On methods for noise filtering of measurement signals

We shall address the problem of measurement noise in measurement systems. At first we shall present the most common types of measurement errors. We shall then give an overview of different digital signal processing methods for data preanalysis and for attenuating the noise level in the signals and also for identifying impulses and outliers in the signal. We shall also introduce how data mining data preparation processes may be adapted to the problem of noise attenuation. Next we shall present a few examples of the usage of the presented methods: attenuation of periodic noise, median filtering and detecting outliers in a noisy signal.