How can I convert multiple images into a 3D mesh using Blender?

Question

Basically, we are a group students, who want to make a video, and convert the main object from the video to a 3D Blender mesh. We make a video, split it into frames, and we want to process all the images, and make a mesh out of them. One of the major problems is that I can't find any way to automate mesh generation, and I would like to know where I can find any useful information regarding my problem, or any way that I could achieve it.

Answer 1

The technique you are thinking of is called photogrammetry .

While there are several options available, you will find most are still basic tools, for example some only map the images onto planes to offer a limited perspective change while moving a camera through the 3D scene.

With some pre-planning, you also have the option of using a 3D scanner instead of pre-existing images. There are some open designs of 3d scanners that combine a laser, a camera and software to automate the creation of 3D models.

Many 3D modelling tutorials show the use of reference images while modeling. This is a common technique that helps to get 3D models that appear real. While this technique does provide a way to make realistic 3D models, it also requires learning 3D modeling and texturing, which I expect you aren't experienced at if you are asking for software to do it for you.