Contents || Index
12/12/2003

 

Tools for measuring thickness of thin film layers from simulated images.

Notes

-----

To read a grf file into a cube:  use 

File / File utilities / Cerius .grf / Show file as a cube

Buttons

-------

Update-vv-plot-files!   Works on the mlx default read directory. 

                        For all .grf files in that directory that DONT have

                        text plot files for the vertical variance of the

                        first image and for the whole cube sum images:

                              1) Reads the grf file into a cube.

                              2) Saves the text file for a plot for the

                              vertical variances for the 1st and sum images.

                        The file names are <image-file-name>-vv-1st.txt, and

                        <image-file-name>-vv-sum.txt.

      These plot files can be read in and examined using

      ***Buttons*** / Profiles / Other Plots / Read as collumns.

Plot series through cube!     Generates a separate plot for each image in the

                              cube.  Plots are vertical variances.

      To tile these plots, make one of them sufficiently small, resize the

others using **Profiles** / Arrange / Size plots, then

**Profiles** / Arrange / Tile plots.  (You don't need to select an image

in the **Profiles** button window, or turn on the profile line in order to use

the 'Arrange' and 'Other Plots' functions.)

Width 01 Chart from plot files !    Prints the original grf image file name and

                                    the width measurement

for each -1st or -sum text file in the default read directory.  Prints to the

text output window, from which measurements can be cut and pasted into Excel

or whatever.  Tabs separate the collumns.

Show width on plot!     Displays a pink band on a plot - indicating the width

                        found by this algorithm.

      Will not work for plots read in as 'other plots / read as collumns' 9/5/2002

Note about the width:

Here is a sample of width mmts from the sum images:

gs00056        72

gs00057        72

gs118-143     -10

gs32-57        72

gs61-79        16

The -10 clearly is a goof, and this can be caught early on in the width

measurement.  But the algorithm needs a better starting point to search.

At the moment, it starts in the middle of the plot, and I think the right

place for the gs118-143 sum plot should be slightly off center.  Anyway,

the other widths should be OK, as you can see from the graphics.

The width could be measured from the sides of the peaks - maybe with more

precision.