I have a proposition for you, College-Now students of Shelby, Ohio. The following video is a snippet of a process that accomplishes a general requirement in some fields of 3D modeling and animation. (Cryptic, huh?) The first person to guess what that process is wins a prize directly related to the video. The video is output from Blender. You can submit your guesses through the IZation Labs contact form. Be sure to include your full name. Have at it!
This is just one of those projects that blows my mind. If you’ve browsed IZationLabs before, you know how appealing computer imaging and CG graphics are to me. Here’s an interesting project that does…. well it just does some mind boggling things, and it does it really well.
This uses algorithms similar to those of motion and camera tracking software packages. By looking at specific points in the video and tracking them from one frame to the next, it can generate a basic 3d model from it. In camera tracking, it uses this information to drive the motion of a digital camera so that you can stick objects on a point and the video and it will stay there. (More about that later.)
It also uses interpolation. Who remembers the slow motion effect from The Matrix? To create that effect, over a hundred photo cameras were set up , all of them focusing on the same spot on a green screen stage. The cameras were then fired sequentially very quickly as the actor made his motions (in this case, the ‘bullet dodge’).
Something to think about. The video we are seeing is the perspective of each camera as it is fired sequentially. After the photos are taken they are stitched together into a video, as if from a moving camera’s perspective. See how it is jittery? Interpolation generates new frames in between each of the actual images that were taken! This results in a video that looks much more smooth, like this: