Unidirectional Scanning for Quantization Matrices in Video Coding

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

The existing H.264 video coding standard's zigzag scanning method for quantization matrices results in an increased amount of code, which affects the efficiency of coding and decoding processes.

Innovation Solution

Introducing a unidirectional scanning method, such as horizontal or vertical scanning, to reduce the amount of code necessary for coding quantization matrices, thereby enhancing coding and decoding efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If zigzag scanning is used to code quantization matrices, then the scanning follows the conventional H.264 standard, but the amount of code increases

Engineering Contradiction:

Improvecompatibility with H.264 standardVSAvoidamount of code

Core Design Contradiction:

Adaptability or versatilityVSQuantity of substance

Solution Approach 1:

The patent changes the scanning parameter from zigzag pattern to unidirectional patterns (horizontal or vertical scanning), which alters how quantization matrix elements are traversed and coded, thereby reducing the code amount while maintaining standard compatibility

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

Instead of following the conventional zigzag scanning approach, the patent inverts the scanning strategy by using unidirectional scanning methods, which fundamentally changes the coding approach and reduces the code amount

Inventive Principle:

Principle #13The other way round (Inversion)

2Productivity

If unidirectional scanning is introduced to reduce code amount, then coding efficiency improves, but deviation from conventional H.264 scanning method occurs

Engineering Contradiction:

Improvecoding efficiencyVSAvoidadherence to H.264 standard

Core Design Contradiction:

ProductivityVSAdaptability or versatility

Solution Approach 1:

The patent introduces dynamic adaptability by allowing selection between different scanning methods (horizontal or vertical) based on the characteristics of the quantization matrix, enabling the system to optimize coding efficiency while maintaining compatibility with the H.264 framework

Inventive Principle:

Principle #15Dynamics

3Speed

If diagonal scanning is used in quantization matrix, then scanning direction is optimized, but the code amount still increases for certain matrix characteristics

Engineering Contradiction:

Improvescanning speedVSAvoidcode amount

Core Design Contradiction:

SpeedVSQuantity of substance

Solution Approach 1:

The patent applies local quality optimization by selecting different unidirectional scanning patterns (horizontal or vertical) based on the local characteristics of the quantization matrix, such that the scanning direction adapts to the specific matrix structure to minimize code amount

Inventive Principle:

Principle #3Local quality

Data Source

PatentEP4075801B1Image coding apparatus, image coding method, image decoding apparatus, image decoding method, and storage medium

Publication Date: 2023.12.06 CANON KK

Figure 1
Figure 2
Figure 3

AI summary

An 8x8 quantization matrix is used to quantize transform coefficients in a two-dimensional block. The quantization matrix is scanned by a unidirectional diagonal scanning method. Difference values are calculated between the matrix elements at successive positions in the scanning order. A first difference value is a difference between the matrix element at the top-left position of the quantization matrix and a predetermined initial value equal to 8.