Signal Processing Method for Noise Suppression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing noise suppression technologies struggle to effectively remove noise from signals when the noise is not significantly smaller than the target signal, leading to poor noise estimate accuracy and increased distortion in the enhanced signal, and require extensive memory storage for various noise types.
Innovation Solution
A signal processing method that generates noise information based on the noise suppression result and uses this information to suppress noise in the degraded signal, allowing for effective noise suppression without pre-storing noise characteristics, using a noise suppressor and noise information generation unit within a signal processing apparatus.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If noise characteristic information is stored in advance to suppress noise effectively, then noise suppression effect is improved, but memory size and manufacturing cost increase
Solution Approach 1:
The system generates noise information autonomously by analyzing the degraded signal itself, rather than relying on pre-stored noise characteristics. The noise information generation unit processes the degraded signal to extract noise characteristics in real-time, allowing the system to adapt to unknown noise types without requiring extensive pre-programmed noise databases, thus reducing memory requirements while maintaining effective noise suppression
Solution Approach 2:
The invention changes the approach from using fixed pre-stored noise characteristics to dynamically generating noise information parameters from the actual degraded signal. By extracting noise characteristics directly from the signal during processing, the system adapts to varying noise conditions without needing large memory stores of predetermined noise patterns, resolving the contradiction between suppression effectiveness and memory capacity
2Reliability
If pre-stored noise information is used for noise suppression, then known noise types can be suppressed, but unknown noise different from stored information cannot be suppressed
Solution Approach 1:
The noise information generation unit analyzes the degraded signal directly to extract noise characteristics, enabling the system to identify and suppress unknown noise types that were not present in pre-stored databases. This self-analyzing capability allows the system to adapt to any noise type encountered, significantly improving versatility while maintaining suppression accuracy through real-time noise characteristic extraction
Solution Approach 2:
The system transitions from static pre-stored noise information to dynamic real-time noise characteristic generation. By continuously analyzing the degraded signal and updating noise information based on actual noise present in the signal, the system becomes adaptable to changing and unknown noise conditions, resolving the limitation of fixed noise type coverage
3Device complexity
If noise estimation is performed on degraded signal alone, then processing is simplified, but noise estimate accuracy deteriorates when noise is not much smaller than target signal
Solution Approach 1:
The invention introduces an intermediary approach where noise information is generated as an intermediate representation from the degraded signal. Rather than directly estimating noise from the noisy signal alone, the system creates a dedicated noise information structure that captures noise characteristics, allowing for more accurate noise estimation while maintaining processing efficiency through structured analysis of the degraded signal components
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Provided is a noise suppressing technology capable of suppressing various noises including unknown noises without storing information relating to a large number of noises in advance. Noises in a degraded signal are suppressed and noise information is generated on the basis of a noise suppression result. The noises in the degraded signal are suppressed using the generated noise information.