yuvalH9/UMERegRobust

[ECCV 2024] UMERegRobust - Universal Manifold Embedding Compatible Features for Robust Point Cloud Registration

37
/ 100
Emerging

This project helps engineers, roboticists, and researchers accurately align 3D scans of objects or environments, even when the scans are incomplete or captured from very different angles. It takes two "point clouds" – sets of 3D data points from sensors like LiDAR – and outputs the precise movement (rotation and translation) needed to make them overlap. This tool is ideal for anyone working with 3D sensor data that needs robust and accurate alignment for tasks like mapping, object recognition, or quality control.

No commits in the last 6 months.

Use this if you need to precisely align two 3D point clouds, especially when they have large rotational differences or only partial overlap.

Not ideal if you are looking for a general-purpose 3D modeling software, as this tool is specifically for point cloud registration.

3D-scanning robotics-navigation LiDAR-processing metrology computer-vision
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 9 / 25
Maturity 16 / 25
Community 12 / 25

How are scores calculated?

Stars

74

Forks

8

Language

Python

License

MIT

Last pushed

Dec 19, 2024

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/computer-vision/yuvalH9/UMERegRobust"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.