Purchased a large die cast Unimog truck from Ebay to practice photo scanning it into a 3D model. The technique is based on the tutorial by Gleb Alexandrov at Creative Shrimp. The process uses free (Open Source) software the only cost is having a camera. However, if you use the latest version of Meshlab note that the Surface Reconstruction Poisson is not contained in the Point Set Filter, but the re-meshing one. I am using the Unimog model to explore the technique and establish how best to produce models, so far I have found that lighting the model correctly is very important as you would expect, also the technique does not like specularity at all.
As you can see from my featured image it also manages to place large furrows in the wooden table. Another issue I have found is the production of more than one model point set which I have manged to combine when there are two, but sometimes there can be a s many as 5.
This is early days but I shall keep trying different processes, I may initially sand the model down to make it more diffuse and remove the specular highlights.