Touch Sensing Device Using FFT and Convolution for Noise Rejection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing touch sensing technologies face challenges in accurately distinguishing between touch inputs and noise components, particularly in noisy environments, which can lead to interference and reduced sensitivity.
Innovation Solution
A touch sensing method and device that utilize a convolution operation of driving signals with different frequencies and a mixing signal, followed by Fast Fourier Transform (FFT) processing to determine touch events by analyzing variations in frequency magnitudes, allowing for the separation of touch and noise components without additional noise processing steps.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional touch sensing methods are used in noisy environments, then noise components interfere with touch detection, but adding separate noise processing steps increases device complexity
Solution Approach 1:
The patent combines noise filtering and touch detection into a single FFT processing operation. By transforming the sensing signal to frequency domain and comparing it with the driving signal spectrum, the system simultaneously achieves noise rejection and touch event detection without requiring separate processing stages.
Solution Approach 2:
The patent introduces frequency domain analysis as an intermediary step between raw signal acquisition and touch detection. The FFT transformation acts as a mediator that separates touch-related frequency components from noise components, enabling accurate touch detection while simplifying the overall processing architecture.
2Reliability
If multiple processing steps are added to filter noise, then noise interference is reduced, but processing time increases
Solution Approach 1:
The patent maintains continuous touch detection capability by performing FFT processing on incoming sensing signals in real-time. The frequency domain comparison with driving signals enables continuous noise filtering and touch event identification without interrupting the sensing operation or requiring batch processing.
Solution Approach 2:
The patent replaces traditional time-domain filtering methods with frequency-domain analysis using FFT. This substitution enables more efficient noise rejection by directly targeting noise frequency components, reducing the computational time required compared to iterative time-domain filtering approaches.
3Measurement precision
If frequency domain analysis is used to separate touch and noise components, then detection accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent extracts only the essential frequency components from the sensing signal by performing FFT and comparing with pre-stored driving signal spectra. This extraction approach focuses computational effort on relevant frequency bands, reducing overall computational complexity while maintaining high detection accuracy.
Solution Approach 2:
The patent performs preliminary FFT transformation and frequency spectrum comparison before making touch detection decisions. By preparing the frequency domain representation in advance and comparing it with expected driving signal characteristics, the system achieves accurate component separation with optimized computational load during actual touch events.
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 enables accurate detection of touch inputs while minimizing noise interference, enhancing sensitivity and speed in touch sensing operations, even in environments with significant noise presence.
Implementation Method 1
a sensing signal sensed through a plurality of sensing electrodes is amplified
Implementation Method 2
the amplified sensing signal is band pass filtered
Implementation Method 3
through first half processing of the FFT on the analog-digital converted sensing signal, a real component and an imaginary component for each frequency of the sensed signal are obtained
Implementation Method 4
a deconvolution operation is performed on the real component and the imaginary component to reconstruct the frequency of the original driving signal
Data Source
AI summary
A touch sensing device includes a touch panel and a touch sensing controller. The touch panel includes driving electrodes and sensing electrodes. The touch sensing controller provides driving electrodes with driving signals having different frequencies from each other, respectively, and performs a fast Fourier transform (FFT) processing on the sensing signals sensed at each of the sensing electrodes to determine whether or not a touch is generated based on a variation amount between frequency magnitude of the sensing signal and frequency magnitude of the driving signal. The touch sensing controller performs a convolution operation of mixing the driving signal or the sensing signal and a mixing signal having a predetermined frequency, and separates the mixing signal from the sensing signal on which the convolution operation is performed to determine whether or not a touch is generated by reconstructing an original driving signal.