Glad it was helpful! It's also worth noting that it's best to include more than one if you can. e.g. use RTK during the capture and then PPK, GCPs or both. The MOST important one though is to use checkpoints. Without that you cannot be sure that everything came out as expected.
How you use checkpoints varies based on the software. I don't use Agisoft so I can't help you there but I'm sure there are tutorials out there for that. Once I have some time I will be doing a series going end to end from setup and capture of a site through processing and delivery of the end product to clients. But...it's going to take a while to get to that.
You're mixing things that are sort of mutually exclusive. If you have a DJI RTK2 and it is correctly placed on a known location, and linked to the RTK on the drone, and you didn't get any errors during the flight then your picture should already be accurate. You can use checkpoints (GCPs not used in any correction) to make sure that everything is where it needs to be - that is how you confirm accuracy. If you want to use GCPs then you can process in software that accepts GCPs. In that case you would upload the location of all of the GCP centers into the software, tag photos with the GCPs in them and identify the center of the GCPs in each photo, and then let the software process. It will adjust the output to place the GCPs in the correct location. If you use GCPs you don't have to use RTK at all, however it will help by reducing the amount of work the software has to do to correct the locations. Hope that all makes sense.
@@aviosmedia Thank you very much for your clear and quik response. I have a situation. I have to map a big area and obtain orthophotos and contour lines. I can't perform this job in one day (around 20.000m2). I'm using Mavic 3E, DJI RTK2 and DJI Terra. I can't run to o many photos in one reconstruction. So, I divided the whole area in smaller blocks and reconstruct it separately and got an orthomosaic for each block. When I open all the orthomosaics together I noticed a gap or mismatch on the border lines (difference in coordinates or elevation). I didn't notice this gap when the RTK is on the same location for two adyacent blocks. So my question is, how do I tie all the blocks together without having gaps? Do I have to make a GCPs network for the whole big area so they match all together? Do I have to put the RTK anthenna in the border of each block so they have a same common point as a GCP? Thanks for your help
@@ElizaldeCC Are you placing the DJI D-RTK on a known point? It sounds like you are just moving it around. The RTK antenna is not a GCP - you are aware of that, right?
@aviosmedia I guess I've been using it wrong. I've placed it on manhole covers, signaling or existing structures where I could check the info any time. But I've change positions every day, so I guess I've been moving it around. What would a known point? And how should I use the RTK antenna? Do you have a video on that? Thanks again.
@@ElizaldeCC Yes, sorry to say but you have been using it incorrectly. If you watch the video you will see that the base station is placed on a known point - and you have to enter those coordinates into the base station when you place it on there. Once the base station knows exactly where it is it can compare that to the information provided by the GPS signals, calculate the difference and then send that correction to the drone (or other RTK rover). If you have just been moving it around then that won't work as you are not giving the base station the correct information. I don't have a video on setting up a base station yet but it's on the list :). However, don't despair! I might have a solution for you if there is an NTRIP station that is reasonably close by as you can use the historic information from that to create a duplicate set of photos with updated GPS location information. This video shows how to do that - although you will need to find your local NTRIP/CORS system. ruclips.net/video/cESXvYiQ5ps/видео.html
Not sure I understand. Does the GNSS receiver (base station) sends RTK correction to the rover in direct or it sends it to the controller then the controller send it to the rover. If second option, with which protocol?
When setting the GCPs you don't use the drone at all. Typically you use two GNSS receivers - one placed on a known location (future video on how to find those coming) and one that moves around. The first GNSS receiver (base) sends corrections to the second GNSS receiver (rover).
Good videos. Whilst many claim RTK makes maps more accurate (they do without using GCPs at all), to get the most accurate maps (what's the point of creating one otherwise) you need to use GCPs. Before RTK drones existed, and now, the most accurate maps are created only by using GCPs. The benefit of RTK is not absolute map accuracy (a common misconception) but increased accuracy of initial image capture location. Think throwing jigsaw pieces out of the box and them being closer to correct final location: that's all. RTK simply makes metadata (exif data) more accurate at time of capture, but doesn't provide the highest possible accuracy. Only GCPs do that. P.S. I guarantee I can create a more accurate map with a non RTK mechanical shutter camera drone (Phantom 4 Pro) with GCPs, than by simply using an RTK enabled Drone. Without GCPs and in particular Checkpoints, map accuracy can never be trusted or proven. IMHO, if you're creating maps without GCPs and Checkpoints, you're not serious about mapping.
Great again Barry... Thanks.
Glad you enjoyed it
Excellent overview, thanks. Interesting that there's so many ways to produce a final deliverable. Lot of options limited only by one's budget!
Glad it was helpful!
It's also worth noting that it's best to include more than one if you can. e.g. use RTK during the capture and then PPK, GCPs or both. The MOST important one though is to use checkpoints. Without that you cannot be sure that everything came out as expected.
@@aviosmedia
great video👍🏾👍🏾, but I have a question please .
how to introduce checkpoints in software such as « agisoft metashape »?
How you use checkpoints varies based on the software. I don't use Agisoft so I can't help you there but I'm sure there are tutorials out there for that.
Once I have some time I will be doing a series going end to end from setup and capture of a site through processing and delivery of the end product to clients. But...it's going to take a while to get to that.
Thanks again- great video.
Has the one explaining how to use and process GCP’s/checkpoints come out?
Is this the one you are looking for? ruclips.net/video/PwXDztweU3I/видео.html
What's the process I have to follow if I have a DJI RTK2 and a Mavic3E with RTK module in order to get accuracy using GCP?
Thanks
You're mixing things that are sort of mutually exclusive. If you have a DJI RTK2 and it is correctly placed on a known location, and linked to the RTK on the drone, and you didn't get any errors during the flight then your picture should already be accurate. You can use checkpoints (GCPs not used in any correction) to make sure that everything is where it needs to be - that is how you confirm accuracy.
If you want to use GCPs then you can process in software that accepts GCPs. In that case you would upload the location of all of the GCP centers into the software, tag photos with the GCPs in them and identify the center of the GCPs in each photo, and then let the software process. It will adjust the output to place the GCPs in the correct location.
If you use GCPs you don't have to use RTK at all, however it will help by reducing the amount of work the software has to do to correct the locations.
Hope that all makes sense.
@@aviosmedia Thank you very much for your clear and quik response.
I have a situation. I have to map a big area and obtain orthophotos and contour lines. I can't perform this job in one day (around 20.000m2).
I'm using Mavic 3E, DJI RTK2 and DJI Terra.
I can't run to o many photos in one reconstruction. So, I divided the whole area in smaller blocks and reconstruct it separately and got an orthomosaic for each block.
When I open all the orthomosaics together I noticed a gap or mismatch on the border lines (difference in coordinates or elevation).
I didn't notice this gap when the RTK is on the same location for two adyacent blocks.
So my question is, how do I tie all the blocks together without having gaps?
Do I have to make a GCPs network for the whole big area so they match all together?
Do I have to put the RTK anthenna in the border of each block so they have a same common point as a GCP?
Thanks for your help
@@ElizaldeCC Are you placing the DJI D-RTK on a known point? It sounds like you are just moving it around. The RTK antenna is not a GCP - you are aware of that, right?
@aviosmedia I guess I've been using it wrong. I've placed it on manhole covers, signaling or existing structures where I could check the info any time. But I've change positions every day, so I guess I've been moving it around.
What would a known point? And how should I use the RTK antenna? Do you have a video on that?
Thanks again.
@@ElizaldeCC Yes, sorry to say but you have been using it incorrectly. If you watch the video you will see that the base station is placed on a known point - and you have to enter those coordinates into the base station when you place it on there. Once the base station knows exactly where it is it can compare that to the information provided by the GPS signals, calculate the difference and then send that correction to the drone (or other RTK rover).
If you have just been moving it around then that won't work as you are not giving the base station the correct information.
I don't have a video on setting up a base station yet but it's on the list :).
However, don't despair! I might have a solution for you if there is an NTRIP station that is reasonably close by as you can use the historic information from that to create a duplicate set of photos with updated GPS location information. This video shows how to do that - although you will need to find your local NTRIP/CORS system. ruclips.net/video/cESXvYiQ5ps/видео.html
Not sure I understand. Does the GNSS receiver (base station) sends RTK correction to the rover in direct or it sends it to the controller then the controller send it to the rover. If second option, with which protocol?
When setting the GCPs you don't use the drone at all. Typically you use two GNSS receivers - one placed on a known location (future video on how to find those coming) and one that moves around. The first GNSS receiver (base) sends corrections to the second GNSS receiver (rover).
@@aviosmedia how does this work with a DJI D-RTK 2 workflow?
@@mekdemltd The D-RTK 2 is the base station
Good videos. Whilst many claim RTK makes maps more accurate (they do without using GCPs at all), to get the most accurate maps (what's the point of creating one otherwise) you need to use GCPs.
Before RTK drones existed, and now, the most accurate maps are created only by using GCPs. The benefit of RTK is not absolute map accuracy (a common misconception) but increased accuracy of initial image capture location.
Think throwing jigsaw pieces out of the box and them being closer to correct final location: that's all.
RTK simply makes metadata (exif data) more accurate at time of capture, but doesn't provide the highest possible accuracy. Only GCPs do that.
P.S. I guarantee I can create a more accurate map with a non RTK mechanical shutter camera drone (Phantom 4 Pro) with GCPs, than by simply using an RTK enabled Drone. Without GCPs and in particular Checkpoints, map accuracy can never be trusted or proven.
IMHO, if you're creating maps without GCPs and Checkpoints, you're not serious about
mapping.
@@alwaystravelling100 100% agree. Love the jigsaw puzzle analogy!