J-xue
commented
3 years ago Thank you a lot for Saijin-Naib's help on figuring out the issue, finding the root cause and propose this feature request. Adding this feature before starting processing the images I think it's very important and helpful on finding the bias data at the first time, saving the total task performance time, improving the efficiency, success rate and accuracy. Integrate this feature to the WebODM instead of separating this step to other software, I think make the WebODM more robust and user friendly. Thank you.
What's your browser and operating system? (Copy/paste the output of https://www.whatismybrowser.com/)
N/A
What is the problem?
Sometimes, images in a dataset can have erroneous EXIF information that can cause reconstruction to fail in ways that seem mysterious and hard to track down (Child Returned 139, etc).
What should be the expected behavior?
I think we should have an extra step after image selection/upload to WebODM that lists the images and their geolocation data (X,Y,Z, and rotation), as well as showing a mini-map. These two things should make it immediately obvious if any images in the set have bad geolocation data and should either be fixed manually or dropped from the reconstruction. I like the mini-map as pictured above in Geotagging, but also appreciate the no-nonsense tabular view of attributes from GPSCorrelate below.
How can we reproduce this? (What steps did you do to trigger the problem? What parameters are you using for processing? Please include multiple screenshots of the problem! If possible please include a copy of your dataset uploaded on Google Drive or Dropbox. Be detailed)
Use a dataset with erroneous data, like the one described above by jixue1.