Wavelet Coefficient Sharing for Signal Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for compressing and decompressing encoded signals in the wavelet domain are inefficient, leading to high memory requirements and potential loss of perceptual integrity in applications like audio and video coding.
Innovation Solution
A system that uses vector quantization to identify and share wavelet coefficients among multiple encoded signals, storing each sharable set only once and allowing for efficient compression while maintaining signal integrity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If wavelet coefficients are stored separately for each encoded signal, then signal integrity is maintained, but memory requirements increase significantly
Solution Approach 1:
The patent merges identical wavelet coefficient sets from multiple encoded signals into a single shared storage location. When multiple signals share the same wavelet coefficients at a given decomposition level, instead of storing duplicate copies, the system stores one copy and maintains references to it from multiple signal representations, thereby reducing memory usage while preserving signal integrity.
Solution Approach 2:
The patent creates a universal wavelet coefficient storage structure where a single set of coefficients can serve multiple encoded signals simultaneously. The system identifies coefficients that are common across different signals and makes them universally accessible, allowing the same coefficient set to be reused for reconstruction of multiple signals without degradation of quality.
2Quantity of substance
If wavelet coefficients are shared among multiple signals, then memory requirements are reduced, but system complexity increases
Solution Approach 1:
The patent performs preliminary identification and grouping of shared wavelet coefficients during the encoding phase. By pre-analyzing which coefficients are identical across multiple signals and organizing them into shared storage structures before decomposition, the system reduces the computational burden during playback and avoids complex real-time analysis.
Solution Approach 2:
The patent uses reference copying mechanisms where instead of storing actual coefficient data multiple times, the system creates lightweight reference pointers or indices that point to the shared coefficient storage location. This copying approach maintains data integrity while minimizing memory usage and simplifying the sharing infrastructure.
3Ease of manufacture
If traditional compression methods are used, then implementation is simple, but perceptual integrity is lost
Solution Approach 1:
The patent changes the fundamental parameter of how wavelet coefficients are organized and stored, transitioning from signal-specific storage to shared communal storage. This parameter change in the data structure organization enables both simple implementation (by reducing data redundancy) and preserved perceptual integrity (by maintaining exact coefficient values needed for perfect reconstruction).
Data Source
AI summary
A system for sharing wavelet domain components among encoded signals receives a set of signals decomposed and encoded according to a wavelet transform. The decomposed and encoded signals each include a set of wavelet coefficients at each level of the decomposition of the encoded signal. Using a vector quantization technique, the system identifies one or more sets of wavelet coefficients that are sharable among two or more of the decomposed and encoded signals at a particular level of decomposition. The system then stores the sets of wavelet coefficients of the decomposed and encoded signals. Each identified sharable set of wavelet coefficients at a particular level of decomposition is stored only once and shared by two or more of the decomposed and encoded signals.