Where the RMSE of Georeferencing is displayed in ArcGIS Pro?

I couldn’t figure out where the RMSE of Georeferencing is displayed in ArcGIS Pro.

Where the RMSE of Georeferencing is displayed in ArcGIS Pro?

I couldn’t figure out where the RMSE of Georeferencing is displayed in ArcGIS Pro.

- 1 person found this helpful
does it appear after you are done the process? or are these just screen shots as you are going along? It isn't clear

The interpreting rms error section here suggests you might be more concerned about the individual points. Perhaps the total has been left out if favor of more detailed information during the collection.

The total error is computed by taking the root mean square sum of all the residuals to compute the RMS error. This value describes how consistent the transformation is between the different control points (links). When the error is particularly large, you can remove and add control points to adjust the error. Although the RMS error is a good assessment of the transformation's accuracy, don't confuse a low RMS error with an accurate registration. For example, the transformation may still contain significant errors due to a poorly entered control point. The more control points of equal quality used, the more accurately the polynomial can convert the input data to output coordinates. Typically, the adjust and spline transformations give an RMS of nearly zero or zero; however, this does not mean that the image will be perfectly georeferenced.

The forward residual shows you the error in the same units as the data frame spatial reference. The inverse residual shows you the error in the pixel units.

. All residuals closer to zero are considered more accurate.*The forward-inverse residual is a measure of how close your accuracy is, measured in pixels*Of course if your intent is to reach zero.. then you may fall into the trap of assuming that the minimum number of points that produces a total of zero is the best... this simply isn't the case

- 2 people found this helpful
Jamal - there is a usability bug for this. It's in the Product Plan to be improved but do not know what future release.

BUG-000112904 : The box that displays information regarding transformation type, number of control points, and total RMS errors while georeferencing a raster dataset is in an inconvenient location.

Thank you for pointing the bug out Robert LeClair. For anybody on this thread affected by that bug, please visit it on support.esri.com https://support.esri.com/en/bugs/nimbus/QlVHLTAwMDExMjkwNA== and click the Subscribe button.

Not only will that provide you with notifications of the bug's status, but it also provides us a more holistic picture of number of users affected. Thank you!

- 2 people found this helpful
Jamal - still in the Product Plan but don't know when. Sorry to be vague.

- 3 people found this helpful
As Jamal has already alluded to, the RMS errors are reported in the Heads up display.

There are 3 types of RMS that we report in both types of software. The Forward transformation is always the default RMS. But if you show all three types of errors in both software programs, you can see they are both available in both sortware.

As you can see the top Image from Pro is the same as the bottom image from ArcMap (slight difference in rounding).

Please let me know if there is still any confusion. There is no real bug; simply an enhancement that some users were requesting it in BOTH the heads-up display and the table.

I couldn’t find any information in the help that explains what these values mean in terms of geometry neither having figures (pictures) that better explain the idea. I would appreciate if you guide me for such integrated resources.

- 1 person found this helpful
I do not have any illustrations, but the last third paragraph of this section should explain it.

It is not a mathematical concept we have created, so perhaps there are illustrations on some other math related sources on the web.

- 1 person found this helpful
Thanks Simon.

Unfortunately, there is no explanation in how the RMSE is calculated. The math and diagrams (pictures) that shows the meaning of RMSE in the geometry is totally absent in the help documents.

In principles, the RMSE refers to the fact that two values are known; the expected and observed. How this can be translated when it comes to the georeferencing? When adding an Image to ArcPro, it is placed in a random place, with georeferencing, a number of points are used to move the image to the correct position. Then where is the expected and observed value in this case?

- 2 people found this helpful
It is the control points you add that will affect RMSE.

If you believe that the RMSE is incorrect for your georeferencing, please log an issue or send us your particular case.

Thanks.

- 1 person found this helpful
Thanks Simon.

Just I wanted to know how the RMSE is calculated mathematically.

In the first screenshot below, the coordinates of the marked point are as shown. With the georeferencing, this point is moved to its correct location. Then what is the meaning of RMSE in this case? Where is the Expected and Observed values in order to be able to calculate the RMSE for this point?

- 3 people found this helpful
One point is not enough for any RMSE. You may notice that RMSE is 0, until the 4th control point is created. It uses these 4 points to figure out what is the expected locations. And each control point after that is added to the calculation of RMSE.

As mentioned, if you believe that the calculation is not correct, please log an incident and we can take a look.

Thanks!

Simon W- 1 person found this helpful
I have no particular incident but wanted to know how the RMSE works.

Well, in the screenshot below, the RMSE with a value of 2.10 appears as the 4

^{th}point is added. Then based on what this value is calculated? Where are the expected and observed values in this case that paves the way for the calculation of RMSE?- 2 people found this helpful
When Auto-Apply is turned on (which is the default), the image is placed in the "expected" location.

The point you choose is the actual point (observed).

- 1 person found this helpful
Does this mean that the first 4 points are considered to be the observed based on which all the points on the image map are calculated?

Once the 5

^{th}point is added, then in this case we have two values for it: the one that is calculated (expected) from the first 4 points and the one that is added. The difference is the RMSE.

forward and inverse are in the control table apparently

http://pro.arcgis.com/en/pro-app/help/data/imagery/georeferencing-options.htm