Picture Encoding With Size-Restricted Inter Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing picture encoding and decoding methods face challenges in enhancing encoding efficiency due to increased processing amounts and dependence relationships between primary and secondary signals, making parallel processing difficult.
Innovation Solution
The technique involves dividing pictures into rectangular blocks of predetermined sizes for both primary and secondary signals, with restricted inter prediction based on the block sizes, allowing for efficient encoding and decoding by optimizing block division.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a secondary signal is predicted from a decoded picture of a primary signal, then encoding efficiency is enhanced, but a processing amount increases and dependence relationship between processing on the primary signal and processing on the secondary signal is generated
Solution Approach 1:
The picture is divided into multiple blocks, and the secondary signal is predicted for each block independently using the decoded primary signal. This segmentation allows parallel processing across different blocks while maintaining the prediction relationship, thereby reducing overall processing complexity and enabling concurrent operations without compromising encoding efficiency.
2Productivity
If a secondary signal is predicted from a decoded picture of a primary signal, then encoding efficiency is enhanced, but dependence relationship between processing on the primary signal and processing on the secondary signal is generated
Solution Approach 1:
By dividing the picture into multiple blocks and performing secondary signal prediction independently for each block, the processing can be parallelized across blocks. This reduces the overall processing time while maintaining encoding efficiency, as multiple blocks can be processed simultaneously rather than sequentially.
3Productivity
If block division is performed for picture encoding and decoding, then encoding efficiency is enhanced, but elongated block shapes may occur which increase memory bandwidth requirements
Solution Approach 1:
The invention applies different block division strategies based on local picture characteristics. By analyzing the content and complexity of different regions, the encoder selects appropriate block sizes and shapes, avoiding excessive elongation in regions where it would increase memory bandwidth requirements, thus balancing encoding efficiency with memory usage.
Solution Approach 2:
The block size and shape parameters are dynamically adjusted based on picture content and encoding conditions. By changing these parameters adaptively, the system can prevent elongated blocks that would require excessive memory bandwidth, while still maintaining high encoding efficiency through appropriate block division.
Data Source
AI summary
Provided is a picture encoding device that divides a picture into blocks and performs encoding for each of the divided blocks. A primary signal block divider divides a primary signal of the picture into rectangles having a predetermined size, and generates a primary signal block. A secondary signal block divider divides a secondary signal of the picture into rectangles having a predetermined size, and generates a secondary signal block. A primary signal predictor predicts a primary signal, and a secondary signal predictor predicts a secondary signal. The secondary signal predictor can perform inter prediction of predicting a secondary signal from an encoded primary signal, and the inter prediction is restricted based on at least one of a size of the primary signal block and a size of the secondary signal block.