School project help : Kinect and skeletal mapping

Hey everyone, I’m currently creating a project for my class that will be an interactive photo booth. I will be using a kinect to detect the users body and then adjust an image of clothing over them to help them fit into the particular scene. Now you know what I’m trying to create I have two problems that I need help with as I’m not the greatest at VVVV but I have a good start to the project.

The first problem is getting the kinect to only show the user instead of the blocky textures that I have on there at the moment. I need the user just to be shown without any of the background.

The second problem is getting the clothing and items to map to the body. I can get them to go to pre determined points on the kinect skeletal mapping but I was curious to find out if there was a way to make it so I can make so that the left shoulder of the tshirt goes to the left shoulder of the user etc, whilst adjusting to size of the image to fit the specific user.

I hope this make sense and I would appreciate if anyone could give me any sort of guidance or advice. I have included my VVVV file with this for people to look at. Thank you in advance.

NewProject.v4p (27.6 kB)

well this thing looks obvious however it’s a little bit complicated to achieve
this is made with specially skinned grid in production software like maya
then u take 2d joint positions from kinect and use them as skeleton for the mesh…
i don’t know in your case maybe would be easier to just post the solution… since the process is bit painfull…

u have also second option to use bezier grid instead and calculate some offsets or bounding box to get corner points, this is much simpler way, pretty sure u can look on help patch see how it works…

Hey, I was wondering if you could expand a little on what you posted. I dont 100% understand what you meant and any help would be greatly appreciated.

I have some experience with maya but I wouldn’t know how to do what you’re suggesting. you mentioned posting the solution and I would really appreciate if you could do that to help me continue with this project.

Thanks in advance

I have managed to half solve one of my problems, that is that of being able to get my image into the renderer instead of the usual blocky texture. However I now have a different problem which is now all that appears is a coloured texture instead of my image. I have started a new patch in order to test and try and figure something out but so far I have no real solution.

If anyone could help me figure this out it would be greatly appreciated.

I’ve included the patch and a screenshot of the renderer image.

Test_patch_1.v4p (15.2 kB)

Can you post also witch kinect you are using, since my stuff is kinect2 it’s not gonna work out of the box for kinect1, also you need to have dx11 pack installed and working in order to play with this…

I’m using the kinect 1 and i’m installing that now

ok i had to mention the skinning rig not gonna work at all on kinect1
so someone would have to produce mesh for it…

i made u little example for the bezier approach u can start to play with

i need to get spread of reference points from kinect1 so i can do the skinning…

kinect.zip (121.0 kB)

Hi antokkhio,

just found this thread after searching for a remapped 3d-to-2d-rigging solution for bitmaps. Hence, I´m interested in your Kinect2 solution but am still lacking the skills to do it on my own. Could you provide the patch for better understanding? Many thanx in advance.