VisualSFM Description
VisualSFM is a user-friendly application designed for creating 3D models through a process known as structure from motion (SFM). This reconstruction software combines elements from several earlier projects I have worked on, including SIFT on GPU (SiftGPU), Multicore Bundle Adjustment, and Towards Linear-time Incremental Structure from Motion. By leveraging multicore parallel processing, VisualSFM enhances speed and efficiency in tasks such as feature detection, feature matching, and bundle adjustment. Additionally, for dense reconstruction, it seamlessly incorporates Yasutaka Furukawa's PMVS/CMVS tool chain, allowing for improved results in the final output. This integration of various tools and techniques makes VisualSFM a powerful resource for those engaged in 3D reconstruction projects.
Integrations
Company Details
Product Details
VisualSFM Features and Options
VisualSFM User Reviews
Write a Review-
Likelihood to Recommend to Others1 2 3 4 5 6 7 8 9 10
Better then Meshroom Date: Apr 01 2024
Summary: To create a point cloud it's quite fast.
Within a few minutes, though I had a 3080Ti not sure how fast on other systems but it was a lot faster than Meshroom.Positive: It works faster and it resolved (Meshroom didn't).
I had a bit difficult object, black, and some shiny paint.Negative: A bit complex to import nvm.
Read More...
You only get a pointcloud not a mesh (but at least that's something).
- Previous
- You're on page 1
- Next