Posted by Guest: Jim Crume on 12/5/2019 to
Drones
Today we have guest author and pilot, Jim Crume, discussing the results he found when testing three DJI drones with and without the use of Ground Control Points (GCP's). Jim runs the very informative Surveying and Mathematics Made Simple website where he offers resource guides ranging from surveying to 3D mapping to drone mapping.
This is a great article that shows why everyone should use a GPS to capture high-accuracy Ground Control Points with any drone you are using. Please visit his site (https://www.cc4w.net/) for more information and to see additional posts like this re-post on GCP's.
DJI Matrice 600 vs Inspire 2 vs Mavic 2 Pro
Equipment As Flown
The DJI Matrice 600 with the Phaseone medium format camera ($30k) does produce awesome results. Drone, gimbal, and camera are around $50k. The Mavic 2 Pro list for about $2000 and the Inspire 2 list for $5000 and up. The cost of the Inspire 2 setup flown in this test is $7,500.
Metashape Pro is used to process the Inspire 2 and Mavic 2 Pro imagery. I use the M600 for survey mapping that will be processed through Summit Evolution software which is a high dollar application as well. EPS Group has a full-scale photogrammetry department that is run by Richard a certified photogrammetrist. I am fortunate to be able to work closely with Richard who has taught me a lot about photogrammetry. Not everyone gets these opportunities which is why I share what I have learned in these email posts and books I have written.
Project Site 1 (13.5 acres)
I had the opportunity this past week to utilize three of my favorite drones on the same project site for a side by side comparison in a real-world scenario.
Nine (9) GCPs were set using GPS methods. GPS methods were used because this is what most GCPs are set with by the surveying community primarily due to budget and time constraints.
Time is Money!
Five (5) of the nine were used to constrain the survey model to with Metashape Pro as shown by the red triangles. Four (4) were used as checkpoints to check the Root Mean Square Errors (RMSE) to the standards set by the American Society for Photogrammetry and Remote Sensing (ASPRS).
The results of the test below verify that the Matrice 600 meets the ASPRS at 95% Confidence Level Horizontal class of 0.63cm and the Vertical class of 1-cm and the 1.25cm H / 2.5cm V classes for the Inspire 2 and Mavic 2 Pro based upon the number of GCPs set for this project.
This project site was flown with the Matrice 600, Inspire 2 and Mavic 2 Pro. The side by side results are shown below as processed through Metashape Pro.
| Matrice 600 with Phaseone Camera | Inspire 2 with x5s | Mavic 2 Pro |
| (55mm) 50mp | (15mm) 20mp | (10.26mm) 20mp |
| 8280 x 6208 (px) | 5280 x 3956 (px) | 5472 x 3648 (px) |
| Sensor 44.0mm x 33.0mm | Sensor 17.3mm x 13.0 mm | Sensor 13.2mm x 8.8mm |
| AGL 350' | AGL 200' | AGL 150' |
| FOV: (w) 279', (h) 210' | FOV: (w) 231', (h) 173' | FOV: (w) 193', (h) 129' |
| Neat area: (w) 195', (h) 84' | Neat area: (w) 92'), (h) 52' | Neat area: (w) 77', (h) 39' |
| Speed: 10.7 mph | Speed 11.8 mph | Speed: 8.8 mph |
| Capture rate 4 sec. | Capture rate 3 sec. | Capture rate 3 sec. |
| GSD: 1.03 cm | GSD 1.33 cm | GSD: 1.08 cm |
| Flight time: 5 min | Flight time: 8 min | Flight time: 11 min |
| Flight app: GSP | Flight app: Litchi | Flight app: Litchi |
| # of images: 70 | # of images: 147 | # of images: 220 |
| 40% side, 70% front | 60% side, 70% front | 60% side, 70% front |
| Drone | Number of GCP's | Y err (ft.) | X err (ft.) | Z err (ft.) |
| Matrice 600 | No GCP's | 1.839594 | 0.684416 | 7.541496 |
| 5 GCP's | 0.024051 | 0.033979 | 0.003777 | |
| 9 GCP's | 0.027185 | 0.042193 | 0.018396 | |
| Inspire 2 | No GCP's | 3.357034 | 3.409264 | 6.352514 |
| 5 GCP's | 0.080126 | 0.056441 | 0.037905 | |
| 9 GCP's | 0.057380 | 0.049887 | 0.038665 | |
| Mavic 2 Pro | No GCP's | 3.227758 | 2.506392 | 5.919997 |
| 5 GCP's | 0.134928 | 0.159709 | 0.022223 | |
| 9 GCP's | 0.102415 | 0.114085 | 0.039212 |
Project Site 2 (180 acres)
The following site (180 acres) was flown with the M600 Pro RTK with the GCPs turned off to test the positional accuracy of the RTK EXIF image tag and alignment processing. The flight spec differences between this and the Project Site 1 include a higher AGL of 400', Flight Time of 45 minutes, Number of images captured 523, and 30% side and 60% front overlap.
Drone Number of GCP's Y err (ft.) X err (ft.) Z err (ft.) Matrice 600 No GCP's 1.424097 1.037848 2.733658
Conclusion
From the above results for Project Site 1; I really like my Inspire 2. It does really well in windy conditions and is a workhorse. The Inspire 2 does really well and is capable of producing survey mapping to acceptable accuracies. SMMS Inspire 2 was used for the above flight. The Mavic 2 Pro is a really good option for starting out with survey mapping that will produce survey-grade mapping if done right. The Mavic 2 Pro holds its own however it needs more GCPs to improve on the accuracies for the project site above. The right number of GCPs in the right location is the trick.
For larger sites and fewer GCPs, the Matrice 600 with the Phaseone camera is the way to go if you can afford it. For Project Site 1, all 9 GCPs were utilized for control. The 3D survey mapping was completed by the EPS Group photogrammetry department for the final deliverable.
For both project sites, the residuals above indicate that the Z values are unreliable without GCPs. The X and Y values are better than the Z values which are close enough for a non-rectified orthomosaic if no horizontal measurements are needed to key features in the ortho.
I am a little disappointed in the RTK results with the M600. I have read and been told that no GCPs are needed with RTK. Each image EXIF tag might be more accurate to adjacent images however when the images are aligned there is still distortion without GCPs.
Without GCPs to a known datum, the processing software block bundle and least squares adjustment has no fixed position to constrain the model to.
Most of the mapping that I do is for 3D design level for engineers. The results have to be within ASPRS accuracy standards and constrained to a known datum.
Does RTK help with the accuracies and eliminate the need for GCPs? I am not convinced that they do, at least for 3D design level mapping. For GIS Mapping RTK would be useful for results to a specific projection.
Is RTK worth the $5k+ price tag, you will have to be the judge on that. I don't think it is worth it. The RTK unit for the M600 sits in a box on my desk. I don't use it.
It really boils down to constraining the model to known positions (GCPs) on the ground to get the most accurate results.
Final Thoughts on Drone Equipment
"Not all drones/cameras are the same nor do they produce the same results."
First determine what your deliverables need to be, then select a camera that will meet the accuracy needed for the deliverables and finally select a drone that will carry the required camera.
The costs of the drone, camera, time and GPS equipment needed to set the required number of GCPs all factor into the cost/benefit ratio for just which drone system will benefit your survey mapping operation.
As nice as the Matrice 600 with Phaseone camera is for survey mapping, it doesn't have video capabilities. The Inspire 2, Mavic 2 Pro and Zoom drones are perfect for capturing videos and oblique imagery.
Every survey operation at a minimum needs two drones. A high-end one for survey mapping and one for videos.
