Hyper-Imaging (Correlating WDS Mapping and EDS Spectrum Image Data)

[John Donovan]

We have released our new Hyper-Imaging feature. It is accessible from the CalcImage Processing | Classify Image window when a quantitative phase map has been loaded and/or re-calculated. Once the WDS quant classification map is loaded, simply check the Load EDS Spectrum Image checkbox and you will be prompted for an EDS spectrum image to load. Currently only the Thermo NSS SI spectrum image format is supported but I would be pleased to add additional spectrum image formats (including CL and EBSD spectrum images) to the feature.



Note in the above screen shot that we can also see markers that indicate the stage extents of the EDS image loaded. If  we now move the mouse over one of the phases we have identified, the EDS spectrum for the nearest neighbor pixel in the spectrum image file is automatically displayed as seen here:



Which is clearly iron sulfide. We can move the mouse cursor to the next phase, which is silica:



and again for copper-iron sulfide as seen here:



The point being that if we have a phase that has been identified by the kmeans phase classification, but is unknown, we can display the full EDS spectra for it by simply moving the mouse to those pixels as seen here:



Please note that to obtain the necessary system files for accessing the Thermo NSS spectrum images, you will need to run the ProbeForEPMA.msi installer. No need to uninstall anything, just run the installer on your existing installation.

A post announcing new information about the Thermo SI file support also has an attachment which provides a reasonable overview of the other "Hyper-Imaging" features.

http://probesoftware.com/smf/index.php?topic=40.msg370#msg370

[jared.wesley.singer]

...Redirected here from John F's recent comments,

http://probesoftware.com/smf/index.php?topic=397.0;topicseen

Since the spectrum at each pixel can be a little ragged with noise, has anyone invented a way to add the EDS spectra by phase classification?  A SUM spectrum feature could produce very high quality EDS spectra for standard-based quantification in NIST's DTSA II.   

If the spectrum is stored at every pixel, in theory we can recalculate the EDS ROI maps that John F is looking for, and handle the EDS background quantification accordingly in DTSA II. 

Jared

[Probeman]

...Redirected here from John F's recent comments,

http://probesoftware.com/smf/index.php?topic=397.0;topicseen

Since the spectrum at each pixel can be a little ragged with noise, has anyone invented a way to add the EDS spectra by phase classification?  A SUM spectrum feature could produce very high quality EDS spectra for standard-based quantification in NIST's DTSA II.   

If the spectrum is stored at every pixel, in theory we can recalculate the EDS ROI maps that John F is looking for, and handle the EDS background quantification accordingly in DTSA II. 

Jared

Hi Jared,
Please see this:

http://probesoftware.com/smf/index.php?topic=400.msg2170#msg2170