Theodolite phone apps - has anybody used them?
-
(still) Interesting
Are you actually planning to use an iPhone with gps and theodolite app to generate/"scan" 3D models from stones?
I would be very interested in reading a "report" on how this works. Somehow I can't see how the (lack of) precision with a standard phone/gps will be able to generate anything close to accurate? I have never used a theodolite, but AFAIK it measures angles, which can be used to calculate distance or elevation? So if the position is accurate to +-10m what will that do to the precision of the models? None of my 4 gps's have been able to stay fixed on a location even when standing still - always a slight wandering. But that may be better closer to equator than up here in Norway?
Any reason why you'll use a phone/theodolite instead of a photogrammetry app? Sounds like a lot of work.. -
The gps function is the least important as we have a pretty good fix on the rocks location. Also we can pinpoint the location on maps of the area using triangulation from know landmarks or survey benchmarks. We are mainly interested in compass heading, level, vertical angle and horizontal angle. Distance can be read with a laser interferometer down to 1/16 of an inch. Level and vertical angle are hignly accurate with phone apps. I think compass heading is also quite accurate but less sure about that. The $3 price is hard to argue with compared to even a used theodolite.
As to location, I can not say until the project is finished.
-
I hope you will keep us posted on this project?
I have not heard of anyone using such techniques/tools to make a 3D model of objects/rocks, and can't see how you're planning to do it. Will you have to put lots of stickers on the rocks for triangulation, or can you measure a 3D point from one single position? How many points will you survey, and how long will it take? Thousands/days? And how will you get that point cloud into a (photo textured?) model with faces?
Lots of questions - as you can see
-
We are only talking about two rocks. So we are talking a couple of hours.
Photo texturing is just a case of painting the texture on the facets and doing a position texture and adding a bump map. Position is achieved with horz angle, vert angle, and distance from a single point.
Will need to build tripod accessory with gimballed mount that allows line of sight to rotate around a single optical center.
-
Is it possible to see an image of the planned project?
-
Rich, I am restrained by by client from releasing specifics at this time, but will be able to post at a later time. FYI in regard to accuracy. The azimuth reading go down to 1/10 of a degree. But being able to give a number to 1/10th of a degree does not mean the instrument is accurate and even if the instrument is accurate it does not mean that the user has sufficient perception and optical acuity to align the instrument properly. I suspect a pixel width may be larger than 1/10th of a degree. However my guess is that if used properly the thodolite app may be reasonably accurate and deliver good value for the money.
-
I'm very interested in this reconstruction type of work, mainly using photogrammetry programs like Canoma, Photomodeler, ImageModeler, and now that amazing PhotoScan SU plugin from tgi3d.
I have downloaded the GeoCam app for Android (filed under Entertainment - no wonder I didn't find it first..
And I have searched the net for good tutorials on how to survey terrain using a theodolite. Not found much of interest though. I remember from my youth that it was a popular summer job to assist surveyers - just standing there in the terrain holding a red/white reference stick
Do you know any good resources for such theodolite tuts?
If it is accurate enough I can see a use for it combined with PhotoScan, to get some fixed 3D reference points to start from, which should make it easier to align the models with the real world, and also in correct scale without any other reference measures?I asked before if you had considered photogrammetry software, but never got a reply.
Would you for instance be able to use a theodolite to reconstruct a sculpture like this one on my local airport, shot when I drove my daughter there a few days ago, and reconstructed in PhotoScan/SU in a few hours. Accurate down to a few mms, and so far untextured.
-
The main reason I don't use PhotoModeler is the price. I am semi retired and money is tight. I could do this project with the one month free trial and build up the experience needed for a better paying second gig which would pay off the software.
Bjorn, see http://www.youtube.com/watch?v=BZi0owCSsso for a good basic survey tutorial.
Many years ago I had a problem absorbing the basics of trigonometry. It has taken me a long time to recognize how ridiculously simple it is. Had we actually done some actual surveying I would have probably gotten it on the spot in just a couple of days.
I was going to use a laser range finder for distances, but one pass through the tutorial has even eliminated the need for that. Pay special attention to stadia techniques and error reduction math.
Thanks for the heads up on the Tgi3D software.
-
Thanks for the tut link. There was a lot of videos on YouTube (never searched for videos...)
Looks like a lot of work even for measuring one point/angle?
Regarding the use of stadia techniques for measuring distances, doesn't that require that you use a meter stick, with measurements on it? Which may be a bit difficult when finding 3D positions of points on a rock?Regarding photogrammetry software it's true that they are quite pricey.
I used mainly ImageModeler (which is no longer freely available) because of its much better "free" modeling vs PhotoModeler where all 3D models had to be built from control points. In PhotoScan there's no need to use any of the control points at all if you don't want to, and you can still easily make a lot more detailed and precise models.I hope you will post your experiences and progress here!
Sounds like a fun project. -
If you have already figured this out forgive me. But here is what is going on in the video:
The stadia is a small target with a center, an upper mark and a lower mark.- First the theodolite is leveled so it can be spun in both directions and stay level.
- Then the stick person raises or lowers the stadia until its center aligns with the cross hairs of the theodolite.
- The instrument operator then takes an angle to the upper and lower stadia marks
-- These angles are quite small and should be equal if the stickman is holding the stick vertically but if not they can be averaged. - Now the operator has a 90 degree angle at the stick and another angle measured from level to the top of the stadia mark and the distance from top of stadia to center of stadia is a fixed distance. So with two angles and a side the operator uses trigonometry to calculate the distance from his instrument to the stick.
- I will uses this line as a base line distance taking a set of reading to points on the rock from each end of the base line.
-- With two sets of readings I can now triangulate the position of every recognizable feature on the rock
-- The stick man will slide the stadia marker up and down at each point to give me a vertical distance below the level plane of the instrument. - If I an measuring a rock, I can now move the stick over the surface of a rock.
- At each point of interest I move the stadia up and down until it matches the cross hairs of the theodolite
- I end up with a point cloud with enough data to describe the surface of the rock.
(Is there an SU plugin to read the data from a spread sheet and spit the data points into a SKP file?) - Now I can stretch photos onto all the major surfaces of the SU rock.
-
So then you measure every point (POI) on the rocks from to stations/locations, and then triangulate/calculate the position related to the stations, with no need for the stick man after first measuring those two points?
Do you put stickers on the POIs on the rocks then, or just remember (or sketch?) where they were? -
I think I would use the stick man because it would be hard to recognize your numbered points from the shallow angle of view of the instrument. I would keep the instrument level and rather than calculate height just read the height measurement from the stick. The angles would be read to the stick as it was moved from point-to-point on the rock. Horizontal angles would be measured from both set up points, but height could be measured once or twice and the height pairs average to increase accuracy of vertical measurement. Even a relatively cheap construction level would do for this project as the accuracy of the shadow marked on rock with a pounding stone is even less than that of a cheap instrument.
-
It would be great if you made a tutorial about how to use surveying techniques, theodolites (and apps) etc to reconstruct objects and landscapes in SketchUp
Have you considered doing that?BTW, would this app be of any help? http://photoephemeris.com/
It's great for calculating sun/moon positions. -
The tutorial may become part of the film animation I am working on.
Check out the moon shot at
http://www.facebook.com/media/set/?set=a.36524649699.43238.548344699&l=eacfeb8a70I just type in time of moonrise into google and don't even pay attention to the site I get the info from.
However, I will check out the photo ephemeris site. Thanks. -
Any news on your petroglyph/theodolite project ?
-
@bjornkn said:
Any news on your petroglyph/theodolite project ?
I let the film project go because the filmmaker did not see like a reliable person to work with, but I may continue with some experiments of my own. I just bought a "Bosch 130" laser interferometer and have done tests to confirm that they are accurate to 1/16" and measure out to the full 130 feet even in intense sunlight.
-
Sorry to hear that! I was looking forward to hear your experiences with that project.
Have you done any tests with your iPhone theodolite and the distance laser too? -
@bjornkn said:
Sorry to hear that! I was looking forward to hear your experiences with that project.
Have you done any tests with your iPhone theodolite and the distance laser too?A couple points of interest. For long distance work with lasers in bright sun I have found some interesting aids. Red laser enhancement glasses are useful. Also red and light blue postIt notes make good laser targets. The light blue is kind or counter intuitive but the laser light coming off it is intense enough to see, but if you are wearing the laser enhancement glasses it will pass the reflected red light while knocking down the color intensity of the blue background. The increased contrast will help you see the target spot.
Also you can fold the postIt to make a shade flap to increase contrast in bright sun.
-
(WARNING/DISCLAIMER - techniques described here could blind you. I do not recommend that anyone try this. The following is for information only)
Laser rangefinders are somewhere between difficult and impossible to use in bright sun. The first thing to do is get red laser enhancement glasses. The red lenses attenuate all but the red wavelengths in light. So your red laser dot will look brighter than all other light with non-red components. But laser dots are quite small and becomes quickly hard to find as distance increases. By attaching my Bosch distance finder to a monocular with a red filter (25A photographic filter) the problem is entirely solved.The only problem is having the laser beam returned to your eye by a shiny surface. Before engaging the laser I use my monocular to scan my target for highly reflective shiny surfaces. Once I am sure the problem does not exist I engage the laser and get a distance reading. Worst case like a mirrored wall, I would have to tape up a paper target. Or just close my eyes just before turning on the laser.
From a practical standpoint there is no problem, but accidents do happen so awareness and an abundance of caution are called for. Next I will need to SketchUp a more permanent rig to attach distance meter to monocular.
Advertisement