Intra Prediction Edge Detection for Image Coding Efficiency

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional video coding standards face inefficiencies in coding sharp linear edges, leading to increased coded data or distortion due to limited prediction directions, which result in suboptimal coding efficiency and increased bit usage for high-frequency components.

Innovation Solution

The method involves detecting the direction of edges in surrounding blocks to generate a predicted block, using a weighted sum of pixel values and selecting appropriate prediction modes, such as edge prediction modes, to reduce prediction residuals and improve coding efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If conventional prediction directions (8 directions) are used for intra-frame prediction, then the coding process is simple and fast, but the coding efficiency is poor for sharp linear edges leading to increased bit usage

Engineering Contradiction:
Improvecoding efficiencyVSAvoidprediction direction complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent implements dynamic prediction by detecting edge directions in the current block and adaptively selecting prediction modes based on detected edges. Instead of using fixed 8 prediction directions, the system dynamically determines prediction directions that match the actual edge orientations in the image, thereby improving coding efficiency for sharp linear edges while maintaining manageable complexity through automated edge detection algorithms

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent changes the prediction mode parameters by introducing edge detection-based prediction directions alongside the conventional 8 directions. When edges are detected, the system modifies the prediction approach to align with detected edge orientations, allowing flexible adaptation to different image content characteristics without requiring a complete redesign of the prediction framework

Inventive Principle:
Principle #35Parameter changes

2Manufacturing precision

If conventional intra-frame prediction with limited directions is used, then the processing is computationally efficient, but distortion occurs in coded images with sharp edges

Engineering Contradiction:
Improveprediction accuracyVSAvoidcomputational efficiency
Core Design Contradiction:
Manufacturing precisionVSProductivity

Solution Approach 1:

The patent applies preliminary edge detection to the current block before performing prediction. By detecting edges in advance and determining their directions, the system prepares the optimal prediction mode selection beforehand, ensuring high prediction accuracy for sharp edges. This preliminary action allows the main prediction process to proceed efficiently with pre-determined parameters

Inventive Principle:
Principle #10Preliminary action

3Loss of information

If conventional prediction modes are used, then the amount of coded data increases for blocks with sharp edges, but implementing edge detection increases processing complexity

Engineering Contradiction:
Improveinformation lossVSAvoidedge detection complexity
Core Design Contradiction:
Loss of informationVSDevice complexity

Solution Approach 1:

The patent implements self-service by having the encoder perform edge detection and determine optimal prediction modes autonomously. The system uses the same edge detection capability at both encoder and decoder sides, allowing the decoder to reconstruct images with high accuracy without requiring additional information transmission. This self-service approach minimizes information loss while keeping the complexity manageable through symmetric processing at both ends

Inventive Principle:
Principle #25Self-service

Data Source

PatentEP2230851B1Image encoding method and image decoding method
Publication Date: 2020.05.27 SUN PATENT TRUST
  • EP2230851B1 patent drawingFigure 1A~1B
  • EP2230851B1 patent drawingFigure 2
  • EP2230851B1 patent drawingFigure 3

AI summary

Provided is an image coding method for performing intra prediction achieving higher coding efficiency. The method for coding image data on a block-by-block basis includes: generating a predicted block by predicting a current block; computing a difference between the current block and the predicted block; coding the difference computed in the computing; decoding the difference coded in the coding; and adding the difference decoded in the decoding to the predicted block to generate a decoded block, wherein the generating includes: detecting an edge in a previously decoded block corresponding to a block adjacent to the current block; and extrapolating or interpolating previously decoded image data corresponding to a pixel included in the block adjacent to the current block, along a direction of the edge detected in the detecting to generate the predicted block.