I don't know, as they are not my targets. I noticed this in the web and thought it might be a good idea.?ÿ
Myself, I have found that a can of paint for targets on pavement and commercial?ÿ 3'x3' iron cross targets work just fine. I can't see how you guys can find those small 1'x1' targets from 200' on large projects. You must fly much lower.
I couldnt agree more about not being able to find those tiny little targets. I've never had good luck finding or processing with them. I always use the commercial 3x3 iron cross targets and they work perfect.
If you can't resolve a 1' target then how can you model any resolution approaching that??ÿ
You could easily "resolve" it. Have you done the processing for aerials before? I find that its significantly faster and easier to spot the target when its a 3x3 target opposed to anything smaller. With hundreds, sometimes thousands of pictures, I want to be able to very quickly discern where the target is from my list in the thumbnail.?ÿ
If you can't resolve a 1' target then how can you model any resolution approaching that??ÿ
Ryan,
I agree with you. Much better working with commercial grade 36" iron cross targets.
I think some of these people are hobbyists?ÿat best, from the comments and looks of these makeshift, low budget targets. It is obvious they have not had to sort through hundreds?ÿof photos at 200 to 300 ft, to find a 12" target, made from 6" black/white tiles.
?ÿ
Lee,
We have a profitable SUAS program that I started late last year. Our 'low budget' targets look like LED lights without even zooming in. Picking true center isnt difficult at all. Having the point number painted upper right and a good control map helps
These targets are one foot square alternating flo blue and white. We fly at around 140 to 180 as our camera is only 20 megapixels (DJI 4 Pro).
I realize thats entry level, but it's no hobby. We are putting out solid products and making good money.
I tried a few of those when we had loose sand. They worked well due to contrast. Im still trying to find them in colors not generally found naturally in our area. Blue is only on some flowers and turkey heads. Not enough of an occurrence to be concerned...
You could easily "resolve" it. Have you done the processing for aerials before? I find that its significantly faster and easier to spot the target when its a 3x3 target opposed to anything smaller. With hundreds, sometimes thousands of pictures, I want to be able to very quickly discern where the target is from my list in the thumbnail.?ÿ
Is this an area the software could become smarter and seek out and find a given target? 'Show' the software a target and set it free to search images.
Sure, there'd be differing backgrounds but with experiment I wonder why it shouldn't work. Pick a particular target type, size, design and stick with it. May need a couple or three for different backgrounds or sites.
With technology smart enough to deduce elevations to sub decimeter I'd argue it should be smart enough to find specific targets. If not then challenge the software manufacturers to do just that.?ÿ
I imagine you would know how many targets are set and if the numbers don't tally then look for the missing ones.
Or does it do that already??ÿ
I imagine you would know how many targets are set and if the numbers don't tally then look for the missing ones.
Or does it do that already??ÿ
Yes, there are some software apps that can automatically identify targets for GCP's, but not Pix4d.
Once Pix4d adds this feature, I see a market for 3'x3' heavy duty targets, printed with a QR code that identifies the GCP ID.
Lee,
We have a profitable SUAS program that I started late last year. Our 'low budget' targets look like LED lights without even zooming in. Picking true center isnt difficult at all. Having the point number painted upper right and a good control map helps
These targets are one foot square alternating flo blue and white. We fly at around 140 to 180 as our camera is only 20 megapixels (DJI 4 Pro).
I realize thats entry level, but it's no hobby. We are putting out solid products and making good money.
Lots of us are not hobbyists; we use large targets when we can but often they need to be collected up afterwards and that costs time. On a large site it might involve another half days walking around to collect them. This site was done in 4 hours flying, 9 flights, 11 hours on site fixing control, 7000 photos, all targets had to be in biodegradable paint. One days work. The flights were with a fixed wing Quest SUA
The link below gives the Google location.
https://www.google.co.uk/maps/ @53.911084,-1.8182477,2521m/data=!3m1!1e3
An extra space got into the link by accident - it should have read
https://www.google.co.uk/maps/ @53.911084,-1.8182477,4211m/data=!3m1!1e3
Sorry.
Looks like rpls is adding in the extra space for some reason. There should NOT be a space at /maps/@53
I'll type it in here longhand and see if that comes out all right.
https://www.google.co.uk/maps/ @53.911084,-1.8182477,4211m/data=!3m1!1e3
Third time lucky.
I imagine you would know how many targets are set and if the numbers don't tally then look for the missing ones.
Or does it do that already??ÿ
Yes, there are some software apps that can automatically identify targets for GCP's, but not Pix4d.
Once Pix4d adds this feature, I see a market for 3'x3' heavy duty targets, printed with a QR code that identifies the GCP ID.
Given the target coordinates, Photoscan will present all the images and just need a confirmation on each of the target centre points - thus you can ignore any image where the target is partly obscured, or poor due to shadow, etc. If you are using road markings for any of the targets then it's probably a good idea for a human to check them in any case (also increases the chance if identifying if a board has been moved). It's a good idea to ground photograph any permanent boards when they are first installed.
I imagine you would know how many targets are set and if the numbers don't tally then look for the missing ones.
Or does it do that already??ÿ
Yes, there are some software apps that can automatically identify targets for GCP's, but not Pix4d.
Once Pix4d adds this feature, I see a market for 3'x3' heavy duty targets, printed with a QR code that identifies the GCP ID.
Given the target coordinates, Photoscan will present all the images and just need a confirmation on each of the target centre points - thus you can ignore any image where the target is partly obscured, or poor due to shadow, etc. If you are using road markings for any of the targets then it's probably a good idea for a human to check them in any case (also increases the chance if identifying if a board has been moved). It's a good idea to ground photograph any permanent boards when they are first installed.
When we collect our GCPs, we take a photo of the set up on the Photo Identifiable (PID) target, usually a stop bar, painted stripe or man hole when there is a good contrast between the concrete and asphalt. We Also take two additional photos(sometimes more) of the surrounding area looking North, and looking East, to show the area around where we taking the GPS point to insure the location is adequate for the project requirements from the client and consistent with the data collection process we agreed upon in the contract and the actual GSD required.
I added the mag nails with flagging, and for all ground points a 60D Nails at grade so we get consistent measurement( could also use a topo shoe but I like swinging the hammer....) and can occupy the same point for any future work.?ÿ
I have printed some targets, a 33" and a 22" and have been adding functionality as I think of it.?ÿ What functionality can you add to a target??ÿ That's a fair question.?ÿ I added viewing slats through the target for help in finding the control point it is over, I added stand offs so that the top and bottom lay parallel, the painted area is slightly raised so that they are easier to change color, wings that can be removed with gloves on .. etc?ÿ The shown 22" is a casing from the printed one and the large one is actually one of the printed ones.?ÿ I am no longer going to print the numbers and will just use "sticky" numbers, which is a shame as the printed numbers went all the way through.?ÿ The cast plastic has a UV protection added to it and is very strong.?ÿ The hardware is all zinc coated and I am not that concerned about the durability of any piece.?ÿ Nearing the end of the prototype stage and might switch to a light grey plastic with black instead of the white, just need more feedback on that.?ÿ It's our rainy season, so, it is a bad time to prototype ...
Very neat, but you would need to go round and collect them at the end of the job - OK for small sites, but a lot of extra walking on large ones - see my comment on June 1st. How are you fixing them down - nail through the centre hole? If you used them as long term markers they would need a "groundsheet" underneath to stop vegetation growing up and quickly obscuring the cross. We use landfill liner sheet for that, cut into 4 ft. squares.
Chris, the fact that they would need to be collected afterwards was always a design parameter.?ÿ These were not designed for the rural survey, but, for a more urban environment or mix.?ÿ As far as fixing, they are not as light as you might be thinking as just the poured plastic is 1Kg on the large ones and a little below 0.5Kg on the small.?ÿ The sighting holes can double for soil nails if they need to be fixed, but, they are heavy enough for most points.?ÿ For the conditions that you are concerned with, they would not work well and you might as well build some simple site specific targets. Different tools for different needs
A 4'x8' sheet of plywood split in half so that it yields two 4'x4' pieces is large enough to be seen clearly from the air and can be painted to any pattern you want.
Drill a hole in each corner to secure it to the ground and one in the middle for a temporary monument. Remove it when done and use on the next project.
You could even install a piano hinge down the middle to make it easier to transport.?ÿ?ÿ
Most survey supply places sell rolls of aerial marking material. It comes in a roll about 2 or 3 feet wide and you can cut it to length and secure it however you wish. Landscape nails work well in loam.
When doing traditional photogrammetry, this is what we did. Paint the top corner of a CB or find something with contrast already existing. This allowed us to get the control before or after the flights were flown. It worked for our highway jobs (ie. Hwy 401 , DVP, 407, in ON and TransCanada #1 in BC, etc.). This allowed us to not have to build/collect targets and still use points that were readily available/visible. It cut costs and allowed us flexibility.