Webcam and mocap

[Dave_Hingley]

realistically, facial mocap cannot be done with a webcam. you would need a minimum of 3 webcams but more likeyl something like 6 or 8 would be better. You would need to find a way to sync up the webcams . It is possible you would need multiple computers each running a webcam and linked to a main PC which would control all of them. This would also need syncronising.

in order to effectivley work out a camera calibration you would need some kind of reference object like an Lframe with markers on. you would need to the use this reference object combined with DOF, Lens size of the camera (etc) to work out its position in space.

Ok if you have made all this then you might be able to capture something. Once captured you would need a routine to extract the marker placement probably by taking the bightest point of the marker, returning its XY position in each video stream and using this information to constuct a point in 3d spce.

In short webcams whilst being cheap and easily accessible would be great for mocap, The practicalities of actually getting something that gave good results makes it very hard ( if not impossible) to do.

A better way whilst not being true mocap might be to either rotoscope a locked off camera or maybe use 2 (front and side) as an aid to rotoscoping.

[CGTalk Moderation]

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.