Entries with tag markerless motion capture.

Scientists Analyze Olympic Swimmers’ Motions

A research team has isolated and analyzed Olympic swimmers’ and divers’ movements above and below the water’s surface using new motion capture techniques. Chris Bregler, a professor at New York University’s Courant Institute of Mathematical Sciences; Courant spin-off company Manhattan Mocap; and the New York Times used motion-capture technology to study the motions of US Olympian swimmer Dana Vollmer and divers Abby Johnston and Nicholas McCrory. For example, the researchers utilized a specially created motion-analysis system called AquaCap to dissect Vollmer’s butterfly stroke and kick. They found that the kick resembles the motion of a dolphin swimming. The researchers were also able to track the divers from video without having to use motion-capture suits. (PhysOrg)(Scientific American)(The New York Times)(Manhattan Mocap) 

Researchers Develop Markerless Motion-Capture Technique

Typically, 3D motion capture requires multiple cameras and actors who don specialized bodysuits laden with markers that help track their movements. However, Disney Research and Brown University scientists have developed a new technique that doesn’t use markers and that needs just one camera. Typically, motion capture is confined to specially-equipped indoor studios, which constrains its use. Eventually, the markerless system could capture motion from existing film or video, without using actors at all. It could also capture difficult forms of motions, such as those in outdoor settings, for use in games or other interactive applications. The technology could also be used for body-mechanics analysis. In addition to capturing motion, the new system also determines an actor’s biped controllers, which are programs that incorporate the physics of a person’s motion and uses the information to generate poses for animators. This will allow let artists more easily reproduce a person’s motions in animations and change the environments in which the action occurs. The researchers admit their method is computationally intensive and not yet as good as traditional motion capture. However, they said the results are promising. They said that human intervention could improve results and that markerless motion capture may eventually use multiple cameras. (PhysOrg)(The Disney Blog)(Brown University)

Showing 2 results.