IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
Image Enhancement for Mineral Identification
CosmicRocker
post Oct 18 2011, 06:55 AM
Post #1


Senior Member
****

Group: Members
Posts: 2207
Joined: 1-December 04
From: Marble Falls, Texas, USA
Member No.: 116



I'm not sure that those filter combinations will be more helpful than the one I posted. We need something that specifically targets the infrared fingerprint of hematite, like this image of Snout in Eagle Crater. I'm waiting for a good R5 so I can make one of those.


--------------------
...Tom
Go to the top of the page
 
+Quote Post
Bill Harris
post Oct 18 2011, 02:15 PM
Post #2


Senior Member
****

Group: Members
Posts: 2343
Joined: 30-October 04
Member No.: 105



I think you've discussed this processing before, but fiddle if I can't find it. From the filename, I presume (assume?) that the processing steps are (R1/R2)(R1)(R5/R7) for the ( R )(G)( B ) values, and that R1/R2 is the value of R1 divided by the value of R2. My image processor doesn't divide and I can't recall what Boolean division is.

Image from Sol 014, BTW.

That spot is fairly bland (compared to recent locales) when looking at L257 and R721 images.

--Bill


--------------------
Go to the top of the page
 
+Quote Post
PDP8E
post Oct 19 2011, 04:38 AM
Post #3


Member
***

Group: Members
Posts: 515
Joined: 10-October 06
From: Maynard Mass USA
Member No.: 1241



here is my shot at the recent (2748) scuff marks
<noise reduction, minor sharping, intensified colors a little>
Attached Image




--------------------
CLA CLL
Go to the top of the page
 
+Quote Post
CosmicRocker
post Oct 19 2011, 06:00 AM
Post #4


Senior Member
****

Group: Members
Posts: 2207
Joined: 1-December 04
From: Marble Falls, Texas, USA
Member No.: 116



The enhanced color images you guys have made definitely help us better see the spherical clasts in this soil. I wonder if we'll get any MIs.
.........................................

Bill: You are right about that file name. It gives the recipe for the R, G, and B channels of the composite image. At first I didn't know how to divide two images, but a friend who is an image processing expert showed me how to do it. I use Photoshop, but I am sure other software has similar abilities. If there is interest in the topic, we can start a thread in "Image Processing Techniques." I'll try to summarize the process here.

For example, to divide R1 by R2, I would create an image with two layers...R1 on top and R2 below. I leave R2's blending mode as a "normal," but set the R1 blending mode to "screen." Next, I insert an adjustment layer between R1 and R2 and set it to "invert." Finally, I merge all the layers together into one flat image that will be R1/R2. The terminology I am using is from my old version of Photoshop. Your software may call these modes and layers by other names.


--------------------
...Tom
Go to the top of the page
 
+Quote Post
Bill Harris
post Oct 19 2011, 10:34 PM
Post #5


Senior Member
****

Group: Members
Posts: 2343
Joined: 30-October 04
Member No.: 105



I'll see how those steps translate to Paintshop Pro.

My observation is that the L257 and R721 may be useful in separatring the various members of te iron (III) oxide-hydroxide family. Hematite tends towards blue while limonite tends towards yellow. Of course, without a spectrophotometer with the MER filters and a set of reference samples, this is just a casual observation.

--Bill


--------------------
Go to the top of the page
 
+Quote Post
JohnVV
post Oct 21 2011, 02:42 AM
Post #6


Member
***

Group: Members
Posts: 444
Joined: 18-November 08
Member No.: 4489



I would not use paint shop

Nip2 would be perfect
http://www.vips.ecs.soton.ac.uk/index.php?title=VIPS
open the images in column A
do the math functions in column B

then turn B into a menu item ( point & click)

after the first you can repeat by opining the images in the correct order then choose this from the menu
simple

Nip / vips has all the math functions built in

in vip's you can do it all from the terminal without the gui overhead
simple bash script would do .
Go to the top of the page
 
+Quote Post
djellison
post Oct 21 2011, 02:57 AM
Post #7


Administrator
****

Group: Chairman
Posts: 13600
Joined: 8-February 04
Member No.: 1



QUOTE (JohnVV @ Oct 20 2011, 07:42 PM) *
I would not use paint shop
....
in vip's you can do it all from the terminal without the gui overhead
simple bash script would do .


And for those without a background in command line syntax and so on, what you describe lies a great big steep chasm of learning away from being possible.

For that vast majority of people, Paintshop Pro, Gimp or Photoshop is far, far more appropriate.

For those with the skills, what you describe may be ultimately faster - but for most, it's folly to suggest.

Go to the top of the page
 
+Quote Post
JohnVV
post Oct 21 2011, 03:00 AM
Post #8


Member
***

Group: Members
Posts: 444
Joined: 18-November 08
Member No.: 4489



nip2 IS the gui front end to the vip imaging library
Go to the top of the page
 
+Quote Post
CosmicRocker
post Oct 23 2011, 03:52 AM
Post #9


Senior Member
****

Group: Members
Posts: 2207
Joined: 1-December 04
From: Marble Falls, Texas, USA
Member No.: 116



But, this is a technical thread, is it not? I'm eager to hear about more efficient alternatives to the labor intensive Photoshop method, especially if we're talking freeware. I understand that command line applications are not for everyone, but I've used such things in the past, even as mainframe batch jobs. It's not my preferred method in this day and age, but I wouldn't be afraid to try it if it saved me time.

I was intrigued by JohnVV's recommendation of an application that has a spreadsheet interface for image processing. I love spreadsheets, and I downloaded vips and nip2 tonight. I haven't been able to try them yet. smile.gif

Getting back to the image at hand, the R5 without dropouts came down today. Using the Photoshop method, here is what I came up with.
Attached Image

I wasn't expecting most of the larger rock fragments to have a hematite signature. The spherical ones and most of the others appear yellow in this IR false color image. I've studied hundreds of Opportunity images created according to this specific algorithm as we've roved across Meridiani Planum. In spite of the fact that they were made from uncalibrated, raw jpegs, the hematite concretions always appear in shades of yellow or occasionally orange. Ignore the purple, as that is how shadows appear in these images.

So, I have to wonder what this might mean regarding the geology and geomorphology of Cape York. Do abundant hematite fragments and spherules mean that the cape was once covered by the younger sulfate sediments with their embedded hematite concetions? Or, might these pieces of hematite have been tossed onto Cape York by meteorite impact processes?


--------------------
...Tom
Go to the top of the page
 
+Quote Post
JohnVV
post Oct 23 2011, 04:18 AM
Post #10


Member
***

Group: Members
Posts: 444
Joined: 18-November 08
Member No.: 4489



nip dose take some getting used to
I have been using it for many years , so at first it will seam a bit " odd"

you might want to have a look at the blog
http://libvips.blogspot.com/

an example of joining images is the Aug 9 joining of the bluemarble earth map

also nip was originally designed to work on 23 layer spectral images of paintings
the hires images were tiles that needing joining

so this DOSE make registered panoramas using "tie points"
an tutorial example
http://www.vips.ecs.soton.ac.uk/supported/...ipguidese2.html
Go to the top of the page
 
+Quote Post

Reply to this topicStart new topic

 



RSS Lo-Fi Version Time is now: 20th April 2014 - 03:12 AM
RULES AND GUIDELINES
Please read the Forum Rules and Guidelines before posting.

IMAGE COPYRIGHT
Images posted on UnmannedSpaceflight.com may be copyrighted. Do not reproduce without permission. Read here for further information on space images and copyright.

OPINIONS AND MODERATION
Opinions expressed on UnmannedSpaceflight.com are those of the individual posters and do not necessarily reflect the opinions of UnmannedSpaceflight.com or The Planetary Society. The all-volunteer UnmannedSpaceflight.com moderation team is wholly independent of The Planetary Society. The Planetary Society has no influence over decisions made by the UnmannedSpaceflight.com moderators.
SUPPORT THE FORUM
Unmannedspaceflight.com is a project of the Planetary Society and is funded by donations from visitors and members. Help keep this forum up and running by contributing here.