Mode Estimation in Pipelined Video Encoders
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In video processing, the lack of availability of mode information from previous pixel blocks in pipelined architectures leads to suboptimal performance, particularly in intra-prediction estimation and motion estimation, resulting in poor picture quality due to incorrect assumptions and increased processing overhead.
Innovation Solution
A video system with a mode decision processor that determines the mode of a first pixel block and uses this information to estimate the mode of subsequent pixel blocks, leveraging contextual information from neighboring blocks to improve accuracy and reduce processing errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If intra-prediction estimation or motion estimation is performed in a differing pipeline stage from mode decision, then parallel processing efficiency is improved, but estimation accuracy deteriorates due to unavailability of mode information
Solution Approach 1:
The patent applies preliminary action by performing mode estimation for the current block before mode decision is complete, using provisional mode information from previously processed blocks. This allows the estimation stage to proceed early in the pipeline while still achieving accurate results by the time actual mode decision information becomes available, thus maintaining both parallel processing efficiency and estimation accuracy.
2Speed
If mode information from previous blocks is not used in estimation, then pipeline processing speed is improved, but picture quality deteriorates due to incorrect assumptions
Solution Approach 1:
The patent implements feedback by using mode information from previously decoded blocks to inform the estimation of the current block's mode. This feedback loop allows the pipeline to maintain high processing speed while improving estimation accuracy, as the estimation can be continuously refined based on actual mode decisions from neighboring blocks, thereby maintaining picture quality.
3Device complexity
If mode estimation is performed without contextual information from neighboring blocks, then processing complexity is reduced, but estimation accuracy deteriorates
Solution Approach 1:
The patent performs preliminary mode estimation using available contextual information from previously processed neighboring blocks before the actual mode decision is finalized. This preliminary action maintains low processing complexity during the estimation phase while ensuring high accuracy by the time actual mode information is available, as the estimation is already based on relevant contextual patterns.
Data Source
AI summary
A video system includes an encoder for generating a compressed bit stream in response to a received video signal. The encoder includes a mode decision processor that is arranged to determine whether the mode of a first pixel block in a first row is an “intra-mode” or an “inter-mode.” The encoder also includes a mode estimation processor that is arranged to estimate the mode of a left pixel block in a second row that is received after the first row in response to the determined mode of the first pixel block in the first row. The encoder also includes a pixel block processor that is arranged to process a pixel block in the second row that is to the right of the left pixel block in response to the estimated mode of the left pixel block.