Contour-Net: a model for tactile contour-tracing and shape-recognition

摘要: We propose Contour-Net as a bio-inspired model for rhythmic movement control of pair of insectoid feelers, able to successively sample the contour arbitrarily shaped objects. Initial object contact initiates smooth transition from large-amplitude, low-frequency searching behaviour local, small-amplitude and high-frequency sampling behaviour. Both behavioural states are defined by parameters Hopf Oscillator. Subsequent signals trigger 180 phaseforwarding oscillator, resulting in repeated object. The local effectively serves contour-tracing method with high robustness, even complicated shapes. Collected data points can be directly fed into an artificial neural network classify shape Given sufficiently large training data-set, tactile recognition achieved position, orientation size -invariant manner. Only minimal pre-processing (normalisation) is required.