I tried out the Split-Merge pipeline on one device since it seemed like the best approach for particularly large datasets, but I’m confused as to how to view the output.
When running the regular OpenDroneMap image from docker on one dataset without splitting it, I get a bunch of .obj, .vec, and other files as output. The terminal also prints out ODM’s logo after processing.
But when I ran the split-merge method with the following command:
docker run -it --rm
-v “/home/explosion/Desktop/ODM_Test/dronedata_test_100overlap/whole:/code/images”
-v “/home/explosion/Desktop/ODM_Test/dronedata_test_100overlap/whole:/code/odm_orthophoto”
-v “/home/explosion/Desktop/ODM_Test/dronedata_test_100overlap/whole:/code/odm_texturing”
opendronemap/odm --split 200 --split-overlap 100
it runs for a bit over an hour, then ends with the following messages on the terminal:
2020-12-21 06:40:55,090 INFO: Reconstruction 0: 383 images, 439886 points
2020-12-21 06:40:55,090 INFO: Reconstruction 1: 159 images, 169534 points
2020-12-21 06:40:55,090 INFO: Reconstruction 2: 13 images, 12693 points
2020-12-21 06:40:55,090 INFO: Reconstruction 3: 7 images, 7713 points
2020-12-21 06:40:55,090 INFO: Reconstruction 4: 7 images, 8751 points
2020-12-21 06:40:55,090 INFO: Reconstruction 5: 7 images, 8607 points
2020-12-21 06:40:55,090 INFO: Reconstruction 6: 7 images, 6923 points
2020-12-21 06:40:55,090 INFO: Reconstruction 7: 7 images, 7148 points
2020-12-21 06:40:55,090 INFO: Reconstruction 8: 6 images, 7084 points
2020-12-21 06:40:55,090 INFO: Reconstruction 9: 6 images, 6753 points
2020-12-21 06:40:55,090 INFO: Reconstruction 10: 6 images, 9915 points
2020-12-21 06:40:55,090 INFO: Reconstruction 11: 6 images, 5956 points
2020-12-21 06:40:55,090 INFO: 12 partial reconstructions in total.
[INFO] Aligning submodels…
But when I check the folders, there aren’t any output files. Is there something I’m doing wrong? The dataset is 621 images by the way.
Thank you!