61–70 of 77 posts
Earlier quoted context omitted.
Impressive. How much time did you invest for that model?
Biggest job was sourcing the images. That took maybe two hours. Of course PC had to compute some hours, but I can do other things while that is going on so not really a time investment as such (do it over the night).
Has anyone tried it? I've been trying for a while to find a way to 3D scans stuff to 3D print but the quality of the final file has always been so crappy that I didn't even want to waste time and filament printing it
What have you tried that's giving you the crappy file? I've used Scandy Pro[1] a few times and was surprised by it's resolution and ease-of-use. Only downsides are it's iPhone only and you get 1 free scan/24 hrs. 1- https://www.scandy.co/apps/scandy-pro
Earlier quoted context omitted.
It seems that there is no photogrammetry software that supports Linux and Radeon at same time.
COLMAP can do CPU-only calculation. It takes a big longer but it’s not horrible.
I've been watching people do photogrammetry for a while now, but the results have seemed really bad. When you remove the texture, the underlying model always seems off. This is important for doing object->mesh->modify->3D-print workflows.
This implementation looks a bit better than the rest.
For those with laptops or who don't have Nvidia GPUs, Meshroom can be used for free in the cloud using Google colab https://github.com/alicevision/meshroom/wiki/Meshroom-in-Goo... . I created this scan using this technique and I only have an integrated Intel GPU: https://sketchfab.com/3d-models/parvati-cantor-arts-center-a...
For those with laptops or who don't have Nvidia GPUs, Meshroom can be used for free in the cloud using Google colab https://github.com/alicevision/meshroom/wiki/Meshroom-in-Goo... . I created this scan using this technique and I only have an integrated Intel GPU: https://sketchfab.com/3d-models/parvati-cantor-arts-center-a...
It looks great, but weighing in at almost 400K triangles and 300K vertices though, so it's going to be difficult to do anything practical with a model that size.
Because I'm trying to get a photogrammetry pipeline working, and able to produce 50M vertices near-term. 1G long-term. All without deferring to chunking. 10fps on the serial parts is a medium goal for throughout.
I'm interested to try it. I've been watching people do photogrammetry for a while now, but the results have seemed really bad. When you remove the texture, the underlying model always seems off. This is important for doing object->mesh->modify->3D-print workflows. This implementation looks a bit better than the rest.
For those with laptops or who don't have Nvidia GPUs, Meshroom can be used for free in the cloud using Google colab https://github.com/alicevision/meshroom/wiki/Meshroom-in-Goo... . I created this scan using this technique and I only have an integrated Intel GPU: https://sketchfab.com/3d-models/parvati-cantor-arts-center-a...
It looks great, but weighing in at almost 400K triangles and 300K vertices though, so it's going to be difficult to do anything practical with a model that size.