Given buildings

  • 1. DEBY_LOD2_4959462

  • 2. DEBY_LOD2_4959460

  • 3. DEBY_LOD2_4959322
  • 4.  DEBY_LOD2_4959323
  • 5. DEBY_LOD2_4906981
  • 6.  DEBY_LOD2_4959458
  • 7. DEBY_LOD2_4959459
  • 8. DEBY_LOD2_4959323_MoFa [link]
  • 9. DEBY_LOD2_4959457 [link]
  • ---
  • 10. DEBY_LOD2_4906970 [link] [x]
  • 11. DEBY_LOD2_4907207 [Chenkun] [link] [x]
  • 12. DEBY_LOD2_4907183 [Yue] [link] [x]
  • 13. DEBY_LOD2_4907184 [Yue] [link] [x]
  • 14. DEBY_LOD2_4907178 [Yue] [link] [x]
  • 15. DEBY_LOD2_4907179 [Yue] [link] [x]
  • 16. DEBY_LOD2_4907180 [Chenkun] [link] [x]
  • 17. DEBY_LOD2_4906965 [Chenkun] [link] [x]
  • 18. DEBY_LOD2_4907173 [Chenkun][link] [x]
  • 19. DEBY_LOD2_4907206 [Chenkun] [link] [x]
  • 20. DEBY_LOD2_4907204 [Chenkun][link] [x]
  • 21. DEBY_LOD2_4907201[Chenkun] [link] [x]
  • 22. DEBY_LOD2_4907196 [Chenkun] [link] [x]
  • 23. DEBY_LOD2_4906966 [Chenkun][link] [x]

---

  • 24. DEBY_LOD2_4907025 [Prabin][link] [ ]
  • 25. DEBY_LOD2_60042 [Prabin][link] [x]
  • 26. DEBY_LOD2_4959320 [Prabin][link] [ ]
  • 27. DEBY_LOD2_4907521 [Prabin][link] [ ]
  • 28. DEBY_LOD2_4907520 [Prabin][link] [x]
  • 29. DEBY_LOD2_4907518 [Prabin][link] [ ]
  • 30. DEBY_LOD2_4907517 [Prabin][link] [ ]
  • 31. DEBY_LOD2_4907516 [Prabin][link] [ ]
  • 32.DEBY_LOD2_4907515 [Prabin][link][]
  • 33. DEBY_LOD2_4907514 [Prabin][link][]
  • 34. DEBY_LOD2_4907507 [Prabin] [link]
  • 35. DEBY_LOD2_4907506 [Prabin][link]


Related files:

Footprints [link]

Settings file [link]

Online buildings viewer [link]

folder with point clouds [link]

Issues

  • DEBY_LOD2_4906970 (Yue)

  • DEBY_LOD2_4907178 (Yue)

  • DEBY_LOD2_4907178 (Yue)

same issues as above - only the balcony on the very top should remain as class 'balcony'. 

  • DEBY_LOD2_4907184 (Yue) 

very good!

  • DEBY_LOD2_4907183 (Yue?)

Is it taken by Chenkun?

  • DEBY_LOD2_4906965 (Chenkun)

  • DEBY_LOD2_4907173 (Chenkun)
  • DEBY_LOD2_4907180 (Chenkun)

  • DEBY_LOD2_4907196 (Chenkun)

  • DEBY_LOD2_4907201 (Chenkun)

  • DEBY_LOD2_4907204 (Chenkun)

  • DEBY_LOD2_4907206 (Chenkun)

  • DEBY_LOD2_4907207 (Chenkun)

  • DEBY_LOD2_4907520(Prabin)

 


Reports

  • Quantitative assessment:
    • IDs of already labeled buildings + approx. time spent
  • Qualitative assessment:
    • Selected pictures of labeled buildings

Working time

Building NoHoursNo of points
116~12.5 mln
218
37
421.5
520.5
6

722.4h
85.5
  • Keine Stichwörter

12 Kommentare

  1. Unbekannter Benutzer (ge35fep) sagt:

    12. Juli 2021

     - Issue 1

    For many small PCD files (around 10M - 100M), it's hard to recognize which kind of elements they are, some cases are shown below

     - Issue2

    The labelling operations are not smooth when processing the large PCD files(more than 200M, for instance, DEBY_LOD2_4959460_00.pcd)

    My machine settings: Dell XPS 13, Intel(R) Core(TM) i7-8550U CPU @ 1.80GHz   2.00 GHz, RAM 8.0G, no GPU

  2. Unbekannter Benutzer (ge35fep) sagt:

    25. Juli 2021

    around 3 hours

    • Install meteor on Windows 10: adding C:\Users\Admin($Username)\AppData\Local\.meteor on system environment variable path, then meteor -v

  3. Unbekannter Benutzer (ge35fep) sagt:

    25. Juli 2021

     

    • Finished the Building DEBY_LOD2_4959462 in around 16 hours.
    • It's more smooth to run the software in Microsoft Edge Browser than Firefox Browser.
    • Distinguish the terrain and ground surface in DEBY_LOD2_4959462.


  4. Unbekannter Benutzer (ge35fep) sagt:

    14. August 2021

    ~ 1h: Refining the results of building DEBY_LOD2_4959462 

    Building DEBY_LOD2_4959460 ~ 18 hours

    ~ 3h

    ~ 1.5h

    ~ 4h

    ~ 2.5h

      ~ 1.5h

    ~ 1h

    ~ 2h

    ~ 2.5h



  5. Unbekannter Benutzer (ge35fep) sagt:

    18. August 2021

    Building DEBY_LOD2_4959322 ~ 7 hours

      ~ 2h

      ~ 2.5h

    ~ 2.5h

  6. Unbekannter Benutzer (ge35fep) sagt:

    05. Oktober 2021

    Building DEBY_LOD2_4959323 ~ 21.5 hours

    ~ 1.5h

    ~ 1h

    ~ 3h

    ~ 1.5h

    ~ 1.5h

    ~ 1.5h

    ~ 2.5h

    ~ 1h

    ~ 2.5h

    ~ 7h


    Building DEBY_LOD2_4906981 ~ 20.5 hours

    ~ 8h

    ~ 1h

    ~ 3h

    ~ 2h

    ~ 1.5h

    ~ 5h


  7. Unbekannter Benutzer (ge47pur) sagt:

    13. Februar 2023

    Building DEBY_LOD2_4959323_MoFa ~ 30.5 hours

     ~ 7h

     ~ 2h

     ~ 4h

     ~ 6h

     ~ 1h

     ~ 3h

     ~ 2.5h

     ~ 5h

    Building DEBY_LOD2_4959457 ~ 35 hours

     ~ 6.5h

     ~ 2.5h

     ~ 4h

     ~ 4h

     ~ 5.5h

     ~ 3h

     ~ 6h

     ~ 3.5h



  8. Anonym sagt:

    10. Juni 2023

    Building DEBY_LOD2_4906970 ~ 24 hours 

    Setting software for windows ~ 10 Hours

     ~ 3h

     ~ 3h

     ~ 6.5h

     ~ 6h

     ~ 5.5h

     ~ 2h

     ~ 2h

     ~ 6h

    1. Unbekannter Benutzer (ge47pur) sagt:

      10. Juni 2023

      From Yue Tan

  9. Unbekannter Benutzer (ge47pur) sagt:

    10. Juni 2023

    Here is the instruction for setting the software semantic segmentation editor (SSE) in windows (take my win11 as an example):

    1. Go to the Github page of SSE: https://github.com/Hitachi-Automotive-And-Industry-Lab/semantic-segmentation-editor, and download it
    2. Replace the setting file with the given setting file
    3. Download the Docker Desktop, here is the link: https://docs.docker.com/desktop/install/windows-install/
    4. Enabel the Hype-V in your computer. You can enable it via Powershell or via settings in your computer, for win11, here is the instruction: https://techcommunity.microsoft.com/t5/educator-developer-blog/step-by-step-enabling-hyper-v-for-use-on-windows-11/ba-p/3745905
    5. Install Docker Desktop. After that, restart your computer. You can start the docker desktop and check if it works well ( for example in your powershell, run the command docker run hello-world )
    6. If not, or there is a warning saying wsl version is not correct, please install or update your wsl, here is the instruction: https://learn.microsoft.com/en-us/windows/wsl/install. Then check if your docker works correctly.
    7. Dwonload the pointcloud you'd like label, and put them under the folder of SSE.
    8. Go into the folder of SSE, open your powershell terminal, and run the following command:
      $env:METEOR_SETTINGS = Get-Content ./settings.json -Raw
      $env:SSE_IMAGES = "D:\_GUG_Master_Program\PointLabelling\semantic-segmentation-editor-master\semantic-segmentation-editor-master\datasets"
      docker-compose -f stack.yml up  
      Here D:\_GUG_Master_Program\PointLabelling\semantic-segmentation-editor-master\semantic-segmentation-editor-master\datasets is the absolute file path of my point cloud, you can switch to yours. 
    9. Open your chrome, and type: localhost:80, enter
    10. Start your labelling! (Zwinkern)  good luck!

    1. Unbekannter Benutzer (ge47pur) sagt:

      13. Juni 2023

      ps: remember to rename your compose file sse-docker-stack.yml as stack.yml, or change the docker-compose -f stack.yml up as docker-compose -f sse-docker-stack.yml

  10. Olaf Wysocki sagt:

    01. Dezember 2023

    For Ubuntu:

    1. Copy the command " curl https://install.meteor.com/ | sh  " to the terminal and run it. If it does not work try: " npm install -g meteor "
    2. Go to: https://github.com/Hitachi-Automotive-And-Industry-Lab/semantic-segmentation-editor/releases, and download the newest version (1.6.0)
    3. Unpack it
    4. Inside the unpacked folder:
      1. Create a new folder for your data, for example "data"
      2. Swap the "settings.json" file with the "settings.json" that I have sent you (make sure that the paths inside the settings file point out to your new "data" folder)
    5. Open the terminal at your home directory (default)
    6. Run the command below, but change the green text to: the global path to point clouds (your new data folder) and the global path to the sse-docker-stack.yml file (which is in your unpacked folder): METEOR_SETTINGS=$(cat ./settings.json) SSE_IMAGES=path_for_pointclouds -f path_for_sse-docker-stack.yml up

      In my case the full command looks like that:

      METEOR_SETTINGS=$(cat ./settings.json) SSE_IMAGES=/home/olaf/Olaf/PhD/3DFacadesLabeling/semantic-segmentation-editor-1.6.0/data docker-compose -f /home/olaf/Olaf/PhD/3DFacadesLabeling/semantic-segmentation-editor-1.6.0/sse-docker-stack.yml up

    7. Open the browser and copy+paste: http://localhost:80


    In case there are furtrher issues when running the software:
    Now, you just need to swap the settings.json file and add the data to the "dataset" folder (/home/ge47pur/project/semantic-segmentation-editor-1.6.0/datasets).

    The only thing you need to do is:

    1. Log in to your account
    2. Go to the folder: /home/ge47pur/project/semantic-segmentation-editor-1.6.0
    3. Open the terminal and type:
      METEOR_SETTINGS=$(cat ./settings.json) SSE_IMAGES=/home/ge47pur/project/semantic-segmentation-editor-1.6.0/datasets docker-compose -f stack.yml up

    4. Open the localhost location 80 or 3000
    5. Start labeling :))



    In case you will get the permission error again:

    1. Open terminal at home location (default location)
    2. Type:   
      sudo chmod 666 /var/run/docker.sock