US 2014/0071095 A1 · SIGNAL PROCESSING SYSTEMS

Abstract

This invention relates to methods, apparatus, and computer programme code for processing acoustic signal data to determine where an object has been tapped with a stylus, finger nail or the like. The method involved storing a set of labelled training data comprising digitised waveforms from a sensor for taps at a plurality of different locations. The labelled training data is then processed to determine mean value and covariance data for the waveforms, which is afterwards used in conjunction with a digitised waveform of a tap at an unknown location to identify the location of the tap. Preferably the covariance is decomposed into a plurality of basis functions for each region each with a respective weighting, which are used to represent captured data for an unknown tap and parameters of the representation are classified to locate the tap.

Claims26

Dependent claims

A method of detecting the location of a tap on an object having at least one acoustic sensor, the method comprising:

storing a set of labelled training data, said labelled training data comprising digitised waveform data of waveforms captured from said sensor for taps at a plurality of different locations on said object in combination with location data indicating for each said waveform the location of the tap;
processing said labelled training data to determine, for each of a plurality of tap-sensing regions of said object, mean value data and covariance data for at least two said waveforms captured from taps in the said region, said mean value data defining a mean of said at least two waveforms and said covariance data defining covariance of said at least two waveforms;
capturing tap data comprising a digitised waveform of a tap at an unknown location captured by said sensor;
and
determining a tapped region of said object from said tap data and from said mean value data and said covariance data from said plurality of regions.

2 more elements

A method of detecting the location of a tap on an object having at least one acoustic sensor, the method comprising:

storing a set of labelled training data, said labelled training data comprising digitised waveform data of waveforms captured from said sensor for taps at a plurality of different locations on said object in combination with location data indicating for each said waveform the location of the tap;
processing said labelled training data using a statistical model having a plurality of different parameters associated with each said region to determine from said waveforms, for each said region, statistical data representing the variability of said parameters resulting from tapping said object in said region;
capturing tap data comprising a digitised waveform of a tap at an unknown location captured by said sensor;
and
determining a tapped region of said object from said tap data and said statistical data.

2 more elements

A method of detecting a tap on an object having at least one acoustic/vibration sensor, the method comprising:

capturing tap data comprising a digitised waveform of a tap on said object captured by said sensor;
processing said captured tap data by applying a tap detection procedure to generate time series tap metric data comprising data providing a detection metric of a potential said tap at a succession of times;
determining a classifier to classify said detection metrics of potential said taps into one or more detected taps;
and
applying said classifier to said tap metric data to detect said tap on said object;
wherein said determining of said classifier comprises identifying a region of background noise in said captured tap data;
and
determining said classifier from data in said region of background noise.

5 more elements

Dependencies are inferred from the claim text.

Signal processing systems

Life of this patent

Signal processing systems

Life of this patent

Abstract

Claims26