Hello @jedeschaud, sorry to disturb you.

We are trying to reproduce the results for ct-icp on the NCLT dataset, and we couldn't succeed so far. How can I get this number?

Checking the implementation, and using the original velodyne_hits.bin file, I also see that you are processing 42764 in total, where the velodyne_sync folder contains only 28127 scans. How did you guys evaluate the results of the system?

Moreover, which ground truth poses were used to carry on the evaluation? According to the implementation, https://github.com/jedeschaud/ct_icp/blob/1ba7ce704e9994d39076089ea3fc0dc4d856fe84/src/ct_icp/dataset.cpp#L151

// Returns the Path to the Ground Truth file for the given sequence
// Note: The name of the sequence is not checked
inline std::string ground_truth_path(const DatasetOptions &options,
                                     const std::string &sequence_name) {
    std::string ground_truth_path = options.root_path;
    if (ground_truth_path.size() > 0 && ground_truth_path[ground_truth_path.size() - 1] != '/')
        ground_truth_path += '/';

    switch (options.dataset) {
        case NCLT:
            throw std::runtime_error("Not Implemented!");
    }
    return ground_truth_path;
}

Thanks a lot in advance

Hi,thanks for your work when I run "./slam -c robust_high_frequency_config.yaml" ,the output is "error while loading shared libraries: libgflags.so.2.2: cannot open shared object file: No such file or directory". I have install gflag in “usr/local/”. And I can't find where link gflags in cmakelists.txt. hope for your reply,thanks a lot.

Hi, I have recenlty built and run ct_icp successfully!
But I have difficulty running another dataset.
I wonder how ct_icp exactly knows about timestamp between frames.
While studying your paper and code, I found something new.

Like this default_config.yaml (permalink highlighted), I found that ct_icp code uses KITTI dataset in the form of frame_####.ply format.
But It cannot work on my own data(*.ply)

Here, I found something new.

In CloudCompare I got to know the ply file of KITTI has x, y, z and timestamp fields.

On the other hand, I have my own data like

I think the difference is whether it has a field of timestamp(PLY_FLOAT32)

Here is my real question!
Could you explain the ply format and how ct_icp uses that format briefly if you can.
Plus, I want to convert my .bag/.pcd file to .ply which has only x, y, z and timestamp file format.
I couldn't find any solution about that.

Best regards.

$ cmake .. -DCMAKE_BUILD_TYPE=Release -DSUPERBUILD_INSTALL_DIR=/home/lingbo/open-source-project/ct_icp/install -DSLAMCORE_INSTALL_DIR=/home/lingbo/open-source-project/ct_icp/install/SlamCore -DCT_ICP_INSTALL_DIR=/home/lingbo/open-source-project/ct_icp/install/CT_ICP -DEIGEN_DIR=/home/lingbo/open-source-project/ct_icp/install/Eigen3/share/eigen3/cmake INFO [Superbuild] -- Successfully found target glog::glog INFO [Superbuild] -- Successfully Found Target Eigen3::Eigen -- Found required Ceres dependency: Eigen version 3.3.7 in /home/lingbo/open-source-project/ct_icp/install/Eigen3/include/eigen3 -- Found required Ceres dependency: glog -- Found required Ceres dependency: gflags -- Found Ceres version: 2.0.0 installed in: /home/lingbo/open-source-project/ct_icp/install/Ceres with components: [EigenSparse, SparseLinearAlgebraLibrary, LAPACK, SuiteSparse, CXSparse, SchurSpecializations, Multithreading] INFO [Superbuild] -- Successfully found target Ceres::ceres INFO [Superbuild] -- Successfully found target yaml-cpp INFO [Superbuild] -- Successfully found target GTest::gtest INFO [Superbuild] -- Successfully found target cereal INFO [Superbuild] -- Successfully found target tclap::tclap INFO [Superbuild] -- Successfully found target tsl::robin_map INFO [Superbuild] -- Successfully found target nanoflann::nanoflann INFO [Superbuild] -- Successfully found target colormap::colormap INFO [Superbuild] -- Successfully found target tinyply::tinyply CMake Error at cmake/external.cmake:13 (get_target_property): INTERFACE_LIBRARY targets may only have whitelisted properties. The property "IMPORTED_RELEASE_LOCATION" is not allowed. Call Stack (most recent call first): CMakeLists.txt:51 (include)

INFO Eigen3::Eigen NOTFOUND INFO /home/lingbo/open-source-project/ct_icp/install/Ceres/lib/libceres.so.2.0.0 INFO /home/lingbo/open-source-project/ct_icp/install/glog/lib/libglog.so.0.5.0 CMake Error at cmake/external.cmake:13 (get_target_property): INTERFACE_LIBRARY targets may only have whitelisted properties. The property "IMPORTED_RELEASE_LOCATION" is not allowed. Call Stack (most recent call first): CMakeLists.txt:51 (include)

INFO tsl::robin_map NOTFOUND INFO /home/lingbo/open-source-project/ct_icp/install/yaml-cpp/lib/libyaml-cpp.so.0.6.3 CMake Error at cmake/external.cmake:13 (get_target_property): INTERFACE_LIBRARY targets may only have whitelisted properties. The property "IMPORTED_RELEASE_LOCATION" is not allowed. Call Stack (most recent call first): CMakeLists.txt:51 (include)

INFO colormap::colormap NOTFOUND INFO /home/lingbo/open-source-project/ct_icp/install/tinyply/lib/libtinyply.so INFO -- [CT-ICP] -- Appending to the INSTALL RPATH the RPATH to the external libraries: [:/home/lingbo/open-source-project/ct_icp/install/Ceres/lib:/home/lingbo/open-source-project/ct_icp/install/glog/lib:/home/lingbo/open-source-project/ct_icp/install/yaml-cpp/lib:/home/lingbo/open-source-project/ct_icp/install/tinyply/lib] INFO [CT_ICP] -- "WITH_GTSAM=OFF gtsam dependent targets will not be built" -- Configuring incomplete, errors occurred! See also "/home/lingbo/open-source-project/ct_icp/cmake-build-release/CMakeFiles/CMakeOutput.log". See also "/home/lingbo/open-source-project/ct_icp/cmake-build-release/CMakeFiles/CMakeError.log".

when I superbuild i met this issue

cmake --build . --config Release Scanning dependencies of target MappingResearchKEU_superbuild [ 12%] Creating directories for 'MappingResearchKEU_superbuild' [ 25%] Performing download step (git clone) for 'MappingResearchKEU_superbuild' Cloning into 'MappingResearchKEU_superbuild'... Already on 'master' Your branch is up to date with 'origin/master'. [ 37%] No patch step for 'MappingResearchKEU_superbuild' [ 50%] Performing update step for 'MappingResearchKEU_superbuild' Current branch master is up to date. [ 62%] Performing configure step for 'MappingResearchKEU_superbuild' -- The C compiler identification is GNU 9.4.0 -- The CXX compiler identification is GNU 9.4.0 -- Check for working C compiler: /usr/bin/cc -- Check for working C compiler: /usr/bin/cc -- works -- Detecting C compiler ABI info -- Detecting C compiler ABI info - done -- Detecting C compile features -- Detecting C compile features - done -- Check for working CXX compiler: /usr/bin/c++ -- Check for working CXX compiler: /usr/bin/c++ -- works -- Detecting CXX compiler ABI info -- Detecting CXX compiler ABI info - done -- Detecting CXX compile features -- Detecting CXX compile features - done INFO [Superbuild] -- [Generation] -- Generating GTest dependency INFO [Superbuild] -- [Generation] -- Generating GLOG dependency INFO [Superbuild] -- [Generation] -- Generating Eigen3 dependency INFO [Superbuild] -- [Generation] -- Generating Ceres dependency INFO [Superbuild] -- [Generation] -- Generating yaml-cpp dependency INFO [Superbuild] -- [Generation] -- Generating cereal dependency INFO [Superbuild] -- [Generation] -- Generating tessil dependency INFO [Superbuild] -- [Generation] -- Generating nanoflann dependency INFO [Superbuild] -- [Generation] -- Generating tclap dependency CMake Error at CMakeLists.txt:527 (add_library): add_library INTERFACE library requires no source arguments.

CMake Error at CMakeLists.txt:528 (target_include_directories): Cannot specify include directories for target "tclap" which is not built by this project.

CMake Error at CMakeLists.txt:468 (install): install TARGETS given target "tclap" which does not exist. Call Stack (most recent call first): CMakeLists.txt:531 (SUPERBUILD_INSTALL_TARGET)

INFO [Superbuild] -- [Generation] -- Generating colormap dependency INFO [Superbuild] -- [Generation] -- Generating tinyply dependency -- Configuring incomplete, errors occurred! See also "/home/user/rosProject/ct_icp/.cmake-build-superbuild/MappingResearchKEU_superbuild/src/MappingResearchKEU_superbuild-build/CMakeFiles/CMakeOutput.log". make[2]: *** [CMakeFiles/MappingResearchKEU_superbuild.dir/build.make:107: MappingResearchKEU_superbuild/src/MappingResearchKEU_superbuild-stamp/MappingResearchKEU_superbuild-configure] Error 1 make[1]: *** [CMakeFiles/Makefile2:76: CMakeFiles/MappingResearchKEU_superbuild.dir/all] Error 2 make: *** [Makefile:84: all] Error 2

My ubuntu is 20.04

great appreciate with opening your great work！ first, can you provide dataset passcord to me， no one relpy in mail， second， And when I see kitti website, the raw point cloud data do not have time information except scan begin and end time, How do you get the time information?

thanks a lot best wishes！

hi，appreciate your great work， but when i use kitti data，which have no distortion in http://www.cvlibs.net/datasets/kitti/eval_odometry.php， that is mean there is no use to Update keypoints when run CT_ICP_GN（） function， and i aslo think we do not need to use pose and velocity constraint in the paper， what should i do to modify the code？

thanks a lot

Hi,thanks for your work. it seems download imgui.git failed, when I run “./ct_icp_build.sh Release "Unix Makefiles" ON ON”.

and the terminal outputs like this: [ 11%] Performing download step (git clone) for 'imgui-populate' Cloning into 'imgui-src'... Permission denied (publickey). fatal: Could not read from remote repository.

Please make sure you have the correct access rights and the repository exists. Cloning into 'imgui-src'... Permission denied (publickey). fatal: Could not read from remote repository.

Please make sure you have the correct access rights and the repository exists. Cloning into 'imgui-src'... Permission denied (publickey). fatal: Could not read from remote repository.

Please make sure you have the correct access rights and the repository exists. -- Had to git clone more than once: 3 times. CMake Error at imgui-subbuild/imgui-populate-prefix/tmp/imgui-populate-gitclone.cmake:31 (message): Failed to clone repository: '[email protected]:pierdell/imgui.git'

hope for your reply. thanks.

thanks for openning your code.

Number of points in sub-sampled frame: 35679 / 126902 Initial ego-motion distance: 0 Elapsed Normals: 35.0737 Elapsed Search Neighbors: 567.692 Elapsed A Construction: 0 Elapsed Select closest: 2.18221 Elapsed Solve: 0.222377 Elapsed Solve: 8.04642 Number iterations CT-ICP : 5 Elapsed Elastic_ICP: 681.094 Number of Keypoints extracted: 4640 / Actual number of residuals: 4180 Trajectory correction [begin(t) - end(t-1)]: 0 Final ego-motion distance: 0.29746 Average Load Factor (Map): 0.324829 Number of Buckets (Map): 16384 Number of points (Map): 39514 Elapsed Time: 683.08 (ms)

here is my log info of one frame, why my Elapsed Search Neighbor cost too much?

best wishes

Hello, thanks for this great work! I saw that the default optimization (in default_config.yaml) is using the Gauss Newton optimization and for the robust configs Ceres is used. The Jacobians for the rotation look for me like an approximation. Is this true, and if yes do you think the approximation error is relevant?

Is Ceres mainly used for the robust loss functions or also to get better Jacobians by the autodiff?

Thanks and best regards, Louis

Hello again, thanks for your contribution. I'm trying to run this code on the MulRan dataset without success.

If you ever ran this code on the dataset, can you please provide some pointers?

I also tried to run with the "PLY_DIRECTORY" with no success:

WARNING: Logging before InitGoogleLogging() is written to STDERR
I20220725 11:23:17.709126 2555704 slam.cpp:351] Creating directory .outputs/
I20220725 11:23:17.713934 2555704 dataset.cpp:278] Found Sequence PLY_DIR
terminate called after throwing an instance of 'std::runtime_error'
  what():  PLY Directory is not supported by read_pointcloud. See the DirectoryIterator.
[1]    2555704 IOT instruction (core dumped)  ./slam -c default_config.yaml

Inspecting the code, I see there is no method to read the PLY directory. https://github.com/jedeschaud/ct_icp/blob/1ba7ce704e9994d39076089ea3fc0dc4d856fe84/src/ct_icp/dataset.cpp#L306

Related to #27, #26, #21

Hello, I am able to run the slam and it works really good, but especially at the beginning of the motion I am facing a possible computation time problem. I am using an Ouster OS0-64, and for what concern the motion on Y an Z I did not notice any problem, but when I walk with my sensor for 10m X remain around the initial value until I reach ~7m. After the output value start to change really fast and reach the correct value (and mantain it for almost all the test). To be more clear:

Real distance(m) | 0 | 0.5 | 1.0  | 1.5  | 2.0     | 2.5  | 3.0  | 3.5 | 4.0    | 4.5   | 5.0   | 5.5 | 6.0   | 6.5  | 7.0 | 8.0 | 9.0 | 10.0 | 10.0 | 10.0 | 
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
/ct_icp/pose(m)  | 0 | 0.1 | -0.2 | 0.015 | -0.03 | 0.3 | 0.40 | -0.2 | 0.10   | 0.15 | 0.16 |  0.2 | 0.10   | 0.15   | 1.0| 3.0 | 5.0 | 7.0 | 9.0 | 10.0  | 

Now my PC CPUs are not in saturation, and them speed is 4.8Ghz so I do not think is an hardware problem, the ram occupation is also low (7Gb/16Gb)

There is a portion of code to modify in order to speed up the computation of the output pose? I have try to change some parameters in ct_icp/include/ct_icp/odometry.h but nothing has changed.

Update Python Bindings, in order to be able to run ct_icp within pylidar-SLAM

TODOS:

SlamCore

• [x] Point cloud API
• [x] Basic types
• [x] IO methods
• [x] Algorithms

CT-ICP

• [x] Datasets
• [x] Map & Neighborhood
• [x] CT_ICP
• [x] Odometry

TESTS, CI, Doc

• [ ] Add test_binding.py to the Github CI
• [x] Documentation on the Readme / Wiki for the python bindings
• [ ] Regression tests on several (shorten) datasets

Hello authors. Thank you all for your excellent work. It's amazing!

I can reproduce the results of KITTI-corrected in your paper (Table I), and I wish to run your ct_icp with pyLiDAR-SLAM (as it produces a loop closure function if I understand your code correctly). However, in your repo and pyLiDAR-SLAM, I cannot find any script to execute pyLiDAR-SLAM with your ct_icp module.

Could you please let me know how to run it or if you could directly upload the pose results of KITTI-corrected (with loop closure)?

I sincerely appreciate your work and consideration! Thanks again!

Hello, I would like to run this slam on my Jetson Nano 4Gb. I have manage to install and build on it, but even if I use Ceres as solver, I can not manage to run the solver on the board GPU. I know that only some type of Ceres option are currently supperted by CUDA:

"CUDA If you have an NVIDIA GPU then Ceres Solver can use it accelerate the solution of the Gauss-Newton linear systems using the CMake flag USE_CUDA. Currently this support is limited to using the dense linear solvers that ship with CUDA. As a result GPU acceleration can be used to speed up DENSE_QR, DENSE_NORMAL_CHOLESKY and DENSE_SCHUR. This also enables CUDA mixed precision solves for DENSE_NORMAL_CHOLESKY and DENSE_SCHUR. Optional."

So, I would like to know which dense linear solver have you use.

Hi, I need to change the configuration for my dataset. Where I can find the parameter meanings? There is an explanation about how they influence the output?

Hi, I have manage to make the slam work on Darpa urban datasets, now I am trying to test it in the tunnel ciruits. As you links say, them al compressed, so I decompress all of them with rosbag decompress file.bag . I have correctly remap the pointcloud topic, lidar model and frequency are the same and at the moment I have not change anything inside the param yaml file. When I launch the simulation the simulation starts, but the odometry topic /ct_icp/pose/odom is not published. As you can see from the following images, also the tf from odom to base_link is not generated.

This image are the TF from the urban datasets:

This one are the TF from the tunnel:

This one is a screenshot from RViz after some seconds: as you can see the position frame are not published even if the robot is already inside the tunnel (the robot stars outside the tunnel)

This one are the topic from rqt:

What I am doing wrong? Can someone please help me?