Video Composition based on motion tracking

    Feb 08 2006 | 3:45 pm
    Hey Friends
    This is the deal:
    im building this "art-project" with MAX MSP, JITTER and SOFT VNS in which a DV camera tracks the presence of a user against a white background, and then according to this info "draws" the silohuette of this presence with a series of (miniaturized) photo stills or video loops. so the user comes in front o the camera and is faced with a shape that resembles his/her own body, but is actually composed of a multitude of videos/images.
    im using SoftVNS to do the tracking. easy money. then i go to jitter where i downsample the (now binarized black and white image) so i can see the silohuette in a more squareish-pixelated way.
    what i have to do now is translate these big chunks of white into images. and i have no clue as to how to do that.
    can anyone point me in the right direction here? i'd be insanely grateful.
    this is what i have so far
    thanks a lot guys federico