Juno Perijove 57, December 30, 2023 Options

[StargazeInWonder]

To zoom in on the problem, here are magnified patches of one of Jason's images, blue filter only, from the dayside and Jupitershine side, normalized to about the same overall brightness.

Obviously, the difference isn't due to Io itself, but the higher share of noise compared to signal in the darker imagery.

Median over a local window (3x3?) is one solution, which of course loses some spatial resolution.

We might beat this by combining information from the three color filters, presuming that the peaks and valleys of the speckles will vary from one filter to another and there is usually not sharp color contrast on the same spatial scale as edges in general albedo. So, maybe average the three filters to produce a grayscale image, apply median (2x2?) to that, and then color it with images that apply a larger median (eg, 4x4) to each of the color filters.

I've gone down a similar road with some astrophotography images, and there may be more wisdom out there in the community that processes images of deep sky objects (galaxies and nebulae). It's almost an entirely different paradigm than the imagery of planets (dayside), where you can always get all the luminance you want.

Attached thumbnail(s)

 

[scalbers]

Yes this may take some doing to set things up in an optimal way. Another example is in the IDL environnment with an FFT related filter. This is more of a random noise situation as I know FFT's can also work with periodic noise.

https://www.nv5geospatialsoftware.com/docs/...ovingNoise.html

[mcaplinger]

Is there really useful color information in the Jupitershine part of the image? A quick look at pj57-031 (the only image with TDI of 6) suggests to me that the green channel by itself is telling you most of what there is to know. Certainly the blue channel is pretty noisy.

We will almost certainly take more nightside images on PJ58, but my inclination is to go with longer TDI and only one channel (green, or perhaps red.)

[Antdoghalo]

Will there be an SRU image for PJ58 as well?

[volcanopele]

Is there really useful color information in the Jupitershine part of the image? A quick look at pj57-031 (the only image with TDI of 6) suggests to me that the green channel by itself is telling you most of what there is to know. Certainly the blue channel is pretty noisy.

Yes, if you do some image stacking. This helps the SNR a bit with blue. While blue can be noisy, green is good enough that it can be useful for distinuishing some albedo units that are muted in the red and near-IR, like the diffuse plume deposit around Tonatiuh. So maybe do green and red?

[StargazeInWonder]

Taking one of Jason's images and applying a 3x3 median to the Jupiter shine side, I noted first that the blue was much brighter than it was on the dayside, no doubt due to greater noise. So, I turned the blue down about 40%.

This was the result. There is real color information at some resolution, but that is binning at a greater level than 3x3.

I guess that reducing the blue in a linear way was incorrect, and it would have been better to reduce it more in the areas where blue was darkest and kept it higher where blue was brighter (the terrain is a more neutral white/gray). As mentioned earlier, the noise is not symmetric because in either dark areas or bright it can push the output brighter, but in dark areas, it can't push the output into negative.

Attached thumbnail(s)

 

[mcaplinger]

So maybe do green and red?

Have to think about that. There are timing/cadence tradeoffs between TDI and number of bands. Could do green on one spin, red on the next, probably.

[volcanopele]

daylit map (includes images 22 to 25)

[volcanopele]

Jupiter shine map, with and without 3x3 median smoothing



For those who need it, this map is in simple cylindrical projection, lat range is -90 to 90 and so is the longitude range. Pixel scale is 2 km/pixel

[chlegrand]

Hello UMSF members,

I follow everyday all your exchanges about lunar and planetary missions.

I'm one of the co-authors of the "Virtual Moon Atlas" (VMA) and "Virtual Planets Atlas" (VPA) freewares.

https://ap-i.net/avl/en/start

https://ap-i.net/avp/en/start

Our goal is to provide for general public and scientists, easy to use basic tools for improving lunar and planetary knowledge.

As you will see if you download and try it, VPA presents telluric planets and Galileo satellites real time maps. For that, we use with permission equirectangular "textures" in the 2:1 ratio buit from various planets maps found on the Internet and publically available.

I would like to add in VPA Juno's maps of the Galileo satellites since presently, there are only Voyagers and Galileo maps used in the freeware.

Using the VPA "double window" functionnality, it will certainly be useful to look for planetary surface evolutions between these various missions for VPA users.

Does anyone could help us and provide such best resolution equirectangular Juno maps for Io (Sure !) but also for Europa and Ganymede that can be used in VPA with permission and thanks ? (Natural or enhanced colors / B&W / 2:1 ratio / jpg format / 4000x2000 pixels minimum or more)

Thank you in advance for your support to our action if you agree.

Ch. Legrand

PS : Sorry for my rather poor english

[kymani76]

I managed to extract preliminary shape model of Io from all the imagery. It might be possible to refine it further,
but I encountered some hardware limitation with processing, which need more time to resolve.


The tall Matterhorn-like mountain really stands out.

Model is available for download by clicking here.

[Decepticon]

The nightside image is a wonderful bonus!

[kymani76]

I refined Io shape model further with more images and added texture overlay.
The mountains are not as prominent as I thought they whey would be, as stereo information is not sufficiently dense.
The resulting DEM is consequently also of low resolution. The model would benefit from the entire sequence of images
from the same hand.



I can also get some cloud points for the nightside (image above), but would need more imagery to model it.

You can download the new model here.

[mcaplinger]

The Jupiter images have been posted to missionjuno. I think we're still waiting on some partials for the approach images, some which are pretty neat (GRS with a distance Io).

[Decepticon]

Does the next flyby cover the same region?