Cometary Coma Image Enhancement Facility

This tool allows you to upload an image to our server and have it be enhanced by a certain enhancement technique. To learn about the different enhancement techniques available, or to download the enhancement programs for offline use, click here to access the help page.

This web Facility was developed for OSX by M. Patrick Martin, and ported to Linux by Gavin Nelson.
This web Facility was rebuilt by Pamela Gay in March 2024 to allow compatibility with Ubuntu and modern compilers.
Corresponding enhancement algorithms and codes were developed by Nalin Samarasinha and Steve Larson.

We appreciate your use of this Facility. Please cite this Facility (and its content) in the following way: Samarasinha, N. H., Martin, M. P., Larson, S. M, 2013. Cometary Coma Image Enhancement Facility, http://cie.psi.edu

First, choose the enhancement technique to apply:

Image to be enhanced:

Image to be enhanced:
The comet image to be enhanced. The image must be *.fits or *.fit.

(Maximum file size: 10.8 MB, Image filename: <40 characters)

(x,y) coordinates of the nucleus/optocenter:

(x,y) coordinates of the nucleus/optocenter:
Where to center the enhancement, usually the brightest part of the image. Round to one decimal place. Note that the origin pixel is (1,1) in this Facility, not (0,0) as it is in some image editing software.

(, )
Note that the origin pixel is (1,1).

Region to enhance:

Region to enhance:
If you wish to only enhance a certain section of the image, fill out this section.

X limits: ,

Y limits: ,

Number of pixels wanted in ρ (rho) axis:

Number of pixels wanted in ρ (rho) axis:
The division by azimuthal average routine converts the input image to (ρ, θ) polar coordinates. Please choose how many pixels are desired in the ρ-axis. The default value is set to the radius of the largest circle completely contained in the image.

Number of pixels wanted in θ axis:

Number of pixels wanted in θ axis:
The division by azimuthal average routine converts the input image to (ρ, θ) polar coordinates. Please choose how many pixels are desired in the θ axis. The default value is set to the circumference of a circle with the radius chosen in the previous box. To limit file size, however, this is capped at 2500 pixels.

How many standard deviations should be accepted
for pixels in the θ axis?

How many standard deviations should be accepted for pixels in the θ axis?
Image enhancement algorithms can be very sensitive to "dead pixels", bright stars, and other anomolies in the image. To account for this, if a pixel is an outlier from its neighbors, it is ignored. Please define outliers by choosing how many standard deviations from the mean pixel value are acceptable. A common choice is 3.

How many standard deviations from the mean should
the minimum and maximum pixel values be?

How many standard deviations from the mean should the minimum and maximum pixel values be?
In azimuthal renormalization pixel values are scaled to a new range determined by the other pixel values in the azimuth. Please define this range by choosing how many standard deviations from the mean pixel values should be within in the enhanced image. A common choice is 3.

Kernel A term:

Kernel A term
The kernel A term is the base distance between the pixel to be enhanced and the edge of its kernel. Increasing this value increases the size of the kernel near the nucleus/optocenter, without significantly impacting the kernel size further away from the nucleus/optocenter. See the explanations page for a more detailed explanation about the kernel. This value is usually on the order of 1.

Kernel A terms:

Kernel A terms
Choose two kernel A terms to be used in the mosaic. The kernel A term is the base distance between the pixel to be enhanced and the edge of its kernel. Increasing this value increases the size of the kernel near the nucleus/optocenter, without significantly impacting the kernel size further away from the nucleus/optocenter. See the explanations page for a more detailed explanation about the kernel. This value is usually on the order of 1.

Kernel B term:

Kernel B term
The kernel B term linearly scales the size of the kernel with the distance from the nucleus/optocenter. Increasing this value increases the size of the kernel across all distances from the nucleus/optocenter. See the explanations page for a more detailed explanation about the kernel. This value is usually on the order of 1.

Kernel B terms:

Kernel B terms
Choose two kernel B terms to be used in the mosaic. The kernel B term linearly scales the size of the kernel with the distance from the nucleus/optocenter. Increasing this value increases the size of the kernel across all distances from the nucleus/optocenter. See the explanations page for a more detailed explanation about the kernel. This value is usually on the order of 1.

Kernel N term:

Kernel N term
The kernel N term exponentially scales the size of the kernel with the distance from the nucleus/optocenter. Increasing this value increases the size of the kernel when it is far from the nucleus/optocenter, without significantly impacting the kernel size near the nucleus/optocenter. See the explanations page for a more detailed explanation about the kernel. This value is usually on the order of 0.1.

Kernel N terms:

Kernel N terms
Choose two kernel N terms to be used in the mosaic. The kernel N term exponentially scales the size of the kernel with the distance from the nucleus/optocenter. Increasing this value increases the size of the kernel when it is far from the nucleus/optocenter, without significantly impacting the kernel size near the nucleus/ optocenter. See the explanations page for a more detailed explanation about the kernel. This value is usually on the order of 0.1.

Transform the input image to log-10 space before enhancing?

Transform the input image to log-10 space before enhancing?
You have the option of transforming the image to a log-10 scale before the enhancement technique is performed. However, the presence of any negative pixels will prevent the transformation from taking place -- you can check the header of the enhanced image for the term "logarithmic image" to be sure the transformation was applied.

Yes No

Note: Because of how browsers display images, your images may not display optimally below. Please download the fits image and adjust the images' brightness, contrast, etc in an image viewer like ds9.
These images will only be saved on our servers for about one hour. Please click "Download" below any images you wish to save.

Unenhanced (input) image	Enhanced x,y image


Unenhanced ρ - θ image	Enhanced ρ - θ image

A=, B=, N=	A=, B=, N=


A=, B=, N=	A=, B=, N=


A=, B=, N=	A=, B=, N=


A=, B=, N=	A=, B=, N=