Home

2015-2016

Main block

Project Title: Home Facial Control Using Facial Recognition and Cloud Services

Group Members: Divine Yuekambe Bawa     

Supervisor: Dr. Robert A. Sowah

Abstract:

Access control refers to any security technique that can be used to regulate who can view, use or access a resource thus selectively restricting the availability of place. Therefore the any such technique used to access the home environment is a Home Access Control technique. Most of the methods currently used in accessing our home environment are physical methods thus requiring the use of typical metallic keys, keycards amongst others. These methods have a few problems where these physical means of access like typical metallic keys and keycards can get misplaced denying legitimate home users access to their homes. Also theses access means can be stolen and used by illegitimate users to access the homes.

In order to overcome this problems, an access control system that eliminates the use of these physical or mechanical means of access whilst ensuring legitimate users are granted access will be designed by incorporating biometric verification of user identity.

This project’s intention is to use facial recognition biometric as an access method for the home environment for authenticating home users, incorporating a phone application to remotely view unauthenticated users and allowing users to remotely grant access to unauthenticated users form the phone application in real-time using cloud-based Facial Recognition API, a microcontroller and web services.

 

 

 

Project Title: Implementation of Mobile Communication Services Using UG Wi-Fi Network

Group Members: Emmanuel Bamfo Okyere & David Komlah     

Supervisor: Dr. Godfrey A. Mills

Abstract:

There are millions of calls made every minute using infrastructure set up by telecommunication companies. The telecommunication industry has undergone many changes from traditional landlines to the advent of mobile phones. The internet has also become an integral part of daily life. Most institutions and companies provide internet access on their premises to their students and employees. The University of Ghana has internet access points at various locations on campus enabling students and staff to surf the internet on their laptops and mobile phones. Students and employees on campus buy airtime to make calls to other students and employees on campus every day. However, the gigabit backbone of the university’s network is under-utilized and most calls made by students and staff are inbound in other words to other students and staff also on campus. The “UG WIFI CALL” app harnesses the university’s intranet to allow students and staff to call each other using the Wi-Fi networks available on campus at no additional cost easing the financial strain on them and improving productivity. Calls made using the app are as audible as those made using traditional phone calls and is therefore a viable communication solution on campus. Before implementation a heat mapping exercise was conducted to identify the network coverage areas where signals are strong and weak. Second, a virtual asterisk PBX server was setup for call switching and routing. Again, an application software was designed, developed and implemented to support voice over the UG intranet infrastructure. Lastly, a unique UG phone number was generated for easy tracking, to create a personalize phonebook and for information broadcasting purposes. The developed mobile communication service was tested at two separate locations. Where users place calls and the quality of voice was good by a voice input and output analyzer. The mobile application was good and a viable option to make inbound calls on campus at a relatively low cost.  The Wi-Fi mobile communication platform built also provides other services which are targeted message broadcasting and a phone directory or phonebook. The system is flexible such that it can be extended to provide other services such as video conferencing and live lecture streaming. The project goes a long way to increasing productivity and encouraging collaboration at the same time reducing the cost of living on campus. 

 

 

 

Project Title: Wireless Local Area Network (WLAN) Server for Smart Home Analysis and Storage

Group Member: Kwabena Manu     

Supervisor: Dr. Isaac Kwadwo Nti

Abstract:

The home environment has been considered the core of human settlements. Certain technologies have been integrated into the home environment which allow it to be aware of the context in which the inhabitant use it. These smart homes provide a valuable avenue to provide context aware services to people in their homes, services which have been shown can improve the level of comfort experienced in the home or even provide automated emergency health assistance. Thus far, most smart home efforts which employ advanced computing capabilities have been research oriented in nature, requiring expensive and unwieldy setup efforts, for monitoring and computing data, and are inapplicable to general home environments. This work sought to create a solution which could enable the creation of a cost effective system for the storage and analysis of smart home data, that would be directly applicable to use in a normal smart home environment.

The solution involved programming a Raspberry Pi computer system in order to be a hub over which generic smart home data could be collected, stored and processed in the home. The smart home data under consideration was binary sensor data. Real data was collected, and unsupervised data analysis algorithms were applied to the data. The results of this analysis were displayed in a web application over the wireless local area network on which the Raspberry Pi was a part to users on the network.

It was demonstrated that the computer system was sufficiently capable of performing the computational analysis required, as well as serving the web application to users over the WLAN, with no need for extraneous internet or hardware support. This showed that a cost effective solution to developing a context aware smart home system is possible. This paves the way for the application of context aware smart home technologies to generic home environments using simple binary sensors.

 

 

 

Project Title: A Framework for Establishing A Node Linked Open Data for Ghana Data

Group Member: Neizer Melvin     

Supervisor: Dr. Wiafe Owusu-Banahene

Abstract:

The Web is known to be a global information space consisting not just of linked documents, but also of Linked Data. In recent times, the web of data has been brought into being by the maturing of the Semantic Web technology stack, and by the publication of an increasing number of datasets according to the principles of Linked Data. The semantic web began with a community called the Linked Open Data (LOD) cloud, which links data which are all structured in the required form called Resource Description Framework (RDF). A lot of countries have got their data published on the cloud and there is no node from which data from Ghana can be published on the cloud. This project applied the Linked data principles into creating a framework for publishing structured data from open data from Ghana into the LOD cloud. It involved creating a web application that converts the traditional csv files into RDF. The RDF files would be stored in an open source virtuoso server to be queried with Simple Protocol and RDF Query Language

(SPARQL), to ensure that they are in triple form. A dataset is created in datahub.io with the RDF files which are later collected and published on the LOD cloud. This project makes data from

Ghana more accessible to applications, help developers integrate data into their applications and by following links in the dataset, discover more useful data linked to data from Ghana.

Keywords: Framework, Semantic web, Resource Description Framework ( RDF), Simple Protocol and RDF Query Language (SPARQL), Linked Data, Linked Open Data (LOD) Cloud, Virtuoso server

 

 

 

Project Title: Intelligence Attendance System Using Facial Recognition

Group Member: Osumanu Ra-Is Silbaway     

Supervisor: Dr. Robert A. Sowah

Abstract:

Intelligent Attendance System using Facial Recognition is an efficient and electronic application which automatically marks the attendance of students in class without any human intervention. More than an electronic device, the system has a fully integrated web application and a mobile application, including a windows 8.1 desktop application integrated with it. It helps each department to manage the students efficiently by knowing the attendees and absentees in class.

The purpose of the project was to create an electronic attendance system to check the attendance of students in a classroom. The system has a hardware part that does the facial recognition, web application, mobile application and windows 8.1 desktop application software. This will help ease the activities of the lecturers who normally pass papers around to check attendance and manually have to go through the attendance list to check for absentees, as they hope to achieve their goal of ensuring a better lecture room management. C++ with OpenCV was used for the face detection and recognition, Ionic framework and Angular JS was used for the mobile application and desktop application. PHP, JavaScript, HTML, CSS were used for the web application.

The hardware part of the system was developed using raspberry pi and the raspberry pi camera module. The web application, desktop application and mobile application was developed to be able to communicate with the raspberry pi. The raspberry pi was also used as the server which hosted the web application, the face database and the attendance database.

Also the system enables lecturers to download the attendance list in excel format. Two mobile applications were developed, one for lecturers and the other for the students. The lectures’ mobile application has the functionalities of knowing the class population, fetching for the attendance records and controlling the raspberry pi to start marking the attendance of the students. The students’ mobile application has the functionalities of students knowing the class population and getting information about the attendance of the class.

 

 

 

Project Title: Utility Fault Login and Fault Management System

Group Member: Eugene Taabazuing

Supervisor: Dr. Robert A. Sowah

Abstract:

Population in cities are increasing exponentially. Due to the increase in population, pressure on our utilities and services provided by these utility providing companies are tremendous. Hence, the need to develop a system for easy utility fault reporting and efficient utility fault management to ensure to ensure effective management of utility services

A good interaction between utility provider companies and citizens in a country is vital in modern smart cities. Computer vision and semantic web technologies can aid in achieving such a goal. A smart utility fault reporting mobile and web platform is presented, to aid citizens in easily reporting common faults in our utilities, such as street faults, faulty electric poles or broken water pipes, and to support the utility providing company in responding and fixing those problems quickly. The tool is based on object detection and recognition algorithms to identify the type of utility fault and a semantic data model designed for the city, which integrates several distinct data sources, opportunely re-engineered to meet the principles of the Semantic Web and linked open data. The platform supports the whole process of road maintenance, water maintenance and electricity maintenance, right from the fault reporting to the management of maintenance activities. The integration of multiple data sources enables increasing heterogeneous and interoperability of information retrieval, thus promoting the development of effective utility fault reporting services.

 

 

 

Project Title: Information Broadcasting System

Group Member: Danso Kwaku Osei

Supervisor: Dr. Robert A. Sowah

Abstract:

Changing notices manually on a daily basis expends a lot of energy and financial cost. The purpose of this project is to devise a means of making this process simple, less time-consuming and less costly in terms of money for usage. In order to achieve this purpose, a system is designed to send by SMS, email and audio broadcast. Only authorized users of any institution or groups will able to send messages from their android application devices to the system. With this automated system, users need not go around posting notices on the notice board and removing them when they are old. Users can be also be anywhere and still send important messages to people.

 

 

 

Project Title: A Web and Mobile Application for Navigation and Notification on the University of Ghana Campus

Group Member: Awura Adwoa Manko Amihere

Supervisor: Dr. Godfrey A. Mills

Abstract:

Getting directions from one place to another can be very exhausting. In the process of finding a venue, one may end up walking all around campus and getting lost a couple of times before arriving at their destination, or even worse, may arrive late at their venue. This is a common problem in University of Ghana campus especially for first year students who need help with locating venues; where their lectures would be held, where their exams and tests would be taken, and where they would have orientation, etc. Not knowing a place beforehand can cause waste of time. In this project, a more suitable way to locate venues using a web and mobile application was developed. This project aims at developing a platform that would enable a system administrator to make modifications to the data from the back-end by allowing for the addition of categories, venues and events that would be happening on campus. This would save the time of the users by showing routes mapping on the hand-held mobile device of the user and displays information concerning events happening on campus. The Rails and android program that implements the stated objectives which run on a GPS-enabled device (such as a phone or laptop) was used, which will not only give the user directions to their destination, but would also send notifications to their mobile device or phone, prompting them of upcoming events such as University of Ghana Required Courses (UGRC) Interim Assessments (IAs), Exams, public events, Career talks and many others. This would save so much time hitherto wasted browsing a notice board for an event and searching for location of the event. The Mobile application also routes the different paths the user can take to reach their given destination. The web application was developed using Ruby on Rails and the mobile application was developed in Android (using mainly Java). The application will help to save time, relieve stress and to do things easier and more efficiently. Future additions to the work can be to enable the users to get dynamic route mapping, where the routes on the maps move as the user moves, and to enable any information concerning an event to be automatically fed into the app.

Keywords: Categories, Venues, GPS, UGRC, IA, Ruby on Rails, Android.

 

 

 

Project Title: ATM Authentication Using Facial Recognition and PIN

Group Member: Armoo Bernard Ekow Sam

Supervisor: Dr. Robert A. Sowah

Abstract:

ATM AUTHENTICATION WITH FACIAL RECOGNITION AND PIN is a means to improve the security involved in performing transactions at the automated teller machine.

The ATM is one of the oldest and most secure piece of technology that is still in use to date.

Advancement in technology has made the system vulnerable to fraudsters who use their idea in technology to steal money. Nonetheless using biometric authentication has been proven to be safer and it will help curb these vices by strengthening the security level at the teller machine.

With this perception in mind, in this project, a facial recognition system was created and a user interface of an automated teller machine system was made as well. The facial recognition system was incorporated into the ATM in addition to a pin for authentication before transactions can be made. A facial recognition system is a computer application for automatically identifying or verifying a person from a digital image or a video frame from a video source. This is achieved by comparing selected facial features from the image and a facial database. Matlab and Java where used in the development of the system and because of the system's user friendly nature it can be used by the young and old. The requirement to using this system is to have a recognised account number and a pin linked to a face recognised by the system (An AND operation system).