Astronomical image processing is a method of cleaning up images taken by space telescopes or highlighting elements of the images so that certain stellar features become more prominent. Image processing technology to do this involves both filters and other built-in telescope technology known as image preprocessing, and work on the imagery afterward using software to increase resolution of objects in space and sharpen other aspects of the image. While image editing varies depending on the focus of the research and what is desired for the end-result of the image, techniques involve several standard approaches.
Routine astronomical image processing first involves a series of foundational steps. Image calibration, alignment, and noise reduction are all important for many types of astronomical images. Calibration requires removing unwanted data or signal recordings from images as they are taken, so that what is being studied can be recorded more clearly.
Alignment and stacking of images atop one another with software by using fixed reference points can be used to increase the quality and density of image data. This involves processes such as that used by the US-based National Aeronautics and Space Administration (NASA) called the Drizzle technique, which works on images taken from the Hubble space telescope. The Drizzle technique sharpens images by stacking multiple samples atop one another to create a resolution with a density of pixels that is higher than any one image alone.
Image processing algorithms in software also facilitate noise reduction. Space-based images can have random noise from radiation effects or light reflections from Earth, and several methods are used to filter this out. A low pass method reduces high-frequency noise, where edge smoothing will eliminate aberrations in an image that look like the edge of objects, but are, in fact, just distortions.
Most astronomical photos are recorded in a series of gray tones using a Charge Coupled Device (CCD), which, nevertheless, contains color data embedded in the image. This necessitates the need for an astronomical image processing mechanism to focus the image on an area of interest. Image visualization techniques do this by employing a wide variety of filters to highlight certain areas of an image and minimize others. These include changing such elements in an image, like its luminance qualities, as well as filters for the primary colors of red, green, and blue light, for hydrogen gas effects in space, and more.
The image filtering used by astronomical image processing is tuned to specific wavelengths of light and usually designed to be broad-band or narrow-band in function. Broad-band filters allow many wavelengths of light to be recorded, such as all of the variations on one color of red in the visible spectrum. A narrow-band filter blocks all light except that of usually one characteristic wavelength that is filtered down to the level of a few nanometers or billionth's of a meter. When studying diverse regions of space such as galaxies, a broad-band filter is chosen, while specific stellar objects like planets, stars, or asteroids might instead be the focus of a particular narrow-band filter.
Many photos of objects in space have undergone a large amount of editing before they are released to the media after astronomical image processing. Since astronomical research works in detail with gray scale images, a true-color representation of the region of space is constructed after the fact by assigning colors based on the wavelengths of light in the image using software tools. As well, often public images can be composed of false colors that are chosen for their ability to enhance the aesthetic or sharp quality of objects in the image.
