Moving image

A group of Stanford University researchers specialising in artificial intelligence have developed an algorithm that makes it possible to embed photographs into video.

The researchers, computer science graduate students Ashutosh Saxena and Siddharth Batra, and Andrew Ng, an assistant professor, see great potential for the technology they are calling ZunaVision.

They say a user of the software can easily place an image on almost any planar surface in a video, whether a wall, floor or ceiling. And the embedded images don't have to be still photos - users can also insert a video inside a video.

Here’s how it works: first, the so-called 3D Surface Tracker Technology algorithm analyses the video, paying special attention to the section of the scene where the new image will be placed. The colour, texture and lighting of the new image are then subtly altered to blend in with the surroundings.

The result is a photo or video that appears to be an integral part of the original scene, rather than a sticker pasted artificially on the video.

Other technologies can perform these tricks, but the Stanford researchers say the existing systems are expensive, time consuming and require considerable expertise to use.

Some of the Stanford work grew out of an earlier project, Make3D, a website that converts a single still photograph into a brief 3D video. It works by finding planes in the photo and computing their distance from the camera, relative to each other.

A hands-on demonstration of the technology can be seen here.