The nonlinear Gaussian filter chain is a procedure which eliminates noisy fluctuations in signals without destroying the coarser signal structures. It works for signals of arbitrary dimensions. The signal passes a chain of nonlinear filters which are derived from linear Gaussian filters. The parameters of the first filter determine coarseness and contrast of the fluctuations which are to be eliminated; the parameters of the following filter steps have to be chosen in a simple prescribed manner in order to obtain optimal performance. The implementation is fairly easy. The algorithm does not use large numbers of iterations and is therefore numerically stable. Considerable speedup can be obtained by obvious parallelization into threads and some other simple modifications. The sequence of images below shows how the filter chain removes gradually the noise.

original image	output of stage 1	output of stage 2	output of stage 3	output of stage 4	output of stage 5