Video Encoder Non-Linear RDO for Perceptual Bit Allocation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video encoding standards rely on rate-distortion optimization (RDO) metrics that are not well correlated with subjective perceptual quality, leading to improper compression decisions and inefficient use of bit rates.
Innovation Solution
Implement a calibration module that provides a non-linear evaluation metric to scale or transform quality measures associated with RDO, adjusting the slope of quality versus bit rate curves to prioritize meaningful areas of the video frame, thereby reducing the effect of compression and improving subjective quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional linear RDO metrics are used for mode selection in video encoding, then the encoding process is simple and fast, but the subjective perceptual quality is poor and bit rate allocation is inefficient
Solution Approach 1:
The patent transforms the linear RDO cost function into a non-linear form by applying a monotonic transformation function to the quality metric. This changes the parameter relationship from linear to non-linear, allowing the system to better correlate with subjective perceptual quality while maintaining computational feasibility through the use of pre-computed transformation curves.
2Productivity
If compression is aggressively applied to reduce bit rate, then bandwidth efficiency improves, but subjective quality deteriorates
Solution Approach 1:
The patent applies different transformation curves to different regions or blocks within the video frame based on their importance. By identifying significant regions and applying appropriate non-linear transformations locally, the system preserves subjective quality in important areas while allowing more aggressive compression in less critical areas, thus improving overall bit rate efficiency.
3Reliability
If linear RDO cost minimization is used, then computational overhead is low, but the correlation with subjective perceptual quality is poor
Solution Approach 1:
The patent pre-computes and stores transformation curves that map linear quality metrics to non-linear perceptual quality measures. During actual encoding, instead of performing complex real-time non-linear optimizations, the system uses these pre-computed curves to quickly transform RDO costs, significantly reducing computational energy while maintaining reliable quality prediction.
Data Source
AI summary
Systems and methods herein are for a video encoder to be associated with a rate optimization distortion (RDO) module and a calibration module, where the RDO module may be to perform RDO for received frames of a media stream and may be to generate at least an RDO output that is based in part on quality measures between the received frames and decoded frames, and where the calibration module may be to provide an evaluation metric that is to scale or transform at least a range of the quality measures, with the scaling or transforming to potentially reduce an effect on a compression performed in the video encoder.