IMAGE PROCESSING
Contributions from David Dench, Dave Feldstein, Jim Ferreira and Jeff Mourron

.
 .
Jeff Mourron uses a Philips VCM 7310 color WebCam camera to captures raw images as seen at left.  The IR filter has been removed from the camera detector to increase detector sensitivity, however the images now records too red.  Too correct the excessive red in the images and to improve image contrast and sharpness, Jeff uses image processing software to separate the red, green and blue components of the image, correcting the red light image for excessive red.
.
Typically he separates the RGB components of 3 raw images, processes them for brightness/contrast, uses an unsharp mask, then combines the 3  red images, the 3 green images and the 3 blue images by stacking resulting in the B&W red light, green light and blue light images seen at left.
.
These images are then converted to the appropriate complimentary colors -- cyan, magenta & yellow, respectively, and combined to produce the full color image a left.  The image is now fully corrected for the excessive red and by stacking several images for each color component, contrast has been improved while image noise has been suppressed.
..
.
..
 
The Saturn image at upper left was made with a PC23C and a 6" aperture telescope, captured to computer with a Snappy frame grabber and processed with un-sharp mask, then adjusted for brightness & contrast in PhotoShop by Jim Ferreira.  Dave Feldstein further processed the image using the Richardson-Lucy routine. 
..
.
 
The Saturn image at left made by David Dench was further processed with 5 interations of Maximum Entropy Deconvolution ( MED ) software.  Dave points out that there is some evidence of the Gibbs Effect at the bright/dark transitions.  Two points that should be noted: 
1.)  8 bit images were used where typically no less than a 12 bit image would be used with this kind of processing. 
2.)  The software, made by one of Dave's students, may not be entirely adequate for the task.
.
.
Balance of contrast and sharpness.
Too heavy use of unsharp mask brings up noise.
Too much contrast, no gain in information.
 
The two Saturn images at right demonstrate too heavy use of un-sharp mask and brightness / contrast processing.  With video capture images bringing up edge sharpness and contrast also brings up image noise and artifacts in the video recording.  Pure white highlights in a too contrasty image tend to be distracting and degrade the aesthetic qualities of the image. 
..
.
Alpine Valley JPEG image made with a 25cm Newtonian using a PC23C camera recording to a Hi8 camcorder.  Image was captured to computer using a Snappy frame grabber as a TIFF file.  Image processing with Adobe Photoshop using unsharp mask and adjusting brightness and contrast. 
.
Same JPEG image processed by David Dench with the Richardson-Lucy routine in the Astro Art image processing program. 

Even just working with a JPEG file, the difference is dramatic with many peaks and craters more clearly defined.  Undoubtably better results would be had working with the original TIFF file.

.
.
 
David Moore used image processing program Megafix to rework a Jim Ferreira image of Saturn.  The 'before' image was processed with PhotoShop only using the unsharp mask routine and brightness/contrast controls.
.
.
 
Frame stacking and image processing
David Dench digitized images from a video of Jupiter made by Peter Drew using a 6 inch f/8 Newtonian, 2x Barlow and red filter.  The first image at left is a single frame captured from the VHS recording.  The second image is the result of stacking 8 frames similar to the first image.  The third is the the stacked image with brightness and contrast adjusted in Photoshop.  The last frame on the right is the result of stacking 16 frames and processing for brightness and contrast again with Photoshop
.
.
Back to VIDEOASTRO home