HiRISE: High Resolution Imaging Science ExperimentThe University of Arizona
New Images Catalog Anaglyphs Stereo Pairs Science in Motion FAQ HiBlog Themes Software Contact Search

Hi-Fi Stereo (the other kind)

Now that all the Phoenix landing excitement has calmed down, we can return to regular old day-to-day operations…. One of those is stereo imaging. (For day-to-day Phoenix operations, check out their website!)

Stereo anaglyph from PSP_001534_1560 On the main HiRISE site, we’ve posted a list of all the stereo pairs we’ve acquired & released. There are 467 of them! Now you can look up whether your favorite image is part of a stereo pair or not. (I don’t think the list is searchable yet, but it’s sorted by the observation ID of the first half.) There is also a PDF document on the page (click here to download directly) with more information about our stereo images. It includes instructions for using these stereo pairs to make your own red-blue anaglyphs (the ones you view with the 1950s 3-D glasses). To the right is one of those anaglyphs. This is part of a delta in Eberswalde Crater from the image PSP_001534_1560, linked on that page. (Search our catalog for ‘anaglyph’ to see even more.) 8)

Wondering how we get stereo imaging?

It’s the same principle as seeing 3-D with your eyes: we need parallax to get the third dimension. To do this with only one “eye” (the HiRISE camera), we observe the same spot on Mars twice, from different directions. Usually the first half of the stereo pair is in a small off-nadir roll in one direction, and then the second half is taken on a larger roll in the other direction.

We can’t take two images at once, so we have to wait for the spot to be viewable twice – that can sometimes happen in the next two-week cycle, but usually it takes several cycles. Sometimes conflicts with other instruments or spacecraft activities prevent us from completing the stereo in enough time, and we have to “abandon” the first half. We’ve started calling those “left at the altar.” It’s very sad. ;)

There’s a time limit on completing stereo because over time, the shadows change with the Sun’s seasonal movement. If the pattern of bright and dark material on the ground changes for any reason, the two halves of the stereo become impossible to correlate. For example, if a dust storm or a dust devil moves dust around, or if frost appears or disappears, the stereo is ruined. So once we take the first half of a stereo pair, the clock is ticking on completing it.

It would be wonderful if we could make every image a stereo observation, but we have to save these for our very highest-priority scientific goals. HiRISE will image less than 1% of the surface of Mars in the primary mission – so every image is precious! When we do stereo, we’re repeating coverage over the same ground. These stereo images have a very high scientific value to the team, but sometimes it’s still hard to think about trading that for an observation of another spot on Mars that has never before been imaged at high resolution. We have to make some tough decisions!

Tags: , ,

4 Responses to “Hi-Fi Stereo (the other kind)”

  1. Fred Calef III Says:

    These ‘left at the altar’ images could perhaps be paired with MOCNA, HRSC or CTX (most likely) to derive the anaglyph (at a lower resolution). Has anyone tried?

  2. DataMiner Says:

    Fred,

    There are some Mars enthusiasts that have made some attempts at creating anaglyphs from mixed sources and posted them on their websites. Some of the anaglyphs *look* pretty good, although I have no idea how accurate they are. So this can be tried by anyone that wants to put in the effort.

  3. HiBlog: HiRISE Team Blog » Blog Archive » Coming soon, to a theater - I mean computer - near you… Says:

    [...] Here’s another blog entry explaining how we take the stereo images that are used to create the anaglyphs: Hi-Fi Stereo (the other kind) [...]

  4. HiBlog: HiRISE Team Blog » Blog Archive » And now for something complete different! Says:

    [...] are created in a three-step process. The first step is to take the two images of the stereo pair and map project them. This helps the pipeline determine which image will be the left image and [...]

Leave a Reply