Picture Encoding QP Modulation via Human Visual System Statistics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional picture encoding methods are inefficient as they do not account for variations in human eye sensitivity to contrast across different areas of a picture, using the same number of bits for areas where the eye is more or less sensitive, leading to suboptimal encoding efficiency.
Innovation Solution
A method that generates macroblock and global statistics to modulate quantization parameters based on Human Visual System characteristics, such as luminance, motion, and edge strength, to adapt quantization step sizes for each macroblock, optimizing bit allocation according to sensitivity variations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the same number of bits is used for all areas of the picture, then the encoding process is simple, but the encoding efficiency is low because it does not account for variations in human eye sensitivity
Solution Approach 1:
The patent applies local quality by generating different quantization parameters for different macroblocks based on their local characteristics (luminance, motion, edge strength). Each macroblock receives a tailored quantization parameter that matches the human visual system's sensitivity in that specific region, rather than using a uniform quantization parameter across the entire picture. This resolves the contradiction by improving encoding efficiency through localized adaptation while managing complexity through automated statistical analysis.
Solution Approach 2:
The patent changes the quantization parameter values dynamically based on macroblock statistics and global picture statistics. The system calculates luminance, motion, and edge strength for each macroblock, then modulates the quantization parameter accordingly - using coarser quantization in less sensitive areas and finer quantization in more sensitive areas. This parameter adaptation resolves the contradiction between simple encoding and efficient encoding.
2Manufacturing precision
If uniform quantization is applied across the picture, then the encoding algorithm is simple, but picture quality is suboptimal due to artifacts in sensitive areas
Solution Approach 1:
The patent implements local quality by computing macroblock-specific statistics (luminance, motion, edge strength) and using these to determine appropriate quantization parameters for each region. Areas with strong edges or high luminance sensitivity receive finer quantization to preserve quality, while less sensitive areas use coarser quantization. This localized approach improves overall picture quality by matching quantization strength to human visual sensitivity in each region.
Solution Approach 2:
The patent performs preliminary analysis of each macroblock's characteristics (luminance, motion, edge strength) before applying quantization. By pre-calculating global statistics and macroblock-specific statistics, the system determines the optimal quantization parameter for each macroblock in advance, ensuring quality is optimized before the actual encoding process. This preliminary action resolves the contradiction between simple and high-quality encoding.
3Manufacturing precision
If more bits are allocated to all areas, then picture quality improves, but the bit rate increases unnecessarily in areas where the eye is less sensitive
Solution Approach 1:
The patent applies local quality by allocating bits non-uniformly across different macroblocks based on human visual system sensitivity. Each macroblock receives a quantization parameter tailored to its local characteristics - areas with high luminance, motion, or edge strength (where the eye is more sensitive) receive finer quantization and more bits, while less sensitive areas receive coarser quantization and fewer bits. This resolves the contradiction by improving picture quality only where necessary while reducing overall bit rate.
Solution Approach 2:
The patent dynamically changes quantization parameters based on macroblock statistics and global picture characteristics. By modulating the quantization parameter according to luminance levels, motion intensity, and edge strength, the system allocates more bits to sensitive areas and fewer bits to insensitive areas. This parameter adaptation resolves the contradiction between maintaining high picture quality and minimizing overall bit rate.
Data Source
AI summary
A method for encoding a picture is disclosed. The method generally includes the steps of (A) generating at least one respective macroblock statistic from each of a plurality of macroblocks in the picture, (B) generating at least one global statistic from the picture and (C) generating a respective macroblock quantization parameter for each of the macroblocks based on both (i) the at least one respective macroblock statistic and (ii) said at least one global statistic.