"Cannot process dataset" - image collection uploaded with gcp.list.txt file

(I checked the forum and github without finding my issue. I bought the Missing Guide, which was a timesaver, however I did not find a resolution. I bought GCP Editor Pro.)

I’ve processed this collection of 979 images, ‘fast orthophoto’, with no GCPs. WebODM produced very good results with no tweaking of parameters. Then I wanted to see the effect of introducing 5 very accurate GCPs.

I guessed about exactly which steps in the GUI to use to upload the gcp file. The uploaded file count changed from 979 to 980. However I continue to get the “Cannot process dataset” message. I’ve tried several variations of parameter settings beyond ‘fast-orthophoto’, with the same result.

Latest version of Chrome; Windows 10; 209 GB of free space.

I am enjoying learning through the ODM project. The resolution to my issue may be as simple as reading instructions which are stated differently. Perhaps I need to use the command line interface with WebODM?

Thanks!

1 Like

Looks like an X/Y column swap is needed.

Are you surveying this area?
Screenshot_2021-11-19_00-19-13

If so, you need to change your X/Y around. See below:

EPSG:4326
-122.438493 48.019965 80.755 1269.27578366631 1506.19183605563 308.jpg gcp1
-122.438493 48.019965 80.755 1351.36791332148 140.434263058208 302.jpg gcp1
-122.438493 48.019965 80.755 1334.06428114327 381.150529465089 303.jpg gcp1
-122.438493 48.019965 80.755 1320.03786348912 627.307011690822 304.jpg gcp1
-122.438493 48.019965 80.755 1313.99326695992 785.704709270321 305.jpg gcp1
-122.438493 48.019965 80.755 1298.10420153588 1024.03115246607 306.jpg gcp1
-122.438493 48.019965 80.755 1285.37330767232 1261.49064804937 307.jpg gcp1
-122.437568 48.020475 82.445 1376.521490174 628.537148145715 524.jpg gcp2
-122.437568 48.020475 82.445 1359.6722443931 867.125336298867 525.jpg gcp2
-122.437568 48.020475 82.445 1345.78663614726 1024.87621331792 526.jpg gcp2
-122.437568 48.020475 82.445 1323.82246227143 1262.94521873339 527.jpg gcp2
-122.437568 48.020475 82.445 1413.25319439063 146.157534481822 522.jpg gcp2
-122.437568 48.020475 82.445 1393.82421188073 397.585410959342 523.jpg gcp2
-122.437568 48.020475 82.445 1276.89676218421 565.731407386462 581.jpg gcp2
-122.437525 48.019895 80.855 1385.40258812096 1473.17913220919 533.jpg gcp3
-122.437525 48.019895 80.855 1491.25335541386 129.896644579123 527.jpg gcp3
-122.437525 48.019895 80.855 1469.58551519785 371.183176871648 528.jpg gcp3
-122.437525 48.019895 80.855 1452.24895183772 614.754968503721 529.jpg gcp3
-122.437525 48.019895 80.855 1435.40984521908 856.53338553262 530.jpg gcp3
-122.437525 48.019895 80.855 1419.65511913666 1012.36549024514 531.jpg gcp3
-122.437525 48.019895 80.855 1405.62762704182 1242.51424516341 532.jpg gcp3
-122.438039 48.016795 79.085 1404.94302157765 1349.37360029204 390.jpg gcp4
-122.438039 48.016795 79.085 1419.45959088029 1495.71023583417 391.jpg gcp4
-122.438039 48.016795 79.085 1299.45247989298 283.537758934412 385.jpg gcp4
-122.438039 48.016795 79.085 1316.15155327394 441.685992644802 386.jpg gcp4
-122.438039 48.016795 79.085 1340.50673422002 676.90222664587 387.jpg gcp4
-122.438039 48.016795 79.085 1357.90912455844 907.258963835125 388.jpg gcp4
-122.438039 48.016795 79.085 1381.47184064319 1132.2954583311 389.jpg gcp4
-122.437633 48.014107 80.325 1271.37799250405 1283.57382920402 257.jpg gcp5
-122.437633 48.014107 80.325 1337.18124183418 182.085002470806 252.jpg gcp5
-122.437633 48.014107 80.325 1327.42804896902 344.084713597855 253.jpg gcp5
-122.437633 48.014107 80.325 1311.7590435528 582.973665470103 254.jpg gcp5
-122.437633 48.014107 80.325 1302.9649469482 809.698324499761 255.jpg gcp5
-122.437633 48.014107 80.325 1286.54006434902 1038.93693069071 256.jpg gcp5
-122.437633 48.014107 80.325 1340.69730582696 752.788850896129 411.jpg gcp5
1 Like

Well, thank you! My early mistake explains why when tagging the GCPs, the “map” which appeared was two-tone, top and bottom, with the markers horizontally along the center line.

Now on to fix the file and re-launch the task… Thanks again!

3 Likes

For anyone else who discovers this thread, I found where/when I made my mistake. It was when I created the stub file (gcp_stub.csv) for uploading prior to tagging the centers of ground targets appearing in images.

If you have installed the GCP Editor Pro plugin, see the pull-down instructions under ‘CSV format specification’, including this step:

“2. The first row (header) must be: GCP Label,Y,X,Z”

I failed to see that the step included “Y,X,Z”, NOT “X,Y,Z”.

(After tagging, the exported file ‘gcp_list.txt’ is also in Y, X,Z sequence.)

4 Likes

Also… I’d like to echo a suggestion by gjoelthiessen79 to use the ‘cameras’ view as a way to select the images most likely to include a particular gcp target or other known/remembered ground point from your time on the site.

“Also, a quick tip for finding images with your GCP’s. I run a fast ortho and then turn the cameras on in the viewer. It’s a lot faster for me than browsing through 1000 images to find a small X.”

4 Likes