False colour composites were something I was introduced to in high school GIS and did a lot of in undergrad remote sensing. It was one of those Eureka! moments for me in terms of developing a better understanding of signals and signal processing, as well as sensors and data collection. Ie. absorbing the reality that the way in which my eyes (or any optical sensor) see the world is not canonical. Vegetation is red, not green!
It was also just the COOLEST THING EVER to swap infrared bands in LANDSAT ETM+ imagery and watch vegetation immediately pop out at you. It's what turned this sort of dull (at least for a teenager) geoscience into a kind of puzzle where you have to figure out how to bring the signal up out of the data.
It's really fun to play with black and white photography using various kinds of color filters and see how different you can get the same scene to appear.
I interpreted that step as "once you click 'Auto Input Levels', drag the center triangle of the 3 at the bottom of the histogram to where you think the median of the histogram is". That middle slider defaults to being the midpoint of the other 2 sliders. I'm not sure if there is a smarter way to move it to the median beyond "looks about right".
I got this as a result https://i.imgur.com/RsXx3y2.jpeg and it certainly doesn't preserve the color of the rest of the image (there are probably more algorithmic steps needed for that) but it does seem to make highlight the same faint patterns much more clearly.
The levels tool maps a subset of the channel range to the full range, effectively changing the brightness and contrast in one step.
The center point controls the gamma correction, so you can map a biased distribution (mostly bright or dark for example) to be more evenly distributed.
That was the purpose of that step - move the midpoint slider to about where the peak in the histogram is, to get an evenly distributed range.
The low/high slider can be moved towards that to increase the contrast.
"Contrast" here will only apply to the A/B channels, which is the color information, so it's increasing the color contrast and unbiasing it, while keeping the luminance information.
Photoshop also can do the same transform, and interactively in the LAB colorspace (no need to combine/uncombine the channels first).
I tried to look for hidden rock art at Ankgor Wat by imaging with multiple bandpass filters and amplifying cross-spectral differences. I didn't succeed, but maybe I was in the wrong place. The shit guards there also didn't like me using a tripod while I changed filters, so it was also hasty.
If it makes you feel any better, I've also never had any luck finding hidden rock art, petroglyphs, or similar using multispectal bandpass filters or decorellation stretching.
I was super excited about DCS based on some of the recovered petroglyph photos online, and integrated it into some software I wrote years ago.[1] Still hoping I find something as interesting with it as the fellow mentioned in the article. It works, but I haven't had a "gosh, I never would have seen that without DCS!" moment.
“No one really knows why people made the rock art they made, at least in many cases,” Harman said
Well, it's obvious they really wanted to though. This isn't like marking on a wall with a Sharpie. They had to work at it to make it happen. The commitment alone is impressive even if it was just a "Grogg was here!"
In reality very interesting work and results, but a little disturbing that it took three or four decades and a self-funded citizen scientist to transfer the knowledge to the application.
"In this paper, we review the theory behind decorrelation stretch and propose some alternative algorithms that resolve the issues of the standard approach."
"A contribution is mandatory before I will send the plugin. It may take me a few days to respond, have patience.
Please note that the plugin is not an app. It has much more functionality than is possible in an app on a smartphone."
"The DStretch [mobile] app is a lightweight version of DStretch intended to give users the ability to enhance rock art in the field. It is not a substitute for the DStretch plugin which has much more functionality."
This is based off a NASA article, which has an expanded version with more details here:
https://spinoff.nasa.gov/Manipulating_Satellite_Photos_Now_R...
The 1996 paper mentioned in the NASA article: https://dstretch.com/DecorrelationStretch.pdf
Way better, thank you
False colour composites were something I was introduced to in high school GIS and did a lot of in undergrad remote sensing. It was one of those Eureka! moments for me in terms of developing a better understanding of signals and signal processing, as well as sensors and data collection. Ie. absorbing the reality that the way in which my eyes (or any optical sensor) see the world is not canonical. Vegetation is red, not green!
It was also just the COOLEST THING EVER to swap infrared bands in LANDSAT ETM+ imagery and watch vegetation immediately pop out at you. It's what turned this sort of dull (at least for a teenager) geoscience into a kind of puzzle where you have to figure out how to bring the signal up out of the data.
It's really fun to play with black and white photography using various kinds of color filters and see how different you can get the same scene to appear.
(Also fun to play with a polarizing filter.)
You can do something similar in GIMP:
1) Colors > Components > Decompose > LAB
2) For the layers of the A/B (chroma) channels, Colors > Levels > "Auto Input Levels" (maximize the contrast), and move the mid point to the median.
3) Colors > Components > Compose > LAB
Can you explain 2 a bit more? I see the Auto input levels but unsure what you mean by "move the midpoint to the median."
The result I got for the original mars image was a washed out red image with some yellow around the craters.
I interpreted that step as "once you click 'Auto Input Levels', drag the center triangle of the 3 at the bottom of the histogram to where you think the median of the histogram is". That middle slider defaults to being the midpoint of the other 2 sliders. I'm not sure if there is a smarter way to move it to the median beyond "looks about right".
I got this as a result https://i.imgur.com/RsXx3y2.jpeg and it certainly doesn't preserve the color of the rest of the image (there are probably more algorithmic steps needed for that) but it does seem to make highlight the same faint patterns much more clearly.
Thanks. I was able to get the same result as you did. Pretty neat trick.
The levels tool maps a subset of the channel range to the full range, effectively changing the brightness and contrast in one step.
The center point controls the gamma correction, so you can map a biased distribution (mostly bright or dark for example) to be more evenly distributed.
That was the purpose of that step - move the midpoint slider to about where the peak in the histogram is, to get an evenly distributed range.
The low/high slider can be moved towards that to increase the contrast.
"Contrast" here will only apply to the A/B channels, which is the color information, so it's increasing the color contrast and unbiasing it, while keeping the luminance information.
Photoshop also can do the same transform, and interactively in the LAB colorspace (no need to combine/uncombine the channels first).
Thanks for the details but I realized I should have said "Pretend I just installed GIMP and my knowledge of image editing is MS paint."
This is what I was after: "move the midpoint slider to about where the peak in the histogram is, to get an evenly distributed range."
It kinda worked! Thank you and sorry for making you explain the gory details though they are appreciated and useful.
I tried to look for hidden rock art at Ankgor Wat by imaging with multiple bandpass filters and amplifying cross-spectral differences. I didn't succeed, but maybe I was in the wrong place. The shit guards there also didn't like me using a tripod while I changed filters, so it was also hasty.
If it makes you feel any better, I've also never had any luck finding hidden rock art, petroglyphs, or similar using multispectal bandpass filters or decorellation stretching.
I was super excited about DCS based on some of the recovered petroglyph photos online, and integrated it into some software I wrote years ago.[1] Still hoping I find something as interesting with it as the fellow mentioned in the article. It works, but I haven't had a "gosh, I never would have seen that without DCS!" moment.
[1] https://www.beneaththewaves.net/Photography/Decorrelation_St... / https://www.beneaththewaves.net/Software/The_Mirrors_Surface...
Is there an imagemagick implementation that I can just throw into the pipeline?
“No one really knows why people made the rock art they made, at least in many cases,” Harman said
Well, it's obvious they really wanted to though. This isn't like marking on a wall with a Sharpie. They had to work at it to make it happen. The commitment alone is impressive even if it was just a "Grogg was here!"
https://smbc-comics.com/comic/rite-on
> "No matter what you do, when we find something weird in ancient history we will decide it must've been part of a fertility rite."
DStretch is also really great at revealing painted ghost signs on old buildings.
See also: Eulerian Video Magnification (EVM) to amplify subtle motions and colour changes:
https://youtu.be/ONZcjs1Pjmk https://youtu.be/_qtzWNZApsw
Always wanted to implement that on a phone, it would be like a superpower!
Amazing. This will change how I take photos
Going to be taking a lot more pictures of rock faces!
I want to try this on faded graffiti. I have a few spots in mind already.
No tricycle-riding penguins yet.
In reality very interesting work and results, but a little disturbing that it took three or four decades and a self-funded citizen scientist to transfer the knowledge to the application.
any repo or link where other people can try?
Matlab - https://www.mathworks.com/help/images/enhance-color-separati...
This paper might also be of interest - https://www.mdpi.com/2227-7390/13/20/3297
"In this paper, we review the theory behind decorrelation stretch and propose some alternative algorithms that resolve the issues of the standard approach."
The app is called DStretch, it's available for iOS and Android
https://dstretch.com/Apps/index.html
Someone needs to vibecode some updated options
Had an llm build something quickly so I could play with it. https://shannietron.github.io/decorrelation_stretch/
"A contribution is mandatory before I will send the plugin. It may take me a few days to respond, have patience. Please note that the plugin is not an app. It has much more functionality than is possible in an app on a smartphone."
"The DStretch [mobile] app is a lightweight version of DStretch intended to give users the ability to enhance rock art in the field. It is not a substitute for the DStretch plugin which has much more functionality."
"Why do we need to spend resources trying to go to space, when there are problems right here on Earth that need solving?"
Weather satellites have done more to increase crop yield than anything since the invention of artificial fertilizer and save more lives.