Posting the way you are using GLIM and related packages.

[koide3]

TL; DR: Sharing that you are using GLIM (and related packages) helps the author receive recognition within his organization and enables continued work on this project.

Long description: The author of this package is NOT a full-time OSS developer but just a researcher who dedicates the majority of his working hours to other assignments. His time for OSS projects is limited, and to sustain his involvement, internal acknowledgment within his organization is crucial. Metrics like the number of stars and forks are often overlooked by executives, who may not grasp their significance. Conversely, stating "this package is used by A/B universities and C/D corporations" can effectively communicate its value. So, if you find this package useful for your work, we kindly ask you to leave a comment in this thread. Just saying hello to the author is appreciated. If you could share about your organization and how you use the package, that would immensely support the author.

EDIT: It would be much appreciated if you could share screenshots or videos of your results at Thirdparty results. I love to see my packages being used in the world.

Related: https://github.com/koide3/small_gicp/issues/3

[robot-Yang]

Hello, koide-san, thank you for the wonderful work, I am interested in using GLIM for an indoor wheelchair navigation system, by the way, I am working at University of Tsukuba.

[South-River]

hello，thank you for your great work. i want to use your work in multi robot collaborative mapping. by the way, the initialization and estimation of accurate relative state and ensuring low-bandwidth(lower than 1MB/s between robots) might be a great idea in improving this work (like kimera-multi). as far as i know, there's not so much people working on this. im working in zhejiang university:).

[koide3]

@robot-Yang Thank you for your message! I'm so glad to hear from a good neighbor!

[koide3]

@South-River Yeah, applications to multi robot settings must be an exciting topic. Actually, a student of mine is now working on multi-robot mapping, and I hope we can share interesting results with the community in the near future. Thanks for your comment!

[changh95]

Thank you for another great contribution to the robotics community!

I'm trying to use GLIM and its related packages for my non-commercial personal project where I do LiDAR mapping of beautiful tourist places in my home country. I'll try it out with a number of LiDARs and mobile platforms, and in the future I will try to integrate into more robotics / autonomous driving projects.

I also plan to share the repo amongst my community of SLAM experts in Korea. There should be many who are working in the field of robotics and autonomous driving, both in academia and industry (I will definitely mention the commercial license).

---

EDIT: There's a typo in the title of this issue. "Your" -> "You"

[koide3]

@changh95 Thanks a lot! I hope this package will help your nice projects!

[oliver-batchelor]

We (the University of Canterbury, New Zealand) attempt to map orchards with RTK GNSS, Mid360, and (12!) cameras. We use forms of photogrammetry (NeRF, Gassusian Splatting) and aim to align the sub-scans (e.g., from two sides of a row) using the LiDAR map (without it, there's a tendency to drift slightly).

[koide3]

@oliver-batchelor Mapping with a dozen of cameras is interesting! Good luck on your work!

[peter-ivarsen]

Hi, and thanks for sharing this project Koide-san. I work with a Norwegian reasearch institution and have tested GLIM in connection with a project we are part of, on robotized ship recycling. So far I am very impressed by both the robustness and accuracy of this system.

[fyandun]

Hi!

I'm from Carnegie Mellon University. I'm using your software for benchmarking different Lidar-inertial SLAM systems. Great work!

[dezhi-l]

Thank you. awesome project !

[hatem-darweesh]

Amazing work, I like the simplicity and easy to understand system of configuration files. I was using hdl_graph_slam for 3 years. currently I am moving to use GLIM. In my work I have used it for creating HD Maps for autonomous driving applications beside other confidential projects ;) I would love to see optimized parameters for indoor mapping.

[se7oluti0n]

Hi!

Thank you for your great work. Among open-sourced 3D Lidar SLAM, I found that your work has the most well designed code base. The module odometry, submapping, global mapping and extension modules. All of that is very clever and make it easy to understand and extend.

Currently I use Glim for outdoor robot navigation. I have added relocalization, localization to GLIM. Also I added wheel odometry factor from @TakuOkawara -san (https://github.com/TakuOkawara/full_linear_wheel_odometry_factor), and it solve the drift in the long narrow bridge (see in the video below)

Here is the video GLIM with localization and wheel odometry - Cong vien Thien Van Hoc

Update: Integrate GPS Factor to align map to NED frame and help correct both horizontal and vertical drift. Video: https://www.youtube.com/watch?v=neGONn7brmI

[koide3]

@se7oluti0n Thanks a lot for your comment and sharing the very nice video! I'll tell Taku that his package has been used in the community, and I'm sure he will be so happy!

[oliver-batchelor]

Hi!

Thank you for your great work. Among open-sourced 3D Lidar SLAM, I found that your work has the most well designed code base. The module odometry, submapping, global mapping and extension modules. All of that is very clever and make it easy to understand and extend.

Currently I use Glim for outdoor robot navigation. I have added relocalization, localization to GLIM. Also I added wheel odometry factor from @TakuOkawara -san (https://github.com/TakuOkawara/full_linear_wheel_odometry_factor), and it solve the drift in the long narrow bridge (see in the video below)

Here is the video GLIM with localization and wheel odometry - Cong vien Thien Van Hoc

This is awesome - we might try to integrate this. We did some tests on a completely flat field to see how the RTK GPS can be used effectively, and the odometry can fail since it all looks the same to the LiDAR. Though this is not our normal use case it seems like it should be useful to add for robustness anyway! (differential GPS could also help odometry too I suppose but maybe this is simpler)

[skohlbr]

Hi! Thanks for making GLIM available! We're currently evaluating it for use with inspection robots and things looks quite promising so far.

@se7oluti0n I was wondering if you plan on making your fork available, as your video looks great and we'd like to test those features, too.

[se7oluti0n]

@skohlbr thanks for interested in my work. I hope I could publish some code soon. Which feature you are most interested in?

[skohlbr]

@se7oluti0n Toss-up between relocalization and odometry integration (slight edge to relocalization as I believe the solution space might be bigger). Would be great to be able to check it out, but your call of course :)

[oliver-batchelor]

@skohlbr thanks for interested in my work. I hope I could publish some code soon. Which feature you are most interested in?

I'd be really keen to try wheel odometry - we'll be working on the same thing here at some point!

[se7oluti0n]

@oliver-batchelor I have published the wheel odometry implementation. you can check glim, glim_ros2 repo in my page.

[mamoto]

i used glim with LiDAR VLP-16, IMU VN-100s, Jetson Orin NX 16G configuration for very good data results, the software is awesome. i want to increase the number of images of the final map and add RGB data from the camera synchronized with the point cloud but i don't know how to configure? can you help me?

[koide3]

@mamoto Thanks for your comment, and sorry, RGB data support is not included in this package. I think you can implement a some post process to generate a colored dense point cloud.

[Efesendil]

@mamoto Thanks for your comment, and sorry, RGB data support is not included in this package. I think you can implement a some post process to generate a colored dense point cloud.

Hello, Thanks for this great work ! I've quickly tested the package with the given sensor data, I'd also like to know more about point cloud colorization tool since you created a chapter in the documentation about it. As you are saying it isn't included in this package means that there's a related package for colorization or is it under paid license ? If not implemented would you have any recommended packages to use in order to achieve that (Like lidar-camera calibration and writing pixels into scans) ? And are you thinking about to create such a package in future ?

[mamoto]

@mamoto Thanks for your comment, and sorry, RGB data support is not included in this package. I think you can implement a some post process to generate a colored dense point cloud.

Hello, Thanks for this great work ! I've quickly tested the package with the given sensor data, I'd also like to know more about point cloud colorization tool since you created a chapter in the documentation about it. As you are saying it isn't included in this package means that there's a related package for colorization or is it under paid license ? If not implemented would you have any recommended packages to use in order to achieve that (Like lidar-camera calibration and writing pixels into scans) ? And are you thinking about to create such a package in future ?

Thank you for your feedback on my question, I'm also looking into it to edit the source code you provided. I will try to finish it soon and send it to you. I think everyone can use it without having to pay. And this is the device I built.

[Efesendil]

@mamoto Thanks for your comment, and sorry, RGB data support is not included in this package. I think you can implement a some post process to generate a colored dense point cloud.

Hello, Thanks for this great work ! I've quickly tested the package with the given sensor data, I'd also like to know more about point cloud colorization tool since you created a chapter in the documentation about it. As you are saying it isn't included in this package means that there's a related package for colorization or is it under paid license ? If not implemented would you have any recommended packages to use in order to achieve that (Like lidar-camera calibration and writing pixels into scans) ? And are you thinking about to create such a package in future ?

Thank you for your feedback on my question, I'm also looking into it to edit the source code you provided. I will try to finish it soon and send it to you. I think everyone can use it without having to pay. And this is the device I built.

Wow, seems great ! Definitly waiting for it ! I am really curious about the results, would be happy to keep in touch. I will see what i can do on my side, currently working on carla-sim.

[koide3]

@Efesendil

As you are saying it isn't included in this package means that there's a related package for colorization or is it under paid license ? If not implemented would you have any recommended packages to use in order to achieve that (Like lidar-camera calibration and writing pixels into scans) ? And are you thinking about to create such a package in future ?

The colorization module is maintained as a closed source code as described here. I'm not planning to release such a package, but I hope someone will create and share it with the community. I designed GLIM to be easily extensible and released it with a permissive license, so I believe anyone can create and release such an extension module by themselves.

@mamoto I'm glad to see you trying to develop and share the module. It is what I want the community be like. Feel free to let me know when it gets advanced or you face some technical troubles.

[se7oluti0n]

@Efesendil Koide3-san also publish the lidar-calibration package here https://github.com/koide3/direct_visual_lidar_calibration . Hope you can use it and create a extension module colorization pointcloud.

@mamoto looks like you are from Vietnam too. Glad to see more Vietnamese working in this field.