Mavic 3e/Matrice 4e vs Mavic 3 Pro accuracy with Orthomosiacs

[mavicpilotsnumpty]

Stupid question. Total newby with mapping. I used my Mavic 3 pro recently to map a forest. Only because that’s the only big drone I have. I have tried to use Pix4d but it has been quite complicated and have not been able to create an ortho mosaic map with the trial to export. Drone deploy took awhile but created an incredibly detailed map being able to see very small detailed points on the. The error rate on DroneDeploy suggests it is 0.6m. Having looked on Google that suggests it has been quite accurate.

Measuring very different control points I put on the ground to rear world objects using annotated points the map seems very accurate.

Given the price of a used Mavic 3e vs new Matrice 4e which is not massively different, I am wondering how much more accurate the RTK enterprise models would be over the Mavic 3 Pro to justify the £3500 price of the Matrice.

 

[Dave Pitman]

If you use a good network of accurately placed ground control points, and use them in the photogrammetry processing, you can achieve centimeter level global accuracy orthophotos and surfaces with the M3P.

If you want to rely on the drone to capture the highest precision photo capture camera positions, it is not possible with the M3P because, A, The gps receiver is not precise. And B. there is no mechanism for the location data to be written into the photo exif/xmp data that occurs with the M3E or M4E.

The potential accuracy difference between the M3E and the M4E is negligible. The consideration would be cost, and what other feature improvement on the M4E over the M3E you find as worth the cost. If cost is not consideration, then the newest model is probably the best bet. If you already have a battery family for the M3P, then that might be another plus to go with the M3E as they use the same batteries.

With the M3E, you would also need to purchase the RTK module, about $300 on Ebay. The module is built into the M4E. The M4E with the built in module can be a liability for tower inspection. But if you don't plan to do that ever, then it is not a factor.

 

[jaja6009]

All great points by Dave Pitman.

Just to add about Drone Deploy and mapping in general.

On Drone Deploy the "Accuracy" you are referring to is not the true accuracy of your map. Long story but here is an short explanation from Drone Deploy itself:

"
The map information panel brings up details about the map. Here, if you scroll down, you'll see the total area, the native resolution, and the accuracy of the map:

Note: This isn't the ACTUAL Absolute Accuracy, but a proxy to this.

Above, we show you where to view the accuracy on DroneDeploy. This isn't the actual accuracy, but a proxy to the Absolute Accuracy of your data. Without Checkpoints, it's impossible to know the precise Absolute Accuracy, but we work out an expectation on the accuracy by comparing the difference of where the drone thought it was (using its GPS), and where it needed to be in order to make the images overlap and stitch appropriately.
As stated above, without Checkpoints one cannot verify the accuracy of their map. This is true whether GCPs were used or not. When we process without GCP's for example, the RMSE number is the 'Optimized Camera Location XYZ RMSE' "


Next, Google Earth itself is not very accurate. Can you use it for ballpark estimates? Yes and its elevations in the Pro version can be suprisingly close (within a meter) in some places.

A major question is what are you using YOUR maps for? Are you doing this for fun? Do you plan to do this professionally or for your job?

One more thing to note is that if you do get an RTK drone, you will need a source of corrections. You will have to either pay for a base, pay for a base service, find a free base (Some states have this), use free PPK (NGS CORS) or some other options out there. RTK equipment needs a corrections source to work their magic.

If you are thinking of doing this professionally, I would start to learn about datums, coordinate systems, elevations, GNSS equipment and GIS software. Today's software does all of the work and anyone can make a map with the most minimum of training.

When you can make a map sub 6cm accuracy and can prove it, in the required datum/coordinate system, with the correct elevation datum, you have set yourself apart from the amateurs.

 

[Meta4]

Stupid question. Total newby with mapping. I used my Mavic 3 pro recently to map a forest.

What were you using to plan and fly the mapping mission?

Given the price of a used Mavic 3e vs new Matrice 4e which is not massively different, I am wondering how much more accurate the RTK enterprise models would be over the Mavic 3 Pro to justify the £3500 price of the Matrice.

Before RTH drones were available, you could produce reasonably accurate maps with the Phantom 4 pro, but easier than with a Mavic 3 pro because you could use apps like DroneDeploy to plan, fly and acquire images.

If you wanted very accurate maps, you needed to use surveyed ground control points.
With GCPs the horizontal accuracy was much higher and you could match the heights to the local height datum rather than just having relative heights based on the drone's launch position.

 

[Vic Moss]

The Mavic 3 Pro isn't a mapping drone, as you found out. Without RTK, it's just not very accurate, and definitely isn't very easy. I fly the Mavic 3 Pros for all of my creative work. But I fly one of my Mavic 3 Enterprise for any mapping for measuring that's needed. And the batteries are interchangeable.