Touch Sensing Device Noise Reduction via Bending Wave Fitting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining the location of an impact on a surface based on acoustic signals are hindered by spurious noise contributions during the calibration phase, making it a tedious and time-consuming process to achieve desired precision.
Innovation Solution
A method that involves fitting sensed signals with functions satisfying bending wave propagation properties, such as the Helmholtz equation and Bessel functions, to suppress noise and improve the signal-to-noise ratio, allowing for more precise calibration and accurate determination of acoustic signatures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If acoustic signals are measured using sensors to determine impact location, then the location identification is achieved, but noise contributions falsify the predetermined acoustic signatures during calibration
Solution Approach 1:
The patent transforms the raw acoustic signal data into the frequency domain using Fourier transform, changing the parameter representation from time-domain to frequency-domain. This allows for frequency-based filtering where noise components can be selectively removed while preserving the characteristic acoustic signatures of different impact locations.
Solution Approach 2:
The patent extracts and removes noise components from the acoustic signals by identifying frequency ranges that contain primarily noise rather than useful acoustic information. Through spectral analysis, noise frequencies are separated and eliminated, leaving only the relevant acoustic signature data for calibration.
2Measurement precision
If noise sources are identified and experimental set-up is adapted to reduce noise, then measurement precision improves, but the process becomes tedious and time consuming
Solution Approach 1:
The patent replaces the mechanical/manual process of noise identification and experimental setup adjustment with an automated signal processing system. The computer automatically performs Fourier transforms, identifies noise frequencies, and applies filtering algorithms, eliminating the need for manual intervention and significantly reducing calibration time.
Solution Approach 2:
The patent introduces an intermediary signal processing layer between the acoustic sensors and the location determination algorithm. This intermediary processing stage automatically handles noise reduction through frequency domain analysis, allowing the system to achieve high precision without requiring time-consuming manual optimization of the experimental setup.
3Measurement precision
If not all unwanted noise contributions can be identified, then noise reduction is incomplete, but more aggressive filtering may remove useful signal components
Solution Approach 1:
The patent employs feedback mechanisms where the processed acoustic signatures from calibration are used to verify the effectiveness of noise filtering. The system iteratively adjusts filtering parameters based on the quality of the resulting acoustic signatures, ensuring that noise is reduced while preserving the characteristic frequency patterns that identify different impact locations.
Solution Approach 2:
The patent applies partial filtering by selectively removing only specific frequency components that are identified as noise, rather than applying aggressive broad-spectrum filtering. This partial action approach removes sufficient noise to improve measurement precision while leaving the essential signal components intact for accurate location determination.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach effectively reduces noise contributions and enhances the precision of acoustic signal analysis, leading to improved calibration data and more accurate identification of touch events on a surface.
Implementation Method 1
fitting a function satisfying bending wave propagation properties to the sensed signals, thereby reducing noise contributions to the sensed signals not satisfying the bending wave propagation properties
Implementation Method 2
fitting a function satisfying the Helmholtz equation to the sensed signal
Data Source
AI summary
The present invention relates to a method for reducing noise in a signal sensing a bending wave propagating in an object comprising the steps of: receiving a sensed signal representative of a plurality of locations of the object, and fitting a function satisfying bending wave propagation properties to the sensed signals, thereby reducing noise contributions to the sensed signals not satisfying the bending wave propagation properties. By doing so, noise contributions not satisfying the wave propagation properties can be suppressed from the sensed signal, thereby improving the signal-to-noise ratio.