Current Projects

Real Time Infrared Detection of Vulnerable Road Users for Automotive Applications

Student:
Patrick Hurney

Supervisors:
Dr. Martin Glavin, Dr. Edward Jones and Dr. Fearghal Morgan

Description:
This project aims to develop an infrared pedestrian detection algorithm for use on an Intel Atom/Reconfigurable hardware hybrid system

Start date:
October 2009, this project is currently active

 Next Generation In-Vehicle Networks

Student:
Shane Tuohy

Supervisors:
Mr. Liam Kilmartin and Dr. Martin Glavin

Description:
A recent trend in automotive technology is the deployment of multiple cameras and associated vision system functionality within vehicles. Traditionally the communication infrastructure associated with this functionality has been based on dedicated proprietary communication technologies. However a recent area of interest is in the use of traditional data networking technologies (specifically Ethernet and wireless LAN) as potential replacement technologies to form the basis of the next generation "networked vehicle".

This research is focussed on the specific demands of using such technology to network the numerous digital cameras and associated back end vision and scene analysis sub-systems that will likely be deployed in the next generation of vehicles. This work will focus on the development of realistic simulation models for the visions system network based on Ethernet and wireless LAN technologies and it will investigate the impacts which network topologies and protocols have on the performance of these networks and on the performance of the vision systems themselves.

Of particular interest in this research will be investigations into the development of realistic models for the noise environment in which these data networks will operate given the relatively high EMI environment found in most modern vehicles and on developing custom Quality of Service based protocols which utilise contextual information derived from the vision systems to manage information rates and flows throughout the system. Additional issues such as security, cost and reliability need also to be investigated in this work.

Start date:
October 2010. Project is currently active.

Project Publications:
S. Tuohy, M. Glavin, C. Hughes, E. Jones, L. Kilmartin. " An ns-3 Based Simulation Testbed for In-Vehicle Communication Networks ", 27th Annual UK Performance Engineering Workshop, UKPEW 2011, University of Bradford, Bradford, July 7-8th 2011. Presentation.

Evaluation and Control of Image and Video Quality in the Automotive Environment

This project will examine the development of methodologies for automotive-specific image/video quality assessment. The project will also investigate the use of these technologies in the context of image and video compression in automotive systems, and will examine how they can be used to help make system level design decisions.

Urban Autonomous Driving : Detection of other vehicle trajectory and intention

This project will examine Urban Autonomous Driving. Investigating the detection of other vehicle trajectory and intention.

Motion-Based Determination of the Automotive Environment for Vehicular Safety

This project will examine the development of methodologies for automotive-specific image/video quality assessment. The project will also investigate the use of these technologies in the context of image and video compression in automotive systems, and will examine how they can be used to help make system level design decisions.

Previous Projects

Distance Determination using a Monocular Camera for an Automotive Environment

Student:
Diarmaid O Cualain

Supervisors:
Dr. Edward Jones and Dr. Martin Glavin

Description:
This project aims to develop a distance determination algorithm for a monocular camera, with particular emphasis with its use in an automotive environment.

Start date: November 2006.

Related projects:
Region Of Interest Determination and Obstacle Detection For Automotive Applications
Vulnerable Road User Detection in Low-Light Conditions
Shadow Removal From Digital Imagery

Project Publications:
Conferences:
D. O Cualain, M. Glavin, E. Jones and P. Denny, “ Distance Detection Systems for the Automotive Environment: A Review”, 15th IET Irish Signals and Systems Conference, ISSC 2007, University of Ulster at Magee, Derry, September 2007. Poster.

Other:
D. O Cualain, M. Glavin, E. Jones, " Distance Determination for Colission Avoidance in an Automotive Environment", College of Engineering and Informatics Research Day 2008, Poster

D. O Cualain, M. Glavin, " Lane Departure and Obstacle Detection Algorithim for use in an Automotive Environment", B.E. Project Report, 2006.

Vulnerable Road User Detection in Low-Light Conditions

This project developed automotive video processing algorithms to detect Vulnerable Road Users (VRU) such as pedestrians and other road vehicles in dark and low-light conditions.

Shadow removal from digital imagery

Student:
Robert McFeely

Supervisors:
Dr.Martin Glavin and Dr. Edward Jones

Description:
Automated Shadow Removal from image scenes. This involves two main stages of shadow detection and shadow correction. Detecting the shadows involves colour constancy, image segmentation, and texture analysis. The shadow correction is based directional smoothing and thin plate reconstruction.

Start date: January 2005. 

Related projects:
Distance Determination using a Monocular Camera for an Automotive Environment

An investigation of objective speech and audio quality assessment techniques, with a specific interest in automotive applications

Student:
Dermot Martin Campbell

Supervisors:
Dr. Edward Jones and Dr. Martin Glavin

Description:
The overall objective of this research project is to conduct research that integrates existing research findings in two areas:
Psychoacoustic based Digital Signal Processing (DSP) algorithms for audio processing and quality evaluation
Automotive Listening Quality Evaluation.
The psychoacoustic DSP algorithms examined are algorithms that are used in general applications to assess the quality of speech and audio signals. These algorithms provide a “score” indicting the quality of the speech or audio signal under investigation. They remove the need for human subjects to manually grade the quality of speech and audio signals based on their average perception. They are typically used in the development of communication networks, the testing of audio codecs eg MP3 and the development and testing of other such communication networks and devices.

The use of speech and audio quality evaluation in the automotive industry is not new. However, the use of psychoacoustic based algorithms to perform this task is quite novel and there appears to be a lack of academic research at this level in the recent past. Speech and audio quality assessment is typically used in automotive applications to test the listening performance for vehicle users in terms of noise interference from outside the vehicle cabin and the audio listening quality of the interior of the vehicle. The evaluation of the quality of the signals is not only relevant to vehicle manufacturers but also to developers of the audio equipment used in such vehicles.

This project focuses on developing DSP algorithmic solutions for general audio quality assessment, and with a further emphasis on applications in the automotive market.

Start date: October 2005. This project has been completed.

Project Publications:
Journals:
D. Campbell, E. Jones and M. Glavin, " Audio Quality Assessment Techniques - A Review, and Recent Developments", Elsevier Signal Processing, 89(8), pp. 1489-1500, August 2009. ScienceDirect

Conferences:
D. Campbell, E. Jones and M. Glavin, " Comparison of Temporal Masking Models for Audio Quality", 15th IET Irish Signals and Systems Conference, ISSC 2007, University of Ulster at Magee, September 2007.

Other:
D. Campbell, E. Jones and M. Glavin "Applications of Psychoacoustic Signal Quality Assessment Techniques", College of Engineering and Informatics Research Day 2008. Poster

Digital Lens and Perspective Correction for the Automotive Environment

Student:
Ciaran Hughes

Supervisor:
Dr. Martin Glavin and Dr. Edward Jones

Description:
This project deals with advanced techniques in the calibration of the intrinsic and extrinsic parameters of wide-angle and fish-eye lenses, and the correction of the distortion inherent in these types of lens cameras. Additionally, with the calibration of the extrinsic parameters, perspective distortion can also be removed. There is particular emphasis on the use of such cameras for displaying blind-zones of vehicles to drivers, including Heavy Goods Vehicles and smaller private vehicles. To this end, a detailed and rigorous mathematical examination of fish-eye perspectives has been carried out. In support of our findings, data has been extensively verified using large sets of synthetic calibration images, and smaller sets of real calibration images.

Start date: September 2005. This project has been completed.

Related Projects:
Automatic Calibration of Fisheye Lenses in the Automotive Environment

Project Publications:
Journals:
C. Hughes, P. Denny, M. Glavin and E. Jones, " Equidistant Fish-Eye Calibration and Rectification by Vanishing Point Extraction", IEEE Transactions on Pattern Analysis and Machine Intelligence, 32(12), pp. 2289-2296, December 2010. IEEEXplore

C. Hughes, P. Denny, E. Jones, M. Glavin, " Accuracy of Fisheye Lens Models", Applied Optics, 49(17), pp. 3338-3347, June 2010. OpticsInfobase

M. Friel, C. Hughes, P. Denny, E. Jones and M. Glavin, " Automatic Calibration of Fish-eye Cameras from Automotive Video Sequences", IET Intelligent Transport Systems, 4(2), pp. 136-148, June 2010. IET Digital Library IEEEXplore

C. Hughes, R. McFeely, P. Denny, M. Glavin and E. Jones, " Equidistant (fθ) Fish-Eye Perspective with Application in Distortion Centre Estimation", Elsevier Image and Vision Computing, 28(3), pp. 538-551, March 2010. ScienceDirect

C. Hughes, M. Glavin, E. Jones and P. Denny, " Wide-angle camera technology for automotive applications: a review", IET Intelligent Transport Systems, 3(1), pp. 19-31, March 2009. IET Digital Library IEEE Xplore

Book Chapters:
C. Hughes, R. O'Malley, D. O'Cualain, M. Glavin and E. Jones, " Trends towards automotive electronic vision systems for mitigation of accidents in safety critical situations", in New Trends and Developments in Automotive System Engineering, Marcello Chiaberge (Ed.), ISBN: 978-953-307-517-4, InTech, (January 2011). INTECHopen

Conferences:
C. Hughes, E. Jones, M. Glavin and P. Denny, " Validation of Polynomial-based Equidistance Fish-Eye Models", 20th IET Irish Signals and Systems Conference, ISSC 2009, University College Dublin, June 2009. Poster

C. Hughes, M. Glavin, E. Jones and P. Denny, " Automotive Blind-Zones: A Review of Legislation and the Use of Close-Range Camera Systems", 1st International Symposium on Vehicular Computing Systems, Trinity College Dublin, July 2008.

C. Hughes, M. Glavin, E. Jones and P. Denny, " Review of Geometric Distortion Compensation in Fish-Eye Cameras", 16th IET Irish Signals and Systems Conference, ISSC 2008, NUI Galway, June 2008. IET Digital Library IEEE Xplore

Other:
C. Hughes, M. Glavin and E. Jones, " Robust and Accurate Principal Point Estimation Using the Five-Point Perspective Model of Fish-Eye Radial Distortion", College of Engineering and Informatics Research Day 2008. Poster

Region Of Interest Determination and Obstacle Detection For Automotive Applications

Student:
Lorcan Browne

Supervisors:
Dr. Edward Jones and Dr. Martin Glavin

Description:
Driven by increasing concerns over automotive safety, the development of on-board automotive driver assistance systems to alert drivers about potential collisions with objects such as pedestrians or vehicles has become a huge area of research. The aim of this project is to present a novel approach for the first processing stage in many of these types of systems. An algorithm is presented which uses a single visual camera to identify Regions Of Interest (ROIs) in front of a moving vehicle, which may contain objects that represent a “threat”. These ROIs are then categorised into different priorities depending on their position in the image, which in turn gives a measure of their threat level.

Start date: November 2005. This project has been completed.

Related projects:
Distance Determination using a Monocular Camera for an Automotive Environment
Vulnerable Road User Detection in Low-Light Conditions

Project Publications:
Conferences:
L. Browne, E. Jones and M. Glavin, “ Region Of Interest Detection For Automotive Applications”, 1st International Symposium on Vehicular Computing Systems, Trinity College Dublin, July 2008. Presentation

Other:
L. Browne, E. Jones and M. Glavin, "Object Detection And Characterisation In An Automotive Environment", College of Engineering and Informatics Research Day 2008. Poster

Automatic Calibration of Fisheye Lenses in the Automotive Environment

Student:
Myles Friel

Supervisors:
Dr. Edward Jones and Dr. Martin Glavin

Description:
This project proposed a technique for calibrating a lens in the automotive environment to compensate for radial distortion introduced by wide-angle or fisheye lenses, without the need for a specific calibration set-up. At present many car manufacturers are endeavouring to provide the driver with views of the car’s surroundings that are not directly visible (“blind spots”). To achieve this, wide angle fisheye lens cameras are fitted to many modern vehicles. However fisheye lenses introduce radial distortion to the resulting images, though this distortion can be compensated for with a suitable compensation procedure. Calibration of the lens is important for compensation, since each lens has different inherent properties which may change with time. However, in situations where recalibration of the lens is necessary, the need for a specific calibration set-up is particularly undesirable as it requires the driver to return to a service centre (e.g. the garage where the car was purchased), and it requires the technical staff to be trained in the calibration procedure. It is proposed in this project that radial distortion introduced by fisheye lenses can be calibrated using only the “every-day” images captured by the camera on a vehicle. The primary focus of this project is to identify and use straight lines in a typical automotive scene as a basis for calibrating for the radial distortion, since straight lines in the scene should remain straight in an image of that scene. The information thus extracted can then be used to determine parameters necessary for image compensation.

Start date: September 2005. This project has been completed.

Related projects:
Digital Lens and Perspective Correction for the Automotive Environment

Project Publications:
Journals:
M. Friel, C. Hughes, P. Denny, E. Jones and M. Glavin, " Automatic Calibration of Fish-eye Cameras from Automotive Video Sequences", IET Intelligent Transport Systems, 4(2), pp. 136-148, June 2010. IET Digital Library IEEEXplore

Conferences:
M. Friel, E. Jones, M. Glavin and C. Hughes, " Comparison of Feature Detection Methods for an Automotive Camera System", 15th IET Irish Signals and Systems Conference, ISSC 2007, University of Ulster at Magee, Derry, September 2007

An Efficient Implementation of Bluetooth Wireless Technology for an Automotive Environment

Student:
Alan Molloy

Supervisor:
Dr. Martin Glavin

Description:
This project implements a low-cost, low-power wireless network for an automotive environment, based on Bluetooth wireless technology. It is envisaged that this network will eb implemented between a device known as a Remote Vehicle Controller (RVC), which is similar in size and aesthetic to a television remote control, and another device situated in the dashboard of a vehicle. These components will be capable of transferring information, which will be used for diagnostic, maintenance and control purposes, and will ultimately facilitate future integration of features to ehance the vehicle owners' experience.

Start date: September 2001. This project has completed.

Project Publications:
Conferences:
P. Doherty, M. Glavin, A. Molloy, F. Morgan, " A Review of Bluetooth Security in the Automotive Environment", 12th IEE Irish Signals and Systems Conference, ISSC 2004, Queen's University Belfast, July 2004. IET Digital Library

L. Kelly, A. Molloy, M. Glavin, F. Morgan, " Error Resilient Image Transmission over a Bluetooth Network", 11th IEE Irish Signals and Systems Conference, ISSC 2003, University of Limerick, July 2003.

Error resilient image transmission over a wireless Bluetooth network

Student:
Lisa Kelly

Supervisor:
Dr. Martin Glavin

Description:
The objective of this project is to achieve efficient, reliable, real-time image transmission over a Bluetooth wireless network. Bluetooth, being an ad-hoc networking standard, is used in a variet of environments, often under difficult conditions, leading to limited bandwidth and significant Bit Error Rates (BER). Image compression and transmission provisions, that avoid catastrophic failure caused by lost, delayed, or errant packets, are cruscial. This project evaluates the transmission of JPEG images through a high BER Bluetooth wireless channel with varying levels of Forward Error Correction (FEC). FEC makes images more robust to errors, however, it adds significant redundant data to an image file. The Error Resilient Entropy Code is presented as a method of modifying the JPEG compressed images to give increased error resilience while still maintaining a high compression ratio.

Start date: September 2001. This project has completed.

Related projects:
An Efficient Implementation of Bluetooth Wireless Technology for an Automotive Environment

Project Publications:
Conferences:
L. Kelly, A. Molloy, M. Glavin, F. Morgan, " Error Resilient Image Transmission over a Bluetooth Network", 11th IEE Irish Signals and Systems Conference, ISSC 2003, University of Limerick, July 2003.