User guide
LBA-USB User Guide Document No: 11294 Page 49
Note: This edit control is repeated in the Beam Display dialog box and is
available in the Display Toolbar.
Reference Subtraction –
Activate this option to subtract the reference
frame from newly acquired or post processed frames of data. The reference
frame location is at
Processing hints:
• If the Set Reference Source item is set to Current Frame, the data
in the currently viewed frame will be copied to the Reference frame.
• If the Set Reference Source is set to Last Gauss, and the Gauss Fit
item in the Computations... dialog box is checked, then the beam
profile resulting from a computed Gaussian fit to the currently viewed
frame will be copied to the Reference frame.
• If the Set Reference Source is set to Auto Gauss, and the Gauss Fit
item in the Computations dialog box is checked, then the beam profile
resulting from a computed Gaussian fit to newly acquired frames will be
automatically copied to the Reference frame. In this mode, setting a
reference frame manually does not make any sense because the next
frame brought into view will automatically update the contents of the
reference frame.
Note: In the last two examples, if the Gauss Fit item is not checked, then no fit
is computed and thus no data will be copied to the Reference frame.
Convolution – The term Convolution indicates a type of two-dimensional spatial
filtering that can be applied to a digitized image. We do not attempt to use this
term in any restricted construct; rather we apply it generally to a variety of area
process transformations. At the time of this release, we are providing just a few
convolution algorithms; and these are primarily aimed at simple low-pass spatial
filtering. We expect this list to grow in time, and welcome suggestions from our
users for additional implementations.
Entries in the Convolution drop down list that begin LPF #, indicate a Low-
Pass filter. This is the most commonly used algorithm for simple image
smoothing operations. The suffix number defines the convolution kernel size in
pixels. For example, 3x3 indicates a 9 pixel kernel 3 pixels by 3 pixels. Other
common sizes include 5x5 and 7x7.
We recommend experimenting with these selections until a process is located
that meets your specific needs.
See Convolution in section 4.27 for more details regarding how these
algorithms are applied.