Post

2 followers Follow
0
Avatar

Manual Georeferencing Surface Slanted

Hello,

 

I am attempting to use the new Manual Georeferencing option, but when I input the elevations they area not representative in the final output data. In fact for some reason there is a significant tilt to the entire surface. 

Please let me know if I am missing somthing or why this might be happening.

 

Thanks 

 

Brandon

Brandon Rivet

Official comment

Avatar

All of your 5 surveyed points have elevations within .4 meters which is probably not true... So that is one source of error. 

What kind of GPS are you using? What is it's vertical accuracy?

The initial model had some curvature to it since you included the images at the ends of your passes. Map Pilot does not do this and gets less curvature in the initial model. The less there is initially to correct for, the better.

The biggest issue here though is the fact that you don't have very many GCPs and the ones you do have aren't spread out around the whole area very well. 

The referencing process takes the GCPs and their errors and tries to correct for it with a polynomial offset (not a linear one). This is an attempt to take out curvature in the DEM. 

For reference, here is your original:

 

Here is the referenced version:

Zane
Comment actions Permalink

Please sign in to leave a comment.

6 comments

0
Avatar

Also, your real coordinates which are the surveyed values don't have much precision. 4 decimal places of precision is 10s of meters.

Zane 0 votes
Comment actions Permalink
0
Avatar

Hi used a TOPCON GPS with a high accuracy, the area is relativity flat which is why having the data, (even the originally processed data without the GCP) being so on a slanted is confusing. 

The ground control points are around the area I am interested in. What are the exact limitations of the GCP. Is there a spacing/frequency that you could recommend.

I have generated more accurate GPS lat ans longs (Base data is in UTM) and have tried reprocessing the data again but it still comes out slanted.

Please let me know if there is anything else i can do to get the ground to be more in line with my control.

 

Thanks,

 

Brandon

Brandon Rivet 0 votes
Comment actions Permalink
0
Avatar

Your most recent one that was made with 7 digits of precision looks pretty good. In what way is it slanted? 

Better GCP distribution is probably required to get the corners to behave better. 

Zane 0 votes
Comment actions Permalink
0
Avatar

When I import the data into CAD and place a point over my control points there is a difference between 1m and 3m from my ground control. This is what is creating the slant I have mentioned. 

Brandon Rivet 0 votes
Comment actions Permalink
0
Avatar

I am assuming you are talking about the elevation being off. We get things lined up pretty well in the photo layer. The elevation data is fit to the differences in the control points so they won't match up directly. We are still working on the elevation processing portion of this based on user feedback. We have to optimize for good data sets that are well spread throughout the survey area. 

Zane 0 votes
Comment actions Permalink