IPB

Welcome Guest ( Log In | Register )

6 Pages V  « < 4 5 6  
Reply to this topicStart new topic
Perijove 1 (PJ1), August 27, 2016
stevesliva
post Sep 20 2016, 07:14 PM
Post #76


Senior Member
****

Group: Members
Posts: 1306
Joined: 14-October 05
From: Seattle
Member No.: 530



QUOTE (Brian Burns @ Sep 20 2016, 07:39 AM) *
This is probably my favorite image from Juno so far - nice colors


Kind of neat how there are clear cyclones and anticyclones on the same image. One type light the other dark.
Go to the top of the page
 
+Quote Post
Brian Burns
post Sep 21 2016, 08:18 AM
Post #77


Junior Member
**

Group: Members
Posts: 54
Joined: 7-July 16
From: Austin, Texas
Member No.: 7991



QUOTE (Gerald @ Sep 20 2016, 08:00 AM) *
To stay independent, I've written another system (also in C++, like ISIS3), with additional calibration capabilities, but specific for JunoCam, and not designed under shipping, usability or extensibility considerations.


That is great - I might wind up having to write something using OpenCV to handle the Voyager image projections also, as ISIS seems to be written with clarity in mind, rather than speed, as far as I can tell - which is fine, but there a lot of images to process, so if OpenCV is faster that will be helpful.
Go to the top of the page
 
+Quote Post
Candy Hansen
post Oct 27 2016, 06:53 PM
Post #78


Newbie
*

Group: Members
Posts: 12
Joined: 13-October 13
Member No.: 7013



QUOTE (xosema @ Sep 4 2016, 02:12 PM) *
Hi all.

I'm not sure if this should be posted in this thread... first post of mine, anyway, after years registered. I think it would be didactic to compare Earth with Jupiter in the polar images and then made this (posted in a blog, but I don't know if it would be nice to put url here).

[attachment=40082:Jupiter_...hoemaker.jpg]

(Credit: NASA / JPL / SwRI / MSSS / NEAR-Shoemaker / Crastinia)


It would be very nice if you could post this on the missionjuno site - it's rare to see the earth's south pole compared to Jupiter's. (-:
Go to the top of the page
 
+Quote Post
Gerald
post Nov 28 2016, 06:46 PM
Post #79


Senior Member
****

Group: Members
Posts: 1765
Joined: 7-December 12
Member No.: 6780



A preliminary collage of a global background map from Marble Movie images, with merged PJ1 maps in the foreground:
Attached Image


The northern part matches roughly, the southern part can be off about 15 degrees longitude.

There are several inaccuracies in the image; some of them make it beautiful. Therefore I'm posting this intermediate version.
Link to the full version.
And here some more making of.

(I'll upload the full version to missionjuno, after preparing some caption.)
Go to the top of the page
 
+Quote Post
Gerald
post Dec 3 2016, 11:35 PM
Post #80


Senior Member
****

Group: Members
Posts: 1765
Joined: 7-December 12
Member No.: 6780



A crop of an enhanced (planetocentric) map derived from JunoCam image #6184:
Attached Image

(image: NASA / JPL /SwRI / MSSS / Gerald Eichstädt)

Reviewing my assumptions about JunoCam's "companding" (triggered by M.Caplinger) helped removing most of the brightness-related color banding.
This elaboration is partially motivated by a discussion among members of the JunoCam team about the nature of the banding.

I think, some linear haze structure is discernible on about the latitude of the northern edge of the two prominent bright vortices (ovals) in the south.
Go to the top of the page
 
+Quote Post
Gerald
post Dec 4 2016, 11:02 AM
Post #81


Senior Member
****

Group: Members
Posts: 1765
Joined: 7-December 12
Member No.: 6780



... Take a crop of a planetocentric map image #6184 in raw colors, i.e. all linear channel weights set to 1.0:
Attached Image

Take 100,000 equally distributed random (Monte Carlo) pixels, and determine mean and standard deviation of the according red->green, and blue->green functions:
Attached File  map_JNCE_2016240_00C06184_rgb_maps.txt ( 18.07K ) Number of downloads: 36

Here a diagram of the according green/red, blue/red quotients:
Attached Image

Apply the red->green, and blue->green functions to the raw map:
Attached Image

This removes most of the color banding, and normalizes the mean image to grey for each brightness.
Approximate de-Lamberting (by dividing by the square root of the cosine of the solar incidence angle), and enhancement by the 4th power results in:
Attached Image

Further manual saturation and brightness enhancemnt returns the map I've used for the crop of the previous post:
Attached Image


This method removes most of the relative color banding. But absolute non-linearity of the green channel leaves some brightness banding.
I'll try to resolve this remaining issue, too. Approximating radiomentric in-flight calibration is a little more tricky, and hence where the fun begins...

Go to the top of the page
 
+Quote Post
Gerald
post Dec 4 2016, 03:52 PM
Post #82


Senior Member
****

Group: Members
Posts: 1765
Joined: 7-December 12
Member No.: 6780



A crop of a map derived from #6180 shows at least three cloud levels, at the top of the image a finely granular layer with some dark ovals, then a layer of turbulent folded filametary regions (FFRs) partly covering one of the dark ovals beneath, then a subtle haze layer with linear (near the bottom of the image) and vortical structures (in the upper half of the image, at the very right), sufficiently translucent to see the layer(s) beneath :
Attached Image

(image: NASA / JPL /SwRI / MSSS / Gerald Eichstädt)

The image is rendered with 30 pixels per planetocentric degree. North is up.
Go to the top of the page
 
+Quote Post
Gerald
post Dec 6 2016, 11:34 AM
Post #83


Senior Member
****

Group: Members
Posts: 1765
Joined: 7-December 12
Member No.: 6780



Applying the three red->green and blue->green maps
Attached Image

returned these three de-Lamberted and exposure-corrected maps for images #6162, #6160, and #6163:
Attached Image

By equalizing the maps, and possibly additional constraints, I'll try to get a better companding correction than just stacking/averaging #6162 and #6160, as in the third image of this gif:
Attached Image
Go to the top of the page
 
+Quote Post
Gerald
post Dec 7 2016, 11:25 PM
Post #84


Senior Member
****

Group: Members
Posts: 1765
Joined: 7-December 12
Member No.: 6780



Using the correct decompanding table simplified this step. Here a gif composed of #6160 and #6162:
Attached Image

The cropped planetocentric maps are decompanded, color-calibrated, exposure-adjusted, approximately de-Lamberted, stretched by gamma=4.0 relative to square-root encoding.
Go to the top of the page
 
+Quote Post
Gerald
post Dec 8 2016, 10:46 PM
Post #85


Senior Member
****

Group: Members
Posts: 1765
Joined: 7-December 12
Member No.: 6780



Selected cropped maps with 30 pixels per planetocentric degree.
(images 6160, 6162, 6163, 6180 (crop), 6184 (crop)).

Color banding is mostly solved with M.Caplinger's decompanding table. (Thanks!)
There is still some brightness variation. I'm going to do related correlation analysis to see, whether this can be reduced in a systematic way. I'm suspecting a mean BRDF other than the cosine of the solar incidence angle (Lambert model) if it's a BRDF at all, and a relevant non-constant camera flat field as among the most promising candidates. But my geometric reference model for de-Lamberting isn't yet quite perfect, too.
Go to the top of the page
 
+Quote Post
Gerald
post Dec 12 2016, 12:33 PM
Post #86


Senior Member
****

Group: Members
Posts: 1765
Joined: 7-December 12
Member No.: 6780



After a more thorough analysis of images #6160 and #6162, I see a very good chance at least, to analyse optical properties of Jupiter's global haze layer quantitatively. Here reduced data (blue) compared to a function (red) which I think, can be interpreted as a horribly simplified haze scattering and extinction model:
Attached Image

The oversimplified model breaks down, where the haze becomes opaque when seen from the Sun or from the observer.
Nevertheless I was stunningly surprised, when I saw the reduced data, and the ease with which a simple function could be fit into the data, even before I had an idea how to interprete this, since real non-lab data rarely fit to a simple function. After I suspected light scattering of Jupiter's haze as a possible root cause, I failed to readily find an according easy formula for light scattering in the web. So I did a quick and dirty back-of-an-envelope calculation, and found a 1st-order simplification which seems to explain the empirically identified type of functions.
Besides the immediate science value, a good model of the haze will help to improve the creation of maps and images of the cloud top below the haze.
The latter is what I'll try to get implemented before PJ3 image data will be published in a few days.
Go to the top of the page
 
+Quote Post
Gerald
post Dec 13 2016, 11:19 PM
Post #87


Senior Member
****

Group: Members
Posts: 1765
Joined: 7-December 12
Member No.: 6780



Heavily enhanced crop of map derived from #6184:
Attached Image


Now going to work on PJ3.
Go to the top of the page
 
+Quote Post

6 Pages V  « < 4 5 6
Reply to this topicStart new topic

 



RSS Lo-Fi Version Time is now: 25th April 2017 - 06:46 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.