Reprocessing Historical Images, Looking for REALLY big challenges? Options

[DDAVIS]

[quote name='mcaplinger' date='May 10 2007, 04:30 AM' post='89888']
http://nssdc.gsfc.nasa.gov/database/Master...g?ds=PSPA-00335

I beleive this refers to the Viking Lander atlas, which I obtained in my USGS days. I used the overhead maps with the fragmentary contours and meter grid to match the grid with landscape features and contours seen in perspective. This is a sample of such work I did in preparation for a shaded releif map I never finished. The entire Viking 1 stereo coverage area took months to chart out in this way.

An automated means to create meshes from stereo pairs from varied formats would indeed be useful. I can paint in gray levels assigned to the contours within each obliquely seen grid section and then rectify the thing section by section but that could take more time than I can presently justify.

Don

[Paolo]

I have been searching for Suisei UV images of Halley, but they seem to be completely forgotten, and the raw data are nowhere to be found (in fact, I even heard that they are lost). Anybody know about it?

[tedstryk]

I have some scans, but no raw images.

[Paolo]

I have a few scans too. I was just wondering where did the raw images go. And beside, there are no Suisei UV images on the net, it seems. The only thing I could find was this http://en.wikipedia.org/wiki/Suisei_probe plus some 2-bit scanned pdf.

[tedstryk]

I have a few scans too. I was just wondering where did the raw images go. And beside, there are no Suisei UV images on the net, it seems. The only thing I could find was this http://en.wikipedia.org/wiki/Suisei_probe plus some 2-bit scanned pdf.

I have looked at that Wikipedia article, and comparing it to what I do have, I am fairly confident that it is not a Susei image. It looks like it was taken with a small telescope with a chromatic aberration problem.

[tty]

I went to CiNii, my favorite place for japanese scientific papers and found these JAXA papers:

http://ci.nii.ac.jp/cinii/servlet/CiNiiLog...d=ART0000606784

"Lyman Alpha Observations of Comet Halley by Suisei"

http://ci.nii.ac.jp/cinii/servlet/CiNiiLog...d=ART0000606972

"Halley's Comet Explorer "Suisei"

The images aren't too hot however, I imagine they are the same You already have,

[tedstryk]

They seem like poorly scanned versions of some of the images that I have. I have made a page with my Suisei images, which can be seen here: http://www.strykfoto.org/Suisei/Suisei.htm

[ChrisP]

I made a preliminary page containing several Pioneer 10 and 11 images scanned from Pioneer: First to Jupiter, Saturn and Beyond (NASA SP-446). Back at the time I acquired this book (years ago) I was surprised the images weren't worse, before I had only seen very crude reproductions of the images.

The page is at http://www.mmedia.is/bjj/misc/pioneer/index.html . I will add more images later.

The Apogee book, Deep Space - The NASA Mission Reports, has that book and 25 others on the attached DVD as PDF files (searchable text and embedded images) which makes it trivial to extract the images. None of these 26 books are available on-line at he NASA History Division and what with the movies and hard copy, this was the best 23 quid I've ever spent!

Last year at work I met someone who had worked on the Pioneer and Viking Lander images for NASA and explained some of the difficulties they went through to try to get decent images out of the downlinked data. Sadly he no longer has any of his notes and documents from that time.

[4th rock from th...]

Here's a WIP Pioneer 10 Jupiter poster. Processing was based on stacking images from different sources, computing a synthetic green channel and correcting for spacecraft motion and planet rotation. The images were placed on Jupiter disks to indicate the incomplete image coverage. I plant to progress to the dataset as possible.

[hendric]

Wow, the GRS really has been shrinking the last few decades! There is a noticeable difference in size between the P10 images and images today:

http://jupiter.cstoneind.com/

[elakdawalla]

Yeah, that's the thing that always strikes me as amazing about the Pioneer Jupiter images, and why it's so important to keep them "alive." Thanks for the great versions, Ricardo.

Emily

[David]

Wow, the GRS really has been shrinking the last few decades! There is a noticeable difference in size between the P10 images and images today:

http://jupiter.cstoneind.com/

Is it really smaller though? I didn't make measurements, but it seems to take up the same amount of latitude judging by the bands. As for the longitude, it certainly looks narrower, but much of that seems to be due to a color change; if you compare the Spot to an eye, the "eyeball" has changed from red to white, so that what seems to be the whole Red Spot is just the internal oval cyclone.

[Bjorn Jonsson]

Its longitudinal extent is getting shorter - this apparent difference is real. If it continues shrinking at the same rate it will become circular around 2040 or so. That's thought to be an unstable configuration so something interesting is going to happen. The big question is - exactly what? There was also some interesting GRS discussion in this thread.

BTW I wish the original digital Pioneer imaging data were available somewhere. Due to the nature of the Pioneer imaging system (spin scan imaging with just a single 'pixel') processing the data is difficult. Still I'm sure that reprocessing it on modern computers would result in something better than the processing done more than 30 years ago. What some of the people here have been doing with ancient data like Mariner 6 and 7 images and Voyager images certainly makes this likely.

An obvious processing idea would be to simply place the spin scan pixels on a cylindrical map, project the map on an ellipsoid and render a perspective view. This was computationally extremely expensive to do in the mid 1970s but takes only a few seconds to render on today's computers. There will be problems at the limb and terminator due to Jupiter's rotation since acquiring the images took about 30 minutes IIRC but that's not a showstopper.

[tedstryk]

Here's a WIP Pioneer 10 Jupiter poster. Processing was based on stacking images from different sources, computing a synthetic green channel and correcting for spacecraft motion and planet rotation. The images were placed on Jupiter disks to indicate the incomplete image coverage. I plant to progress to the dataset as possible.

Great work! I am also working with that dataset. I have put most of my time into trying to find digital data, but I am beginning to to think it futile. Also, while there is slight improvement, I have found that really good reproductions can be largely reconstructed from the best versions in print. Due to the low number of pixels/high level of enlargement and 6-bit encoding, the original graylevels can be pretty well reconstructed (although it is by no means truly photometric).

[4th rock from th...]

Modern processing will benefit in having a better green channel (remember that the Pionners only had a Red and a Blue color channel) and in geometrical correction.
From what I've made out of the imaging system, the images were scanned using the spacecraft spin (spin scan) and movement relative to the planet. So let's say, the X axis was given by the spacecraft "horizontal" movement as it passed by Jupiter, and the Y axis by it's own rotation.
So the distortions are enormous and the images tool a lot of time to be built. We are talking several minutes here. Yet there are some good quality images taken at interesting angles. Closest approach pictures do reveal nice details, and yes, the GRS photos are unique.

I've made some progress in producing a cylindrical map out of the dataset, but it's somewhat inaccurate. So it is possible to "complete" each individual image coverage with those taken before and after.