I noticed when scanning objects using photogrammetry or the AI-based Neural Radiance Field (NeRF) in apps such as Kiri Engine that the original size is not retained in the object.
How can I actually retain the original size, or set the proper sizes when or after scanning objects like that? So that, when 3D printed it is the same size as the original?
When scanning something, including a simple reference geometry, such as a cube, is best. This can then be used to scale the final model to the correct size and as a reference for if the item has the correct angles. Besides cubes, coins (esp. matte ones) make handy markers and scale references.
If your scan contains color values, a simple photo scale angle could suffice. That item would be taken directly from photography. There it is sometimes required to contain size information and a pointer to the angle the photo was made at. This is made by using a square or angle with scale markings and distortions of the item can be solved later till the angle appears properly and to the right scale.
