In Digital speckle-pattern interferometry the interferograms are corrupted by strong speckle noise having high amplitudes. The simplest method of speckle reduction is two-dimensional lowpass filtration by spatial averaging. The implementation of such filters is simple and they run very fast, but they always result in image blurring, which affects the following segmentation procedure negatively if the fringes are dense. For the digital filtration of speckle noise in digital holografic and speckle pattern interferometry, we have achieved good results with a modified WIENER-filter to improve the filter process. The filter algorithm based on Fast Fourier Transformation (FFT) was implemented in a micro-computer system with a fast grafic memory. This method will be described and demonstrated by practicle examples.