Monday, May 17, 2010

Using Excel to calculate the RMSE for LiDAR vertical ground control points

The height accuracy of the collected LiDAR data can be verified by comparing with independently surveyed ground control points on hard, flat, open surfaces. It is essentially just calculating the height differences for all the control points and then determining the height root mean squared error (RMSE) or differences. Most LiDAR processing software have the reporting function built-in. However, plain Microsoft Excel can also do the job (except for extracting the elevation from the LiDAR data.

Assuming that you are able to calculate the height differences for all the control points and place in a spreadsheet as shown in the figure below. I have a column of delta Z values in column A.


Then to calculate the RMS value for the elevation differences, I can do the following.

  1. In a cell, type in the formula:

    = SQRT(SUMSQ(A2:A18)/COUNTA(A2:A18))

    where A2:A18 are the values from cell A2 to A18 in the spreadsheet. Simply replace these with the actual locations on your spreadsheet.

  2. Press RETURN.

    The RMSE value is calculated.



No comments: