Hello,
we have a problem when generating ortho photos under a little bit specific conditions, and I wanted to ask if someone could help.
we do lidar + aerial photos at the same time, and flying height is normally determined by the lidar configuration rather than the image requirements. So sometimes we have acquisitions where the image GSD is much better than needed. For example the client wishes an ortho photo with 5 cm resolution, but our native GSD is around 2 cm. So when we calculate with this 2 cm images an ortho photo of 5 cm, we noticed a loss of quality in the images.
After some investigation, we are pretty sure that metashape uses not the original images for the ortho creation, but the next image pyramid (which would be around 4 cm in this example). This is in theory not a problem as the resolution is still superior than what we need. The problem is that this pyramid is highly compressed, and in the final ortho photo, strong compression artifacts are visible, and we lose a lot of colors (see attached images).
Our workaround is to calculate the ortho photo close to the native resolution (2.5 cm) in this example, and then do the exports in 5 cm resolution for the final product. But for bigger projects, this explodes a) processing time, and b) the metashape project size.
So the question is: is there a way to either prevent metashape from using the pyramids or if not, control the quality of the pyramid creation in the orthorectifing process?
Any help would be appreciated!
Thanks,
Johnny