Bit Stream Quality Assessment Using Encoding Complexity Metadata
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing bit stream quality assessment methods fail to accurately account for the complexity of the encoding process, leading to inconsistent accuracy in bit stream quality scores across different encoder setups, even when using the same data settings.
Innovation Solution
A method that calculates bit stream quality scores based on complexity-related information extracted from the encoded bit stream, which is then used to select appropriate data settings for improved accuracy or adapt the decoding sub-system, considering parameters like macro block partition size, motion vector resolution, and reference pictures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If bit stream quality assessment is performed using conventional methods without considering encoding complexity, then the assessment process is simple and fast, but the accuracy of quality scores is inconsistent across different encoder setups
Solution Approach 1:
The patent extracts complexity-related information from the encoded bit stream before performing quality assessment. By pre-analyzing encoder setup characteristics (such as macro block partition size, motion vector resolution, and reference picture usage) and storing them as metadata, the system prepares complexity data in advance. This allows the quality assessment to use pre-computed complexity metrics rather than recalculating them during assessment, thereby improving accuracy while controlling the complexity of the assessment process itself.
2Manufacturing precision
If encoder complexity features are increased to improve video quality, then the quality of encoded video improves, but the computational power and encoding time increase
Solution Approach 1:
The patent enables dynamic adjustment of encoder complexity features based on decoder resource availability. The system extracts complexity-related information from the bit stream and uses it to adaptively select appropriate data settings for quality assessment. This allows the encoder to optimize the balance between video quality and encoding speed by adjusting complexity features (such as macro block partition size, motion vector resolution, and reference picture count) according to the target decoder's capabilities, rather than using fixed high-complexity settings for all scenarios.
3Measurement precision
If the same data settings are used for bit stream quality assessment across different encoder complexities, then the assessment process is consistent and simple, but the accuracy varies with encoder setup
Solution Approach 1:
The patent applies local quality by extracting specific complexity-related information from different parts of the encoded bit stream corresponding to different encoder features. Instead of using a single uniform assessment method, the system analyzes local characteristics such as macro block partition size, motion vector resolution, and reference picture usage separately. Each complexity feature is evaluated with appropriate data settings tailored to its specific characteristics, allowing the overall quality assessment to accurately reflect variations in encoder setup while maintaining adaptability across different encoding scenarios.
Data Source
AI summary
In a method for extracting information from an encoded bit stream, the extracted information may be used for improving bit stream quality assessment, or for adapting a decoding sub-system. From an encoded bit stream, received from a bit stream source, complexity related information that is indicative of the complexity of the encoding of said bit stream is extracted. On the basis of said extracted complexity related information one or more data settings are then selected, after which the data settings may be used as an input for improving bit stream quality assessment at a decoding device or a network node, or for adapting a bit stream decoding sub-system at a decoding device.