3D scanning is currently a bit difficult on the new 400 Series cameras, as commercial 3D model software packages do not yet support it. So to make a 3D model with the 400 Series, you would have to scan the body as a 'Point Cloud' type of scan (an image made up of dots) and then convert the point cloud file into a solid model using a software package such as MeshLab.
Body scanning is much easier with the older R200 camera model, as it is well supported by commercial 3D model software packages and can take a full solid scan of the human body instead of having to convert a point cloud. The R200 excels at scanning large objects and the human body, and is still available from Intel's online store as part of the Robotic Development Kit bundle deal of an R200 and a single-board computer called an Up Board.
As the R200 is a retired product that no longer receives updates though, the 400 Series cameras may be a better choice if you need future-proofing and ongoing software updates. the 400 Series cameras also have a sample program for measuring dimensions called 'Measure'.
MeshLab is a standalone piece of software. If you google for 'convert point cloud to mesh c++' then you can get some leads about how to approach writing your own integrated solution. Also see the Wikipedai entry about point cloud conversion. the 'See Also' section in this link also provides useful links for methods of conversion, such as the open-source Point Cloud Library (PCL).
Creating a 360 degree point cloud does not seem to be an easy thing to do. The link below was the one that emerged from my research of your question as the most potentially useful.
I remember at the Intel IDF 2015 conference, Intel were demonstrating the scanning of a room with a RealSense device whilst walking around it. I believe they were using a cancelled RealSense product called the Smartphone Developer Kit, which was an Android phone with a ZR300 camera and Google Project Tango AR in it (Google cancelled Tango later on and replaced it with ARCore).