Inter-Frame Boundary Prediction Using Adjacent Block Motion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video encoding and decoding technologies face inefficiencies due to inaccurate prediction values for boundary pixel points when the block boundary does not fit the contour, as current methods like OBMC fail to fully consider motion differences between blocks, leading to reduced encoding and decoding efficiency.
Innovation Solution
An inter-frame prediction method that determines first and second prediction values based on motion information of adjacent blocks, using difference values to correct prediction values for boundary pixel points, fully considering motion differences between blocks to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If OBMC technology is used to correct prediction values of boundary pixel points, then the accuracy of prediction values is improved, but the computational complexity increases due to weighted processing of multiple motion information
Solution Approach 1:
The patent applies different processing strategies to different regions: for boundary pixel points, it uses difference value based correction with selective weighting, while for non-boundary pixels, it uses standard prediction. This local differentiation improves accuracy where needed without unnecessarily increasing complexity across the entire block.
Solution Approach 2:
The patent dynamically adjusts the weighting parameters based on the magnitude of difference values. When difference values are small, it uses simplified weighting; when difference values are large, it applies more aggressive correction. This adaptive parameter adjustment resolves the contradiction by making complexity proportional to actual need.
2Measurement precision
If difference values are calculated and used for correction, then the accuracy of boundary pixel prediction is improved, but the encoding overhead increases due to additional data to be transmitted
Solution Approach 1:
The patent calculates and transmits difference values only for boundary pixel points rather than all pixels in the block. This partial action provides the necessary correction accuracy at the critical boundaries while minimizing the additional data overhead, directly resolving the contradiction between accuracy improvement and overhead increase.
Solution Approach 2:
The patent extracts only the essential correction information (difference values for boundary pixels) from the full prediction data, separating the critical correction needs from the redundant information. This extraction reduces the quantity of data that must be encoded and transmitted while maintaining the accuracy benefits.
3Adaptability or versatility
If motion information of adjacent blocks is used for prediction, then the adaptability to different motion patterns is improved, but the reliability of prediction decreases due to potential motion differences between blocks
Solution Approach 1:
The patent uses difference values as feedback to assess the reliability of using adjacent block motion information. When difference values indicate significant motion discrepancies, the system adjusts its correction strategy accordingly, allowing it to adapt to different motion patterns while maintaining reliability by responding to actual prediction errors.
Solution Approach 2:
The patent dynamically adjusts the influence of adjacent block motion information based on calculated difference values. Rather than statically applying fixed weights, the system adapts its correction strength according to the actual motion differences observed, resolving the contradiction between adaptability and reliability by making the prediction process dynamic and responsive to actual conditions.
Data Source
AI summary
An inter-frame prediction method and a terminal and pertains to the field of video encoding and decoding technologies. The inter-frame prediction method according to embodiments of this application includes: obtaining first motion information of a first block and second motion information of a second block; determining first and second prediction values corresponding to each pixel point in a first pixel region associated with the first block; and determining a target prediction value corresponding to each pixel point in a second pixel region of the first block based on target information; where the first block is a block to be encoded, and the second block is an encoded block; or the first block is a block to be decoded, and the second block is a decoded block.