By Evan Ackerman
Like it or not, most of the time when you take a picture, a significant portion of the image is (for lack of a better word) useless. Or rather, there are lots of areas of the image that aren’t conveying any important information. Obviously this excludes you artsy-picture types, but if you’re just trying to show and/or explain something (which most of the images on the Internet, excluding porn, are probably trying to do) wouldn’t it be useful if you could make images smaller by selectively removing the least important pixels? Ariel Shamir of the Efi Arazi School of Computer Science in Israel has developed some software to automatically do just that, in realtime no less. Watch and be amazed:
So it’s obviously still a work in progress, but it’s already pretty clever and seems to have a working automated mode as well as a decent user interface. What with the population of itty bitty mobile devices capable of Internet browsing exploding like a rabbit warren underneath a Viagra factory, a little piece of software that can work behind the scenes to dynamically re-size images (in the same way that HTML dynamically re-sizes) could have a huge range of applications.
[ Content-Aware Image Resizing (PDF) ] VIA [ Neatorama ]
