问题
We have a multiple 2D planar image of an object scanned from a fan-beam perspective. An example is in Fig 5 below. We have multiple grainy dotted planes to scan the whole image.
The issue with these images is that they cannot be directly mapped to a 3D plane due to the fan beam deformation.
Is there correction algorithms/methods that can be recommended so that these planes can be correctly mapped to 3D plane and an object can be reconstructed properly?
回答1:
Depending on how you store your data, there might be various approaches. Guessing that you store the data as points ("grainy dotted planes"), you can do an interpolation of the corresponding points in the consecutive planes and thereby get the scan of the entire object. It does require the points to be in the same frame, so you might have to do some kind of transformation to find the parameters of each plane to a global framework.
Another procedure might be use of a least square fitting of each plane which can then be used to map together the object. You might find some helpful approaches of scanning 3d objects using 2d methods. Hope this helps.
来源：https://stackoverflow.com/questions/18903542/convert-2d-planes-to-3d-model