Minimizing Image Distortion With Anti-Aliasing

During the digital image processing, anti-aliasing minimizes distorted artifacts. Anti-aliasing is usually required when a high-resolution image is being shown at a lower resolution than when it was originally taken. It is a process not restricted by photo processing, but it is also common in computer graphic design and audio production. Lines, bands or other rough noise in an image are all types of aliasing artifacts. These occur when the image is shrunk to a new size and can be cleaned up with anti-aliasing techniques.

The first type of anti-aliasing focuses on the idea that a perfect image has infinite details. It is shown as a function f(x,y) where x and y define coordinates of a given image. Since a computer can only display a finite number of pixels, the function is divided into f(x,y) and g(x,y) in order to compensate for the the display. The task at hand is to find the proper way to reduce the image in size without losing its quality via artifacts. You reduce both the f(x,y) at the overall level and then the g(x,y) at the individual pixel level to find the best contrast ratio to display the high-resolution image at the desired lower resolution.

Many image editing programs, such as Paint Shop Pro and Photoshop, allow you to reduce the image size automatically. The program compensates and uses said functions to automatically find the appropriate contrast ratio without creating aliasing artifacts. While you can also do this yourself, using programs like that make it easier to reduce photo size without damaging the photo's clarity in the process.

Another approach to anti-aliasing is signal processing. Each pixel coordinated is treated as an individual signal. Therefore, it assumes that the filtered signal being displayed is what the brain is perceiving to be the original quality image. As long as the brain has an idea of how clear the original image is, changing the dimensions and reducing it should still produce the same result, since the brain is already assuming it will be displayed as such.

The Fourier transform is the most accepted tool for tricking the brain in signal processing. It seeks to transform the image into a set of wave frequencies for processing in the equation of cos(2jπx)cos(2kπy). J and K serve as non-negative integers for the equation. They form the frequency of the wave, whereas J is the frequency in X's direction and K is the frequency in Y's direction. The Nyquist-Shannon theorem specifies that in order to find a unique signal of no more than N frequencies, you need at least 2N points. Since the eye sees more lower frequencies, we eliminate the higher frequencies and only keep the ones that we can see with the naked eye. This idea is an approximation, since it depends on the pair of eyes looking at given photo to determine the frequency at which it is viewed.

Practical anti-aliasing ideas create points, lines and triangles within 3D imaging such as software or hardware displays. This creates fuzzy edges by creating a form of low-level anti-aliasing. Certain shapes can cause this type of anti-aliasing to fail. Mipmapping, the approach used for texture mapping, creates a lower resolution version of a given texture map for display. When the image is rendered, the mipmap is chosen and filters the pixels out at a lower resolution on screen.

Full-scene anti-aliasing comes in handy when you have a high-end graphics card installed. Most of these cards support fullscreen anti-aliasing. This helps smooth out your operating system's graphic user interface as well as any tweaks you make to your monitor resolution to give you a clear, crisp picture across the whole display as opposed to focusing clarity on specific areas, such as the menus or graphical icons. This type of anti-aliasing, however, can cause problems when displaying single images that may not be meant to be completely free of artifacts at a lower resolution. Wallpapers and screensavers can be adversely affected by the use of full-scene anti-aliasing when they are not meant to be run or used at certain resolutions.

Object-based anti-aliasing created images based on the primitive shapes, mostly polygons, from a given image. These images are recreated and shaped based on their silhouettes of the objects in a photo. Opacity levels of the polygonal images change these edges and silhouettes until the image is clearer and crisper at a lower resolution. This technique was behind the original Microsoft Xbox graphic design to create a smooth movement-based image generation platform.

Anti-aliasing can be a very helpful tool in creating lower resolution images when what you begin to work with is at a higher resolution. Tinkering with your digital camera settings and learning how to use an image editing program can help you find the right compromise between image sizes. Only by diving in and editing your photos will you learn how to avoid image artifacts and clean up photos that might be affected by them.