Tuesday, April 12, 2005

Huygens Flat-fielded MRI images

Rene Pascal has posted online a zip file contain all non-duplicated medium resolution imager frames from DISR taken during Huygens descent in January. Each image has been flat-field, a image-processing procedure where camera and/or CCD defects are removed from an image by ratioing the image with an image that contains just the defects. Rene created that file by summing several bland MRI images together to produce an average bland file. This bland file thus contains only the defects and no contrast due to surface features. This procedure produces far cleaner and scientifically useful images than the raw images that were released.

6 Comments:

Blogger Bruce Moomaw said...

They are indeed marvelously improved in quality -- I hope we see something even better eventually from the official DISR team. (By the way, the final returned pre-landing MRI photo was taken at about 800 meters altitude. By contrast, the final returned Side-Looking Imager frame and the final full-frame image from the HRI were taken at only about 400 meters -- and they got a a handful of half-frames from the HRI down to about 250 meters. Those would have been full frames, except that Channel A was supposed to return half of each of the very last HRI frames.)

One thing I notice is that we are not just looking at narrow channels of dark material -- in many places we see equally fine lines and filigrees of light-colored surface against a mostly-dark background (which are made much more obvious by Pascal's processing). I presume this confirms once again that what we're looking at in these scenes is dark mud deposits rather than current surface bodies of dark liquid.

4/12/2005 10:32:00 PM  
Blogger Bruce Moomaw said...

A bit more on the altitudes of the last returned images from each of Huygens' cameras:

MRI: 875 meters

SLI and full-frame HRI: 405 meters

4 HRI half-frames returned from below that-- the last one at 258 meters. (I'm not sure that any of these have even been publicly posted yet -- and, according to both Jason and Jonathan Lunine, there was at least some interference in their clarity due to mist backscatter from Huygens' floodlight, which was turned on at roughly 500 meters.)

4/12/2005 11:24:00 PM  
Blogger Jason said...

Indeed, Rene has done a great job processing these relatively difficult data set. With flatfielding, many of the higher altitude images can be more accurately contrast enhanced.

In terms of the post-lamp HRI images, it does indeed look like many of them are a wash out due to the back scattering of the haze. While it is an unfortunate result for surface science, given that in the last 100 meters, Huygens flew over a 5 meter tall cliff, but it definitively shows that the haze extends all the way to the surface.

4/13/2005 11:05:00 AM  
Anonymous paxdan said...

Impressive processing, the site also contains the best Huygens mosaic i've seen yet. Any more info on the 5m cliff, this is the first i've heard of it.

4/14/2005 03:24:00 AM  
Blogger Jason said...

The Sonar experiment on the SSP instrument on Huygens detected at double ech oin the last 100 meters before landing. The shape and intensity of this double peak, and its evolution from 100 meters alitude to the ground, is consistent with Huygens having flown over a 5 meter tall cliff.

4/14/2005 03:27:00 PM  
Blogger Bruce Moomaw said...

Aha! So THAT'S the explanation of that double echo! Up to now I haven't seen it mentioned by anyone besides myself, although it was very clear on the animated version of the echo sounder's graph during the first Huygens scientific news conference. This may be because the ESA didn't put that part of the graph on its webpage of graphs from that news conference -- you had to be watching the animated version, which ran very quickly, to see it, although if you did it was (as I say) extremely clear.

The theory I had was that the sounder might be detecting a second echo off the floor of a layer of mud below the capsule. I presume that the post-landing data from the sounder rules this out?

4/14/2005 07:43:00 PM  

Post a Comment

<< Home