SEM: how much larger are bright particles than dark? …

[michael shaffer]

Having installed a FEG image analyzer and comparing its results with our tungsten SEM, we are seeing a slight bias with respect to brighter particles measuring smaller (possibly more accurately) with the FEG.  The image is backscattered electrons and we use a fixed background threshold for isolating particles from the epoxy.  It occurred to me that clipping the edge at the same background value would make brighter particles larger depending on how "fuzzy" the edge is.  It would even be possible to predict this "bright versus dark" bias if the SEM's spot size was determined.  Some difference should be associated with each SEM's magnification calibration, but I'm accounting for that by first relating calibration to the darkest minerals (quartz which is abundant).

I thought before I re-invent the wheel and get involved with the math and all the other possible variables (e.g., pixel resolution, irregular perimeters, ...) that I'd ask my peers if this has been done before(?) ... leastwise, I haven't yet found a reference ...

TIA & Cheerios from the Avalon
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Michael Shaffer
SEM-MLA Research Coordinator
Bruneau Centre for Research & Innovation
Memorial University
St. John's, Newfoundland, Canada
Information: http://www.mun.ca/creait/maf/
Scheduling: http://www.mun.ca/creait/maf/SEM-MLA/calendar.php
        [phone] (Ofc)
        [phone] (Lab)
cogito ergo ZzoooomM