Mapping Accuracy with IP

[Ranger Bob]

Hi bluelight (and everyone else!)

Thank you for your quick reply; apologies for my tardy response but I was working my way through another thread!

I work for a nature conservation organisation and most of the perceived work would be overflying nature reserves or protected areas to update aerial maps of those sites. Often the accuracy won't need to be that expected of surveyors but we use ArcMap 10.2.2 and it would be great to create a Feature layer on that to populate with the imagery (so other staff could use it too). For that reason a reasonable level of accuracy will be required. We look like we may be getting an eBee+ fixed wing too but that will be controlled centrally and the idea of a local aircraft is a more convenient, faster and more reactive deploy when needed.

Some of the jobs envisaged are huge (estuaries and harbours along the Dorset / Hampshire coast for instance) and I imagine the eBee is more suited to these. However there will also be sites up to a couple of hundred hectares to map and it is these 'smaller' jobs that I hope to cover.

I originally only had funds for a P4P but a bit more stretching to the basic I1 Pro / I2 basic set-up may be possible. So I've basically spent this weekend trawling the Phantom & Inspire Pilot forums (plus others) trying to figure out the best suitable combo I can afford to get! Not easy!!

To start with I thought the P4P would suffice (everyone agrees the optics for mapping are superb) but I've been put off a bit by accounts of wind affecting it too much so then I started looking at the Inspires. The I1 Pro would be perfect if the batteries lasted longer like in the I2, as the sensor is great but I'm unsure if the glass in front of it is as good. The I2 scores on flight time but I can't find any reference to it being used for mapping and posters hint that it is much better for video than still photography.

With either Inspire I could stretch to two extra batteries but as the P4P is much cheaper I could afford even 3 or 4 more meaning I could fly the P4P all day (therefore making my time on site as efficient as possible) but I'd be restricted to less time with the Inspires (maybe an hour tops). However, I might have to raise funds with the setup in the future and it is generally agreed that the Inspires look professional while the P4P look like 'toys' and clients aren't impressed if you suggest their use.

So many decisions! Any advice or help / experience greatly appreciated!!

Thanks

Bob

 

[bluelight.support]

I started mapping with a Phantom 4 and with the inbuilt GNSS could get accuracy to around 2m. Ground resolution was around 2-3cm/px at around 150ft and you could get around 14-18 minutes flight time. Lager areas could be facilitated by going to a higher altitude - however this lost resolution. (Remember resolution is not accuracy its the picture quality - which in effect can effect other factors, just to confuse you!). I decided to move the the Inspire 1 Pro Black which i had the X5 camera and 12mm Olympus lens. I got rid of the 15mm DJI one as the Olympus has the best press for imaging. I got amazing results with this at 100-150ft getting up to 1cm/px and 16mp images, which made for very high quality mapping and data. This coupled with a few ground control points with a trimble unit gave us both 1cm/px resolution, and around 1-3cm position accuracy. However you could argue in Arcgis both of these position resolutions (either phantom or inspire) would have overlaid the basemap very well!!

I was so impressed by the Inspire 1 Pro, I sold it and bought the Inspire 2 for a number of reasons. Better mp sensor, I could still use my 12mm lens, 2 batteries for redundancy and slightly longer run time, and i loved the I1 and wanted the latest and best. I will probably look at the M200 (not 210RTK as that is useless for positional geolocations (currently)) as it also offers a longer flight time and weather resistance, I would prefer it more if it also had rotor redundancy as well, oh well!

You mention wind, however both the Phantom, Inspire 1, Inspire 2 and M200 are all rated at 10m/s wind which is pretty crap ! This however, is much better than the eBee+. The eBee does have the advantage of time and distance, however in the UK that makes no difference as the 500m line of sight etc and i can do that with a multi rotor anyway. Also the eBee at £15k basic for a bit of polystyrene and a bastardised freeware flight control system is a little high for me.

You could always prove concept with the P4P which would be an ideal starting point with the 1" sensor, however i know from my experience. i wanted more very quickly and moved through all of them in 3 months! Your also using ESRI which is 15k or so itself so as you know - you get what you pay for!

 

[LJCF]

In reference to the M210RTK, why exactly is it that the accuracy of RTK is not available to the final orthomosaic? Obviously some basic accuracy geolocational data is within the image - to perhaps a couple meters, so can't (or isn't) the RTK data upgraded for use in the final product and eliminate ground control points? Maybe I am not asking this quite right, but hopefully you get my point....trying to understand here...

 

[bluelight.support]

In reference to the M210RTK, why exactly is it that the accuracy of RTK is not available to the final orthomosaic? Obviously some basic accuracy geolocational data is within the image - to perhaps a couple meters, so can't (or isn't) the RTK data upgraded for use in the final product and eliminate ground control points?

Stop looking at the DJI RTK system to work in the way you would expect it to within the mapping / survey industry. I have written loads of posts of this in different sections if you do a search. DJI in their wisdom do not transmit the RTK data through the flight control system into the camera - therefore the geo location data on the EXIF info on the JPG is not cm accurate. Apparently if you put in exact co-ordinates into ground station pro the system will fly to that co-ordinate, however, no accurate details are added to geo-location which makes the system useless for our purposes. Im sure DJI will come up with a survey solution and add many $$$$$ to it!

Ground control point, or cross reference with previous ground survey points for accuracy is the only way to go!

 

[LJCF]

OK, thanks - it is hard to understand the benefit (for me at least) to have RTK precision on your craft and not pass that through to what you are imaging???? But, there are many things I don't understand....

Bluelight - any comments of Direct Georeferencing in the drone world?

 

[bluelight.support]

any comments of Direct Georeferencing in the drone world?

In what context ?

 

[LJCF]

In what context ?

Ok, yes that was a bit vague....
1. Is anyone in the drone world doing it?
2. Is it comparable to using ground control points (cm level) with traditional aerial photogrammetry?

I understand it has been around for quite awhile, but have not heard of it yet in the context of flying drones.

 

[bluelight.support]

1. Is anyone in the drone world doing it?
2. Is it comparable to using ground control points (cm level) with traditional aerial photogrammetry?

Good questions, people are trying it now, however they are looking at drones of £25k+. I can built an add-on for around £3k to give CM Accurccy, however with my workflow, its just as easy to use map pilot on an Inspire 2 and the add GCPS and post process to 1-2cm accuracy. Systems like eBee and the Here+ will only go to 6-8cm even processed - but they dont tell you that!!

If you already have ground control points produced for traditional surveys and the known locations of the points, then adding the co-ordinates in post process is an easy thing to achieve also.

 

[Plingboot]

Just out of interest, what kind of cost are we talking about for a handheld device for accurate ground control points?

I'm about to start doing some mapping tests with my P4 and I1/X5 as I've been asked about doing some aerial surveys for a developer. He's currently paying £800 for a chap with a tape measure and a clipboard (sic) and he was wondering whether an aerial survey might be more useful.

If I offer to do a survey I want to make sure my 'product' is as accurate, if not more so than what he's currently getting - with the added benefits of it being much more 'visual'.

 

[bluelight.support]

I'm about to start doing some mapping tests with my P4 and I1/X5 as I've been asked about doing some aerial surveys for a developer. He's currently paying £800 for a chap with a tape measure and a clipboard (sic) and he was wondering whether an aerial survey might be more useful.

tape an clipboard will be accurate however, what does your client want?

With a drone you will be able to provide him full georeferenced imagery, DEMs, 3d model, and the ability to mature volumes and distances from for his plot.

What is he looking for from this chap - simple measurements?

What accuracy is he looking for - 1m, 1ft, 1cm, this is all relevant, and can be the difference between £2k and 15k!!!!

 

[Plingboot]

From what I understand these aren't big plots - we're talking anything from 2-10 houses - redeveloping brownfield land most of the time.

It seems that there have been times when a traditional survey has come up short - for whatever reason.

I honestly don't know what accuracy he's expecting - I will find out.

 

[bluelight.support]

It seems that there have been times when a traditional survey has come up short - for whatever reason.

I honestly don't know what accuracy he's expecting - I will find out.

Where in the UK are you ?

 

[Plingboot]

South West London. But the project in question isn't around here.
I fired off an email to see what kind of accuracy he'd be looking for…

I also need to satisfy myself that I can produce a quality product.

 

[Ranger Bob]

Learning a lot here!

What still evades me, however, is an opinion on the suitability of the DJI 15mm lens which comes with the X5 (I1 Pro) for mapping? I understand that the Olympus lenses are superior but I'm not sure I can stretch to one yet. Is it a waste of time to use the DJI one for this?

On the same theme the forum seems silent on the suitability of the stock X4S for mapping (I haven't found any answered posts on this!) - anyone out there have experience of this? I ask for the same reason as above; if I can convince the bean counters to come up with enough for an I2 then it will be the 'lowest spec' one available and additional cameras / lenses are unlikely to be forthcoming for a while.

As ever, thank you in advance for your help.

Bob

 

[bluelight.support]

I like the 12mm as its a wide angle lens (more so than the 15mm standard one), hence why its popular for mapping. Its also an easy lens to add as with a suitable lens filter you don't need a counter weight.

The X4S would be fine for mapping think having a sony rx100 on your craft and you'll see its a good combination.

I prefer the x5s as i advised earlier as i like operating at different heights and resolutions.

Either combination X1Pro and x5 or I2 and X4s will work well for you - however also look at the complete workflow, what will you be using after the image capture ? That has a massive effect on the quality. Also what accuracy is required ?

 

[Ranger Bob]

Thanks bluelight for your reply. From that and other posts on these forums it would appear I need to find out a few more things from our mapping bods about the final product they would be happy with, and what additional kit they might supply that I could use to improve the accuracy of the acquired data (as I won't have the funds myself). So, for the moment anyhow, a decision still pending! All the best,

Bob

 

[Plingboot]

tape an clipboard will be accurate however, what does your client want?

With a drone you will be able to provide him full georeferenced imagery, DEMs, 3d model, and the ability to mature volumes and distances from for his plot.

What is he looking for from this chap - simple measurements?

What accuracy is he looking for - 1m, 1ft, 1cm, this is all relevant, and can be the difference between £2k and 15k!!!!

They're currently using a topographical survey to get measurements of the site and boundaries - accuracy to 1cm.

Levels need to be taken across the site(s) to establish finished floor heights and set drainage - accuracy to 5mm.

Not sure what an aerial survey will bring to the party here - given I'd probably need an expensive piece of gps hardware and the traditional skills and experience of a surveyor.

I can see how it might be an useful addition to their existing process - though that's also adding a cost, but I don't think it's a replacement.

 

[RMartin]

DJI 15 mm(30 equivalent on full sized sensor)

When you say Simactive ...is that Correlator 3D?
I'm very interested in Correlator 3D. Propeller gave us very useful information and accurate but when I inquired about pricing I lost interest. Data Mapper allows you to enter GCPs individually or import as a csv. File.however we have to convert metric elevations to feet. You don't have that problem.
It's all good.
I'll try the Agisoft and Simactive. Thanks for the info.

You can go into the setting in Correlator3D and change between meters and feet as a unit of measurement. The only drawback to Correlator3D and AgiSoft is that your GPU MUST support at least OpenCL 1.1. Pix4D only requires OpenGL compatibility. If not, my cost analysis over a five year period still shows that it is cheaper in the long run to go with Correlator3D and the FirePro W7100 GPU over Pix4D. You could even go with the W9100 and save a little.

 

[RMartin]

Municipal Park and Recreation Boards like to have Orthorectified maps of their parks for planning purposes. I place a white 4foot by 10 foot white plastic table cloth(you can get at Lowes or a party store on a roll) out in an area easily seen from the air and use it for scale in the finished photo. There are a number of Apps to set up the flight. I use Dji GSPro, Drone Deploy and MapPilot. If under 200 images averaging 12 mp each, Maps Made Easy does a good job of stitching images together free. When Maps Made Easy can't do it then Drone Deploy Can. Both of these can be upgraded to accept GCPs and then add Data Mapper, Pix 4D and Correlator 3D and you have lots of choices out there for finishing higher accuracy maps. 250 feet AGL is a sweet spot for a decent Ortho Map with an I2/x5s combo. 100 to 150 feet for cm accuracy(.4 inch). some software will go as small a GSD as you can provide. Some are limited in their output.

We use 24" X 24" plywood panels with a hi-vis paint job. I drilled a pilot hole in the center of each for the range pole to lock into for a more consistent measurement. I would use DroneDeploy but $75.00USD per package to use GSPs is not going to happen. We produce hundreds of maps per year and the cost would never get past the board. I do my own processing in-house. More fiscally responsible and more capital in the chain floating around to be earmarked for more toys.

It has been fun trying them out. I had to get used to the bird doing everything on its own. The last 3 flights, after set up , I tapped once to fly the whole mission without touching any controls on the RC. At first I felt compelled to land it myself but not any more. The I2 has worked flawless for me so far.

I know it is a lot more accurate to let the software control the flight, and that is how we do it as well. It does take a bit of the fun out of it standing around waiting for something to go wrong though...

 

[RMartin]

They're currently using a topographical survey to get measurements of the site and boundaries - accuracy to 1cm.

Levels need to be taken across the site(s) to establish finished floor heights and set drainage - accuracy to 5mm.

Not sure what an aerial survey will bring to the party here - given I'd probably need an expensive piece of gps hardware and the traditional skills and experience of a surveyor.

I can see how it might be an useful addition to their existing process - though that's also adding a cost, but I don't think it's a replacement.

If horizontal accuracy is set to 1cm then they are at mapping grade; one step below survey grade. Matching 5mm of accuracy is going to be more challenging and is probably only going to be possible via a surveyor with a total station. GPS is excellent for measuring northings and eastings but has some limitations when it comes to heights.

We are using a UAS to save time gathering horizontal data but we are also backstopping it with conventional survey methods and equipment. Both to spot check existing orthomosaic data and provide accurate height/elevation data. I do not think that traditional survey methods will be replaced by UAS flights in the near future. You still, after all have to set GCPs by the old tried and true method. I do think that where extreme accuracy is not an issue, or not a big issue, that UASs can provide a cheaper alternative vs conventional boots on the ground. (The preceding was my opinion. I am legally entitled to have them. My opinions hold no sway, influence any government, or amount to a hill of beans. But they are mine and mine alone.)