In the olden days there was something called Facegen which let you upload a front and a side profile of a head, click some pointers, and then it would generate the 3d model of the head. That was 15 years ago. How come that thing isn’t INSIDE Blender by now?
There is open source Mesh Room. I have not tried it in a while but people are getting good results from it. I think incorporating photogrammetry to the level users would expect and demand it to be now would be a very big job. Right now photogrammetry is made in stand alone and very fast evolving apps. The tech is still quite fluid I think and moving forward at a very fast pace. My preference would be to have more tools within Blender to work with and refine the scan data itself. It is quite far along in this direction anyway. The bulk of the work comes in working on and refining the data for whatever uses it is required for.
None of the other major 3D content creation apps have a photogrammetry application built in yet to my knowledge. But who knows where this will go as the tech becomes more standardised and widely used. And Blender is not a standard 3D content creation app either and more a self contained platform so I guess never say never. I can think of many development prioritys though that I think most users would be wanting in the shorter term such as true layered texture painting and continually improved sculpting performance. Which actually would help so much in working with raw scan data as well all within Blender.
Facegen was generating avatars with less details than what would be expected nowadays.
And it was not an open source technology.
People making Free Open Source Software were more inclined in developing tools like Makehuman to allow people to create characters able to be avatars or pure inspiration.
There was an attempt to integrated photogrammetry tools, some years ago.
Taking into account a lot more details than what facegen was doing.
2.8 refactor impacted the project.
And as Toka said, Mesh Room alternative showed up and become popular.
There are free open source alternatives to Mesh Room.
Regard3D is a simple to use without nodes.
Colmap is more rough to use. But it is GPU accelerated.
1 Like
So why don’t they put it back inside?
Because programming stuff like that takes time and effort, and is probably not a prioritized think for the Blender devs.
Personally I would recommend 3DF Zephyr for photogrammetry, it’s free for most uses and does a better job than Meshroom afaik.