Imagine you're an archaeologist at a dig in an inhospitable, remote location and you want a model of the site and any relics to work on back at the lab.

Or you're a forensics expert called on to help find the cause of death of someone lying inconveniently in a busy street.

Or you're a software developer working on a video game that you'd like to populate quickly with an enormous cast of characters and props, without having an art department to call upon.

Industrial Research (IRL) senior scientist Robert Valkenburg and his team are working on technology that could potentially help out archaeologists, detectives, the entertainment industry and others.

They're developing a handheld scanner that produces a 3D image in real time. The prototype, consisting of a $10,000 laser in a shoebox-sized unit, is connected to a trolley laden with the equivalent computer power of 10 PCs.

Once refined, the goal is to create a scanner that is truly portable - the trolley of electronics shrunk to a size that can attach to the operator's belt - and whose images are photo-realistic.

Then, all manner of commercial possibilities beckon.

Initially IRL, a Crown Research Institute, is offering paying customers use of the equipment at its Parnell site, advertising it as a "scene scanner". But Valkenburg doesn't rule out licensing the technology as another means of getting a return on the cost of its development.

The project began several years ago and has been funded by a Foundation for Research, Science and Technology (FRST) grant, made for the development of "a digital content creation system". The grant is helping pay the way for a five-strong research team.

Valkenburg says what they set out to do isn't unique, but their approach is. The challenge when making large 3D scans is that it's impossible to record all of the image in one pass of the scanner. That means coming up with a way of piecing together multiple scans in a precise way.

One approach is to fix the scanner to a rig that allows it to make multiple passes over the object being digitised, accurately recording the position of the scanner as it goes. The positional information is then used to stitch the scans into one image. The disadvantage of such a system is that it's not portable.

Valkenburg's team has come up with another way of recording the scanner's position. They distribute dozens of small LED beacons around the site or object being scanned. As the laser beam is passed over the target, six cameras attached to the scanner track where it is relative to the beacons.