Transform Block Scan Order Selection for Video Coding

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing video compression techniques require processing non-zero coefficients across multiple superblocks to select an optimal scan order for transform blocks, which is computationally burdensome and inefficient, especially when all non-zero coefficients in neighbor superblocks need to be processed for a current superblock.

Innovation Solution

A method for transform block-level selection of scan order, where cost values are determined for candidate scan orders based on the number of zero-value coefficients before the end of block position, allowing for the selection of the most efficient scan order for encoding or decoding, which reduces computational resources and memory usage.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If scan order selection is performed at the superblock level by processing all non-zero coefficients in neighbor superblocks, then the optimal scan order can be determined, but computational complexity and processing time increase significantly

Engineering Contradiction:
Improvescan order selection accuracyVSAvoidcomputational complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent divides the superblock into multiple transform blocks and performs scan order selection at the transform block level independently. This segmentation allows each transform block to be processed separately, reducing the computational burden compared to processing all non-zero coefficients across the entire superblock while still achieving optimal scan order selection for each block.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies different scan order selection methods to different transform blocks within a superblock based on their specific characteristics. Each transform block is analyzed independently to determine its optimal scan order, allowing localized optimization without requiring global processing of all neighbor superblock coefficients.

Inventive Principle:
Principle #3Local quality

2Reliability

If all non-zero coefficients in neighbor superblocks are processed for scan order selection, then comprehensive information is available, but memory usage and processing time increase

Engineering Contradiction:
Improvescan order selection reliabilityVSAvoidprocessing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

By segmenting the superblock into transform blocks and processing each independently, the patent reduces the amount of data that needs to be stored and processed in memory at any given time, thereby reducing memory usage and processing time while maintaining reliable scan order selection for each block.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent performs scan order selection only for transform blocks that have non-zero coefficients, rather than processing all transform blocks in the superblock. This partial action approach reduces processing time and memory requirements while still ensuring that all necessary blocks are processed for optimal scan order determination.

Inventive Principle:
Principle #16Partial or excessive action

3Productivity

If transform block-level scan order selection is implemented, then computational resources are reduced, but the complexity of managing multiple candidate scan orders increases

Engineering Contradiction:
Improveencoding/decoding efficiencyVSAvoidscan order management complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent pre-determines a set of candidate scan orders for each transform block based on the characteristics of non-zero coefficients. By preparing these candidate lists in advance, the patent simplifies the subsequent selection process and reduces the complexity of managing multiple scan order options during encoding/decoding operations.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent uses the transform block's own characteristics (distribution and position of non-zero coefficients) to automatically determine the optimal scan order without requiring external processing or complex management systems. Each transform block essentially selects its own optimal scan order based on its intrinsic properties.

Inventive Principle:
Principle #25Self-service

Data Source

PatentEP3725076B1Transform block-level scan order selection for video coding
Publication Date: 2023.08.02 GOOGLE LLC
  • EP3725076B1 patent drawingFigure 1
  • EP3725076B1 patent drawingFigure 2
  • EP3725076B1 patent drawingFigure 3

AI summary

A scan order for encoding or decoding coefficients of a transform block is selected on a transform block-level. A set of candidate scan orders is processed by identifying end of block positions within the transform block for candidate scan orders. Cost values are determined for each of the candidate scan orders to reflect a number of the coefficients of the transform block that are located before the respective end of block positions. In particular, a cost value for a candidate scan order reflects the number of zero-value coefficients located before the end of block position for that candidate scan order. One of the candidate scan orders is then selected based on those cost values. The selected scan order is used to scan the coefficients in the transform block, such as for encoding those coefficients to a bitstream or for decoding those coefficients to an output video stream.