avoid complete blindness. Effective treatments for DR are available though it requires early diagnosis and the continuous monitoring of diabetic patients. Also, many physical tests like visual acuity test, pupil dilation, and optical coherence tomography can be used to detect diabetic retinopathy but are time consuming. The objective of our thesis is to give decision about the presence of diabetic retinopathy by applying ensemble of machine learning classifying algorithms on features extracted from output of different retinal image. It will give us accuracy of which algorithm will be suitable and more accurate for prediction of the disease. Decision making for predicting the presence of diabetic retinopathy is performed using Deep Learning
Detection of HCV Based on Machine Learning
(MSA university Faculty of Computer Science, 2020) Mahmoud Mohamed, Mohamed
The Project uses data analysis to early detect the presence of HCV type and the methodologies of treatment, The project will be helping the medical workers in Egypt to have a comprehensive study with realistic data to predicting the infected patients of HCV in Egypt depending on their medical history and their geographical data as well . Nowadays, Data analysis are widely using in the world for both scientific and trade wises to detect best results for future visions Moreover, Earlier Predicating for (Possible to be infected) cases according to the past pathological records. This Thesis , Might be helping the continuous efforts for the Egyptian government journey against HCV in Egypt which started with 100 Million Seha Campaign and still running in all governmental hospitals Viruses and Infection centers.
A dissertation submitted in partial fulfillment of the requirements for the degree of Bachelor of computer science
(MSA university Faculty of Computer Science, 2020) Rafaat Mohamed Abdelrahman Kishk, Abdelrahman
Farmers have a problem in detecting plant diseases from the beginning but according to their work ,they could detect the disease after a certain time .This problem will lead to throwing the farms ‘products away and with that I decided to solve this problem using deep learning to help farmers to detect the diseases of plants in any level in order to recover its infection easily .The project will be implemented using deep learning and we will work on spyder 3.7 to train the plant using its leaves .The results that will be occurred in training and testing accuracies are more than 50%
Face Aging Using Generative Adversarial network
(MSA university Faculty of Computer Science, 2020) Ahmed, Ahmed Zakaria Abdelrahim Abdelmalik
If we provide someone with a large amount of faces in different ages which is not paired and asked him if he can generate a photo of his face at older or younger age his answer would be probably no. Providing someone with an older or younger face of him is a complex task. Nowadays most face aging models are required paired sample as the dataset contains someone face and an older face of the same person to learn the traversing between the two age. Collecting a paired samples dataset is a complex task to achieve and have an availability limitation. In this paper, the model did not require a paired sample it only require a dataset for every group of age from 0 – 100 (group for each ten years). The model learns the face details between each group of age. In addition, that the model not only generate an age progression and regression faces, but it also saves the personality of the face during the aging process. The face personality is saved during aging according to the use of encoder which down sample the image to save the high-level face feature. The model structure consists of encoder and generator to generate progression and regression faces. Also, it consists of two adversarial networks respectively to force generating real faces.
Fake watermark detection using deep learning
(October University for Modern Sciences and Arts, 2020) Anany, Sheriff Mahmoud
Water mark recognition contained many theories on how to be detected, historical water mark was only done on small scale as there were no public dataset, and the only model was done only comparing historical water mark according to type (Egyptian, Greek … etc.). the problem in historical water mark was that it can’t be recognized by naked eye like other water mark as it contains noise that cover the water mark and some colors are removed which has been dilemma for archeologist. This model used large dataset for historical watermark with the comparison of real and fake ones which hasn’t been made before. This make archeologist work easy and prevent any fake historical watermark to be involved in history.

Browse

Browsing Faculty Of Computer Science Graduation Project 2019 - 2020 by Title