Discussion of Color Imaging, Image Corrections and Sky Noise


In order to create a color image using CCD technology, we have to digitally combine the 3 images of the object taken through the 3 different filters. Generating a 'True Color' image of the greatest accuracy requires that these 3 images should be as equal as possible. In the process of acquiring each of these images, the following factors affected the quality of each image: different sky conditions, different filter sensitivities, and camera response to different wavelengths. The next few activities guide you through the process of making these images equal to one another, such that the negative effects listed above are corrected from the images. This is analogous to controlling variables in the traditional scientific method.


CCD cameras of the type used in astronomy can only record simple black and white digital images. In order to generate an accurate, 'true' color image, it is necessary to begin with three images of the same object. Each image is done through a different color filter, either red, green (visual), and blue. These images are then color-coded and digitally combined using image processing software such as HOU-IDL. This is why you requested images done through 3 filters in the previous activity.

In order to combine these images in the most scientifically accurate way possible, you will be asked to make several corrections to the images in order to make up for inadequacies in the image gathering process. One necessary correction is to subtract sky noise from the image. Background sky 'noise' adds clutter to telescopic CCD images in an equivalent manner to static on a radio. While the image was being created this sky noise was ADDED to the image, so it will be necessary to SUBTRACT the noise from the images. Each raw image will be processed to have its sky noise removed. This is necessary because each image will have different amounts of sky noise.


The procedure for this activity is as follows:

1. Open your 3 raw color filter images in HOU-IDL. Notice the black areas on the images. Here the counts per pixel should be nearly zero.

1. Move your cursor over the red-filtered image in the black areas that appear to be empty space. Watch your count values. Likely you will notice the counts are not near zero as one might expect. Confirm this for the other two images.

2. Use the HOU-IDL tool 'Analysis Tools: Sky' in order to determine the amount of sky noise in your red image.

The results window will display 3 values for this operation:

sky calculation of (image name) : sky mode : (#) sky standard deviation : (#)

The sky mode value indicates the average background sky noise across your red image.

3. Subtract the sky value from the red image

a. Use 'Data Manipulation: Subtract'

b. Choose to subtract 'Number of Counts'

c. Leave checked the box labeled 'Display results in new window'

4. Save your sky corrected red image with a new name indicating the sky noise has been subtracted. For example 'NGC 2419 Red Sky.fts'

5. Repeat for the other two images.


None or clarify students' understanding of the value of correcting these images.