DJI Matrice 210 RTK

Hi,

we would like to inform you that the DJI Matrice M200 series is now officially supported on Pix4Dcapture.

The list of supported drones can be found here: Which drones are supported by Pix4Dcapture

Best,

1 Like

Great news.  Thank you for that information.

Hi

What does the DJI M210 RTK support actually means? Does it mean that the RTK basestation position information will be utilized which gives better absolutely position or even replaces the need of CGP`s or something in between?

Pix4D capture still does not support two cameras (X4S attached to the 1st and XT attached to the 2nd gimbal) using dual gimbal. Only the camera at the 1st gimbal is supported by the app. It makes no difference whether using iOS (iPad) or Android (CrystalSky).

I loaded pix4d capture on my crystalsky. I am not getting background maps for mission planning. Location services are set on the CS. I see DJI Go4 app uses MapBox for background maps as well. Any help appreciated. I can still use ios device if needed.

1 Like

Does pix4d take in the RTK data from the M210 RTK when processing the data?

@Jan and Benjamin, our latest understanding is that DJI’s M210 RTK does not report an RTK-grade coordinate position of the camera in the metadata of each image. I anticipate that the most likely reason the image’s metadata does not report a coordinate measurement that is RTK-grade is that the drone must accurately account for the offset, or lever-arm, between the RTK GPS’ measured position and the position of the camera, which is hanging below the drone on a stabilized gimbal.

I have spoken with a handful of people who have experience with DJI’s M210 RTK, and they have each reported that their images’ geotags were not as accurate as what would they expected if the RTK GPS was measuring the camera’s position.

You can test the consistency of your DJI M210 RTK geotags by creating a Pix4D project that includes a handful of RTK-grade, or similar, ground control points. The Absolute Geolocation Variance and Relative Geolocation Variance section of the Quality Report will give you a basic idea of how well the image’s initial position, according to the coordinates in the images’ metadata, aligns with the images’ computed position that Pix4D computes according to your project’s ground control points.

@Karsten, you are correct that Pix4Dcapture does not support DJI’s dual downward-facing gimbal at this time. You can stay up to date about the status of dual downward-facing gimbal support in Pix4Dcapture on this post: DJI M210/M210(RTK) Dual Gimbal Support

@Brad, Pix4Dcapture is not officially supported on DJI’s CrystalSky at this time, so I recommend that you consider taking advantage of your mobile iOS device. I see that you have already found the corresponding post on our Community so you can stay informed about the status of Pix4Dcapture support on DJI’s CrystalSky.

@All, if you would like to stay up to date about Pix4Dcapture support on DJI’s CrystalSky, or contribute to the discussion, you can follow this post: Pix4Dcapture on DJI CrystalSky

Andrew thanks for your response…

If the RTK base is not set on precise ground control point there is no RTK positioning so it would be a waste of the system.   I see this mistake done even in regular RTK ground surveys.

I contacted DJI since I have been trying a DJI M210 RTK with Pix4D Capture and they responded:    _ It sort of sounds to me like Pix4D isn’t capable of supporting the RTK configuration. Users of their software weren’t able to get accurate geotags. That’s a Pix4D issue, not a DJI issue.      _

So maybe there is a secret coding of the M210 RTK positioning or geotagging that DJI is not sharing with Pix4D and other platforms?

The DJI RTK system is not a survey grade system.

It uses the RTK technique to accurately hover.

I think DJI misled many people into thinking it is a survey grade system by using the RTK moniker. RTK in itself is a technique of resolving positioning ambiguity. Commonly used by surveyors and some survey grade drone manufacturers.

The DJI system ground station does not mount to a tribrach, does not have a levelling check for height above known point.

It’s aimed at the user that wants to accurately hover in place while capturing data. It does not provide accurate geotagged images nor does it provide a processing solution to use the RTK tags with the non georefeenced images.

If you are looking to carry out aerial surveys then go for a fixed wing solution designed for survey and mapping.

The danger comes when non-surveyors pick up a few buzzwords like ‘RTK’ and ‘control point’, proceed to convince their client that they can provide highly accurate surveys using their so called ‘RTK drone’ and of course the reality is very different indeed. What you end up with is people not really understanding what they are trying to achieve, not understanding the achievable accuracy and not understanding the limitations of the equipment they have invested in.

As mentioned, RTK is a technique - just because you see it on a drone doesn’t mean it makes the drone a survey instrument. Caveat emptor as always.

https://en.wikipedia.org/wiki/Real-time_kinematic

 

@Alexander, thank you for sharing the comments that you received from DJI technical support. For the record, I disagree with the comments that you received from DJI technical support. Unfortunately, I am unaware of anyone who is acquiring cm-grade image geotags using an original DJI M210 RTK. My understanding is that the differential GPS measurements are only available for navigating or staying on a flight path.

@Philip, thank you very much for sharing your insights.

Regardless of the drone that you use to capture your images, I strongly encourage that you independently verify the accuracy of your results for all of your projects. Not only to gain a better appreciation for what your hardware is capable of in conjunction with Pix4D, but also as a way of verifying that all systems are behaving as you expect.

@Andrew DJI themselves have told me the metadata contains RTK positioning and they have reported camera/gimbal offsets as of pretty recently.

1 Like

@Benjamin - the drone is positioned using RTK technique for accurate hovering. But without a base station thats accurately located and you are entering the known coordinates of the base station in terms of Lat, Long and eliipsoidal height, there is no way the drone can know what it is positioning itself against. The DJI base is not a survey grade base station and provides no option to enter its position over a known point, including the antenna height above the control point.

In order to have accurate RTK positioning for survey purposes you need to either have a base station with known coordinates in continuous link with the drone or receive the RTK corrections from a CORS or VRS reference station via the internet.

 

 

@Philip Yeah, I’ve heard this before, however, it was framed like DJI had an issue they were working through to correct it. If that’s the case and the base station allows the manual entry of positioning in the future, paired with the data being reported in the metadata, should we theoretically see more precise survey data?

1 Like

@Benjamin

 

The base station is not a survey grade GNSS antenna. It doesn’t mount to a tribrach which you would need in order to be able to level it over a control point.

 

Second, they don’t provide any software that permits the input of the coordinates of a benchmark in terms of lat, long, ellipsoidal height and antenna height above the benchmark.

I don’t think they are targeting the professional surveyor with this system and they have caused a lot of confusion amongst users who think by purchasing an RTK system they are getting a full blown survey system.

The systems is intended to be used as an inspection drone thats capable of remaining in an accurate hover whilst imaging say power lines where any drift could be catastrophic. Thats all. It seems that DJI support are also in the dark re the capabilities as well.

Its a flying camera, not a photogrammetry instrument. 

1 Like

@Phillip Angell

It can be mounted to a tribrach as mine is and the software does allow benchmark coordinates to be input.  But I do agree with you the RTKs main purpose is only to increase hovering accuracy.

1 Like

Tribrach mount

1 Like

@Adam 

So you are saying that you set up your base over a known control point, enter that lat, long ellipsoidal height of that point, you then measure the height to the phase center of the base antenna and you receive an ‘RTK Fixed’ message during use? If you lose any signal you would see an RTK Float or Standalone message somewhere as well.

After you have gathered images, you then are able to somehow view the quality of the position in terms of delta easting, delta northing and delta height?

Could you provide some screenshots of this as I would be keen to see how this works?

1 Like

@Andrew DJI  is there any way for Pix4D to check the 210 RTK log file and confirm if the geotags of the images have the RTK coordinates or not??     I would assume that if you set the DJI RTK base on a tripod over a know geodetic control point monument, input  these control coordinates and fly the 210 RTK the metadata would show the positioning and residuals of RTK (cm level) or the autonomous GNSS positioning (meter level) or not?

Unless you can figure out how to related the phase center of the antenna to the point on the ground, it is pretty much useless. It’s not impossible, but the fact that there is no assurance of the antenna being plumb, and that the antenna is offset from the center of the mount leaves you with an infinite amount of error combinations. Might as well drop a plumb bob from the base of the antenna and measure up.

1 Like

Maybe using a survey grade GNSS RTK base receiver installed (Trimble R10, etc) with RTCM 3.x message protocol can be used with DJI Matrice 210 RTK device??   Has anyone tried?

1 Like