HomeTipsCorrect Shaky Home Videos With Voodoo Camera Tracker

Correct Shaky Home Videos With Voodoo Camera Tracker

Computer scientists at Adobe, University of Wisconsin and Madison have developed a technique rather a software that can correct the shaky and wobbly home videos to give them a big-budget movie kind of feel.

The software uses 3D modeling technique to guess the intended look and then warps the original video to adjust. The process uses an off-the-shelf software called Voodoo Camera Tracker that can reconstruct a camera’s path through 3D space from a video sequence.

Well here is a list of the process and features that the new software will offer :-

Warped thinking

With the voodoo camera tracker you can easily reconstruct the cameras path through 3D space from a video sequence.

# Optical illusion

According to the research scientists at Adobe it is practically impossible to translate the camera in a physically correct way. So they took a different approach and tried to trick the eye into seeing a stable video by carefully warping salient parts of the scene onto stable paths.

shaky ground

# Ghosts in motion

The existing 2D techniques reduce the effect of camera shake but in no way they can read the cameras path through 3Dspace. This isn’t a problem for the new technique because it distorts each frame in isolation, instead of combining multiple frames.

# Wave goodbye

Well the software is having some limitations of its own like that it can be only applied only after the entire film is shot and the resulting product will need a little bit of cropping but the final product is actually good.

P.S:- To conclude i would like to say that now you can easily shoot your home videos and present them in a professional way without spending big bucks. It  looks like a pretty great tool for amateur filmmakers, and should be available in about two years.

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Recent Articles

3,016FansLike
0FollowersFollow
32,600FollowersFollow
2,270FollowersFollow
18,100SubscribersSubscribe