Multiview Video Encoding Weighting Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video signal coding techniques face inefficiencies in reducing data transmission and improving prediction accuracy for multiview video signals, particularly in handling spatial, temporal, and inter-view redundancies.
Innovation Solution
Deriving weighting-value predicting information for a current view texture block based on that of a neighboring view texture block, using a weighting-value predicting information flag to determine usage, and modifying this information considering the average and differential average values of pixels in both views to enhance prediction accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If weighting-value predicting information is transmitted separately for each view block, then prediction accuracy is improved, but data transmission amount increases
Solution Approach 1:
The patent merges the weighting-value predicting information from neighboring view blocks to form the prediction information for the current view block. Instead of transmitting separate weighting values for each view, the system combines information from multiple neighboring blocks (left, upper, upper-left neighbors) to derive the weighting values, thereby reducing the total data transmission requirement while maintaining prediction accuracy.
Solution Approach 2:
The patent uses neighboring view blocks as references to copy and adapt their weighting-value predicting information for the current view block. The decoder copies weighting information from neighboring blocks and modifies it based on local picture characteristics (average and differential average values), eliminating the need to transmit redundant weighting information for each block.
2Quantity of substance
If weighting-value predicting information is derived from neighboring blocks, then data transmission is reduced, but processing complexity increases
Solution Approach 1:
The patent segments the derivation process into distinct, manageable steps: (1) acquiring weighting information from left, upper, and upper-left neighboring blocks, (2) calculating average and differential average values separately, (3) modifying the neighboring block information using these calculated values, and (4) generating the final prediction. This segmentation simplifies the overall complexity by breaking down the complex derivation into modular operations.
Solution Approach 2:
The patent performs preliminary calculations of average and differential average values of pixel data before modifying the weighting-value predicting information. By pre-calculating these statistical measures from the current and neighboring picture data, the system prepares the necessary modifiers in advance, making the subsequent weighting derivation more efficient and less complex.
3Measurement precision
If weighting-value predicting information is modified using picture characteristics, then prediction accuracy is improved, but processing steps increase
Solution Approach 1:
The patent applies local quality modification by using picture-specific characteristics (average and differential average values of the current picture and neighboring picture) to adaptively modify the weighting-value predicting information. Instead of using a universal weighting scheme, the system tailors the weighting values to local picture characteristics, improving prediction accuracy for different picture contents while maintaining a systematic approach through the defined modification formula.
Data Source
AI summary
The video signal decoding method according to the present invention involves acquiring weighting-value predicting data of a neighboring view texture block corresponding to the current view texture block, deriving weighting-value predicting data of the current view texture block by using the weighting-value predicting data of the neighboring view texture block, and subjecting the current view texture block to weighting-value compensation by using the derived weighting-value predicting data.