High speed 3D & 4D data capture for specialised research project

COMPANY:

University of Edinburgh School of Informatics

INDUSTRY:

Research

CHALLENGE:

To investigate methods to improve robot sonar through the capture and analysis of bat movements and behaviour using high speed 3D video.

SOLUTION:

DI4D PRO Custom System

RESULT:

Validation of using a DI4D PRO Custom System to evaluate bat movements and behaviour.

Acquisition and assessment of stereo sequences of four different bat species.

Research published in several high profile journals and major Computer Vision conferences.

BACKGROUND

The Research Team at the School of Informatics investigates a variety of topics in Computer Vision. Concentrating on topics within 3D (still) data and 3D video (“4D”) data capture and analysis, they are involved in interesting application projects, all of which have led to new vision research as well as application results.

THE CHALLENGE

Led by Professor Robert B. Fisher*, Chair in Computer Vision at the School of Informatics, University of Edinburgh and a pioneer in the field, the Team had a requirement for high resolution 3D video data capture to analyse the movement and behaviour of bats - in particular, of the bats in flight while they undertook foraging activities. Not only did they require a System that could capture data at 500 frames per second (FPS), but they also needed a solution able to work in low light to mimic the bats natural behaviour.

THE SOLUTION

After extensive research by Professor Fisher, it was evident no ‘off-the-shelf’ solution offered the high frame rate 3D data capture required by the project. In order to meet the team’s stringent requirements and be capable of validating their research, it was necessary to source a custom built system.

Professor Fisher said,

“I knew I could trust Dimensional Imaging to build a System for the project to deliver the kind of capabilities we needed. The kind of System we were looking for didn’t exist in the marketplace….only this custom solution could acquire the very high frame rate of 3D data we needed.”

He adds,

“The FPS captured by the cameras was crucial in allowing the team to analyse every precise movement by the bats.”

“One of the great things about working with Dimensional Imaging is their eagerness and ability to adapt their existing products and systems to produce the specialised custom system we needed for our very particular application…,and they work really well.”

Professor Robert B. Fisher, Chair in Computer Vision, School of Informatics, University of Edinburgh

THE RESULT

The team has made significant advances in studying bat movements and behaviour with the Custom DI4D PRO System.

Several articles describing the specialised research carried out have been published in high profile academic journals and major conferences.

Professor Fisher’s work is renowned worldwide having published and edited 9 books and around 260 peer-reviewed scientific articles focused on his specialised research work.

 

Recent published paper around research-

Y. Xiao, R. B. Fisher, M. Oscar, Performance Characterization of a High-Speed Stereovision Sensor for Acquisition of Time-Varying 3D Shapes, Machine Vision and Applications, Vol 22(3):535+, 2011.

 

*Prof. Robert B. Fisher received a B.S. with Honors (Mathematics) from California Institute of Technology (1974) and a M.S. (Computer Science) from Stanford University (1978). He received his PhD from University of Edinburgh (1987), investigating computer vision. Since then, Bob has been an academic at Edinburgh University, now in the School of Informatics, where he helped found the Institute of Perception, Action and Behaviour. His research covers topics in high level and 3D computer vision, focusing on reconstructing geometric models from existing examples (of industrial parts, buildings and people). More recently, he has also been researching video sequence understanding, in particular attempting to understand observed behaviour. He also has a passion for on-line teaching and research resources for the computer vision community, leading to several well-used text and interactive exploration resources. He has published or edited 13 books andaround 260 peer-reviewed scientific articles. He has been the principal investigator on approx £5 million pounds of research funding, primarily from EPSRC and the EC. Currently, he is also the Dean of Research in the College of Science and Engineering. He is a Fellow of the Int. Association for Pattern Recognition (2008) and the British Machine Vision Association (2010).

“The high fidelity of data delivered from their 4D system and quality of facial performance capture, ensured an extremely successful end result.”

Michael Illingworth, Owner, Vine FX

Experience high fidelity facial performance capture

Request a Quote
Demo>>>>>>>>