Filed under: research
Videotrace is the name we gave to an approach to interactive image-based modelling that we published in SIGGRAPH 2007. It was quite popular at the time and generated a lot of hits.
The software has been licensed to a company called Punchcard to commercialise it. It’s currently in beta and you can subscribe to be a part of the testing process at the Punchcard web site.
We’ve continued to work on it of course, and particularly on its application to modelling for Augmented Reality (AR). That work was called Jiim and was published at ISMAR’09, there’s a short post about it here, but the video from the paper gives you a pretty good idea about what it does:
Our latest work in image-based modelling is in reducing the level of interaction necessary. We’re aiming for a one button interface, the kind of thing that might go into a video camera, or a phone. We’ve just submitted the work to a conference, so we can’t tell you much more about it, apart from the fact that one of the most compelling applications is in copying and pasting objects in live video. This means you can effectively pick an object up and move it, synthetically, in live video, while you are capturing it. You can inset 15 copies of your neighbour’s Ferrari into your driveway, and take the video. More news on this as soon as it is published.
Leave a Comment so far
Leave a comment