Computers that see
Capturing the movement of the human body on a computer has been a goal in many areas of science and entertainment, including film making, computer games, medicine, surveillance and sport, for over 20 years. Recently Oxford Brookes has been working with OMG plc, an Oscar-winner and world leader in the field, on an exciting KTP which is taking the technology to new levels of sophistication.
The project is advancing the technology of ‘markerless motion capture’ - digitising motion without the use of special markers on the subject - with the aim of creating a system to achieve this in real time.
OMG was established in 1984 and has recently introduced its sixth generation of motion capture systems. OMG’s ‘mocap’ is applied across a range of markets and applications, from visual effects in epic blockbuster movies (Star Wars, Gladiator and Van Helsing) to virtual environment engineering (BMW, Peugeot, and Renault) to hospitals, universities and research institutes. The technology is scalable and has been sold in systems as large as 200 cameras, capable of capturing simultaneous data from multiple people including detailed face and hand movements. OMG’s technology has been recognised with an Emmy award and a technical Oscar.
The KTP draws on the unique expertise of Philip Torr, Professor in Computer Vision and Machine Learning at Oxford Brookes. After research at Oxford, he worked for Microsoft Research, first in Redmond USA in the Vision Technology Group, then in Cambridge UK, founding the vision side of the Machine Learning and Perception group. His Computer Vision Group at Oxford Brookes now includes two KTP Associates, Manish Jethwa and Morné Pistorius, who are leading the project at OMG.
Andrew Stoddart, Chief Scientist at OMG, and a keen advocate of KTPs said: “The KTP programme has proved to be an effective way of building a bridge between the academic and industrial worlds. It is an excellent way of anchoring a company R&D project.”