Drone Mapping

Prior to starting our p4p program we used a p3 consultant. The results weren't bad, and you get experience. Good luck

You will need to fly lower and have more GCP's to get the same results as a 20MP w/ Mechanical Shutter. Most likely the 12MB camera has a digital shutter. This creates "Rolling Shutter" syndrome, which Pi4D can attempt to correct if you enable the feature under camera calibrations.?ÿ

Like Lee said, fly lower. 100ƒ??AGL or less if possible.

In reference to what Lee said about rolling shutter I would suggest you stop the UAV for each pic, or reduce the flight speed to a minimum to reduce distortion. Although Pix4D compensates for the rolling shutter effect I havenƒ??t extensively experimented with a lower resolution sensor with digital shutter. Therefore, like He said be sure you get your camera settings into the software correctly. I believe Pix4D will automatically recognize the sensor from the Phantom 3 Advanced/Pro. Pix4D has pretty good support if you have any issues although itƒ??s usually a few days of emails back & forth even for simple answers.

Of course you could also just upload your images to the cloud platform from Pix4D or DroneDeploy on either of their trials.

?ÿ

Thanks for the input!?ÿ I was planning on flying at 150' but will run an additional flight at 100' for comparison (it gets a little nerve racking at that height but there should be enough wiggle room above any obstacles).?ÿ Additionally I was planning on a speed around 10 mph.?ÿ I'm not sure how to stop the flight to take a pic using the drone deploy app but I will look into it.?ÿ So far my camera settings will be as follows: ISO=100, shutter 1/1000 or close to that depending on the over exposure warnings and histogram results.?ÿ Practicing still shots I noticed that the camera HDR setting produces crisp results but I'm not sure if it is possible or a good idea to perform aerial mapping with it on.?ÿ One suggestion I had, since it will be a 12mp camera, was to use the pix4d capture app and fly a double grid or a separate perpendicular flight in drone deploy.?ÿ Am I missing anything else? ?ÿ

I started off with Phantom 4 standard (12mpix rolling shutter camera) and the results was very good compared to ground surveys. The workflow i use is to put GCP's every 50 to 70m and then fly slow speed (5m/s max) 30 to 60m AGL single or double grid depending on the project.The problem is that the P4 is not productive. Recently I upgraded to Phantom 4 Advanced and the quality of the camera is much better. I haven't done any project yet but will soon.

Posted by: @leegreen

You will need to fly lower and have more GCP's to get the same results as a 20MP w/ Mechanical Shutter. Most likely the 12MB camera has a digital shutter.

Lee, I want to ask how many GCP's do you place on the ground in case of the Phantom 4 Pro/adv? Every 100m is ok? With the standard P4 more GCP's are needed for sure.

@gto234

I currently use the P4RTK. Yesterday I flew a private air strip 3200'x100' with 4GCP's. They are more for ground truthing, not for adjusting control. I use both RTK via the DJI RTK-2 base, and PPK for backup via Topcon Hiper-VR. The DJI base is a bit crude in the method needed to properly enter the base coordinates.  Which is why many people fail at getting sufficient results.

Thanks Lee,

before the P4RTK did you had some rule about GCP's distribution? I have a 3km linear project (road) and planning to use the P4A with GCP's and i need the minimum number of GCP's.

With the P4 I was planning to place about 110 GCP's...

We have been trying to do drone work for topo stuff for awhile, but we kept getting mired in the learning curve.?ÿ We have a pro4, like mentioned above, and tried several different programs.?ÿ Recently we hired a drone pilot who also cleaned up the data, processed the GCPs, etc.?ÿ He was using a 12mp camera, in a mavic.?ÿ He now uses out pro4, and just started using drone deploy to process the photos and GCPs and the photos are noticeably cleaner.?ÿ As for how much tighter, I cannot really say.?ÿ I know we are able to compare points located on the ground to points located by DES/DEM usually less than a tenth both horizontally and vertically.?ÿ I used to put out about 10 panels per job, this pilot puts out about 7.?ÿ So far we have been really happy with the results.?ÿ I also am using a program called virtual surveyor to help me work through the data that is collected, turning the DES/DEM into usable points to take into AutoCAD/Carlson.

@gto234

With the P4Pro or P4Pro 2.0, I would place GCP's every 500ft to 700ft around the perimeter of site and add points at high /low points throughout the site.  I have found that you must encompass the site with control when not using RTK or PPK. Test have shown that data just 75ft outside the permiter would have distorted vertical outside our tolerances. Flying the road project you mentioned is very tough with this type of drone. You would need control on both sides of the and fly in a grid pattern.

@leegreen

Hi Lee,

I was wondering if you could give your opinion on the P4RTK vs the BAAM Tech setup you were using here:

https://surveyorconnect.com/community/photogrammetry-lidar-uas/ppk-results-with-phantom-4-pro-v-2/

From the outside looking in, it seems as though the P4RTK has more options (RTK & PPK), but that the BAAM Tech saves you a few thousand $$ with a less polished look.

Any thoughts/inputs on your experience with them greatly appreciated.

Cheers.

@jaccen

The Baam system is a good value, but the biggest limitation is the reduction in flight time due to added weight and changes in aerodynamics. The P4RTK is the most stable and most accurate I have used. I always have a base collecting static data, and use the DJI RTK-2 running at same time. RTK is same accuracy as PPK, and much faster processing. RTK has no post processing. With BaamTech I used Drone Deploy and enjoy that much better than DJI GS that I am forced to use on the P4RTK. 

@leegreen

Thanks.  Very much appreciated.  I often find reports "from the trenches" are worth far more than most review websites.

Well, the results are in and . . . not too bad. Definitely surpassed my expectations in regards to accuracy and precision. I did fly a double grid using drone deploy's add on feature and set the gimbal at 70?ø per advice I read online. The only drawback is the number of images to process due to flying at such a low altitude (100' AGL). Pix4d worked very well and selecting the rolling shutter option for the camera produced a very good point cloud and orthomaisaic. Next I'll try the same grid at 200' and compare the differences.?ÿ A 12mp camera definitely could work for small sites around a handful of acres but I don't think I would push it for any more than that.?ÿ Unless there is a way to process the data in increments and merge it which I seem to recall reading.

We have had good success. Different focal lengths, altitude, and flight patterns help. You can check our web site for examples:

https://digital-mapping.net/

The Rockspan example is a wooded area. We flew it during leaf off to pick up the topography. We also flew it during leaf on for comparison:

https://digital-mapping.net/rockspan/

?ÿ

?ÿ

?ÿ

Hi Mr Lee Green or others.
I have a BaamTech base and Phantom 4 Prop with Baamtech blackbox PPK?ÿ and antenna.
My problem is how to create a RINEX file on the base station.?ÿ
Do I have to have to capture images on the drone with Maps Made Easy (say) to have the base RINEX file created ?

My failed workflow:
I set up the base outside (simply on the ground as this is a test only), waited 5 minutes for good satellite coverage , started the drone motors for 5 minutes, did not fly it, turned off drone, turned off base and found I had no RIXEX file on the base. I did not create any images on the drone.
Maybe this is the issue ??
Cors Stations are available nearby for later PPK processing.

Thanks for your advice.. my hair is falling out with worry about this matter!!