The methods described till this point worked on a single pixel intensity, and are called point processing methods. Sometimes it is helpful to look at the neighborhood of a pixel rather than only single pixel intensity. This are called neighborhood processing techniques. The neighborhood can be 3 x 3, 5 x 5, 7 x 7, and so on and are matrix-centered at a particular pixel. Image filtering is an important neighborhood processing technique.
Filtering is an important concept in signal processing where we reject a certain band of frequencies and allow a certain band of frequency to pass. How is frequency measured in images? If gray levels change slowly over a region, then it is a low-frequency region. If gray levels changes drastically, then it is a high-frequency region. Normally the background of an image is considered a low-frequency region and the...