Georeferencing Total Station Data - ends up looking "squished."

[LesliePerkins]

I'm not sure if this is the right place to post this.

Let me start by explaining the data I'm working with:

My professor has a mapping project that she's working on in Belize. We walk a transect, taking various points on our Mobile Mapper, 4 Garmin GPS units and a Total Station. The Total Station data is arbitrary and needs to be georeferenced (so it shows up in the right location along the transect). I tried linking various points that are in the TMS to ones that are taken with the Mobile Mapper, effectively trying to "move" the TMS data to overlay with the other GPS data.

The problem:

When do the usual georeferencing procedure (linking control points etc) - the TMS data ends up moving, but it appears "squished" - somewhat like an image you might open in MS paint, shrink it by the corners and make it distored - etc. The points MOVE, but they're no accurate, like GIS is stretching or squishing them for some reason.

I don't use GIS a lot, and I have never taken a class. I'm only trying to help my professor based on tutorials I can find online, and what she has taught me - but we're both stumped!

I'll try and play with the data a bit and show an example picture...but my shp files got a little corrupted and I have to fix them. I guess what I'm really looking for is just a general idea as to what would make GIS squish the data. I was told we may need to get more points to use as control points. If that's the case, we can do that this summer when we go back down to Belize. But if that's not really the issue, it could save us time in the field if anyone had an idea how to fix the problem within GIS.

[TimothyHales]

Georeferencing does stretch and rotate the image when adding control points.  It sounds like your control points are not very well spaces across your image and are focused in one area.  It is best to space them out in each of the corners.

Fundamentals for georeferencing a raster dataset: "If possible, you should spread the links over the entire raster dataset rather than concentrating them in one area. Typically, having at least one link near each corner of the raster dataset and a few throughout the interior produces the best results."

[KellyWhitehead]

Hello,

I work on a project that involves the use of a total station to make high resolution topographic maps of streams for salmon habitat studies. We encountered this same problem with georeferencing the survey data. The problem is due to the higher accuracy of the total station compared to the accuracy of the gps coordinates collected for the control points. When you try to "lock on" to all of the control points, you end up distorting the survey data.

To solve this problem, we developed an Add-In that our field crews use to project total station survey data in to a real-world coordinate system:

http://ctt.joewheaton.org/home

I hope this helps!

[LesliePerkins]

Georeferencing does stretch and rotate the image when adding control points.  It sounds like your control points are not very well spaces across your image and are focused in one area.  It is best to space them out in each of the corners.

Fundamentals for georeferencing a raster dataset: "If possible, you should spread the links over the entire raster dataset rather than concentrating them in one area. Typically, having at least one link near each corner of the raster dataset and a few throughout the interior produces the best results."

That's what we were thinking - that we'd need to take some more points out in the field to space out the control points better. I'll take a look at the data and make a list of good spots to take points.

kmwhitehead
Hello,

I work on a project that involves the use of a total station to make high resolution topographic maps of streams for salmon habitat studies. We encountered this same problem with georeferencing the survey data. The problem is due to the higher accuracy of the total station compared to the accuracy of the gps coordinates collected for the control points. When you try to "lock on" to all of the control points, you end up distorting the survey data.

To solve this problem, we developed an Add-In that our field crews use to project total station survey data in to a real-world coordinate system:

http://ctt.joewheaton.org/home

I hope this helps!

Hey thanks! I'll try that tonight or tomorrow when I have some spare time! :]

[LesliePerkins]

Hello,

I work on a project that involves the use of a total station to make high resolution topographic maps of streams for salmon habitat studies. We encountered this same problem with georeferencing the survey data. The problem is due to the higher accuracy of the total station compared to the accuracy of the gps coordinates collected for the control points. When you try to "lock on" to all of the control points, you end up distorting the survey data.

To solve this problem, we developed an Add-In that our field crews use to project total station survey data in to a real-world coordinate system:

http://ctt.joewheaton.org/home

I hope this helps!

I like the idea of the add on, but I'm having trouble using it - even when following the instructional video. I get all the information set up properly, then when I finish choosing the coordinate system and loading the 3 benchmarks - when I go to select the output workspace as the geodatabase I've created, the "select inputs" button stays grayed out, and I'm not sure why. I thought I followed all the steps properly, but maybe I'm not?

I think it may be a coordinate issue. We work in Belize and take coordinates in NAD 27 UTM 16N. I'd select that as the coordinate system, and leave the transformation drop down to the default (NAD27 to WGS 84 1) - but the TMS needs to be projected into NAD27.

I'm so new to tall this stuff so I have no idea what I might be doing wrong.

[KellyWhitehead]

The version available is a new release, so there might still be a bug or two. I will work with our team to see what the issue could be.

[ChrisGarrard]

Can you tell me exactly what you entered and in what order in order to get this problem. I tried replicating it and had no luck. If I can replicate it, I can probably fix it.