RC-Astro Processing Console
This is a program that provides some new image processing functions to
MaxIm. In addition, it allows some of the native MaxIm image processing functions
that operate on only one image to operate on multiple images in a batch fashion.
Click one of the following to jump to that section, or just read through
the page:
New Image Processing
Functions
MaxIm Functions
that are Automated
- Hot Pixel Filter
- Dead Pixel Filter
- Double Size
- Crop Images
- Remove Bad Pixels
Price
For the time being, you may download this program and use its full functionality
free of charge. If you do find it useful, feel free to send me some bucks
to help defray the cost of development.
Contact me
for the mailing address. To make a contribution online via PayPal, click the
following button:
Download/Installation
Instructions
- If you do not have Microsoft's .NET framework version 1.1 installed,
get it here.
- Click here to download the
RCConsoleSetup.zip file.
- Unzip the contents and save them in a convenient directory.
- Run the Setup.exe program and follow the installation instructions.
Make sure the installation directory is where you have MaxIm installed.
The plug-in files must be in the same directory as MaxIm_DL.exe.
- That's it... it should now appear as "RC-Astro Processing Console" in
MaxIm's Plug-in menu.
If you ever want to un-install the plug-in, don't use MaxIm's remove plug-in
function. Instead, use the Add/Remove Programs function in the Control Panel,
the same way you would remove a normal application.
Operation
RCConsole's user interface is hopefully very intuitive. The images you
want to process must first be open in MaxIm. Start the plug-in by selecting
it's entry in MaxIm's plug-in menu. Choose which images you want to process
using the selection list buttons, and then click which processing operation
you want to run. Set any options needed, and then click the OK button. When
you are finished processing the images, click the Exit button on the main
dialog form.
Here are the details on the new functions:
Sigma-Reject Combine
This is very similar in operation to the original
SigmaReject Combine plug-in.
The most significant change is the addition of an option to compute the
average time of exposure of the set of images being combined. This feature
inspects the DATE-OBS and TIME-OBS keys in the FITS headers of each source
image. From these it computes an effective mid-point date and time of the
whole set of exposures, and places this information in the DATE-OBS and TIME-OBS
keys of the combined image. Since these keys are by default supposed to refer
to the start time of the exposure (according to the
FITS standard),
comment fields are added to explain the new meaning.
Other enhancements include better defaulting of the options based on what
kind of images are being combined.
NOTE: A known problem exists if you try to close MaxIm without first
closing the combined image. Make sure you save and/or close it before you
attempt to exit MaxIm, otherwise you may cause trouble with future operation
of MaxIm. If you do accidentally do this and notice trouble, you will need
to re-boot your computer.
Normalize Histograms
This function attempts to make the histograms of the selected images match
each other. First, the background levels of the images are equalized (set
to the same value as the first image). Then, the pixel values of each image
(other than the first one) are scaled in an attempt to make its total brightness
range equal to that of the first image.
This function is useful prior to certain kinds of combination operations
to obtain best results, and potentially has other uses. This function is the
same as is optionally implemented in the Sigma-Reject Combine function.
Sigma-Reject Noise
Filter
This uses the sigma-reject statistical technique to try to repair pixels
in an image. Each pixel in the image is evaluated in the context of its neighbors.
If it is too bright or too dim, according to the threshold set in the dialog
box, the pixel value is replaced with a sigma-rejected mean of its neighbors.
The "kernel size" parameter tells the filter how many pixels around a given
pixel should be considered its neighbors.
This filter has a couple of uses. First, it is an excellent hot and dead
pixel filter. For this purpose, it should be applied using a 3x3 kernel to
individual calibrated images prior to alignment or combination.
It can also be used as a noise filter of sorts. If dim portions of an image
are noisy, this filter can have the effect of reducing the noise somewhat
without the loss of contrast that a blurring function would cause. Try different
kernel sizes and rejection thresholds, but be aware that too large a kernel
size will begin to treat the centers of stars as pixels that need repairing.
Anisotropic Gaussian
Blur
Anisotropic is a fancy word that simply means "not the same in all
directions."
A Gaussian blur filter is commonly used in image processing to reduce noise
in an image. It is very effective at this, but unfortunately it also causes
a loss of contrast.
A Gaussian blur filter modifies each pixel by looking at its neighbors
and computing a "weighted average" of their values, with more weight being
given to closer neighbors. The standard Gaussian blur filter found in most
image processing programs is isotropic: it blends pixels values equally
in all directions. The neighbors to the north and south of a given pixel have
just as much weight in the blurring operation as the neighbors to the east
and west.
What this new filter tries to do is to impart some directionality to the
blurring action, based on the characteristics of the image in the vicinity
of the pixel being processed. In this way, blurring is effectively accomplished,
reducing the noise level in the image, but fine details in the image tend
to be preserved, much more so than with a standard Gaussian blur filter.
Try different kernel settings to see which works best on your image. Larger
settings cause more blurring.
I find this filter particularly useful for suppressing the noise in an
image that has had its histogram pushed hard or has been deconvolved or otherwise
sharpened.
Repair Unreliable
Pixels
This function uses a map of "unreliable pixels" created when dark exposures
are combined to form the master dark frame. If you notice black pixels in
your calibrated object exposures, you may be the victim of "unreliable pixels."
Get the full scoop on this phenomenon, as well as how to use RC Console to
correct for it, here.
Enjoy!
Please send any suggested improvements or bug reports to
me by email.
Version History
- 1.6
- made dialog boxes remember user-entered settings
- added automation of MaxIm's "remove bad pixels" function
- fixed unreliable pixel fix crash with low threshold settings
- fixed anisotropic Gaussian blur crash with large kernel settings
- 1.3 -- beta release -- 29 December 2003
- added Fix Unreliable Pixels function
- added Crop Images automation
- 1.2 -- beta release -- 6 August 2003
- simplified installation (registers itself in MaxIm's plug-in menu)
- 1.1 -- alpha release -- 4 August 2003
- Converted to a plug-in.
- Implemented Sigma-Reject combine and Normalize Histograms in C++ for
faster operation.
- 1.0 -- alpha release -- 31 July 2003
Known Problems
Here is a list of currently known problems with the plug-in. These will
be addressed in future versions.
- Automation of "remove bad pixels" function requires that the map name
be set to one of MaxIm's default map names (i.e., "Map1" through "Map5"),
even if the real map name is different.
- If the plug-in is invoked after a "split color" command, two of the
three split images will disappear from MaxIm.
Disclaimer
I've made every effort to make sure this program operates properly and
does not cause loss of data or other such trouble on the host computer. That
being said, it could conceivably cause just that. By downloading, installing,
and running it, you are assuming responsibility for any consequences thereof. |