|Steve Belleguelle||3D primitive shape extraction and replacement for ray-traced images|
|Steve Benford||A really useful system for module selection|
|Edmund Burke||An interface for the rapid backcalculation of pavement layer properties|
|Dave Elliman||Secret fingerprinting of images and documents|
|Chris Gilbert||The year 2000 problem|
|Chris Greenhalgh||Spatial text conferencing system|
|Colin Higgins||Real-time control of a model railway|
|Graham Hutton||Dynamic retiming of traffic lights|
|Mark Jones||A proof editor|
|Stergios Maglaras||The smart house software system|
|Paul Newton||WebBook: managing pages of interest on the world wide web|
|Mark O'Brien||A language tutor|
|Marcus Roberts||A cooperative approach to web indexing|
|David Snowdon||The interface for a portable memory aid|
Ray tracing is a method of producing very realistic looking pictures with shadows, reflections and highlights not possible with many other rendering methods. There are a number of ray tracing packages available, some commercially and some freely available. A package called PovRay is one such freely available package that runs on many platforms and supports many of the features of the very expensive commercial packages. PovRay works from text scene description files which fully describe the objects in the picture. These objects are often combinations of very simple primitive shapes such as spheres, discs, boxes and the like but putting them all together allows complex objects to be generated. Other packages can be used to create scene files for PovRay from scratch or convert files from other formats into PovRay files. The problem is that these often represent all of the objects in the scene as meshes of triangles rather than using the primitive shapes, the upshot being that the files can become huge where a much smaller one could produce an almost identical result by using primitives.
The aim of this group project is to write a filter that will read as its input a PovRay file that contains meshes and create as its output another PovRay file that replaces many of the meshes with more space efficient primitives. The filter will be able to recognise and replace meshes that represent planar polygons, boxes, discs, cylinders, cones and spheres with their equivalent primitive shape descriptions. As well as the obvious geometry involved and the structure handling that this will entail, this project will require reasonably complex file parsing and creation.
Choosing an appropriate combination of modules is a difficult task. Do I have all of the pre-requisites? Which modules open up most future choices? Do I have the right number of credits? Do I have a balance of exams and assessed courseworks? Is this module heavily subscribed or not? Are there any timetable clashes? And so on.
The aim of the project is to design a new module information system (starting from a blank page) complete with a suitable, intuitive, and easy to use user interface.
The Automated Scheduling and Planning group are involved in a joint project with the department of Civil Engineering to apply neural networks and genetic algorithms and, in particular, hybrid techniques to pavement engineering. The focus is on rapid calculation of pavement layer properties (stiffness moduli, layer thicknesses etc) from surface deflection measurements collected using standard Falling Weight Deflectometer (FWD) equipment. The FWD is a Department of Transport approved test method for providing detailed information on the structural response of a pavement to simulated traffic loading. Pavement evaluation is of key importance to enable maintenance to be carried out at the most advantageous time and in the most cost-effective manner. A thorough structural investigation is always essential where there may be a need for strengthening, particularly when major expenditure is being considered.
The PC based system will apply hybrid neural networks/genetic algorithms to deflection readings, the load, temperature of pavement and the thickness of various layers to generate the modulii of various layers. The purpose of this group project is to build a user friendly interface to the system.
Document theft, copyright violations, leaks from government departments and tampering with documents and photographs used in evidence are all big issues today. Highwater Designs in Cheltenham have developed some interesting and powerful methods for concealing a "fingerprint" in images and can detect this even after low-quality printing and re-scanning of the image. They are also able to detect any tampering with the picture using a related method. I have some ideas on how this could be done for binary images such as in digital photocopiers. It would be useful to hide a photocopier number and key number in all copies. Michael Hesseltine would love a copier like this in his Department!
The project will involve using the Highwater technology and checking the robustness of the fingerprinting to print quality, scanner resolution, and skew. These parameters will be investigated in as scientific a way as we can think of. The other part of the project will be to brainstorm, code, and test my binary image ideas, and see how well these stand up to multiple copying, skew and so on. The group will also be encouraged to come up with their own ideas, and to meet and discuss the project with staff at Highwater, and to demonstrate their work to the company.
Project description to be supplied.
Computer networking provides an extremely powerful basis for supporting communication and collaboration between geographically dispersed users, with current technologies ranging from email to networked virtual reality. It has emerged from examination of real-world working practice that the situation of collaboration and communication within a working space can be highly significant. For example notions of overhearing, peripheral awareness, chance encounters and making activity visible at a glance all rely on this situated and spatial character of collaboration. Also, it is very clear from working with networked virtual reality systems that a computer-based communication tool should also take account of the parallel interactions that are occuring in the physical world, such as answering the telephone or leaving the office.
The goal of this project is to create a system to support text-based communication and collaboration over the Internet from standard workstations or X terminals without high-performance graphics. The distinctive features of the system are that it: provide support for the situation or contextualisation of activity in a virtual spatial framework; and assist users in coordinating and understanding the relationships between physical and electronic activities. For example, collaboration might be situated within a shared 2D or 3D graphical environment, and there might be facilities to indicate real-world interruptions and to catch up with events on returning.
Several commercial systems are available for monitoring and controlling model railways. These all need some intrusive method for monitoring the position of the trains and specialist hardware and software for controlling them. It will be interesting to design and build a such a system from scratch without the aid of specialist equipment.
The aim of this project is to take the existing hardware and train set in the department and build both the hardware and software necessary for controlling the system. While the track layout is fixed, the method of integrating the hardware has yet to be determined. In order to detect the position of the train (in this case only a single engine will be allowed on the track at any one time), a video camera will be placed above the track. Thus the project involves a variety of computer science areas including real-time systems; image processing; hardware control and interfacing; compilation techniques; programming; etc. An exact specification of the minimum system to be produced and enhancements that are desirable will be produced after discussion with the group at the start of the project.
Within a network of roads, traffic lights are used to manage the flow of vehicles across busy road junctions, and to allow pedestrians to cross roads safely. Synchronising a network of lights to achieve a high throughput of traffic is a tricky problem, compounded by the fact that traffic flow is not constant, but changes due to factors such as the time of day, special events, road closures, and accidents. One approach to this problem is to use a system that continuously monitors traffic flow, and automatically changes the timing of lights to take account of any changes in traffic flow, however they may be caused.
The aim of this group project is to design and implement a computer-based system for dynamically retiming traffic lights. Since the system can't be tested on real roads, part of the project is to build a graphical simulator for a network of roads and traffic lights. Such a simulator will allow traffic flow to be modified, and the results of any changes in the timing of lights to be observed. For simplicity, the system will be tailored to a specific network of roads and traffic lights within the Nottingham area.
In the first year mathematics courses, we saw that proofs can be constructed in a very precise manner, manipulating symbols and propositions according to a carefully specified set of rules. Computers are particularly good at handling such rules, and the goal of this project is show how we can take advantage of this, using the machine to help build proofs.
There are a number of different issues that you will need to address, including, for example:
Previous experience with mathematical proofs will be useful, but this is not required; the goal of this project is to focus on the computer science, not the mathematics.
Part of the challenge in introducing computers into man's every day life is revolutionising the dwelling environment of people. Several projects towards that direction involve the construction of a computer controlled household. Such a system would simplify many aspects of house maintenance and security. Additionally, it would provide the user with an integrated environment for monitoring the use of several household appliances for the purposes of economy and safety. The continuously rising number of computers available in households points to the technical feasibility and financial viability of such a system.
Therefore, the purpose of this project will be the development of an integrated software environment which simulates the behaviour of a computer controlled house management system. The main focus of the project will be on the user interface issues (design and implementation) of the software and the availability of an adequate range of functions as well.
Due to the practical difficulty in obtaining the necessary hardware for implementing a working prototype (digital signal processing cards, sensors etc.) the system's functionality could be demonstrated with the use of an appropriate test files. However, should the group wish to implement some of the system's functions as a fully operational hardware supporting function, that will be encouraged.
The Department stores scanned pictures of students and staff, both present and past, online in /cs/faces/..., ordered by year and username. This data is a relatively un-tapped resource offering many possibilities for exploitation. In addition, several limitations of the data have already come to light, that need to be addressed. Firstly, when students or staff leave, their username can be re-used, so it is not a unique name for old pictures and its becoming harder to identify images of people after they have left. Secondly, some people change their appearance within a year and their image can rapidly become unrecognisable. At present, some member of staff has to wheel out a camera and pc each year, to update the scanned images. A photobooth that anyone can use, at anytime, might be better, providing some security and practical issues can be sorted out.
The aim of this group project is to form a research and development team, to investigate the potential of the /cs/faces/... data, to develop tools for its exploitation, and present these as useful products for future students and staff to use. This might include brain-storming sessions, graphics and image processing, using existing tools or programs that you develop, data-base construction, and possibly, if some members of the group with some engineering/practical background, the construction of a secure "photo" booth.
The internet has seen an explosion in popularity over the past few years, much of this due to the success of the World Wide Web. The WWW provides a distributed hypertext environment which allows users to access pages of information from around the world, and publish their own pages for others to view. The web has received much media interest and is now used by thousands upon thousands of users; most people will have access to the web, and many already have their own home page. However, as the web continues to grow at an alarming rate, the sheer size of the web can become unmanageable, as the user becomes overloaded with information. Many web browsers offer a partial solution to this, by providing a bookmark facility, whereby users can store pages of interest in a list, then view them at a later date. This can be a very useful feature, yet the functionality of the bookmark service varies greatly between web browsers, and often becomes unsatisfactory as the bookmark list grows. A user is forced to adopt a particular bookmark format for each browser, so that a user cannot swap between browsers while maintaining the same set of bookmarks. Also, the bookmark entries can become out of date so that a user may find that some links are no longer working.
The aim of this project is to design a more flexible method of managing references to World Wide Web pages. The group will be required to look at existing bookmark tools and resource management systems which already exist on the web, to identify areas which could be improved. For example, a user should be able to access their bookmarks from any any browser, the system may offer more advanced search facilities, and support multiple bookmark sets so that the user can maintain parallel bookmark lists. A more advanced bookmark system may also check to see if any bookmarks are no longer valid, or notify the user if any of their bookmarked pages have been updated since they were last viewed. The group is free to choose the tools to implement the project, but will be required to use and understand the ideas behind the World Wide Web and related areas.
Learning a new language is difficult. On the one hand the learner must have the practical ability to form words and phrases so that simple ideas can be communicated. While on the other a firm grasp on the grammar of a language must be obtained so that sentences can be correctly constructed. With the low cost of personal computers which have the ability to combine sound and text, it would appear that new aids to language learning can be provided using this technology.
This project will build a PC based language tutor. The details of the language to be used will be provided using a simple text book with an associated tape, together with other materials which will be collected as the project progresses. While one language will be chosen as an exemplar to be implemented, the difficulty in implementation will be to make the system as generalised as possible so that it forms a template into which other languages can be inserted. Nevertheless, the ultimate aim is to produce a working and robust system that can be distributed on CD ROMs.
The World-Wide Web has experienced explosive growth in the number of resources available through it. The Alta Vista search engine boasts an index of 30 million documents. Yet every day it seems more difficult to locate resources which are actually useful. The problem with systems such as Alta Vista stem from their approach to data gathering. Automated agents known as Web robots iteratively follow links from one site to another, blindly indexing every document seen. Whilst the resulting data set is impressively large, it is difficult to obtain useful results from it. A simple query can often return thousands of matching documents. However, many of these will be irrelevant or only have a small mention of the topic of interest. Many other documents will no longer exist. The most useful source of resources is usually found on individual users' home pages. Nearly every home page on the Web has a page of links collected by the author covering their favourite topics. These pages tend to be well researched and organised, kept up to date and create a useful collection of links on a particular topic in one place.
The aim of the project is to bring the information found in these pages into a single easy to use system. The group will be required to create a system which supports a cooperative approach to building a database of useful resources. The aim is create a single database of links, to which users can add when they discover a resource they find personally useful. The system should allow the links to be organised by subject, and to have additional information associated with each link, such as a short description. Any user may recommend a resource by adding it to the database. Each entry will show who submitted the recommendation and their rating of it. The system must support multiple users, such that many people may be browsing and editing the links database at one time. The group will need to investigate the best approach to security, such that maximum freedom for updating the database is provided, but with a minimum risk of malicious use. The choice of tools and platform used for implementing the system are left to the group. However, the browsing interface should be provided through a Web page, requiring an (eventual) understanding of HTML, forms and CGI scripts.
People have difficulties in recalling events and retrieving "lost" information. The modern world places great demands on human memory and to a large extent computer systems add to the demand rather then relieve it. They ask users to recall command sequences, for example and to remember the names under which they have filed away documents, perhaps many months or even years ago. As people come to rely on more and more systems, and store more amd more information electronically, the burden on their memories must increase.
Design a memory aid which offer richer contextual information for retrieving files, to enable people to recall events and retrieve information more easily, including events ands inforamtion that they may not have realised they might need to access again. This system should not only store the information to which the user might need access - documents, phone numbers, online help - but which will also record the context in which access was made, for example, the phone number dialled after a document was retrieved.