Audio Signal Spectral Estimation via Morphological Peak Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for estimating spectral information of audio signals in mobile communication systems fail to consider the energy values of small peaks, resulting in lost information.
Innovation Solution
The method employs morphological operations using a structuring set size (SSS) determined by pitch information, peak extraction techniques like hitting, mid-point, and pitch-based methods, and interpolation to identify true peaks spectra, ensuring accurate extraction of spectral information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If conventional peak estimation methods are used that focus on large peaks, then processing simplicity is maintained, but information of small peaks is lost resulting in incomplete spectral information
Solution Approach 1:
The patent segments the spectrum into multiple regions based on peak magnitude, categorizing peaks into large peaks and small peaks. This segmentation allows the system to apply different processing strategies to different peak types, ensuring that small peaks are not overlooked while maintaining manageable processing complexity through structured analysis.
Solution Approach 2:
The patent applies local quality by using different evaluation criteria for different regions of the spectrum. Large peaks are evaluated using traditional energy-based methods, while small peaks are evaluated using relative position and local energy characteristics. This localized approach ensures that each peak type is assessed with appropriate metrics, preserving spectral information without requiring uniformly complex processing across all frequencies.
2Measurement precision
If morphological operations with optimized SSS are applied, then noise peaks are removed and spectral estimation accuracy is improved, but processing complexity increases
Solution Approach 1:
The patent performs preliminary action by conducting morphological operations and peak extraction before final spectral estimation. The structuring set size is predetermined based on pitch information, and peaks are pre-identified and categorized. This preliminary processing organizes the data in advance, enabling more accurate spectral estimation while managing complexity through structured pre-processing steps.
Solution Approach 2:
The patent introduces an intermediary structure through the use of morphological operations with a specifically designed structuring set. This intermediary processing step acts as a mediator between raw spectral data and final spectral estimation, filtering noise peaks while preserving meaningful spectral information. The structuring set serves as an intermediary tool that bridges the gap between complex raw data and accurate spectral analysis.
3Productivity
If pitch-based SSS determination is used, then processing efficiency is improved and small peaks are preserved, but complexity increases compared to fixed SSS methods
Solution Approach 1:
The patent applies dynamics by making the structuring set size adaptive rather than fixed. The SSS is dynamically determined based on pitch information extracted from the audio signal. This dynamic adjustment allows the processing efficiency to vary according to the characteristics of the input signal, improving performance for different types of audio while managing complexity through adaptive rather than static processing parameters.
Data Source
AI summary
An apparatus and method for estimating audio signal spectrum information. The method including the steps of performing a morphological operation on a received audio signal, extracting peaks by using various peak extraction methods and extracting a remainder signal region from the extracted peaks, selecting a high-order peaks spectrum from the extracted remainder signal region. In addition, spectral envelopes are detected by performing an interpolation operation on the high-order peaks spectrum.