In this paper, we developed a fully textile sensing fabric for tactile touch sensing as the robot\nskin to detect human-robot interactions. The sensor covers a 20-by-20 cm2 area with 400 sensitive\npoints and samples at 50 Hz per point. We defined seven gestures which are inspired by the social\nand emotional interactions of typical people to people or pet scenarios. We conducted two groups of\nmutually blinded experiments, involving 29 participants in total. The data processing algorithm first\nreduces the spatial complexity to frame descriptors, and temporal features are calculated through\nbasic statistical representations and wavelet analysis. Various classifiers are evaluated and the feature\ncalculation algorithms are analyzed in details to determine each stage and segments� contribution.\nThe best performing feature-classifier combination can recognize the gestures with a 93.3% accuracy\nfrom a known group of participants, and 89.1% from strangers.
Loading....