Figure 6
From: Texture recognition based on multi-sensory integration of proprioceptive and tactile signals
The illustration of the experimental setup comprises a robotic system with three degrees of freedom, three tactile sensors, and naturalistic textures. (a) Schematic of all preprocessing stages and converting recorded data to the appropriate input current to be applied to the Izhikevich neural model to produce spike trains. (b) Ten different naturalistic fine textures. (c) The experimental setup with the custom-built robot and sensors. A hydro-elastomer sensor to mimic muscle spindle (a proprioception sensor) for encoding the speed and position of the biomimetic finger. An 8 × 8 pressure tactile and piezoelectric sensors were embedded within a soft biomimetic fingertip to encode texture details during manipulation. (d) Scheme of the interface circuit for data collection.
