I thought I should share this because it’s helped my workflow and output significantly.
Prior to processing images in webodm, I load everything into lightroom, cull out of focus or aberrant images, and then eliminate shadows (as much as possible), and finally even out the exposure. I’ve found that this improves both the output and the processing time. It’s especially useful in wintertime in the northeast US, since the low winter sun creates harsh shadows.
Cheers!
I’ve thinking on eliminating shadows but I’m concern of impacting the accuracy.
Do you mean increased point cloud Density? or improved/modified geometry of the point cloud?
It improves the geometry of the outputs. Shadow can confuse distance when stitching images.
Cool. Any favorite settings? I have some snowy, sunny day shots that’d be interesting to improve.
Those are always challenging!
Contrast adjustment is your best friend in snowy and bright scenes. Lacking a polarization filter on the camera itself, contrast is the most impactful slider.
You can use dehaze (sparingly) to bring out details in snowscape, and the same with clarity. You’ll know you’ve gone too far when you see chromatic aberration appear, so look for rainbows and back it off.
Finally, you can adjust the white balance. I know this sounds counterintuitive, since snow is “white,” but it really isn’t. It’s clear, right? So the white you see with your eyes isn’t the same white your camera will see, which is bleach-y and blinding. You can warm a scene slightly to see more detail, and it’s aesthetically pleasing to the human eye, because we find the colors of fire and the sun to be comforting in some primal way.
Snow is also a giant bear to map in photogrammetry software. I’m not huge on the programming side, but it’s like mapping sand dunes or grass. GCPs become pretty important!