Developed by five students from Tsinghua University in China and the National University of Singapore, PhotoSketch is a nifty online tool that allows you to create photomontages from simple stickmen-style sketches
The programme works with a simple tool palette on which you draw representations of an intended image, which are then tagged up by the user with keywords.
Using some neat technology explained in the demo video (see below), thousands of images are pulled in from Flickr, Google and Yahoo, from which 20-100 candidate images are produced. The most well-matched of these are processed (processing time is roughly 15 minutes for the items in the picture, plus a few minutes for the background), resulting in an almost seamless—lighting isn’t adjusted—composite picture.
Unfortunately the splurge of interest in the programme has brought the university’s website to a stand-still (hopefully it will be up again soon), but the research paper detailing how PhotoSketch works is available to download.
The potential for producing quick storyboards without the need for any artistic talent is potentially boundless, although we’re not sure about the copyright implications of such a tool. The next question is, how long will it be before PhotoSketch’s talented creators turn their attention to video?
Originally posted on Mashable