Weighted Prediction Video Coding Local Brightness Variation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional video compression standards like H.264 struggle with coding efficiency when dealing with significant local brightness variations, such as illumination changes or camera flashes, due to limitations in the number of weighting parameters available for weighted prediction.

Innovation Solution

A method and apparatus for weighted prediction video coding and decoding that uses a localized weighted function to determine weights for local brightness variations, derived from previously encoded and reconstructed neighboring pixels in the source picture and corresponding motion predicted pixels in the reference pictures, without explicit coding.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If conventional weighted prediction with limited parameters is used, then global brightness variation can be handled, but local brightness variation cannot be sufficiently addressed

Engineering Contradiction:

Improvebrightness variation handling capabilityVSAvoidcoding efficiency

Core Design Contradiction:

Adaptability or versatilityVSLoss of information

Solution Approach 1:

The patent applies local quality by deriving separate weighting factors for different regions within a picture. Instead of using a single global weighting factor for the entire picture, the encoder derives region-specific weighting factors based on local brightness characteristics. This allows each region to be adapted to its specific brightness variation pattern, thereby improving the handling of local brightness variation while maintaining coding efficiency.

Inventive Principle:

Principle #3Local quality

2Adaptability or versatility

If more reference pictures are used to handle local brightness variation, then weighting flexibility improves, but device complexity increases

Engineering Contradiction:

Improveweighting parameter flexibilityVSAvoidreference picture management complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent changes the parameters by deriving weighting factors directly from picture data rather than storing multiple reference pictures. The weighting factors are calculated based on local brightness characteristics of the current picture and reference picture, allowing flexible adaptation without requiring additional reference picture storage. This reduces device complexity while maintaining weighting flexibility.

Inventive Principle:

Principle #35Parameter changes

3Measurement precision

If explicit coding of weighting parameters is performed, then precision improves, but bit rate increases

Engineering Contradiction:

Improveweighting parameter precisionVSAvoidbit rate

Core Design Contradiction:

Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent applies self-service by having the encoder and decoder both derive the same weighting factors independently from the picture data. The derivation process uses locally available information (current picture blocks and reference picture blocks) to calculate weighting factors without needing to transmit them. This eliminates the need for explicit coding of weighting parameters, maintaining precision while avoiding bit rate increase.

Inventive Principle:

Principle #25Self-service

Data Source

PatentUS8284837B2Video codec with weighted prediction utilizing local brightness variation

Publication Date: 2012.10.09 INTERDIGITAL VC HOLDINGS INC

AI summary

There are provided video encoders, video decoders, and corresponding encoding and decoding methods for video data for a picture, wherein the video data has local brightness variation. The video encoder includes an encoder for inter-coding the video data using a localized weighted function to determine weights for the local brightness variation. The weights for the localized weighted function are derived without explicit coding.