Multiwavelet Transform for Digital Data Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current digital data compression methods, such as JPEG and MPEG, suffer from limitations like blocking artifacts, motion artifacts, and high complexity, especially in handling large image sizes and high-motion scenes, and lack flexible bit allocation for scalable compression.
Innovation Solution
A method using multiwavelet transformations to generate vector-valued datasets, which are then transformed into multiwavelet coefficients and entropy-coded, employing orthogonal, biorthogonal, or non-separable filters for efficient compression and decompression of digital data like images and videos.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If DCT-based compression (JPEG/MPEG) is used, then compression efficiency is achieved, but blocking artifacts and motion artifacts occur
Solution Approach 1:
The patent transitions from DCT basis functions to wavelet basis functions, changing the fundamental mathematical parameter of the transform. This allows for multi-resolution analysis where the image is decomposed into different frequency bands at multiple scales, eliminating the fixed-block processing that causes artifacts while maintaining compression efficiency.
Solution Approach 2:
The invention introduces the dimension of scale or resolution by applying wavelet transform at multiple levels. Instead of a single-frequency analysis like DCT, the wavelet transform analyzes the image at different scales (coarse to fine), allowing artifacts to be distributed and managed across multiple resolution levels rather than appearing as blocking artifacts in the original image space.
2Object-affected harmful factors
If MPEG compression with motion estimation is used, then video quality is improved, but computational complexity increases
Solution Approach 1:
The patent extracts the essential features of video content through wavelet decomposition, separating the video signal into different frequency subbands. This allows for efficient representation of motion and texture information without requiring complex motion estimation algorithms, thereby reducing computational complexity while maintaining video quality.
Solution Approach 2:
The wavelet transform naturally segments the video signal into different frequency bands and resolution levels. This segmentation allows for independent processing and compression of different subbands, reducing the overall computational complexity compared to full-frame motion estimation while preserving important visual information.
3Quantity of substance
If JPEG compression is used, then data size is reduced, but image quality deteriorates under high compression ratios
Solution Approach 1:
By changing from DCT to wavelet transform, the patent achieves better energy compaction properties. The wavelet transform concentrates image energy into fewer coefficients, especially at coarser scales, allowing for more efficient quantization and compression while preserving visual quality even at high compression ratios.
Solution Approach 2:
The multi-resolution nature of wavelet transform provides an additional dimension for quality control. Important visual information at coarse scales can be preserved with higher precision while less critical fine-scale details can be compressed more aggressively, maintaining perceived image quality while achieving high compression ratios.
4Adaptability or versatility
If MPEG scalable profiles are used, then spatial scalability is achieved, but bitrate overhead increases
Solution Approach 1:
The wavelet transform inherently segments the image into different resolution levels and frequency subbands. Each level can be independently encoded and transmitted, providing natural spatial scalability without requiring separate encoded versions of the image. This eliminates the bitrate overhead associated with transmitting multiple full-resolution images at different qualities.
Solution Approach 2:
The multi-resolution wavelet decomposition creates a nested structure where coarse-resolution approximations are embedded within fine-resolution details. This nested representation allows for efficient scalable coding where lower-resolution versions can be extracted from the same coefficient set, reducing bitrate overhead compared to independent encoding of multiple resolutions.
Data Source
AI summary
Methods and apparatuses for compressing and decompressing digital data. The method for compressing digital data comprises a number of steps: a) generating a vector-valued dataset according to the digital data, b) transforming the vector-valued dataset into multiwavelet coefficients, and c) entropically coding the multiwavelet coefficients. The method for decompressing digital data is substantially made up of the same steps as the method for compressing digital data but functioning in a reverse manner.