BoingBoing already picked up on this incredible video that basically shows a technique for enhancing video that is similar to what is already familiar in “photoshopped” still images¹

They’ve worked out a way to extract spacial information from video frames and then identify spacial features through time (or at least through video frames). I haven’t read the research papers but I wonder if the same technological method that compresses digital video by only changing on screen what changes on the image as time progresses could be (or is being) used to identify different parts of the image through time.

Anyhow, what the video demonstrates is absolutely unbelievable and potentially brings awesome new power to amateur filmmakers to enhance their video. It may start with integrating high resolution photography to make any film HD, but there’s no reason why this can’t be used to radically change videos.

One potential legal issue that the technology presents is to change the value of video evidence. It is already well established that photographs can be modified and should be carefully considered before admitting them into evidence and giving them much weight, but I suspect the attitude toward video has been much more confident. With the technology demonstrated here, that may need to be seriously reconsidered.

¹ although it’s ironic that I would use “photoshop” as a verb and then link to the site that’s making the awesome online Photoshop killer called Aviary (↑)