Hi-index | 2.89 |
This paper applies multi-fractal spectrum theory into surface roughness evaluation in order to eliminate its heavy dependence on sampling position and sampling size. Based on AFM images of various substrates, their partition functions and multi-fractal spectra are calculated. The calculation indicates that the multi-fractal spectra differ significantly from each other for different samples, but are quite similar for the same sample in spite of the random change of sampling position and size, so it can reflect surface feature more objectively and accurately.