3
$\begingroup$

Link to more info: https://www.youtube.com/watch?v=MnX6pHC9VwE

I saw the Crayola Easy Animation Studio for about $17. It would be cool to hack this to drive Blender animations. It looks to have about 16 potentiometers or the equivalent packed onto a manikin that you can position.

I can see that this would make a great, widely useful hack. Either by writing a (presumably small program) to get the values off of all those potentiometers, or by linking it directly into Blender, which would be a much, much bigger job.

$\endgroup$
3
  • $\begingroup$ Depends on how easy it is to get values from the potentiometers into blender (probably via blender's python api in some way) $\endgroup$
    – gandalf3
    Feb 22 '16 at 5:57
  • 2
    $\begingroup$ I don't see any evidence of potentiometers, or other support for wireless communications (e.g., a battery port cover), and I don't think spatial orientation and wireless communication could be provided at the pricepoint I see. I do note that each exterior surface of each element of the manikin seems to have a unique pattern in green, so that first guess is that it's working from optical recognition. If so, then it becomes simpler, perhaps, as that would be needed is to have optical tracking of the manikin. $\endgroup$
    – brasshat
    Feb 22 '16 at 8:10
  • $\begingroup$ So the answer to your question, or at least the title of it is to use Python and OpenCV (Open Computer Vision) opencv.org and there are tons of books and tutorials available. $\endgroup$
    – MrMowgli
    Nov 20 '16 at 8:21
6
$\begingroup$

The mannequin in the package is simply a few bits of plastic that clip together, there are no electronics in it. The included app that does the animation takes a photo of the mannequin and detects the location of each piece to then position the armature used for animation.

While any mannequin could be used for the same thing, you will find the odd shapes printed on this one will be used to get error free location detection of each piece.

Someone could write an app that detects the mannequin pieces and exports the armature data needed to match it. It will be about shape recognition within an image not reading any data being output by the mannequin.

$\endgroup$
2
  • 3
    $\begingroup$ we already have this software: its called blender ;) (have a look at the motion tracking component of the video editor - see blenderguru.com/tutorials/introduction-to-camera-tracking for example) $\endgroup$ Feb 22 '16 at 14:08
  • 3
    $\begingroup$ Not the same thing. Blender's motion tracking is used to follow a point as it moves in a video. Using the mannequin you aren't tracking it's motion in a video, you have a still image and want to automatically locate each body part in that image, then repeat for another image and interpolate between them. Where motion capture is designed to find points that have moved a small amount from one frame to the next, you can expect large jumps between key poses with the mannequin posing that will break (automatic) motion tracking. The app used is closer to OCR than motion tracking. $\endgroup$
    – sambler
    Feb 23 '16 at 14:40

Not the answer you're looking for? Browse other questions tagged or ask your own question.