3 ms·

Does anyone know what kind of processing these images go through? Does Ceres receive enough light for the satellite to see it plain, or is this the result of X-

by beaner 11y ago

Does anyone know what kind of processing these images go through? Does Ceres receive enough light for the satellite to see it plain, or is this the result of X-ray vision or something similar? And the bright spots - I can understand those being the result of ice reflection, but they seem to absolutely glow in these images. I imagine that's the result of some filter or processing making them pop. Is that accurate?

opcvx 11y ago
The image is brightened. This makes the spots stand out. The surface of Ceres is effectively black, but the image shows it as grey to light grey. Consequently the spots shine, but in reality they are just white.
geuis 11y ago
I asked a similar question a couple days ago over at reddit and got this reply: http://www.reddit.com/r/space/comments/3ag085/live_sources_of_nasaesa_imagery/csdretg http://www.reddit.com/r/space/comments/3ag085/live_sources_o... "Spacecraft data is downlinked to Earth as "telemetry packets" - basically big bundles of data formatted/compressed in such a way as to maximize the efficiency of transfer. There are typically separate science (e.g. images/data) and housekeeping (e.g spacecraft health, voltages, temperatures) telemetry packets. These data are nearly always "open" but we generally don't make them public because it's kinda meaningless. Instead, as soon as we receive the telem files, we unpack them and process them into useable data files. Most often these will be saved as FITS or HDF5 files and placed into a database for scientists to access immediately. At the same time - and where applicable - the data are also turned into jpg/png files that are placed on websites for the public and scientists to view. (For science analysis, we generally use the FITS/HDF5/etc data but we use the jpgs for general viewing) tl;dr Yes the raw data is generally available but formatted weirdly."
Mithaldu 11y ago
Damn good question. Short answer: The images are brightened visible light. The wikipedia article has more information: https://en.wikipedia.org/wiki/Bright_spots_on_Ceres https://en.wikipedia.org/wiki/Bright_spots_on_Ceres Notable points: In infrared the spots are darker or equal than surroundings, thus cooler or equal. In lower resolution pictures the bright spots are considerably larger than their now apparent size, thus they're throwing out massive amounts of light, likely reflection of sunlight. They seem to reflect equally well from any given viewing angle, thus whatever their material is seems to exhibit some kind of cat's eye effect. ( https://en.wikipedia.org/wiki/Cat's_eye_(road) https://en.wikipedia.org/wiki/Cat's_eye_(road) )
- mturmon 11y ago
  About where the images come from, don't forget that there are two cameras on Dawn: -- Framing Camera (FC) which is high-resolution and has a 7-position filter wheel (which allows for specific wavelengths to be chosen) as well as allowing white-light images. FC is used for navigation and general mapping of the asteroid geometry. -- A Visible and Infrared Spectrometer (VIR) which is lower resolution, but better calibrated radiometrically. It produces spectrally-resolved images to determine mineral composition. More at http://dawn.jpl.nasa.gov/technology/science_payload.asp http://dawn.jpl.nasa.gov/technology/science_payload.asp The images you're seeing in this press release seem to be from FC. I say this because they have high resolution, and some other press releases with similar images say outright that they used FC. The FC spec is 94 micro radians/pixel which, at a distance of 4400 km, works out to be about 410 meters per pixel, and that seems compatible with the feature sizes of ~10km noted in the press release. But, some of the images in the wikipedia page you linked, which is very good, are in infrared, and are of lower resolution, and probably therefore come from VIR. The temperature plot also seems like VIR (you can get a temperature if you have a spectrum). Finally, about processing, I notice that the images in the press release seem to have a lot of pixels. This is a clue that they might have composited a lot of individual images to get this one image. If so, there had to be a geometric correction (at the very least) because of course the camera is in motion while the images are taken. This requires very good geometry information. Additionally, there might have been a projection into a standard coordinate system (so that "north" is up). You might be interested to know that there is controversy about asteroid coordinate systems (for Vesta, https://en.wikipedia.org/wiki/4_Vesta#Coordinate_systems https://en.wikipedia.org/wiki/4_Vesta#Coordinate_systems). About the white areas -- yes, probably they saturated the detector.
- CamperBob2 11y ago
  In infrared the spots are darker or equal than surroundings, thus cooler or equal. We can't really say this, though, without knowing the material's emissivity. And since we don't know what it is...
  - mturmon 11y ago
    This is true in the abstract, and a good point. But in this case, they have a pretty densely sampled and well-calibrated spectrum from one of the instruments (http://dawn.jpl.nasa.gov/multimedia/vir_gallery.asp http://dawn.jpl.nasa.gov/multimedia/vir_gallery.asp -- there appear to be ~100 spectral samples), so they would be able to make a black-body assumption, derive a temperature, and then validate the extent to which that is correct. In other words, it's not just a plain IR image, it's a whole spectrum.
codezero 11y ago
geuis's comment pulled from reddit sounds right. I used to work with satellite data that looked at visible light, but my dataset was close to Earth. I can imagine that the relative brightness of the surface of Ceres is very very low compared to what I've worked with, so keeping that in mind, any spot that spikes above some threshold is going to saturate your image. My data set was in black and white and had an integer brightness range of 32767. The raw data hovered between 20000 and 32767 (unitless intensity) because the instrument I worked on was old, and had suffered degradation, leaving a stronger "dark current" that needed to be removed. General steps for processing involved: 1) subtracting noise from the raw data, which came via a number of different paths 2) median filtering raw image frames into a single composite image, this may involve some masking 3) removing large scale noise (in our case, stars, planets and the moon) 4) adjusting the data range for presentation (this is only done for looking at the images, not for actual processing) The last step is probably what's causing this bright spot to seem so bright. With a different range of values, Ceres would look like a dark dark grey disc, with a lighter grey spot where the very bright spot is visible. I'd like to know the actual surface brightness of that spot, if it looks to be the same brightness as the Sun's reflection off of ice at that distance, it would clear things up pretty fast. Ice could be subsurface ice, or it could be remnant from the actual impacting object, if it were a comet.
- Mithaldu 11y ago
  Would ice exhibit the property of reflecting from any angle as seen in this gif? https://en.wikipedia.org/wiki/File:Ceres_spots_animation_May_4_2015.gif https://en.wikipedia.org/wiki/File:Ceres_spots_animation_May...
  - Symmetry 11y ago
    If it's not smooth ice sure. If you look at an iceburg, for instance, it'll be white and be equally bright from any angle.

by beaner 11y ago

opcvx 11y ago

The image is brightened. This makes the spots stand out. The surface of Ceres is effectively black, but the image shows it as grey to light grey. Consequently the spots shine, but in reality they are just white.

geuis 11y ago

I asked a similar question a couple days ago over at reddit and got this reply: http://www.reddit.com/r/space/comments/3ag085/live_sources_of_nasaesa_imagery/csdretg http://www.reddit.com/r/space/comments/3ag085/live_sources_o... "Spacecraft data is downlinked to Earth as "telemetry packets" - basically big bundles of data formatted/compressed in such a way as to maximize the efficiency of transfer. There are typically separate science (e.g. images/data) and housekeeping (e.g spacecraft health, voltages, temperatures) telemetry packets. These data are nearly always "open" but we generally don't make them public because it's kinda meaningless. Instead, as soon as we receive the telem files, we unpack them and process them into useable data files. Most often these will be saved as FITS or HDF5 files and placed into a database for scientists to access immediately. At the same time - and where applicable - the data are also turned into jpg/png files that are placed on websites for the public and scientists to view. (For science analysis, we generally use the FITS/HDF5/etc data but we use the jpgs for general viewing) tl;dr Yes the raw data is generally available but formatted weirdly."

Mithaldu 11y ago

Damn good question. Short answer: The images are brightened visible light. The wikipedia article has more information: https://en.wikipedia.org/wiki/Bright_spots_on_Ceres https://en.wikipedia.org/wiki/Bright_spots_on_Ceres Notable points: In infrared the spots are darker or equal than surroundings, thus cooler or equal. In lower resolution pictures the bright spots are considerably larger than their now apparent size, thus they're throwing out massive amounts of light, likely reflection of sunlight. They seem to reflect equally well from any given viewing angle, thus whatever their material is seems to exhibit some kind of cat's eye effect. ( https://en.wikipedia.org/wiki/Cat's_eye_(road) https://en.wikipedia.org/wiki/Cat's_eye_(road) )

mturmon 11y ago
About where the images come from, don't forget that there are two cameras on Dawn: -- Framing Camera (FC) which is high-resolution and has a 7-position filter wheel (which allows for specific wavelengths to be chosen) as well as allowing white-light images. FC is used for navigation and general mapping of the asteroid geometry. -- A Visible and Infrared Spectrometer (VIR) which is lower resolution, but better calibrated radiometrically. It produces spectrally-resolved images to determine mineral composition. More at http://dawn.jpl.nasa.gov/technology/science_payload.asp http://dawn.jpl.nasa.gov/technology/science_payload.asp The images you're seeing in this press release seem to be from FC. I say this because they have high resolution, and some other press releases with similar images say outright that they used FC. The FC spec is 94 micro radians/pixel which, at a distance of 4400 km, works out to be about 410 meters per pixel, and that seems compatible with the feature sizes of ~10km noted in the press release. But, some of the images in the wikipedia page you linked, which is very good, are in infrared, and are of lower resolution, and probably therefore come from VIR. The temperature plot also seems like VIR (you can get a temperature if you have a spectrum). Finally, about processing, I notice that the images in the press release seem to have a lot of pixels. This is a clue that they might have composited a lot of individual images to get this one image. If so, there had to be a geometric correction (at the very least) because of course the camera is in motion while the images are taken. This requires very good geometry information. Additionally, there might have been a projection into a standard coordinate system (so that "north" is up). You might be interested to know that there is controversy about asteroid coordinate systems (for Vesta, https://en.wikipedia.org/wiki/4_Vesta#Coordinate_systems https://en.wikipedia.org/wiki/4_Vesta#Coordinate_systems). About the white areas -- yes, probably they saturated the detector.
CamperBob2 11y ago
In infrared the spots are darker or equal than surroundings, thus cooler or equal. We can't really say this, though, without knowing the material's emissivity. And since we don't know what it is...
- mturmon 11y ago
  This is true in the abstract, and a good point. But in this case, they have a pretty densely sampled and well-calibrated spectrum from one of the instruments (http://dawn.jpl.nasa.gov/multimedia/vir_gallery.asp http://dawn.jpl.nasa.gov/multimedia/vir_gallery.asp -- there appear to be ~100 spectral samples), so they would be able to make a black-body assumption, derive a temperature, and then validate the extent to which that is correct. In other words, it's not just a plain IR image, it's a whole spectrum.

codezero 11y ago

geuis's comment pulled from reddit sounds right. I used to work with satellite data that looked at visible light, but my dataset was close to Earth. I can imagine that the relative brightness of the surface of Ceres is very very low compared to what I've worked with, so keeping that in mind, any spot that spikes above some threshold is going to saturate your image. My data set was in black and white and had an integer brightness range of 32767. The raw data hovered between 20000 and 32767 (unitless intensity) because the instrument I worked on was old, and had suffered degradation, leaving a stronger "dark current" that needed to be removed. General steps for processing involved: 1) subtracting noise from the raw data, which came via a number of different paths 2) median filtering raw image frames into a single composite image, this may involve some masking 3) removing large scale noise (in our case, stars, planets and the moon) 4) adjusting the data range for presentation (this is only done for looking at the images, not for actual processing) The last step is probably what's causing this bright spot to seem so bright. With a different range of values, Ceres would look like a dark dark grey disc, with a lighter grey spot where the very bright spot is visible. I'd like to know the actual surface brightness of that spot, if it looks to be the same brightness as the Sun's reflection off of ice at that distance, it would clear things up pretty fast. Ice could be subsurface ice, or it could be remnant from the actual impacting object, if it were a comet.

Mithaldu 11y ago
Would ice exhibit the property of reflecting from any angle as seen in this gif? https://en.wikipedia.org/wiki/File:Ceres_spots_animation_May_4_2015.gif https://en.wikipedia.org/wiki/File:Ceres_spots_animation_May...
- Symmetry 11y ago
  If it's not smooth ice sure. If you look at an iceburg, for instance, it'll be white and be equally bright from any angle.