System allows gigapixel photo navigation using hand gestures
A digital imaging student has devised a method for navigating gigapixel photos projected on a large screen using hand gestures alone.
The system, which employs Microsoft’s Kinect motion capture device, allows users to first select an image, to scroll and zoom using sweeps of the hand and finally to print out specific sections of interest.
Although the original concept is intended as an art installation, the technology could be useful for navigating detailed astronomical images and microscopic slides from biomedical research.
‘What I wanted to do was get people immersed in images,’ said Samuel Cox, a masters student in digital imaging at Lincoln University. ‘If you go to a normal photographic gallery, you look at images on the wall and it’s quite a passive experience — you don’t really interact with them.’
Current technology for creating gigapixel images — those containing one billion (109) pixels or more — usually involves making mosaics of a large number of high-resolution digital photographs.
Cox uses a 16-megapixel digital SLR camera with a long-lens zoom mounted on a robotic tripod setup that can take grid co-ordinate inputs. He takes between 200 and 350 photographs per scene, overlapping them by around 30 per cent, which takes around 45 minutes.
He then merges and edits the images in post-production software, taking account, for example, of the movement of people to create one large image.
While a number of such gigapixel images have been made available to the public on the internet — generally of panoramic views of city skylines — they employ basic interface software with mouse functions to navigate.
Cox says this is cumbersome, restrictive and does not make full use of the information within the image.
‘It’s one thing taking them [gigapixel images], but it’s another viewing them in an experience that is meaningful. I wanted to break out of the screen and get people more involved.’
His solution was to make use of the motion capture abilities of Microsoft’s Kinect — originally intended as a gaming peripheral, but increasingly ‘hacked’ by programmers for practical, real-world applications.
While Cox said gigapixel photographs are still out of the question for hobbyists, consumer cameras are increasingly including features such as panorama sweep, which stitches together multiple images — thereby potentially lending itself to his navigation method.
‘Maybe one day it will become the normal technology — you’ll buy cameras that are in gigapixels, not megapixels,’ he said.