Goals and prerequisites

The lesson deals with basic issues related to digital image filtering. Lesson 2 should be carefully studied before starting this lesson. Previous knowledge concerning one-dimensional filters - as presented in classical digital signal processing courses - is also helpful.

Filters

Filters are electrical circuits or software algorithms which change frequency characteristics of a given signal in a some previously designed way. Ideal filter transmits some part of frequencies with no distortions (pass band) while fully suppressing others (attenuation band). According to the passed band, filters are classified as low-pass, high-pass, band-pass or band-stop.

Filters are also classified according to its impulse response. There exist Finite and Infinite Impulse Response filters, called FIR and IIR respectively. IIR filters are realized as a circuits/algorithms with memories, where some returning paths from output to input exist. According to this fact, they are also called recursive filters. IIR filters are usually not used in digital image processing. FIR filters used in digital image processing exhibit relatively short impulse response and are realized by convolution, as described on next pages.