What would you like to see next?


#1

An exciting new year is ahead for OpenDroneMap! :slight_smile:

I’m opening this thread to ask a simple question:

What would you like to see next?

In short, what’s important/valuable to you that is currently missing?

This feedback will help guide development decisions going forward in the year.


pinned globally #2

#3

I was recently taking a look at the GCP interface. And I had some trouble understanding how to use it properly. I think it could maybe use a look-over by someone who has good UI/UX experience? Or usability feedback from a variety of people trying to work with it for the first time?


#4

Hi, new on this forum but not new to the ODM.

  1. There is feature to spread jobs on different nodes, i wish there was a spread a single job/load to multiple nodes. (i dont have a beefy computer)
  2. An option to swap/overwrite existing coordinates in exif with coordinates in a different file. (e.g this open issue on github https://github.com/OpenDroneMap/WebODM/issues/282) This is a highly wanted feature in the PPK World.

#5

3.And option to correct distored images based on camera and lense level (camera calibration). Ref, this one on github https://github.com/dakotabenjamin/CameraCalibration
4. Lever arm adjustment. This is more as an extended feature of nr 2. Apply an relative XYZ offset between the camera and the GNSS antenna that is fixed for the entire job. Ref. this description https://community.emlid.com/t/does-reach-need-to-be-level/6221/4?u=tb_rtk

Had to split my post as new users cant have more then 2 links i a post :grimacing:


#6

Hi Piero - my number 1 request is georeferencing point clouds before making meshes, ie:

  • create point cloud
  • georeference
  • done - no mesh generation or any other things.

(but I am also maybe missing some documentation which says how to do that already)

  1. depth filtering (found in some commercial packages) - sort of like ‘confidence thresholding’ for MVE in: https://github.com/OpenDroneMap/ODM/pull/921

  2. general point cloud/depth map alignment improvements, I’m sorry I don’t have an exact recommendation - point density of smvs with geometric quality of MVE? (yes, asking for all the things - I want to see branches on trees!).

I second @TB_RTK ’s number 2 - being able to provide a list of camera centres which over-ride exif would be awesome. I’m not sure it’s ODM’s job to add lever arms, but really, I guess it could. It’s cheap math compared to everything else going on.

  1. MOAR TIME for me to play with and test ODM things instead of chasing money :smiley:

#7

Hi, new I. The forum. I would like to see a quality report of the project. Thank you


#8
  • GCP interface: Maybe its possible to include WMS Services?
  • get a quality report
    Thanks!

#9

Hello pierotofy,
i am new to the forum but would like to add that i agree with the others in updating the GCP tool.

i would also like to see the projections reviewed. in particular this setting. proj (PROJ4 string).
living Australia output to UTM North 55 is what i have been getting yet it should be MGA94 Zone 55.

other than a few minor settings i really appreciate being able to use it. Thanks a lot.


#10

(Good news: the Southern Hemisphere issues are being fixed: https://github.com/OpenDroneMap/ODM/pull/936)


#11

I’d like to have an option to add a custom base layer (something other than Google Satellite) for the ‘View Map’ screen, and also turn off the background base layer.


#12

Thanks a lot smathermather-cm. Sorry i was not aware of this.


#13

Definetly true NDVI based on Micasense or Sequoia.

Thats really missing for all us in agriculture or forestry.

Together with ardupilot that would complete the sequence.


#14

Pierro,

From my view the following would be great (in no particular order!);

1.Improved GCP interface which offers the nearest images to the GCP initially, then refines the location once enough have been marked to centre in on approx location. I’d prefer it in the workflow rather than as a plugin, ie calc sparse point cloud, then mark GCPs, and optomise sparse point cloud and continue processing. This is perhaps quite an ask!
2. Use EPSG codes rather than proj4 strings, as few people use proj4 in the real world.
3. Accuracy report
4. Ability to load autocad dwgs as an overlay in WebODM
5. Ability to use WebODM to view other drone data (raster/point cloud/etc) even if processing has not taken place, ie upload data processed elsewhere, or previously saved and removed from WebODM.
6. Timeline view for a project with multiple tasks, perhaps tabbed, so progression can be shown.


#15

Embed Camera calibration in the work flow:
As a start, allow one to specify the location of a camera calibration file and, if present, undistort the images prior to beginning other steps.

A tool to extract the camera calibration. It could be from images of a checkerboard pattern, or even a method like the one proposed at this link which describes (at a high level) a strategy for automatically extracting the calibration.

Finally, I second the wishes of others to improve the GCP interface. In addition to making it easier to find/match GCPs, there should be an interface to specify known lengths/heights of objects in the scene.