Printable Version of Topic

Posted by: um3k Aug 16 2005, 07:33 PM

I'm not sure if this is the right forum...

How can I do calibration (radiometric and geometric) of Voyager images? Does anyone know?

Posted by: Malmer Sep 21 2005, 12:43 PM

I have done some simple work with voyager images. I mainly did darkcurrent subtraction and flatfielding. I found a software that did geometric correction but it wasnt very good so i only used it as a reference. (it was only 8 bit...) That was enough for me to be able to do what i wanted. (im not a scientist, I just make pretty pictures)

I dont have any camera response curves but you can make one if you find images of the same target but with different exposure settings. Just reproject the images to a common viewpoint and then load them into hdrshop and let it estimate a camera response curve. Then you can linearize the pictures as well.

Mattias

Posted by: edstrick Sep 22 2005, 10:20 AM

The Voyager datasets contain "CAL" or some such image sets. Most of them are either dark frames of various sorts or flatfield images.

Voyagers had a flat sandblasted aluminum plate mounted on the spacecraft that could be viewed by the remote sensing instruments, primarily the cameras and the infrared radiometer/spectrometer (they measured the plate's temp for thermal cal, I think)

They'd orient spacecraft so sunlight would fall on the plate and then they'd take image series of the plate through different color filters with different exposures from very short to overexposed.

Dark field images depended on camera, whether lightflood was on or off, on scan readout rate and on whether the dataprocessor holds his mouth crooked. I experimented with Voyager 1 Titan wide angle images and found that I could achieve almost perfect dark field correction including the corners of the image, but I had to have a matching set of dark field images to cancel shading in the data image. Some image sets that were *SUPPOSED* to be equivalent clearly are not. There's some other camera parameters I can't identify that causes repeatable dark field shading across the image.

After cleaning dark fields from the Titan stuff, I had pretty clean data, but I had little luck working with the flatfield images. I needed to register then better, deal with dropped lines and bit-hit noisy pixels, saturated areas in partially overexposed images etc etc etc. My best flatfield corrected images were worse than just dark correction. I can do image processing in 32 bit real numbers in multi-frame image "cubes" up to 2 gigabytes of data, but some of the commands I have available are uselessly rudimentary for parts of the task.

Posted by: Bjorn Jonsson Sep 25 2005, 04:45 PM

There is some interesting information on this topic here (part of the ISIS documentation):

http://isis.astrogeology.usgs.gov/Isis2/isis-bin/pdfs2.cgi?voycal

In particular, the formula for linearity correction appears on this page.

Posted by: Tayfun Öner Sep 26 2005, 05:21 PM

Hi, don't forget that Voyager cameras are vidicon tubes and have very large geometric distortions, especially near the edges. These distortions are different in every frame and are affected by image content, electromagnetic environment it was taken, etc. Using reseaux, these distortions are calculated and corrected for each frame. Without correcting this geometric distortions it is not possible to make good mosaics.

But if the image is uniform and does not have many regions which are very different in brightness you may use a single matrix for correction, one for each of the four Voyager cameras, which I can send you if you wish.