Video Compression via Frequency Domain Block Matching

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing video compression techniques face challenges in achieving low complexity while maintaining high quality, particularly in motion estimation, as they require extensive computational overhead and are not effectively implemented in the frequency domain.

Innovation Solution

The method involves block matching in the frequency domain, where blocks are transformed into a frequency domain representation, and a sorted list is used to select reference blocks based on energy indices and bit rate metrics, allowing for efficient prediction coding and reducing the number of coefficients that need to be processed.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If motion estimation is performed using conventional block-based techniques in the spatial domain, then video compression can be achieved, but computational complexity and processing overhead increase significantly

Engineering Contradiction:
Improvevideo compression efficiencyVSAvoidcomputational complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent replaces the conventional spatial domain block-based motion estimation with frequency domain processing. By transforming video blocks into the frequency domain using DCT (Discrete Cosine Transform), the invention substitutes complex spatial comparisons with simpler frequency coefficient operations, reducing computational overhead while maintaining compression effectiveness.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The invention changes the domain parameter from spatial to frequency domain. This parameter transformation allows motion estimation to be performed on frequency coefficients rather than spatial pixel values, fundamentally altering the processing approach and reducing the computational complexity associated with traditional block matching algorithms.

Inventive Principle:
Principle #35Parameter changes

2Manufacturing precision

If more reference frames are used for inter frame prediction, then compression quality improves, but processing complexity and memory requirements increase

Engineering Contradiction:
Improvecompression qualityVSAvoidprocessing complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent transforms the reference frame matching process into the frequency domain, where blocks are compared based on their frequency coefficients rather than spatial pixel values. This parameter change enables more reference frames to be processed efficiently, as frequency domain operations are computationally less intensive than spatial domain block matching.

Inventive Principle:
Principle #35Parameter changes

3Ease of operation

If block matching is performed in the spatial domain, then motion estimation can be conducted, but the number of coefficients to be processed increases computational overhead

Engineering Contradiction:
Improvemotion estimation capabilityVSAvoidprocessing overhead
Core Design Contradiction:
Ease of operationVSDevice complexity

Solution Approach 1:

The patent substitutes spatial domain block matching with frequency domain coefficient comparison. By applying DCT to video blocks and performing motion estimation on the resulting frequency coefficients, the invention replaces computationally intensive spatial operations with more efficient frequency domain operations, reducing processing overhead.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The invention changes the operational domain from spatial to frequency, transforming how motion estimation is performed. This parameter change reduces the number of operations required for block matching, as frequency coefficients provide a more compact representation that requires less processing power for comparison and analysis.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP2522141B1Data compression for video
Publication Date: 2019.04.24 SKYPE
  • EP2522141B1 patent drawingFigure 1
  • EP2522141B1 patent drawingFigure 1a~1b
  • EP2522141B1 patent drawingFigure 2

AI summary

A method of compressing video data, the method comprising: providing image data for each of a plurality of image portions; determining an index value for each of said image portions based on the respective image data; for a target image portion to be encoded, identifying a matching image portion having an index value within a search threshold of the index value of the target image portion; and generating difference data for the target image portion relative to the matching image portion, and including the difference data in an encoded video signal.