Google algorithm allows video of loose Street View photo




Scientists at Google have developed an algorithm that can make smooth video images of the individual photos in Google Street View. In a demo video show scientists how the 24fps footage ultimately look like and how Street View pictures they are based.

Google’s algorithm is a simple principle. The scientists take a series of Street View images, remove a frame, and then let reconstruct their algorithm frame. Then they let the algorithm comparing the reconstructed frame with the original frame, which can pull the algorithm teaching. For example, an image sequence of about 2fps made a series of 25fps.

Scientists Google had to use 100,000 photo series from Google Street View to the algorithm, called Deep Stereo, sharp enough to get it can deal with issues such as poor visual fields by obstacles. “Eventually you get results that are difficult to distinguish from the original pictures,” says Google researcher John Flynn at MIT Technology Review.

For now users will not use a video version of Street View. At this time, it takes about 12 minutes to allow one frame in each set by the algorithm. Therefore, the Google team is still working on further optimization of the algorithm. The work so far Flynn has written a report and published


In: A Technology & Gadgets Asked By: [19122 Red Star Level]

Answer this Question

You must be Logged In to post an Answer.

Not a member yet? Sign Up Now »