I assume it should be possible to use Photoshop to do image analysis of rocks, e.g., modal %'s and color index from photographs of rock slabs, scans of thin-sections, etc.).  When I use the eye-dropper tool to select light-colored areas, it picks up lots of reflections from cleavages. I tried messing around with the contrast, but this didn't really work.  Also, how could all the variations in color through the sample be simplified to a few colors in preparation for modal analysis?  Anyone have any experience with this?  I'm sure there is a published paper or two out there, but just some handy tips would be great.

Thanks.

Kent Ratajeski

Dr. Kent Ratajeski

Lecturer and Dice Mineralogical Museum Director
North Hall 081

Department of Geology, Geography, and Environment

Calvin University

3201 Burton St. SE
Grand Rapids, MI  49546
(616) 526-6769
https://calvin.edu/directory/people/kent-ratajeski

Hi Kent,

There's software called imagej that's designed for scientific image analysis. It will probably do a much better job at what you're interested in than Photoshop, with the benefit of being lightweight and open source.
Download here: https://fiji.sc/
There are many tutorials online on how to operate it.

Cheers
Michael

Michael Anenburg
Experimental Petrology
Research School of Earth Sciences
ANU College of Science
142 Mills Road
The Australian National University
Canberra ACT 2600
michael.anenburg@anu.edu.aumailto:michael.anenburg@anu.edu.au

From: Kent Ratajeski via MSA-talk msa-talk@minlists.org
Sent: Wednesday, 5 February 2025 12:28 AM
To: msa-talk@minlists.org
Subject: [MSA-talk] using Photoshop to find modal %'s and color index

I assume it should be possible to use Photoshop to do image analysis of rocks, e.g., modal %'s and color index from photographs of rock slabs, scans of thin-sections, etc.).  When I use the eye-dropper tool to select light-colored areas, it picks up lots of reflections from cleavages. I tried messing around with the contrast, but this didn't really work.  Also, how could all the variations in color through the sample be simplified to a few colors in preparation for modal analysis?  Anyone have any experience with this?  I'm sure there is a published paper or two out there, but just some handy tips would be great.

Thanks.

Kent Ratajeski

Dr. Kent Ratajeski
Lecturer and Dice Mineralogical Museum Director
North Hall 081

Department of Geology, Geography, and Environment

Calvin University
3201 Burton St. SE
Grand Rapids, MI  49546
(616) 526-6769
https://calvin.edu/directory/people/kent-ratajeski

Hi Kent,

I’ve never  been able to get photoshop to do what you are asking — there’s just too much color variations etc to have it work automatically.

Rather, I’ve used ImageJ (from the NIH web site — ). It’s java based and works on all platforms. You need to outline the minerals by hand (magic want and shift to do multiple areas) — if you can tell the minerals apart in a photomicrograph then you are good to go -- then a single measure command will give you the area. It’s a bit tedious, but if you don’t have too many samples, it’s not too onerous.

Cheers,

Frank

Frank Spear
Professor and Edward P. Hamilton Distinguished Educator Chair
Department of Earth and Environmental Sciences
Rensselaer Polytechnic Institute
110 8th St., Troy NY 12180
518-276-6103
spearf@rpi.edumailto:spearf@rpi.edu

On Feb 4, 2025, at 8:27 AM, Kent Ratajeski via MSA-talk <msa-talk@minlists.orgmailto:msa-talk@minlists.org> wrote:

CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender and know the content is safe.
I assume it should be possible to use Photoshop to do image analysis of rocks, e.g., modal %'s and color index from photographs of rock slabs, scans of thin-sections, etc.).  When I use the eye-dropper tool to select light-colored areas, it picks up lots of reflections from cleavages. I tried messing around with the contrast, but this didn't really work.  Also, how could all the variations in color through the sample be simplified to a few colors in preparation for modal analysis?  Anyone have any experience with this?  I'm sure there is a published paper or two out there, but just some handy tips would be great.

Thanks.

Kent Ratajeski

Dr. Kent Ratajeski
Lecturer and Dice Mineralogical Museum Director
North Hall 081
Department of Geology, Geography, and Environment
Calvin University
3201 Burton St. SE
Grand Rapids, MI  49546
(616) 526-6769
https://calvin.edu/directory/people/kent-ratajeski

MSA-talk mailing list -- msa-talk@minlists.orgmailto:msa-talk@minlists.org
To unsubscribe send an email to msa-talk-leave@minlists.orgmailto:msa-talk-leave@minlists.org

Kent,

For optimal results with ImageJ (and its successors) use chemical maps generated on the microprobe. You do low-resolution stage scans of a thin section/probe mount in 45-60 minutes, keeping spectometers fixed on single element. Should be able to do same thing with SEM and EDS system, only you scan beam and merge the tiles into single image.

The powerful part is you can use image math in ImageJ to add element images together to highlight specific phases. We used Ca+Si to differentiate olivine, Opx, Cpx, and spinel in serpentinized peridotites. The pixel counts give you modes, and you should be able to do grain size measurements also.

If course, you need an SEM or EMP to do this.

Best, John

John W. Shervais                      Professor Emeritus of Geosciences
Department of Geoscience          Office Phone: (435) 797-1274
Utah State University                Main Office:  (435) 797-1273
4505 Old Main Hill                      FAX:            (435) 797-1588
Logan, Utah 84322-4505          email: john.shervais@usu.edu

On Feb 5, 2025, at 08:11, Spear, Frank via MSA-talk msa-talk@minlists.org wrote:

Hi Kent,

I’ve never  been able to get photoshop to do what you are asking — there’s just too much color variations etc to have it work automatically.

Rather, I’ve used ImageJ (from the NIH web site — ). It’s java based and works on all platforms. You need to outline the minerals by hand (magic want and shift to do multiple areas) — if you can tell the minerals apart in a photomicrograph then you are good to go -- then a single measure command will give you the area. It’s a bit tedious, but if you don’t have too many samples, it’s not too onerous.

Cheers,

Frank

Frank Spear
Professor and Edward P. Hamilton Distinguished Educator Chair
Department of Earth and Environmental Sciences
Rensselaer Polytechnic Institute
110 8th St., Troy NY 12180
518-276-6103
spearf@rpi.edumailto:spearf@rpi.edu

On Feb 4, 2025, at 8:27 AM, Kent Ratajeski via MSA-talk <msa-talk@minlists.orgmailto:msa-talk@minlists.org> wrote:

CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender and know the content is safe.
I assume it should be possible to use Photoshop to do image analysis of rocks, e.g., modal %'s and color index from photographs of rock slabs, scans of thin-sections, etc.).  When I use the eye-dropper tool to select light-colored areas, it picks up lots of reflections from cleavages. I tried messing around with the contrast, but this didn't really work.  Also, how could all the variations in color through the sample be simplified to a few colors in preparation for modal analysis?  Anyone have any experience with this?  I'm sure there is a published paper or two out there, but just some handy tips would be great.

Thanks.

Kent Ratajeski

Dr. Kent Ratajeski
Lecturer and Dice Mineralogical Museum Director
North Hall 081
Department of Geology, Geography, and Environment
Calvin University
3201 Burton St. SE
Grand Rapids, MI  49546
(616) 526-6769
https://calvin.edu/directory/people/kent-ratajeski

MSA-talk mailing list -- msa-talk@minlists.orgmailto:msa-talk@minlists.org
To unsubscribe send an email to msa-talk-leave@minlists.orgmailto:msa-talk-leave@minlists.org

MSA-talk mailing list -- msa-talk@minlists.org
To unsubscribe send an email to msa-talk-leave@minlists.org