Hi All,
Using Archival aerial oblique images from the early to mid- 20th century, I am building SFM models of large mountain walls (up to 2000 meters in height) for scaling and volume change calculation against modern TLS scans.
None of my images (all are scanned from either slides or prints) have any information about the cameras used. Ofter there is EXIF data which is simply the data from the scanner or camera used for reproduction, and has no relation to initial camera specs.
All of my images come up as NC, and I'm wondering if there wha is the best way to use these images in Metashape, in terms of camera calibration, and secondly, the optimise cameras steps. There are many options during the optimise cameras steps, and the resulting sparse and therefore dense point cloud shape seems be to altered accordingly, however I have no idea what is the best option.
One advantage I have is I can directly set "GCP"s as markers taken from my modern LIDAR scans, so I can effectively enforce a strictly measured geometry in this process, however I am unsure if I am even doing this correctly. For example should I be leaving the GCP markers unchecked during photo alignment, then checked for optimisation? And if so, which optimisation options are best for this situation.
Also take into account that each scanned image has different pixel dimensions (sometimes quite a bit different). Therefore will metashape consider these different cameras?
Any advice very welcome!