Toby Breckon

Automatic/Assisted Target Recognition and Classification of People and Vehicles in EO/IR imagery : Identify, develop and demonstrate real-time techniques for the automated detection and identification of people and vehicles in realistic, cluttered scenes in optical/thermal (EO/IR) imagery.

[ Demo ] [ Related Publications ]

Grant Reference: CDE/STELLAR/PV/11657
Research collaborators: MoD CDE, Stellar Research Services.

Threat Image Projection (TIP) for 3D CT-based Baggage Screening Operations: Automatic insertion of 3D threat items for Threat Insertion Projection (TIP) imagery for on-line operator training using automatic image understanding and noise filtering techniques.

[ Demo ] [ Related Publications ]

Grant Reference: HOSDB IRC2010/ITT036
Research collaborators: Home Office Scientific Development Branch (HOSDB / CAST), TRANSEC (Department of Transport), Centre for the Protection of National Infrastructure (CPNI), Metropolitan Police Service, US Dept. of Homeland Security.
Funded under the Innovative Research Call in Explosives and Weapons Detection (2010), a cross-government programme sponsored by a number of government departments and agencies under the CONTEST strategy.

Extended Autonomous Threat Detection : Further development of the cross-spectral computer vision based threat detection capability to consider aspects of behavioural nuance, temporal tracking and preliminary work on 3D sensor integration.

From prior related work: [ Related Demos ] [ Related Publications ]

Grant Reference: SELEX ATD/1/001
Research collaborators: Selex Galileo.

SATURN: Further development of the award-winning Team Stellar SATURN system.

SATURN (Sensing & Autonomous Tactical Urban Reconnaissance Network) - A connected sensing network of autonomous UAVs and Unmanned Ground Vehicle (UGV) to provide imagery for automated threat detection within the complex urban environment.

[ Related Demos ] [ Related Publications ] [ Press Coverage ]

Grant Reference: SELEX/SATURN/1/007
Research collaborators: Stellar Research Services, Blue Bear Systems Research, Marshall SDG, TRW Conekt, Selex Galileo, DCMT Shivenham - Guidance and Control Group

MoD Grand Challenge: Involvement with the Team Stellar UK MoD Grand Challenge project team in the automatic target recognition phase of the team's work to "create a system with a high degree of autonomy that can detect, identify, monitor and report a comprehensive range of military threats in an urban environment."

R.J. Mitchell Trophy (Grand Challenge) Winners 2008

[ Demo ] [ Related Publications ] [ Press Coverage ]

Grant Reference: STELLAR/GC/1/005
Research collaborators: Stellar Research Services, Blue Bear Systems Research, Marshall Specialist Vehicles, TRW Conekt, Selex SAS, DCMT Shivenham - Guidance and Control Group

Automatic Screening of CT Baggage Imagery for Unaccompanied Baggage and Package Screening: The project aims to investigate the use of automated object segmentation and volume based shape recognition for integration into existing Computed Tomography (CT) based Unaccompanied Baggage and Package Screening (UBPS) security procedures. This involves both the exploitation of existing techniques utilised in the medical domain for CT segmentation, their adaptation to the UBPS problem domain and the integration of established 3D image recognition and classification techniques into the UBPS process.

[ Demo ] [ Related Publications ]

Grant Reference: HOSDB 02/08/07
Research collaborators: Home Office Scientific Development Branch (HOSDB), TRANSEC (Department of Transport), Centre for the Protection of National Infrastructure (CPNI), Metropolitan Police Service.
Funded under the Innovative Research Call in Explosives and Weapons Detection (2007), a cross-government programme sponsored by a number of government departments and agencies under the CONTEST strategy.

Novel imaging methodologies for high throughput cell nuclear segmentation in bright field histological imagery: Joint research project with the School of Dentistry at the University of Birmingham to investigate the use of Haar-classifier based real-time object detection techniques, common to the defence/security domain, to the medical imaging problem of cell nuclei detection.

[ Related Demos ] [ Related Publications ]

Grant Reference: STFC ST/F003374/1
Research collaborator: Gabriel Landini, University of Birmingham.

Investigation Of Techniques For Obstacle Detection In Real Time Using Low Cost Monocular Video Camera Technology: Joint research project with TRW Conekt to investigate obstacle detection for real-time automotive applications.

[ Demo ] [ Related Demos ] [ Related Publications ]

Grant Reference: EPSRC CASE/CNA/07/85
Research collaborator: TRW Conekt

3D Surface Completion & Geometric Texturing: Pursuing the goals of Amodal Volume Completion via a novel Non-parametric Completion technique for 3D surface completion, hole-filling and occlusion resolution.

In addition investigating the realistic 3D texturing of artificial surfaces with surface relief that is captured, extracted and then transferred from a real-world 3D surface. Offering 3D texturing by example for future 3D content generation.

Related Demos [ 3D Surface Completion] [3D Geometric Texture Transfer ] [ Related Publications ]
Research collaborator: QinetiQ PLC