So is this motion capture only producing an animation sequence from one view perspective? Unlike the Ronda Rousey example in the next slide which has cameras at all angles, it seems we can only capture one distinct perspective of Shaq.
I think motion capture usually uses more than one camera not only to get multiple perspectives on motion but also to be able to more accurately determine 3D position. It seems like single camera motion capture is more experimental/low-budget so you'd opt for a more traditional motion capture setup for something like ShaqFu.
For motion capture techniques such as this one, how does the camera maintain a consistent record of which ball is which? Are balls just placed far enough apart so that it's always possible to tell which one is which by which ball from the previous frame it was closest to? Does it use the number or relative position of neighbors? How does it hold up with sudden or extreme contortions?
As a follow-up to the comment above^ what is the rule of themb for how many markers are needed, and where markers are place on a person's body?
I was curious about why they choose the background color to be green and found this: https://www.quora.com/Why-is-a-scene-shot-with-a-green-background-for-animation-or-adding-special-effects. That's very interesting.
Here's a good article about how motion tracking works: https://teslasuit.io/blog/motion-capture/motion-capture-what-it-is.
Thanks Lyynnn for posting the link. I was curious why the background is green lol.
can you put these same markers on facial muscles as well to improve realism in terms of facial movement and expressions?
Here's a really cool presentation on motion matching: https://www.youtube.com/watch?v=KSTn3ePDt50.
This requires minimal animators to clean up the animation captured from mo cap suits.