4

I attended the North Eastern Astronomy Forum this year and one of the speakers was mentioning that JunoCam will not be working because of the angle of the spacecraft. He was giving a lecture on what amateurs could do to replicate its functionality.

During this lecture he mentioned that JunoCam is using a very simplistic banding on its camera sensor. Instead of going the route that a lot of DLSR's go which utilizes Debayering they decided to do a banded sensor approach. I suppose this is analogous to having multiple monochrome cameras that detect different wavelengths of light. The wavelengths that they chose were, red green and blue with 2 additional NIR spectrum filters. So, basically, their camera takes single shots capturing:

  • The top X rows of pixels as red light
  • The next X rows of pixels as blue light
  • The next X rows of pixels as green light
  • The next X rows as near infrared
  • The next X rows as Methane (?)

enter image description here

The end result of capturing each of these spectrum in monochrome is an image like this:

Example JunoCam Image

So, essentially, to get a full-color image, you need to slice these images on the lines multiple times and put the data into the correct channels. Then you need to stitch each channel back together to get the full image and proceed to do additional processing on each piece to correctly align them. I'm not even mentioning planetary derotation and orbital motion compensation as this is already a decently complicated process.

My question is:

Why did they choose such a complicated methodology of image capture, when they could've used an easier method? The whole banding of channels is really bizarre for a camera, I've heard of nothing quite like it before (let me know of other instances where they've done this). I understand JunoCam was indeed an afterthought, but if you know why this specific design choice was made I'd love to hear more about it.

Also, if you're interested in helping out the JunoCam project, head over here to see what NASA needs from amateurs during the JunoCam downtime.

Here's the main site: https://www.missionjuno.swri.edu/junocam

Magic Octopus Urn
  • 9,387
  • 3
  • 31
  • 82
  • 2
    Ah! I knew this sounded familliar. Please have a look at @DavidHammen's most excellent answer to the question How is JunoCam different from a normal CCD camera? If it answers your question, then it might be better to close your question as duplicate to that one. – uhoh Apr 08 '19 at 23:40
  • 1
    You could also consider writing a second answer there, based on the paper you're discussing here. – uhoh Apr 08 '19 at 23:44
  • 1
    @uhoh I don't have a hammer :(. But YES that is the PERFECT answer, I didn't know it alternated between the methane band and the others. This is definitely a dupe then :). Great detective work as always. – Magic Octopus Urn Apr 08 '19 at 23:48

1 Answers1

6

update: There is a much better explanation in @DavidHammen's answer to How is JunoCam different from a normal CCD camera?

Why did they choose such a complicated methodology of image capture, when they could've used an easier method?

That is probably false premise, you certainly haven't demonstrated how they could.

To get the best, single pixel resolution, you don't mess with Bayer filters. And a Bayer-like filter with five colors is a nightmare.

You can instead...

  1. use a filter wheel and use all pixels of your CCD for each color, or
  2. put different filters over different chunks of the CCD sensor and do it the way you describe in the question.

That's why; and it's not... so... odd after all!

Look at GAIA's sensor arrangement, which works on a somewhat similar principle.

both below: Gaya's CCD array, from Spaceflight 101, image credit: ESA and Astrium respectively.

enter image description here

enter image description here

uhoh
  • 148,791
  • 53
  • 476
  • 1,473
  • I'm guessing the filter-wheel was a no-go because it's mechanical moving parts. I did a poor job of explaining alternatives, because I'm not sure of alternatives used in other spaceflights. The gaia CCD's is a great addition though I'll research that for more information, because that looks like it wasn't an after-thought. – Magic Octopus Urn Apr 08 '19 at 17:48
  • 2
    My biggest confusion is why they didn't write software to stitch these images back together at home, they're asking amateurs to do the stitching for them (because, obviously, they're busy and on a budget). Just seems this method did not enable ease-of-use. I guess I worded everything poorly as usual; the presenter was the one saying "this is a weird way of doing it" and I was looking for confirmation of what he had said. – Magic Octopus Urn Apr 08 '19 at 17:49
  • @MagicOctopusUrn that sounds like a great follow-up question! – uhoh Apr 08 '19 at 22:32
  • check out the picture I added from this paper I found it's actually quite informative. – Magic Octopus Urn Apr 08 '19 at 23:19
  • 1
    +1, the original Shuttle cameras used a mechanical filter wheel. – Organic Marble Apr 08 '19 at 23:20
  • @organicMarble really? That's pretty darn cool! I can't afford a filter wheel for my telescope yet, but it's on my todo list after a 2-speed focuser. Seems the way to go for short missions. Have they used it on any long missions? – Magic Octopus Urn Apr 08 '19 at 23:22
  • 1
    @MagicOctopusUrn "the field sequential color format uses a three-color, six-segment rotating filter to process color scenes." - July 1988 CCTV training manual. Don't know about other missions. – Organic Marble Apr 08 '19 at 23:33
  • 2
    Looks like Apollo did too. https://www.hq.nasa.gov/alsj/Niemyer-Paper.pdf – Organic Marble Apr 08 '19 at 23:39
  • Wow that's cool! Thanks again OrganicMarble :). I'm reading that now. I can't find the July 1988 CCTV Training Manual, do you have a link for that too? – Magic Octopus Urn Apr 08 '19 at 23:41
  • 1
    Sadly, I don't think that one is online. I'll put it on my list to scan in. – Organic Marble Apr 09 '19 at 00:08