E.V. Egorova - Ph. D. (Eng.), Associate Professor, Department «Telecommunication systems», MIREA. E-mail: calipso@dubki.ru

Introduced the concept of a generalized linear filtering and signal processing that can be represented as a product or convolution of the component signals. Application of the approach is based on a generalization of the principle of superposition, which allows it to apply swarm for a wide class of nonlinear systems. Revealed that the transmission coefficient filters equation equalization energy spectra more than the Wiener filter, but does not lead to an increase in noise, both in the reverse filter, the reconstructed image contains a large number of small parts associated with high-frequency component of the that the Wiener filter usually suppresses. The noise level increases, but under normal conditions the operator agrees to put up with the increase in noise when it receives additional useful information about fine structure of the image. Respectively given cases of homomorphic filters for dynamic range compression and increase in image contrast.