This AOI had RGB and BGNIR collected on 5/9/2017.
Prior to the flight, ground control targets were placed.  Following the flight, these targets were surveyed and removed.
The evening following the flight we began to see that stitching the imagery in the NW corner would be difficult. The trees have an ambiguous pattern and a very dense canopy.
As a result, we determined a re-flight with a wider collection area should be completed to provide all avenues for full stitching in this tree area.
The RGB re-flight was successfully collected on 5/10/2017.
Processing the two RGB flights together proved difficult and resulted in a poor product because of different times of day at which they were collected.  Thus, it was determined to use the second flight for the deliverable image.  However, as the GCPs had been removed following the first flight, processing required going outside our normal procedure.  
We used the first flight with GCPS to produce an accurate orthoimage (albeit dirty in the NW corner).  From this image we extracted locations of features which were in both flights.  These included: scientific boardwalks, the LiDAR control target, and the NOAA checkpoint targets.  Our accuracy assessment of the first flight had the accuracy between 6-10 cm.  
Once we had processed the second flight, we did a thorough comparison between the two datasets and determined the second flight dataset met project accuracy requirements.