I’m looking for a body scanner solution which is able to bring human persons within Unreal Engine. Do you have any suggestions?
Furthermore, it would be great if the scanned body would be rigged too, so to be used as a Skeletal Mesh in Unreal. Are you aware of anything like that?
Thanks in advance for your replies.
PS: I hope that I’m posting in the right section of the forum.
You can look for scanning services that might be nearby. They’ll have a set up where you can have a person stand in the middle of a bunch of cameras and do an instant capture. However, the 3D models you get from scanning will need to be cleaned up to be usable, they’ll be higher detail than you need and for something like an animated mesh (characters) you will need to retopologize the model so that it animates well. More than that, things like a person’s face will be incomplete because hair won’t scan well and eyes won’t scan well either due to the reflective properties. So there’s a bit more involved with getting a 3D scanned head as a clean game asset. The 3D scanning company might be able to do some of the mesh cleanup for you, but you might need to hire someone else to do that and rig it for animation.
thanks for your reply. I’m already aware of these kind of solutions (third-party companies). However, I was looking forward for an in-house solution, a scanner that I am able to purchase for myself.
In terms of cleaning the model, I have resources that might take care of that so this is not an issue. Ideally though, I would prefer something that can come as close as possible to automatizing this process. In fact, the final result is not a game, but rather an application, so it is fine for the model to have slight imperfections. Do you know of any 3D scanning solutions which if do not suit, offer something close to these needs?
It depends on how much money you have, a cheap solution is something like a Kinect sensor, which can do a body OK but not a face (if you want skin detail)
But like I said, for something like a human face you aren’t going to be able to automate a cleanup, eyes and hair just don’t scan well and you have to do some work to fix those areas. The rest of the body comes out pretty well but if you need to animate them then you’d need to get a mesh with proper topology.
We are aware of the Kinect solution, and we are evaluating it. However, we were also looking into other ways to bring the body in. For instance, by using sensors to register some measurements of the body, and use those to build a 3D model, and then apply the textures on it. We aren’t looking for achieving photorealism, but enough to have a virtual avatar who looks like the user.
Either way, we are open to any hack. So, even though there is no (a proper) body scanning per se, some system (of sensors) able to create a virtual avatar for the user who looks similar enough to them will work for us.
That sounds like some custom software you’d have to develop–for example you could use the Kinect to take some general body measurements and then fit a pre-made mesh to that, but it’s something you’d have to have developed for you.