I’m a jeweller, who is new to Blender. I want to create animations of the pieces I’ve made; mainly physics simulations, melting effects, undulations et cetera.
Problem: I need to first produce a mesh of these objects. What is the most efficient way to produce a mesh from a photograph, without having to manually construct it (if possible).
I want it to be an exact replica of the photo, as much as possible. I have looked into converting it into an SVG, but I feel like there is a better way to achieve this.
Take a look at Photoscan http://www.agisoft.com/ Photoscan Standard has the features you would need, the Pro version is mostly for geographic applications.
There is a methodology for taking the photos, and the objects should not be shiny, although you could apply dulling spray. You need lots of pictures to get a good, hi res model. Ideally you would mount the object so you can move the camera completely around it. Its possible to photograph sections of an object (like a top and bottom) and join the results, but this lengthens the process, and you would have to use additional software to register the sections.
Also, if the object is not rigid, any difference in the shape between the scans would be an issue, for example if you took some pictures and turned it over to get the other side, the object might not be in exactly the same shape it was when you photographed the first side.
