Skip to main content »

Trinity College Dublin

Dublin University Crest Computer Science Department > Department Details > Personnel > Dr. Kenneth Dawson-Howe Trinity College Crest
 Home
 Teaching
 
 CS1010 Programming
 CS4053 Vision
 Student Projects
 
 Organisation
 Support
 Resources
 Previous Projects (from tcd.ie only)
 Research
 Qualifications
 Contact points
 Personal
 
 What's important
 Travel the world
 Banoffi Pie
 Humour
 

Previous Projects details

This page provides details of some projects which have been previously done under my supervision. Note that this list was originally created in 2004 and does not attempt to list all projects before that date. Where possible the full project report has been provided in PDF format. These are for information only and should not be distributed to any other party.

For information about projects currently on offer click here.

Project Details. Illustration.

Tracking with Height Estimation in an Autonomous Surveillance System, by Sinead Plunkett (2015) (Report)
This Masters dissertation presents a system for estimating pedestrian height in surveillance video using automatic calibration from any bicycles that appear in the scene.

Project Details. Illustration.

Mobile phone bus time app, by Brendan Hegarty (2015) (Report)
This final year project thesis presents a system which automatically presents live bus time information once given a snapshot of the bus stop.

Computer Vision in User Interfaces, by Darragh Hickey (2015) (Report)
This final year project thesis presents an investigation into the use of computer vision within user interfaces.

Identifying Abandoned, Moved and Removed Objects in Automated Surveillance Systems, by Jack Fitzsimons (2014) (Report)
This final year project thesis presents a system for identifying abandoned and removed objects and distinguishing them from the more common local movement of an object within a scene. This project was the Overall Winner in the Computer Sciences & Information Technology category of the Undergraduate Awards. It was also published in a journal article:
Jack Fitzsimons and Kenneth Dawson-Howe, Abandoned, Removed and Moved object classification, International Journal of Pattern Recognition and Artificial Intelligence, 30, (1), 2016.

Frontal Detection of Backpacks in Surveillance Videos, by Ian Beatty-Orr (2014) (Report)
This Masters thesis presents a surveillance system for detecting backpacks where the person is walking directly towards the camera. This work was subsequently published:
Ian Beatty-Orr, and Kenneth Dawson-Howe, Frontal Detection of Backpacks in Surveillance Videos, Irish Machine Vision and Image Processing Conference (IMVIP), Derry-Londonderry, 27-29 August 2014, edited by Sonya Coleman, Bryan Gardiner, Dermot Kerr , 2014, pp116 - 121

An Assessment of the current state of background model evaluation and a proposal for a fair and comprehensive evaluation framework and methodology, by Sarah Conway (2014) (Report)
This Masters thesis presents an Assessment of the current state of background model evaluation and a proposal for a fair and comprehensive evaluation framework and methodology!!

Augmented Annotation of Real-Time Video on a Mobile Phone, by Chris Fenlon (2014) (Report)
This Masters thesis presents a system which attempts to annotate video on a mobile phone using information from Street View and Google Maps.

Weather Analysis using Computer Vision, by David Young (2014) (Report)
This final year project thesis presents a system for automatically analysing the weather (wind, rain, sun, etc.).

Distinguishing Abandoned and Removed Objects, by Danny Murphy (2013) (Report)
This final year project thesis presents a system which distinguished between abandoned and removed objects.

Ground truth specification for video, by Colin Power (2013) (Report)
This final year project thesis presents a system attempts to aid in the production of ground truth for video.

Door Modelling, by Alexander Leonov (2013) (Report)
This final year project thesis presents a system which locates doors in video through an analysis of the moving edges in the scene.

Handling Changing Illumination, by Ronan Fahey (2013) (Report unavailable)
This final year project thesis presents a system which attempted to locate moving objects in the presence of major illumination changes.

Illustration unavailable

Tracking of a Basketball using a Single Camera, by Brian Byrne (2012) (Report)
This final year project thesis presents a system for tracking the ball during a basketball game. This is a particularly hard task as the ball is small, moves quickly, is sometimes occluded and is similar in colour to the skin of the players.

Open Windows, by Liam Stanley (2012) (Report)
This final year project thesis presents a system to detect whether windows are open or closed. This project was intended to support efforts to better understand airflow and heating within large buildings.

Identifying ghosts in the background, by Robert Conlon (2012) (Report)
This final year project thesis presents an attempt to locate ghost objects in a background model. Ghosts are changes to the background which have not yet been updated in the model.

Detecting abandoned and removed objects, by Sean Cronin (2012) (Report)
This final year project thesis presents an attempt to distinguish between object abandonment (a serious security problem) and object removal (which is not a security risk but might be theft!!).

Viewer-Aware Dynamic Advertising, by Sean Bruton (2011) (Report)
This taught MSc thesis presents an advertising system which could adapt/change the advertisements based on the viewer gender.

Real-time Speed Limit Sign Recognition System, by Wen Liu (2011) (Report)
This taught MSc thesis presented an application on a mobile phone to recognise speed limit signs reliably and feedback this information to the driver of the vehicle (to warn them when they were speeding).

Mobile Phone Paintball: A New Multiplayer Gaming Concept, by Dongfan Kuang (2011) (Report)
This taught MSc thesis presented a multi-player mobile phone based paintball game where players could be shot by taking pictures of them. Custom hats were used together the the GPS and compass information from the phones.

Smart Traffic Lights, by Dermot MacSweeney (2011) (Report)
This taught MSc thesis aimed to build a system to improve traffic flow at quiet junctions through the analysis of videos of each road leading to the junction in order to predict the arrival times of vehicles and pedestrians. Where there was no or little competition for the use of the junction the sequence of the lights would be altered so that the vehicles and pedestrians were not stopped.

PS3 type motion controller, by Emma Louise Lynch (2011) (Report)
This final year project attempted to emulate the Playstation Move controller using a webcam together with a coloured wand.

Camera Phone Sign Translator, by Darren Mason (2009) (Report)
This final year project developed a working system to translate signs using a mobile phone camera. An image acquired on a mobile phone was transmitted to a server which then read the text (using the Tesseract OCR package) and translated it (using Google translate). The project was awarded joint Best Project Prize in D stream engineering and also was selected to represent Trinity in the "Innovative Student Engineer Award 2009".

Advertising Replacement Using Computer Vision, by Timmy O'Mahony (2009) (Report)
This final year project developed a system which automatically located advertising hoarding and then allowed the user to specify an alternative sign which was overlaid on top of the original hoarding in the video. This concept allows the advertisements to be customised to the audience.

Sporting Enhancements for Television, by James Collins (2009) (Report)
This final year project worked on developing a system to automatically show the distance from of a football from the goal.

Adaptive Skin Selection Hand Tracker for Use with Augmented Reality, by Orla Codyre (2008) (Report unavailable)
This final year project developed a skin tracking system for tracking hands. The system made use of face detection (based on grey scale images only) in order to automatically adjust the skin tracker to work with the current lighting conditions.

Illustration unavailable

Vision for Advertising, by Declan McGuire (2008) (Report)
This final year project developed a system to automatically emulate some adverts which were created for a HP Printer ad a few years ago. In the ads a white frame is used and pictures of what is behind the frame can be virtually snapped and then moved elsewhere in the image.

Mailbox Monitoring system, by Mark Kelly (2007) (Report)
This final year project developed a real-time system to monitor physical post-boxes using an IP camera and automatically determined whether there was post in each box. If there was an appropriate icon was then displayed on the screen of the owner of the postbox (assuming the owner had registered and was logged in).

3D Structure from Motion from an Uncalibrated Camera by Brian Cullen (2007) (Report)
This final year project developed a system to extract a 3D model from a uncalibrated video sequence (which is the state of the art in this field). It succeeded in computing the model up to a projective reconstruction.

A Software Face Tracker Using a Webcam by Eoghan McCabe (2006) (Report)
This final year project developed a real time (15Hz) face tracker as a user interface and demonstrated the usefulness of the interface through two working applications. This project was awarded best project prize in 2006. There is a video of the working system available here.

Watching Snooker on a Mobile Phone by Michael Pearce(2006) (Report)
This final year project investigated the automatic analysis of images of a game of snooker in order to transmit these in icon form to a mobile phone (hence significantly reducing the required bandwidth).

"Missed Knock", A Computer Vision System for Logging Missed Visitors by John McVann (2005) (Report)
This M.Sc. in Ubiquitous Computing project monitored a corridor and logged the best face image of those people who knocked at office doors. These were transmitted to an application on the users desktop showing who called while they were out. There is a video of the working system available here.

Building Recognition using Computer Vision by Ken O'Hanlon (2005) (Report)
This M.Sc. in Ubiquitous Computing project analysed images of buildings and attempted to extract descriptions of the windows visible in order to allow the buildings to be recognized.

Real-time Face Tracking for interaction with 3-Dimensional Worlds by Barry O'Moore (2005) (Report)
This final year project processed live images from a webcam to determine the users position relative to the screen and then used this information to control a 3D environment.

Blackjack Tracking System by Wesley Cooper (2004) (Report)
This final year project monitored play of blackjack from a low resolution overhead camera in real-time.

Automated Red-Eye Detection & Correction by Hua Yuen Hui (2004) (Report)
This final year project took JPEG images, automatically located any eyes which were exhibiting red-eye and corrected the red-eye - producing much better photographs.

Video Stabilization by Mark McDonnell (2004) (Report)
This final year project processed video sequences from a handheld video camera with the aim of stabilising the sequence. A demo is available.


Click image for demo
.

Faking Group Photos by Andrew O'Sullivan (2004) (Report)
This final year project created realistic composite photos from two images of the same group of individual - with the aim of creating one good picture of everyone.

Automated Real Time Sport Analysis and Distribution with Computer Vision and Mobile Phones by Ryan Sherlock (2004) (Report)
This final year project analysed video footage of a game of football and relayed the game (in terms of where the players and ball are located) to a mobile phone.

Direction of Camera based on Shadows by Darren Caulfield (2003) (Report)
This final year project analysed video from a surveillance camera on a sunny day and determined the camera compass orientation based on the shadows of individuals walking through the scene. This project was jointly awarded best project prize in 2003.

Using computer vision techniques to count and track humans in surveillance video by Scott Tattersall (2002) (Report)
This final year project analysed video of moving people in order to count and track them. The counting was intended to be used by marketing firms in order to allow them to determine the number of people who pass a particular advertising. The system was demonstrated in a variety of real environments and was able to deal with groups of people and with shadows. In fact a new technique for shadow detection was developed as part of the project.

Email  Kenneth.Dawson-Howe @ cs.tcd.ie