Dr Adrian Dorrington from the School of Engineering at the University of Waikato is conducting research in the field of optoelectronics. His work is involved with finding out how 3D cameras can be improved to provide higher-quality images and more accurate information.

How we see 3D

The idea behind 3D is that we see stereoscopically. 3D TV or films work by showing two pictures at the same time, with each picture delivered to individual eyes – we see one picture with one eye and the other picture with the other eye – so it’s tricking the brain into seeing a three-dimensional image.

3D camera technology

Adrian’s research is looking at time-of-flight 3D cameras, which are different from stereoscopic 3D TV cameras. While three-dimensional cameras have been in use for a few years, the technology is still relatively new and needs to be refined.

A time-of-flight 3D camera records the amount of light arriving at each pixel in the image and also measures the distance between the object and the camera, providing information about depth and making it possible to create a 3D image of what has been photographed.

There is a problem that occurs often in current commercial 3D cameras – the multi-path problem – where the light takes a secondary route back to the camera, making the distance measured incorrect. For example, this can happen if the reflected light bounces off something else on its way, which can result in inaccurate readings. The 3D image produced in such a way contains incorrect distance measurements and is not very useful.

The way Adrian tries to correct this problem is to take two measurements instead of one and to change the encoding method for each measurement. The information received still needs to be refined, which is achieved by applying mathematical formulae that address inaccuracies that are being measured and allow Adrian and his team to work out the real distance.

Adrian and his team run these mathematical simulations as well as practical experiments to help solve this problem, and their work is heavily maths based.

3D camera applications

Improving the accuracy of 3D cameras allows people to use them for interactive gaming and gesture control, as well as in medicine, for example, by using 3D images for more accurate and less invasive detection of tumours.

3D imaging can also be used for security purposes, and Adrian and his team’s advances should help to make these more successful.

    Published 19 March 2012