Block Importance Mapping for Low-Complexity Video QP Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video encoders lack an efficient method for dynamically selecting Quantization Parameter (QP) values, relying on slow pre-encoding passes to determine future picture usage, which is computationally intensive and inefficient.
Innovation Solution
An algorithm that estimates future importance based on a subset of neighboring pictures, using calculations from an encoder's temporal filter to compute error values and adjust QP values for blocks, reducing complexity while maintaining visual quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a slow pre-encoding pass is used to determine all future pictures where information from the current picture is likely to be used, then the QP value selection becomes more accurate, but the encoding complexity and processing time increase significantly
Solution Approach 1:
The patent extracts only the necessary information from the full pre-encoding analysis by focusing specifically on error values from temporal filters of neighboring pictures (at distances of 1 and 2 frames). This selective extraction maintains QP selection accuracy while avoiding the computational burden of analyzing all future pictures.
Solution Approach 2:
Instead of performing a complete pre-encoding pass through all future pictures, the patent applies partial action by computing error values only for a limited set of neighboring pictures (up to 2 frames away). This partial computation provides sufficient information for accurate QP selection without the excessive complexity of full analysis.
2Manufacturing precision
If different QP values are assigned to different blocks based on hierarchical importance, then the visual quality of dependent pictures is improved, but the computational overhead for determining block importance increases
Solution Approach 1:
The patent merges the QP determination process with the existing temporal filter computations already performed during encoding. By reusing the error values (V and SSD) calculated by the temporal filter for motion-compensated reference pictures, the system avoids separate importance calculations while still achieving block-specific QP optimization.
Solution Approach 2:
The temporal filter's existing computations serve dual purposes: both motion compensation and QP determination. The error values already computed for temporal filtering automatically provide the importance metrics needed for differential QP assignment, making the system self-sufficient without additional computational overhead.
3Device complexity
If a single QP value is used for all blocks in a picture, then the encoding complexity is reduced, but the visual quality and bitrate efficiency deteriorate
Solution Approach 1:
The patent applies local quality by assigning different QP values to different blocks within the same picture based on their specific importance metrics. Blocks with higher importance (lower error values in temporal filters) receive lower QP values for higher quality, while less important blocks receive higher QP values. This localized differentiation maintains high visual quality where needed while managing overall bitrate efficiency.
Data Source
AI summary
There is provided a method for encoding a picture. The method comprises, for each picture on either side of the current picture with a distance of one or two away from the current picture, computing a first error value, E, wherein the first error value, E, is computed for each block of a set of blocks of a first block size of the current picture and is based on the respective block of the current picture and a corresponding block in the picture. The method comprises, for each picture on either side of the current picture with a distance of one or two away from the current picture, aggregating the first error value, E, over a block of a second block size for each block of the set of blocks of the first block size contained in the block of the second block size to generate a second error value, E′, corresponding to the picture. The method further comprises computing an importance score for the block of the second block size in the current picture based at least in part on the second error value, E′, corresponding to one or more of the pictures on either side of the current picture with a distance of one or two away from the current picture. The method comprises computing a change in quantization parameter, QP, value for the block of the second block size in the current picture based on the importance score. The method comprises signaling the change in QP value to an encoder.