Abstract: As a nonlinear filtering tool, morphological filtering method has good characteristics such as noise suppressing and low-pass characteristics. Accurate description of the mathematical morphological filter’s characteristics supplies the theoretical basis and prerequisite for its application. The frequency response characteristic of the combined morphological filter is presented by excitation signal inputted into the given morphological filter, and the normalized cutoff frequency table with structure elements of different width is given. The research indicates when the sampling frequency is fixed, as the width of the structure element increases, the cutoff frequency decreases. The cutoff frequency gets smaller as the width of structure element grows, and the relationship between them is nonlinear. When the width of the structure element is fixed, as the sampling frequency increases, the cutoff frequency varies directly with the sampling frequency. The simulation result shows that the morphological filter has better noise suppressing and phase keeping ability. Finally, through processing the fault signal by the morphological filter, the effectiveness of the conclusions is verified. The results supply the theoretical basis for the application of mathematical morphology in vibration signal processing.

Key words: Cutoff frequency, Filter, Frequency response, Mathematical morphology, Vibration