OpenMVG + OpenMVS pipeline multi sources

[elit...@gmail.com]

I have a drone video, from which I extract image frames with VLC (using scene filter).

I also have a few photos taken with the same drone.

When mixing both video extracted JPEGs (without EXIF data) with the photo JPEGs (with EXIF data) in the same directory and run the OpenMVG pipeline, the drone photos never gets calibrated and are exluded from the model. I know I managed to include them with COLMAP, but how can I include them with OpenMVG, and later in the pipeline OpenMVS?

This is my windows bat file:

.\openmvg\openMVG_main_SfMInit_ImageListing.exe -i .\images -d .\openmvg\sensor_width_database\sensor_width_camera_database.txt -o .\output -f 4915

.\openmvg\openMVG_main_ComputeFeatures.exe -i .\output\sfm_data.json -o .\output -p HIGH 

.\openmvg\openMVG_main_ComputeMatches.exe -i .\output\sfm_data.json -o .\output -r 0.8 

.\openmvg\openMVG_main_IncrementalSfM.exe -i .\output\sfm_data.json -m .\output -o .\output

Thank you very much for your support!

[cDc]

This is really a problem that should be addressed to OpenMVG developer, OpenMVS will happily process any image calibrated by the SfM pipeline. But in few words, you need to make sure when u run OpenMVG it is aware that the scene is composed of images coming from different cameras. In any SfM pipelien there should be a way to group images taken with same camera, pls ask Pierre Moulon (https://github.com/pmoulon) for OpenMVG details.