Breath Sound Detection Using Zero-Crossing Interval Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional breath blowing determination methods require significant memory capacity to store various sound patterns and incur high processing loads due to the need for accurate recognition of breath sounds, which affects the success rate and efficiency of the recognition process.
Innovation Solution
A method that uses a processor to detect sound inputs, measure times between zero crossings in the waveform data, and determine whether the sound is made by breath based on characteristic time intervals and frequency distributions, reducing the need for extensive memory storage and complex calculations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sound elements forming breath sounds are stored in advance in memory for comparison, then breath sound recognition accuracy is improved, but memory capacity increases
Solution Approach 1:
The invention extracts only the essential characteristic (zero-crossing time intervals) from the complex breath sound waveform, storing only these interval patterns in memory rather than complete sound elements. This extraction approach maintains recognition accuracy while significantly reducing the memory data volume.
Solution Approach 2:
The invention transforms the breath sound recognition problem from comparing complete waveform patterns to comparing zero-crossing time interval sequences. By changing the parameter from full waveform to interval sequence, the memory storage requirement is reduced while preserving the essential recognition characteristics.
2Reliability
If various patterns of sound elements are stored in memory to recognize different waveform patterns, then breath sound recognition success rate is improved, but memory capacity increases
Solution Approach 1:
The invention segments the breath sound waveform into discrete zero-crossing intervals, creating a sequence of time measurements. This segmentation transforms continuous waveform variations into discrete comparable units, allowing different breath patterns to be recognized through interval sequence comparison rather than storing complete waveform variations.
Solution Approach 2:
By changing the representation parameter from complete waveform patterns to zero-crossing time interval sequences, the system can recognize various breath patterns with minimal stored reference data, improving reliability without increasing memory capacity.
3Measurement precision
If fast Fourier transform is used to determine frequency band distribution for breath recognition, then breath detection precision is improved, but processing load increases
Solution Approach 1:
The invention extracts only the zero-crossing time interval information from the sound waveform, discarding the need for complete frequency spectrum analysis. This extraction provides sufficient breath detection precision while avoiding the computationally intensive FFT process.
Solution Approach 2:
The invention replaces the complex mathematical FFT processing with a simpler time-domain zero-crossing detection method. This substitution maintains adequate detection precision while dramatically reducing computational complexity and processing load.
Data Source
AI summary
A game apparatus includes an operating switch and a microphone. A player operates a player object through intuition by the operating switch or inputting a sound. The number of zero crossings contained in waveform of a sound input through the microphone is detected, and also individual interval times between the zero crossings are detected. Then, it is determined whether or not the distribution of the interval times, i.e. the frequency distribution matches the distribution of interval times (frequency distribution) related to a breath sound stored in advance. If there is a match between the two, the input sound is recognized as a breath sound, and a game process based on the breath (wind) is carried out. For example, a game screen depicting the breath or wind is displayed on an LCD.