RTV Tools

Monday, September 16, 2013

Aligning Point Clouds without known coordinates

Not a great position to be in (having point clouds with different coordinate systems and no known common point), but here is some code that may help:
andymiller/FeatureBasedAlignment · GitHub

From GibHub:
The goal is to accurately register two point clouds created from different image spectrums (e.g. visual to infrared), in different coordinate systems of different scales.
Current approach is: 
- Take two point clouds 
 - Estimate normals for each point 
- Find FPFH Features for each point 
- Find likely correspondences between two models (based on FPFH) 
- Use top features to estimate scale, 
 - Run RANSAC to estimate transformation, and eliminate correspondences 
- Use remaining correspondences to run ICP for finer alignment

No comments:

Post a Comment