It exist a software called MeX from company Alicona which now owns to Bruker, wich allows to make stereographic 3D images in SEM by taking images under different tilt angles of the sample.
I have made a comparison of roughness evaluation using AFM and SEM&MeX. You find the results in the pdf document.
The SEM results are generally a little bit smaller than AFM results, but it is unknown, whether this is due to SEM or AFM measurement. Further comments you find in the file.
You have to make some assumption about the z-dimension unless you have confocal capability. Sample preparation techniques e.g. slicing may be useful but everything obtained in this manner is an artifact and not representative of the whole.
It exist a software called MeX from company Alicona which now owns to Bruker, wich allows to make stereographic 3D images in SEM by taking images under different tilt angles of the sample.
I have made a comparison of roughness evaluation using AFM and SEM&MeX. You find the results in the pdf document.
The SEM results are generally a little bit smaller than AFM results, but it is unknown, whether this is due to SEM or AFM measurement. Further comments you find in the file.
MeX is a good software (spot measurements, profiles, 3D pictures, roughness calculations). I believe there are some other programs to work with stereopairs. By the way, calculations and measurements are simple. If just a few points are of interest, it is possible to measure with no software at all (but then you have to read about the method).
The only thing I do not understand, what kind of 3D reconstruction of thin films you want to do. There is another method (dealing with brightness of BSE image), not as dependable, but may be better suited for your needs.