Outside of GCP Polygon Limits - Part 3
 Part 3 - Distortion outside of GCP Polygon.
 Here is the link to Part 1: https://conta.cc/2Ut13eH Here is the link to Part 2: https://conta.cc/3lyaq8N
 Objective: Is to determine just how many GCPs are needed for various camera/sensors/lens combined with the above ground level (AGL) flying heights for 3D mapping.
 Testing site: A 5 acre rectangular parcel that contains brush, trees, buildings, dirt and paved roads.

Drone Platforms: Mavic 2 Pro, Inspire 2 using an X7/16mm lens and X5S 15mm/lens

AGL: 100', 200' & 300'

Overlap: Front 70%, Side 70%

For more details on the drone flight parameters see book "Flight Parameters" at www.cc4w.net.

﻿Processing software: Metashape Pro by Agisoft

For more details on using Metashape Pro for 3D mapping, see books "Metashape" and "3D Road Intersection" at www.cc4w.net.
 Five (5) GCPs were placed in specific locations on the 5 acre test site. One in each corner and one in the middle.
 A good solid mathematical triangle is critical to constrain the 3D model. See Part 1 for a discussion on the proper placement of the GCPs: Here is the link to Part 1: https://conta.cc/2Ut13eH See Part 2 for a discussion on the RMSE values of multiple Check Points: Here is the link to Part 2: https://conta.cc/3lyaq8N
 As discussed in Part 1, understanding the image Neat Area (NA) is important to get good solid overlap to minimize the lens distortion for each image.

﻿For more detailed information on the NA, review books "What's My Mission?" and "Time to Fly" at www.cc4w.net.
 Part 3 takes a look at the amount of distortion when venturing outside of the GCP polygon. The question is, just how far can you go before the distortion is beyond acceptable limits?
 The 5 acre test site was divided into two areas for this testing purpose.
 West Side
 East Side
 Five GCPs were held on the west side in a geometrical position for the best possible aerial triangulation solution. The east side was then used to determine the amount of distortion as shown in the tables below for Checkpoint 'S' at 300' outside of the GCP polygon.
 The next table shows the RMSE values with 5 GCPs on the west side to constrain the model.
 The next table shows the results for only 4 outside GCPs to emphasize the importance of the center GCP to build a solid aerial triangulation model.
 Conclusion The above charts clearly show the importance of a good GCP polygon and one in the center for a solid aerial triangulation solution. Five (5) GCPs at a minimum are required. Venturing outside of the GCP polygon will work for a little ways with the right GCP configuration. The sensor combined with the right AGL for that sensor will produce solid results. Flying lower or higher does not equate to better results. This will be demonstrated in an upcoming post. There are two training videos available on "3D Mapping" located at https://vimeo.com/ondemand/smms/
 www.cc4w.net "I fly Drones for a living mapping the earth one (1) acre at a time and write DIY surveying math/mapping books/videos for fun!"
 Check out the New video series on "Survey Mapping Made Simple".