Authoring New Haptic Textures Based on Interpolation of Real Textures in Affective Space

Abstract

This paper presents a novel haptic texture authoring algorithm. The main goal of this algorithm is to synthesize new virtual textures by manipulating the affective properties of already existing real-life textures. To this end, two different spaces are established: two-dimensional (2-D) “affective space” built from a series of psychophysical experiments where real textures are arranged according to affective properties (hard-soft, rough-smooth) and 2-D “haptic model space” where real textures are placed based on features from tool-surface contact acceleration patterns (movement-velocity, normal-force). Another space, called “authoring space” is formed to merge the two spaces; correlating changes in affective properties of real-life textures to changes in actual haptic signals in haptic space. The authoring space is constructed such that features of the haptic model space that were highly correlated with affective space become axes of the space. As a result, new texture signals corresponding to any point in authoring space can be synthesized based on weighted interpolation of three nearest real surfaces in perceptually correct manner. The whole procedure including the selection of nearest surfaces, finding weights, and weighted interpolation of multiple texture signals are evaluated through a psychophysical experiment, demonstrating the competence of the approach. The results of evaluation experiment show an average normalized realism score of 94 $\%$ for all authored textures.