Video Encoding Mode Decision with Temporal Error Integration

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current video encoding algorithms do not adequately account for the impact of potential distortion over time, leading to persistent errors and artifacts due to the lack of temporal integration in error propagation distortion estimation.

Innovation Solution

Incorporating temporal integration into the error propagation distortion estimation by using a factor that weights past losses over time, ensuring that even small persistent distortions trigger intra coding modes, thereby improving the trade-off between bitrate and distortion.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of energy

If inter frame encoding mode is used to reduce bitrate, then compression efficiency is improved, but error propagation distortion accumulates over time leading to persistent artifacts

Engineering Contradiction:
ImprovebitrateVSAvoiddistortion estimation accuracy
Core Design Contradiction:
Loss of energyVSReliability

Solution Approach 1:

The encoder performs preliminary estimation of error propagation distortion for each macroblock before making the encoding mode decision. By calculating the potential distortion that would result from inter-frame encoding and comparing it against a threshold, the system proactively identifies regions where inter-frame encoding would cause unacceptable artifact accumulation, thereby preventing poor quality outcomes before they occur.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system implements a feedback mechanism where the estimated error propagation distortion from previously encoded frames is fed back into the current frame's mode decision process. The distortion map from frame n-1 is used to inform the encoding decisions for frame n, creating a continuous loop that adapts to accumulating errors and adjusts encoding strategies accordingly to maintain quality.

Inventive Principle:
Principle #23Feedback

2Reliability

If intra frame encoding mode is used to reduce distortion, then encoding robustness is improved, but bitrate increases

Engineering Contradiction:
Improveencoding robustnessVSAvoidbitrate
Core Design Contradiction:
ReliabilityVSLoss of energy

Solution Approach 1:

The system applies different encoding modes to different regions of the video frame based on local characteristics. By evaluating error propagation distortion at the macroblock level, the encoder can selectively apply intra-frame encoding only to specific regions where it is most needed (high motion, high distortion risk areas) while using more efficient inter-frame encoding in stable regions, thereby achieving robustness where required without unnecessarily increasing overall bitrate.

Inventive Principle:
Principle #3Local quality

3Measurement precision

If temporal integration is applied to error propagation distortion estimation, then persistent errors are detected earlier, but computational complexity increases

Engineering Contradiction:
Improvedistortion detection precisionVSAvoidcomputational complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system applies temporal integration selectively rather than uniformly across all macroblocks. By using a distortion threshold to determine when intra-frame encoding should be triggered, the system performs partial temporal integration only where necessary to correct persistent errors, avoiding the full computational overhead of exhaustive temporal analysis across the entire frame while still achieving improved detection precision where it matters most.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP2712481B1Mode decision with perceptual-based intra switching
Publication Date: 2019.09.18 SKYPE
  • EP2712481B1 patent drawingFigure 1a
  • EP2712481B1 patent drawingFigure 1b~1c
  • EP2712481B1 patent drawingFigure 2~3

AI summary

A method comprising: for each target image portion to be encoded, selecting a preferred one of a set of encoding modes by optimizing a function comprising an estimate of distortion and a measure of bit rate required to encode the target image portion, the encoding modes comprising at least an intra frame mode and an inter frame encoding mode; encoding the target image portion into the encoded video stream using the selected encoding mode; and transmitting the encoded video stream over a lossy channel. The estimate of distortion is based on a contribution from possible new loss over the channel comprising possible loss of the target block, and on a contribution from possible past loss over the channel comprising possible loss in a history of the encoded video stream upon which decoding of the target portion depends. The method comprises accumulating the contribution from past loss over time, such that the estimate of distortion grows irrespective of the contribution from new loss.