Category: CSE Project Reports

Project Title: The use of Genetic Algorithms and neural networks in object detection.

Project Description:

Genetic algorithms are powerful algorithms that can The aim of this project is to develop and apply a ‘genetic algorithm’ to the task of automated object recognition. Neural networks are very useful devices that can be trained to perform many different tasks.

The human vision system is capable of locating and identifying complex objects quickly on cluttered scenes. The genetic algorithms themselves can be used to breed many solutions to a problem and reject/accept various solutions depending on their fitness. The intention of this project is to develop a genetic algorithm that can breed and train many generations of neural networks capable of recognising simple objects on sets of images, either captured by the student or supplied by the tutor.

Automatic object recognition is an interesting and demanding area, very useful in many areas, ranging from robotics to security systems.

Product Details:

Image Processing Analysis software

Resources Required:

Images (supplied), Matlab, C++, C# (all supplied).

Project Title: Object recognition on x-ray images of concealed weapons

Project Description:

Develop and test object recognition software capable of spotting key hazardous items such as scissors and knives on x-ray images of baggage. One of the state-of-the-art object recognition techniques is the Scale Invariant Feature transform (SIFT) which works well on simple items in well lit conditions, and can generally locate fairly simple objects with a degree of certaintly. Your aim will be to investigate and improve this technique, or apply others to work well on noisy and textured images.

Product Details:

Image Processing Analysis software

Resources Required:

Images (supplied), Matlab, C++, C# (all supplied).

Project Title: Multimedia services in mobile networks

Project Description:

Mobile communications has been deployed in recent years on two major fronts: cellular mobile and infrastructure-based wireless local area networks (WLAN’s). The former supported fast mobility over large distances, while the latter supported high data rates in a relatively fixed setting with a short wireless hop. Both have provided steadily increasing data rates over the last decade. Mobile telephony added low- to medium-speed data services (SMS, emails, etc.) in 2G. There is now an increased demand to transmit broadcast music and low bandwidth video over cellular networks, hence recently deployed 3G networks can support higher data rates (384 kb/s to 2 Mb/s). WLAN’s can offer much higher data rates, enough to offer streaming voice, video, and multimedia services. It is considered crucial that both these networks are able to support all possible services: voice, data, video, and multimedia. Providing multimedia services over such networks is a challenging area.

Aims and objectives:-

The project should consider how a large corporation (e.g. multimedia based) could best engineer a transformation to a wholly mobile network. Multimedia services over wireless/mobile ad hoc networks should be a priority, with special emphasis on multimedia streaming. Envisaged research areas are to consider whether a mixture of UMTS (3G mobile) and IEEE 802.11 WLAN’s could provide a possible solution.

Product Details:

The primary aim of the project is use OPNET’s ITGURU modelling software to model and analyse the performance of a large scale wireless network, with particular emphasis on the deployment of multimedia services.

Project Title: 3D Scene Capturing System

Academic Research Areas: Design & Manufacture

Project Description:

This project relates to the design and build of a compact and low-cost, 3D scene capturing system i.e. one that captures depth.

Product Details:

The project will involve the design and build of a 3D scene capturing system. The requirements are for it to be portable and low cost.

Resources Required:

CAD software, digital camera/scanner, image recognition software.

Project Title: Product design for the elderly.

Project Description:

The elderly are becoming a major consumer group. Their life expectancy is increasing, and they represent a considerable economic market. Most of today’s products however are not designed for them, and they are faced with considerable problems buying and using simple things like radios, cash dispensers, or even opening a packet of coffee.

Product Details:

The project will focus on the user group to identify a particular problem. The student will then design a product to answer some of the user’s difficulties and aspirations.
The scope of the project can cover domestic consumer products, or products or equipment which are used by the public. The project can focus on the design of hardware and/or software.

Resources Required:

Digital still and video camera. Modelmaking facilities. 2D and 3D Computer aided design software.
Possibly software to simulate man machine interface designs.
Possibly rapid prototyping facilities.

Industry based:

The student could work with a manufacturer if they establish contact sufficiently early in the project.

Project Title: Product design and obesity.

Project Description:

We have a growing population of overweight people. As product designers we can’t do anything to stop this phenomenon, but we can identify areas where today’s products need to be redesigned to be more suitable.

Project Title: Effect of mesh characteristics on the numerical simulation of the flow around a vehicle

Project Description:

This project will investigate the numerical simulation of the airflow around a vehicle, using a Computational Fluid Dynamics (CFD) software package, FLUENT. Of special interest to this project will be the effect of the various meshing possibilities on the aerodynamic results on the vehicle. The vehicle will need to have:

(i) a complex geometry file which can be directly imported into the CFD software
(ii) readily available experimental data for subsequent comparison with numerical results.
Finding such a suitable vehicle will be part of the project.

Accuracy of the results will be analysed for a large number of meshing scenarios with the goal of producing a set of meshing constraints for a reliable simulation. The merits of such numerical simulations will also be assessed.

Product Details:

Resources Required:
FLUENT software package, currently available within the Department

Project Title: Oscillation of a circular cylinder immersed in a current

Project Description:
This project aims to investigate the characteristics of the interaction between a fluid current and a flexible circular cylinder immersed in it. The investigation will be of a numerical nature involving the coupling between the structural software ANSYS and the CFD (Computational Fluid Dynamics) software CFX. The initial case of the flow around a fixed cylinder will be reproduced before making the cylinder flexible.

The complex and often intricate nature of the interaction will be studied by varying the current velocity and observing the resulting relationship between the various frequencies (natural, oscillation, vortex shedding).

Product Details:

Resources Required:
ANSYS and CFX software packages, currently available within the Department

Project Title: Learning in Autonomous Robots using Artificial Neural Networks

Project Description:

This project will investigate the use of Artificial Neural Networks for learning in making robots autonomous. The project will use learning methods based on ANNs that will use sensory information from a robot in a controlled environment (e.g., a maze) to implement and verify the efficacy and extent of learning to enable a robot to be autonomous.

Aim: To apply Artificial Neural Networks (ANNs) for learning about the environment using sensory data from a roving robot.

Objectives: It is expected that the student will be able to
(i) Critically review the learning algorithms used in autonomous robotic applications;
(ii) Investigate the nature and characteristics of sensory data typically used in autonomous robotic applications;
(iii) Critically review and select one ANN architecture for investigation;
(iv) Implement and validate an algorithm for incremental learning using sensory data;

Product Details:

The project will develop and validate a Matlab-based suite (with C/C++/JAVA if necessary) that can take the sensory data from a roving robot and build a picture of the environment. The final stages of the project may involve compiling code to run independently on a programmable chip mounted on a robot.

Resources Required:

1. Purchase of some necessary sensors and actuators for a robot may be necessary.
2. Matlab/Simulink software.

Project Title: Contour detection and classification using Artificial Neural Networks

Academic Research Areas: Image Processing, Medical diagnosis

Project Description:

This project will investigate the use of Artificial Neural Networks for detection of contours from X-ray and CT scan images and compare them with at least one traditional image processing method.

Aim: To apply Artificial Neural Networks (ANNs) for the detection of Region of Interest (contours) in medical images to automate medical diagnosis.

Objectives: It is expected that the student will be able to
(i) Investigate the traditional image processing techniques for contour detection;
(ii) Investigate the use of ANNs for contour detection applications;
(iii) Implement one algorithm based on ANNs using medical images;
(iv) Implement at least one technique of contour detection used in image processing;
(v) Compare the effectiveness of the two approaches.

Product Details:

The project will develop a Matlab-based suite with appropriate GUI for the selection of test images, choice of algorithm to use and display all the results in appropriate graphical and numerical formats.

Resources Required:

1. Access to medical images (some available on the Web and some as part of a research project within the department).
2. Matlab/Simulink software

 

Project Title: Image transmission and processing using the Web

Academic Research Areas: Image Processing, Process Tomography

Project Description:

This project will investigate the transmission of a set of images obtained using a Capacitance-based tomographic imaging system over the internet to a remote client. Processing of the images at the server/client side also will be investigated using appropriate image processing software.

Aim: To investigate techniques and transmission of real-time images over internet to remote monitoring station with application to Process Tomography.

Objectives: It is expected that the student will be able to
(i) Investigate the requirements and protocols for efficient transmission of images over internet;
(ii) Implement and demonstrate the selected techniques;
(iii) Implement a suitable HCI for supervisory functions.

Product Details:

The project should come up with a software suite that will have a “client” and “server” modules that can be plugged into any existing image database or images obtained by real-time measurements. The client module should be able to request transmission of images based on a selection of choices provided by the server module.

Resources Required:

1. Access to real-time images from a tomographic system (available within the department).
2. Matlab/Simulink software