Video Coding Unit Prediction Using Neighbor Information
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing high-efficiency video coding technologies face challenges in improving compression efficiency and reducing redundancy of information between coding units (CUs) during the coding and decoding process.
Innovation Solution
A method and apparatus for predicting and coding information on a current coding unit by using neighbor information, including generating prediction information and determining its similarity with neighboring CUs within the same frame or reference frame, and transmitting appropriate flags or differences when necessary.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of substance
If video information is coded using traditional H.264/AVC standards, then compression is achieved, but compression efficiency is insufficient for UHD resolutions requiring twice the compression efficiency
Solution Approach 1:
The video frame is divided into larger coding tree units (CTUs) that can be recursively split into smaller coding units (CUs). This hierarchical segmentation allows the encoder to adaptively select the optimal block size for each region, improving compression efficiency by matching the segmentation to the actual content characteristics rather than using fixed block sizes.
Solution Approach 2:
Different prediction modes and transformation methods are applied to different coding units based on their local characteristics. The encoder selects from multiple intra prediction modes (planar, angular, DC) and inter prediction modes (merge, AMVP, GB) to match the local content, achieving better compression efficiency by treating each region according to its specific properties.
2Loss of substance
If coding unit information is independently coded without using neighbor information, then coding simplicity is maintained, but redundancy between neighboring CUs is not removed reducing compression efficiency
Solution Approach 1:
The encoder performs preliminary prediction of coding unit information using already-coded neighboring CUs before actual coding. By predicting CU mode, prediction mode, and other parameters from neighbors, the encoder prepares expected values in advance, allowing it to only code the differences when actual values differ from predictions, thus removing redundancy efficiently.
Solution Approach 2:
The encoder copies information from neighboring coding units to predict the current CU's parameters. When neighboring CUs have similar characteristics, their mode information, prediction mode, and other parameters are copied as predictions, reducing the number of bits needed to encode the current CU and improving compression efficiency.
3Productivity
If all coding unit information is transmitted without prediction, then decoding simplicity is maintained, but compression efficiency is reduced due to inability to exploit neighbor information
Solution Approach 1:
The decoder uses feedback from already-decoded neighboring CUs to predict the current CU's parameters. By comparing predicted values with actual decoded values, the decoder can identify when prediction was accurate (no transmission needed) versus when differences exist (transmission required), enabling efficient reconstruction while maintaining relatively simple decoding operations.
Data Source
AI summary
A method for coding image information includes generating prediction information by predicting information on a current coding unit, and determining whether the information on the current coding unit is the same as the prediction information. When the information on the current coding unit is the same as the prediction information, a flag indicating that the information on the current coding unit is the same as the prediction information is coded and transmitted. When the information on the current coding unit is not the same as the prediction information, a flag indicating that the information on the current coding unit is not the same as the prediction information and the information on the current coding unit are coded and transmitted. At the generating of the prediction information, the prediction information is generated by using the information on the coding unit neighboring to the current coding unit