Analysis of fingertip/textile friction-induced vibration by time-frequency method

Tactile textures of textiles depend on the non-linear friction vibrations which are generated by the fingertip sliding across textile surfaces. It is relatively difficult to understand these complex vibrations, since skin and textile are viscoelastic and their vibration spectra are too redundant. Currently, the method of handling such complex vibrations in the field of tactile evaluation and tactile rendering usually adopts Fourier analysis. Unfortunately, only Fourier analysis can neither trace the multi-scales surface textures nor delete the redundant information. This paper proposed a time-frequency analysis, which extends the recorded 1-D vibration signals to 2-D time-frequency spaces to realize the multi-scales decomposition and dimension reduction. By applying this method to four typical kinds of texture surfaces, such as grille and textiles, the results demonstrated that the time-frequency analysis can accurately capture the major textural features from friction-induced vibration signals and decrease the dimensionality of complex signals. Considering the merits of dimension reduction, the time-frequency analysis could use in the texture synthetic of tactile virtual rendering and the tactile design of textile products.