Noise Suppression Device Harmonic Structure Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional noise suppression methods degrade voice quality by excessively suppressing low-frequency regions, especially in high-noise environments, and require significant computational and memory resources.
Innovation Solution
A noise suppression device that analyzes the harmonic structure of input signals, calculates weighting coefficients based on voice/noise determination and signal information, and adjusts suppression coefficients to maintain the harmonic structure of voice, avoiding excessive suppression.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If conventional noise suppression methods use SN ratio-based suppression amounts, then noise suppression is achieved, but voice quality degrades due to excessive suppression of low frequency regions
Solution Approach 1:
The patent applies local quality by differentiating suppression strategies for different frequency regions. Voice components are identified through harmonic structure analysis (detecting periodicity and harmonic relationships), and suppression is applied selectively: strong suppression for noise-dominated regions, gentle or no suppression for voice-containing regions. This resolves the contradiction by maintaining voice quality in low frequency regions while still suppressing noise in other regions.
Solution Approach 2:
The patent changes the parameter used for suppression decision from simple SN ratio to a composite parameter incorporating harmonic structure analysis (periodicity detection, harmonic component identification). This parameter change enables the system to distinguish voice from noise more accurately, preventing excessive suppression of voice components while maintaining noise suppression effectiveness.
2Manufacturing precision
If low frequency region signals are generated and recovered using harmonic analysis, then voice quality improves, but computational complexity and memory requirements increase significantly
Solution Approach 1:
The patent extracts only the essential voice characteristics (harmonic structure, periodicity) from the input signal using efficient spectral analysis. Instead of generating and processing complete low frequency signals as in conventional methods, it extracts voice presence indicators and applies them to control suppression amounts. This extraction approach maintains voice quality improvement while significantly reducing computational and memory requirements.
Solution Approach 2:
The patent performs preliminary harmonic structure analysis and voice/noise determination before applying noise suppression. By pre-identifying voice components through periodicity detection and harmonic relationship analysis, the system prepares suppression control parameters in advance, avoiding the need for complex real-time signal generation and filtration processing required by conventional methods.
3Object-affected harmful factors
If suppression amounts are calculated using power spectra ratios, then noise suppression is achieved, but negative suppression amounts cause incorrect suppression and voice degradation
Solution Approach 1:
The patent applies preliminary anti-action by using harmonic structure analysis to pre-identify and protect voice components before noise suppression is applied. Through periodicity detection and harmonic component identification, the system anticipates where voice signals exist and pre-determines appropriate suppression amounts, preventing the occurrence of negative suppression amounts that would cause voice degradation.
Solution Approach 2:
The patent incorporates feedback mechanisms where the detected harmonic structure and periodicity information continuously inform the suppression amount calculation. The system monitors spectral characteristics, identifies harmonic relationships, and adjusts suppression amounts based on this feedback, ensuring suppression amounts remain positive and appropriate for the actual signal content, thereby improving reliability.
Data Source
AI summary
A noise suppression device includes: a power spectrum calculator converting an input signal of time domain into power spectra of frequency domain; a voice/noise determination unit determining whether the power spectra indicate voice or noise; a noise spectrum estimation unit estimating noise spectra of the power spectra; a period component estimation unit analyzing a harmonic structure constituting the power spectra and estimating periodical information about the power spectra; a weighting coefficient calculator calculating a weighting coefficient for weighting the power spectra; a suppression coefficient calculator calculating a suppression coefficient for suppressing noise included in the power spectra; a spectrum suppression unit suppressing amplitude of the power spectra in accordance with the suppression coefficient; and an inverse Fourier transformer converting the power spectra output by the spectrum suppression unit into a signal of time domain to generate a noise-suppressed signal.