Better accuracy and consistency when processing large dataset

Hi everyone,

We did 4 flights to cover the whole study area using drone and multi-spectral cemara (Micasense Altum). Now we want to generate a map of the whole study area and we’re not sure how to process these datasets.

Due to our computer performance, we can only process 1 flight at a time and then merge them. So which workflow below is better?

a): 1) do step 1 for individual flights; 2) merge 4 flights into one project; 3) do step 2 and 3.for dsm and reflectance maps.
b): 1) do step 1, 2, 3 for individual flights; 2) merge dsm and reflectance maps in Arcgis

We tried both a) and b). a) will lose some accuracy because all the images from 4 flights are calibrated together. Results from b) should be theoretically accurate, but the outputs from individual flight did not show consistent in overlaping area when we want to merge them.

So what would be your suggestion, a) or b)? Or do you have any new suggestion on processsing?

Thanks in advance!

Hey, can anyone give any suggestions :smiley:

Hi,

If the weather (sunlight) was similar between the flights, I would recommend using the targets after flight 2/before flight 3 for workflow a. For workflow b, you can use MTPs/GCPs (have whiteboards in the field) so that you have geometric accuracy.