Video Encoder Parallel Processing Tile Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current video encoding and decoding processes face increased computational complexity due to advanced compression algorithms, leading to slower performance and the need for accelerated methods to reduce encoding and decoding time.

Innovation Solution

The implementation of parallel processing techniques and algorithmic optimizations, such as content-dependent early termination rules and multi-pass encoding, to prune the encoding search space and optimize video encoder performance, along with the use of parallel processing approaches like wave front parallel encoding and lockstep execution to accelerate video encoding and decoding.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of substance

If advanced video compression algorithms are used to achieve higher compression ratios, then compression efficiency is improved, but computational complexity increases leading to slower encoding and decoding performance

Engineering Contradiction:
Improvecompression ratioVSAvoidencoding and decoding speed
Core Design Contradiction:
Loss of substanceVSProductivity

Solution Approach 1:

The patent divides the picture into multiple tiles that can be processed independently in parallel. Each tile is further divided into coding tree units (CTUs) that can be encoded simultaneously by different processing threads, enabling parallel processing to overcome the computational complexity of advanced compression algorithms while maintaining high compression ratios

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary classification of CTUs into different complexity groups before encoding. By analyzing CTU characteristics in advance and organizing them into processing queues, the system prepares the data structure to enable efficient parallel processing, reducing the overall encoding time despite using computationally intensive compression algorithms

Inventive Principle:
Principle #10Preliminary action

2Manufacturing precision

If the encoding search space is extensively evaluated to improve encoding quality, then encoding precision is improved, but encoding time increases

Engineering Contradiction:
Improveencoding qualityVSAvoidencoding time
Core Design Contradiction:
Manufacturing precisionVSLoss of time

Solution Approach 1:

The patent applies early termination rules that evaluate only a subset of possible encoding modes for each CTU. By using content-dependent criteria to determine when sufficient evaluation has been performed, the system achieves acceptable encoding quality without exhaustively searching the entire encoding space, thus reducing encoding time while maintaining quality

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The patent applies different encoding strategies to different regions of the picture based on local content characteristics. Complex regions receive more thorough evaluation while simple regions use faster encoding methods, optimizing the balance between encoding quality and time by adapting the evaluation depth to local requirements

Inventive Principle:
Principle #3Local quality

3Productivity

If parallel processing is implemented to accelerate encoding, then productivity is improved, but device complexity increases

Engineering Contradiction:
Improveencoding speedVSAvoidprocessing system complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent divides the encoding task into independent tiles and CTUs that can be processed in parallel by multiple threads. This segmentation enables productive use of multi-core processors while keeping each processing unit relatively simple, as each thread handles a discrete, self-contained portion of the image data

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent creates a universal processing framework that can handle different tile configurations and CTU sizes through the same encoding logic. The parallel processing architecture uses standardized interfaces and data structures that work across different processing configurations, reducing the complexity that would otherwise arise from handling multiple specialized processing paths

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP3777168B1Accelerating video encoding and decoding
Publication Date: 2023.07.05 FRAUNHOFER GESELLSCHAFT ZUR FORDERUNG DER ANGEWANDTEN FORSCHUNG EV
  • EP3777168B1 patent drawingFigure 1
  • EP3777168B1 patent drawingFigure 2~3
  • EP3777168B1 patent drawingFigure 4

AI summary

A video encoder for encoding pictures of a video into a data stream, the picture being partitioned into an array of coding blocks, is configured to process coding blocks in a first line of the array of coding blocks by a first thread. Coding blocks in a second line of the array are processed by a second thread while processing the coding blocks of the first line of the array is still in progress, wherein processing a coding block in the second line of the array requires that reference coding blocks in the first line have been processed. Upon a specific event, processing coding blocks in the second line of the array by the second thread is interrupted and coding blocks of a third line of the array by the second thread are processed.