Adaptive filter theory (2nd ed.)
Adaptive filter theory (2nd ed.)
| Hi-index | 0.00 |
The authors present a preprocessor to mitigate the effects of the impulse noise. In contrast with previous work, the proposed preprocessor requires no information about the correlated background process and can adapt as the background process changes. An adaptive nonlinear filter (preprocessor) introduced mitigates the effects of the impulses when the nominal process (background noise) is a correlated Gaussian process. The preprocessing procedure consists of identifying the impulsive component, eliminating it and replacing the samples with an estimate. Comparison of techniques for suppressing impulse noise is given.