I have an extremely large amount of photos that are geotagged. There are ground control points spread unevenly throughout the project area so I was hoping to manually split the images into subprojects taking into account the following parameters:
- by quantity (<2000 images according to the “processing large datasets” Pix4D page
- Coverage of the GCPs
- Overlap of different flights
Currently, I only see the option to split the images automatically by limiting the quantity per subproject, which takes into account the number of images and overlap, but does not appear to take into account the location of the GCPs.
Is there a way to manually split images into subprojects even if they are already geolocated?
Also, creating subprojects is a necessity because the project area is extremely large (5,500 acres). I found in this article (https://support.pix4d.com/hc/en-us/articles/202558529-Merging-Projects) under step 11 the caveat that one can continue adding more projects to merge “if the processing resources can handle multiple projects.” The specs I will be using to process this data consist of an Intel Xeon CPU E5-1607 v3 @ 3.10GHz with 32 GB of RAM. With that I have two more questions:
- Is there a certain number of images I should have as a goal to process at one time given those specs? As of now, my goal was to keep it under 2000 images according to this article (https://support.pix4d.com/hc/en-us/articles/202558579)
- After completing Step 1 of the subprojects, is there a certain number of subprojects I should limit the merging to? Meaning, should I actually complete several different merges and then merge all of those “submerges” at the end? Would this even speed up processing and would it even work?