Multi-scale metric-based encoding for video bit allocation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional multimedia encoders inefficiently allocate bits to pictures due to inappropriate quantization parameters, leading to unnecessary resource consumption or reduced video quality, as the effect of quantization on bit allocation and perceptual quality is not readily apparent.
Innovation Solution
A pre-analysis module assesses pixel activity at multiple spatial scales and dynamic ranges to generate a multi-scale metric, which is used to select an optimal quantization parameter and bit allocation for each block, ensuring efficient bit usage and maintaining perceptual quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If a lower quantization parameter is used, then video quality is improved, but bit usage and computing resource consumption increase
Solution Approach 1:
The patent applies different quantization parameters to different blocks within a picture based on their local characteristics. The encoder analyzes each block's complexity and assigns appropriate QP values, allowing high-quality encoding for complex blocks while using lower bit rates for simple blocks, thus resolving the contradiction between overall video quality and total bit usage.
Solution Approach 2:
The patent dynamically adjusts quantization parameters based on real-time analysis of block characteristics such as gradient magnitude and variance. This dynamic adaptation allows the encoder to optimize the balance between quality and bit usage for each block according to its specific content, rather than applying a static QP to the entire picture.
2Quantity of substance
If a higher quantization parameter is used, then bit usage is reduced, but video quality deteriorates
Solution Approach 1:
The patent prevents quality deterioration by applying higher QP values only to blocks with low complexity (simple regions), while maintaining lower QP values for complex blocks that require higher quality. This selective approach reduces overall bit usage without sacrificing perceptual quality in important regions.
Solution Approach 2:
The patent changes the quantization parameter based on measured block characteristics such as gradient magnitude and pixel variance. By adapting QP to the actual content complexity, the system achieves efficient bit allocation that reduces total bit usage while maintaining acceptable quality where it matters most.
3Device complexity
If uniform quantization parameter is applied to all blocks, then encoding complexity is reduced, but bit allocation efficiency decreases
Solution Approach 1:
The patent segments the picture into multiple blocks and analyzes each block's characteristics independently to determine appropriate QP values. This segmentation approach increases bit allocation efficiency by matching QP to local content requirements, while the use of simple metrics like gradient magnitude keeps the additional encoding complexity manageable.
Solution Approach 2:
The patent performs preliminary analysis of block characteristics (gradient calculation, variance computation) before the actual encoding process. This preliminary action enables informed QP selection that optimizes bit allocation efficiency, while the analysis uses computationally efficient methods that do not significantly increase overall encoding complexity.
Data Source
AI summary
A processing system analyzes pixel activity levels of blocks of a picture at a plurality of spatial scales and/or dynamic ranges to generate a multi-scale metric that indicates how bit allocation or assignment of a given quantization parameter (QP) will affect the perceptual quality of the block. Blocks that have similar multi-scale metrics are likely to be visually similar and to benefit from similar bit allocations or QPs. Based on the multi-scale metric, an encoder encodes each block of the picture with a QP and/or a number of bits.