Intelligent motion capture without markers

Normally motion capture recordings associated with a lot of effort: the actors have to special, provided with markers, wear suits and be filmed from a variety of cameras from all angles. With this information, their movements are recorded and digitized as spatial vectors which are then used to animate virtual characters in a natural way. The Max Planck Institute of computer science has developed a method to minimize this expense - there are only needed a few cameras and no marker more, the actor no longer have to be in the skin-tight Spezialanzuege (and her play disabling) constraints . PIC 2: traditional motion capture with Spezialanzuegen on the set of Lord of the Rings Possibly this success is by a kind of knowledge of the systemusing a model of skeletal movement of the actors - it can continuously make use of the current 2D images, the 3D model into line with it. In real time the virtual characters are animated so realistic - and also several people who hide in part mutually can so be captured digitally (along with her clothes), even the folds of a ball gown to be properly recognized. In the future it will also be possible to make motion capture shooting outside the studio in the wild. For more knowledge-wishers that (as usual in scientific work) very technical paper here: Spatio-temporal motion tracking with unsynchronized cameras Bilder zur Newsmeldung:  more infos at bei www.uni-saarland.de deutsche Version dieser Seite: Intelligentes Motion-Capturing ohne Marker