Wind Noise Detection via Autocorrelation Gradient Analysis

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods for detecting wind noise in electronic devices, such as mobile terminals and hearing aids, require signals from multiple microphones, increasing complexity and component count, whereas a method to detect wind noise from a single microphone is needed to simplify the process.

Innovation Solution

The method involves generating autocorrelation coefficients from a microphone signal, determining gradient values, and using these values to detect the presence of wind noise by analyzing their smoothness and threshold satisfaction, allowing for wind noise detection in a single microphone signal.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If signals from multiple microphones are used to detect wind noise, then detection reliability is improved, but device complexity and component count increase

Engineering Contradiction:

Improvewind noise detection reliabilityVSAvoidmicrophone array complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent extracts the wind noise detection capability from a multi-microphone system and implements it within a single microphone by analyzing autocorrelation coefficients of its own signal. This separates the detection function from the need for multiple sensors, reducing system complexity while maintaining detection reliability through mathematical analysis of the single microphone's signal characteristics

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent introduces autocorrelation coefficients as an intermediary mathematical construct that mediates between the raw microphone signal and wind noise detection. By computing autocorrelation coefficients and analyzing their gradient values, the system can detect wind noise characteristics without requiring multiple physical microphones, thus reducing hardware complexity while preserving detection capability

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS7305099B2Electronic devices, methods, and computer program products for detecting noise in a signal based on autocorrelation coefficient gradients

Publication Date: 2007.12.04 SONY GROUP CORP

AI summary

An electronic device can be operated to detect noise, such as wind noise. A microphone signal is generated by a microphone. Autocorrelation coefficients are determined based on the microphone signal. Gradient values are determined from the autocorrelation coefficients. The presence of a noise component in the microphone signal is determined based on the gradient values