error multipleviewprogress

[RickMaruyama]

Hello! I tried bash "multipleviewprogress.sh (youe dataset name)". Then got this error after trial of bundle adjustment. please give me advice what I should check.

Especially I'm concerned about following. I cannot find /colmap_tmp folder anywhere. :(

・[Open3D WARNING] Read PLY failed: unable to open file: ./colmap_tmp/dense/fused.ply ・Finished running COLMAP, see ./colmap_tmp/colmap_output.txt for logs

Thank you!!

Bundle adjustment report
------------------------
    Residuals : 304
   Parameters : 237
   Iterations : 101
         Time : 0.039079 [s]
 Initial cost : 2.5143 [px]
   Final cost : 0.232259 [px]
  Termination : No convergence

  => Filtered observations: 0
  => Filtered images: 0

==============================================================================
Finding good initial image pair
==============================================================================

  => No good initial image pair found.

Elapsed time: 0.017 [minutes]
ERROR: failed to create sparse model
cp: cannot stat './colmap_tmp/sparse/0/*': No such file or directory

==============================================================================
Reading reconstruction
==============================================================================

F0622 04:09:03.531672  9854 reconstruction.cc:745] cameras, images, points3D files do not exist at ./colmap_tmp/sparse/0
*** Check failure stack trace: ***
    @     0x7757424fdb03  google::LogMessage::Fail()
    @     0x7757425059d1  google::LogMessage::SendToLog()
    @     0x7757424fd7c2  google::LogMessage::Flush()
    @     0x7757424ff78f  google::LogMessageFatal::~LogMessageFatal()
    @     0x59fe584e06f2  colmap::Reconstruction::Read()
    @     0x59fe5843987a  colmap::RunImageUndistorter()
    @     0x59fe58423499  main
    @     0x775740229d90  (unknown)
    @     0x775740229e40  __libc_start_main
    @     0x59fe584263e5  _start
multipleviewprogress.sh: line 11:  9854 Aborted                 (core dumped) colmap image_undistorter --image_path ./colmap_tmp/images --input_path ./colmap_tmp/sparse/0 --output_path ./colmap_tmp/dense --output_type COLMAP
ERROR: Dense stereo reconstruction requires CUDA, which is not available on your system.

StereoFusion::Options
---------------------
mask_path: 
max_image_size: -1
min_num_pixels: 5
max_num_pixels: 10000
max_traversal_depth: 100
max_reproj_error: 2
max_depth_error: 0.01
max_normal_error: 10
check_num_images: 50
use_cache: 0
cache_size: 32
bbox_min: -3.40282e+38 -3.40282e+38 -3.40282e+38
bbox_max: 3.40282e+38 3.40282e+38 3.40282e+38

Reading workspace...
F0622 04:09:04.685953  9862 misc.cc:310] Check failed: file.is_open() ./colmap_tmp/dense/stereo/fusion.cfg
*** Check failure stack trace: ***
    @     0x70cb13788b03  google::LogMessage::Fail()
    @     0x70cb137909d1  google::LogMessage::SendToLog()
    @     0x70cb137887c2  google::LogMessage::Flush()
    @     0x70cb1378a78f  google::LogMessageFatal::~LogMessageFatal()
    @     0x5a0794afcdf0  colmap::ReadTextFileLines()
    @     0x5a0794a5fec1  colmap::mvs::StereoFusion::Run()
    @     0x5a0794b17786  colmap::Thread::RunFunc()
    @     0x70cb118dc253  (unknown)
    @     0x70cb11494ac3  (unknown)
    @     0x70cb11526850  (unknown)
multipleviewprogress.sh: line 13:  9861 Aborted                 (core dumped) colmap stereo_fusion --workspace_path ./colmap_tmp/dense --workspace_format COLMAP --input_type geometric --output_path ./colmap_tmp/dense/fused.ply
RPly: Unable to open file
[Open3D WARNING] Read PLY failed: unable to open file: ./colmap_tmp/dense/fused.ply
Total points: 0
[Open3D WARNING] Write PLY failed: point cloud has 0 points.
Cloning into 'LLFF'...
remote: Enumerating objects: 774, done.
remote: Counting objects: 100% (26/26), done.
remote: Compressing objects: 100% (21/21), done.
remote: Total 774 (delta 7), reused 20 (delta 5), pack-reused 748
Receiving objects: 100% (774/774), 31.94 MiB | 14.51 MiB/s, done.
Resolving deltas: 100% (409/409), done.
Requirement already satisfied: scikit-image in /home/ubuntu/miniconda3/envs/Gaussians4D/lib/python3.7/site-packages (0.19.3)
Requirement already satisfied: networkx>=2.2 in /home/ubuntu/miniconda3/envs/Gaussians4D/lib/python3.7/site-packages (from scikit-image) (2.6.3)
Requirement already satisfied: PyWavelets>=1.1.1 in /home/ubuntu/miniconda3/envs/Gaussians4D/lib/python3.7/site-packages (from scikit-image) (1.3.0)
Requirement already satisfied: pillow!=7.1.0,!=7.1.1,!=8.3.0,>=6.1.0 in /home/ubuntu/miniconda3/envs/Gaussians4D/lib/python3.7/site-packages (from scikit-image) (9.5.0)
Requirement already satisfied: scipy>=1.4.1 in /home/ubuntu/miniconda3/envs/Gaussians4D/lib/python3.7/site-packages (from scikit-image) (1.7.3)
Requirement already satisfied: packaging>=20.0 in /home/ubuntu/miniconda3/envs/Gaussians4D/lib/python3.7/site-packages (from scikit-image) (24.0)
Requirement already satisfied: numpy>=1.17.0 in /home/ubuntu/miniconda3/envs/Gaussians4D/lib/python3.7/site-packages (from scikit-image) (1.21.6)
Requirement already satisfied: tifffile>=2019.7.26 in /home/ubuntu/miniconda3/envs/Gaussians4D/lib/python3.7/site-packages (from scikit-image) (2021.11.2)
Requirement already satisfied: imageio>=2.4.1 in /home/ubuntu/miniconda3/envs/Gaussians4D/lib/python3.7/site-packages (from scikit-image) (2.31.2)
Need to run COLMAP
Features extracted
Features matched
Sparse map created
Finished running COLMAP, see ./colmap_tmp/colmap_output.txt for logs
Post-colmap
Cameras 5
Images # 3
ERROR: the correct camera poses for current points cannot be accessed
Done with imgs2poses
cp: cannot stat './colmap_tmp/poses_bounds.npy': No such file or directory

[guanjunwu]

just wait for a week, my friend is in his final weeks :) Additionally, it seems like colmap progress failed. There are no matching points in the first frame?

[RickMaruyama]

It looks so. I think the problem is caused by few features point to be matched of my photos. To solve the problem, I'd like to change the feature extract as well as feature matching algorithm used in colmap. can you advise me how to change it? probably, should i use other tool like pycolmap to do it and use the output for 4D gaussian?

I am considering ALIKED for feature extraction and lightglue for feature matching, for example.

Thank you for your help in advance!

[kshonu]

Take a look at your COLMAP installation. Using apt-get won't work. Build from source. I had the same error with COLMAP not being CUDA-enabled.