IS IT JUST ME??

Probably not as far-fetched as it sounds. It may be currently technically possible, but I doubt the cost/benefit is realistic yet. Given things like project Tango on the horizon I would say even approaches like this will superseded soon.

> Or does this sound squirrely:
>
> "Looking for a few good land surveyors, engineers or architects who want to learn how to use a regular digital camera for surveying and mapping at 1-2 cm accuracies. Please join me for this free seminar. You can register at http://ericcolburn.com/events/ or https://datumate.eventbrite.com. I look forward to seeing you there."
>
>
> So all we now need is a hand-held Garmin unit and a Nikon CoolPic and we are ready to do topos???

Imagine you are back in the early 80's or late 1970's. Some guy comes along telling you that one day you will be able to carry around a stick and know exactly where you are on the earth to a survey grade accuracy. Sound squirrelly? Very. That dude is a nut and out of his mind. Yet here we are.

I have no idea of this Eric Colburn guy can do what the ad hints. But I don't doubt that some day someone might be able to do just that. Who knows, today could be that day.

Larry P

Having worked for several years in phogrammetry, I still think it is pretty far fetched. Considering such things as depth of field, lens curvature corrections and calibration problems, I just cannot see it.
PLUS they are talking about tying a photo to a known ground elevation. That's tough enough on a stereo plotter, but for software to be able to do so is a bit of a streatch in my mind. (How good is the software's stereo vision??)
ALL of those corrective factors listed above would apply to even a digital camera, as would the link to ground control. I just cannot see it.
May the Good Lord save us from GIS "experts". I'd hate to have to take that data into a court room and try to prove my case.

Perhaps if we strap all that gear on a drone....????+o(

7 years ago... http://www.topconpositioning.com/sites/default/files/articles/alaska_photogrammetry.pdf

I looked into this a bit. You still need to accurately survey several control points common to the multiple photos you take. Just like aerial photogrammetry, but turned on it's side, kind of. Once the 3-D model is created you manually draw the vectors. The program they're hawking is being offered at a "special discount" of $5000.00 to those who attend their seminars. It's called Datugram 3D. There's a Youtube video if you Google it.

This technology has been in use in other industries for a while now...especially in the 3D animation/design areas.

I saw a demo of this at the NYS Conference. It is pretty slick and seems like it works pretty well. You have to take shots with the total station to reference stuff but if you are doing things like building faces it would work pretty good. I'm still trying to think of the right applications for this technology and how they fit my business.

It does work and it works pretty good!

Tom

I've never worked in photogrammetry much, but I've talked at length some people I would consider "Gurus". Apparently the high-resolution close-range photos that are now available is what is making it possible. From what I have seen it's just not possible "enough" to be a money maker, or to produce results that I would seal.

You may be right, it may all be very pie-in-the-sky.

As I have stated before, surveyors are wholly unfamiliar with the rapid development that can take place when consumer funding is thrown into the mix. Google may likely develop something over the next year that we wouldn't have seen from Trimble/Leica or others within the next decade.

This sort of thing is currently being sold as a system by Trimble. I've seen a "quickie" demo, and it's quite powerful stuff, given certain limitations. It's faster and capable of higher accuracies than some scanners.

The main engine is the ability to autocorrelate conjugate imagery. When the software can do that with 3 to 5 different highly convergent images, it's really going to make a whole lot of urban topo a snap.

I'm not yet convinced of multiple image correlation done "automatically" when it's not targeted.

Maybe someday soon ...

Good article. Powerful stuff! It makes me wish I was a lot younger.

:good:

Cliff -

I wish I was still at NEI (did I just say that??) just so I could show you what Trimble is doing now with the V10 and the latest release of TBC; I'd love to get your take on it. We've done a "real world" demo with it and the results were impressive.

Lee,

That's the system I was referring to.

I agree, I would love to be just starting out. Alas, I am just wrapping up.

However, I do not think that anything will replace a good shovel and a sharp mind!

Joe

:good:

This is totally doable, and has been for a while. Photomodeler ( http://www.photomodeler.com/) has been around for years, and all you need is a digital camera. I used it back in 1998 or 1999, and have pulled measurements form the photos that were within 2cm.

In terms of modern Photogrammetry (UAS, Trimble V10), anyone wondering how it all works, Google "bundle block adjustment".

Eric Colburn is a good egg.

"...I'm not yet convinced of multiple image correlation done "automatically" when it's not targeted...."

Google Blaise Aguera y Arcas and Photosynth

I've used the demo version and it seems to do what it says. Eric Colburn was helpful and genuine.

Great conversation and (legitimate) questions about this new technology. I think you'll be surprised at how far and how fast this new technology has advanced. To show you just what I'm talking about, I'd like to personally invite you to my free DatuGram™3D Webinar on Thursday, June 5th from 2-3pm EDT. You can register here >> http://eepurl.com/UI1uf