Skip to content

I modified the colmap,when it reconstructs from known pose ,only let it optimize rotation ,fixing position!

Notifications You must be signed in to change notification settings

yuancaimaiyi/Mapping-base-lidar-pose-or-vslam-pose

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

10 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Mapping-base-lidar-pose-or-vslam-pose

I simply modified the colmap,when it reconstructs from known pose ,only let it optimize rotation ,fixing position!
Why do I do this, because when reconstructing from a known pose, if you use the rtk position, there is no need to optimize the position during BA fine-tuning. You can only fine-tune the rotation(3dof +3Dpoints optimize), but this is a very loose coupling, if you want to use rtk to constrain BA, you can look at openmvg ---"sfm_data_BA_ceres.cpp". You can implement it in colmap immediately!!!!!!!

Reconstructing from a known pose reference : https://colmap.github.io/faq.html?highlight=known%20pose#reconstruct-sparse-dense-model-from-known-camera-poses

first step

(1) base/cost_functions.h replace your colmap code

(2) controllers/bundle_adjustment.cc replace your colmap code

(3) optim/bundle_adjustment.cc replace your colmap code

second step

example folder

..........model

................images.txt // your piror pose

................cameras.txt

................points3D.txt //empty file

..........trans.py // trans colmap tvec to camera center and trans camera center to colmap tvec(-R.Ttvec=C and tvec=-RC)

..........run 3D_from_prior_pose.sh

About

I modified the colmap,when it reconstructs from known pose ,only let it optimize rotation ,fixing position!

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages