Encoder Adaptive Quantization for Bitrate Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing bitrate control methods in encoders, particularly the R-λ model, are complex and difficult to implement in hardware circuits, leading to challenges in maintaining image quality when encoding images with high complexity, especially under constant bitrate constraints.
Innovation Solution
A simplified bitrate control method is introduced, which includes a frame level processing circuit and a CTU level processing circuit that calculate and adjust quantization parameters adaptively to ensure consistent output bitrate, using formulas to determine average bitrates and energies for each frame, and employing adaptive quantization to improve image quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If the R-λ model is used for bitrate control, then the image quality can be optimized, but the calculation complexity increases significantly
Solution Approach 1:
The patent extracts and removes the complex exponential and logarithmic calculation components from the R-λ model, retaining only the essential quantization parameter calculation functionality. This simplification eliminates the hardware implementation difficulties while preserving the core image quality optimization capability.
Solution Approach 2:
The patent replaces the complex R-λ calculation model with simpler, more computationally efficient algorithms that require fewer computational resources. This substitution maintains the bitrate control functionality while significantly reducing the hardware complexity and implementation cost.
2Productivity
If constant bitrate encoding is used, then transmission over limited bandwidth channels is facilitated, but image quality is sacrificed for high complexity images
Solution Approach 1:
The patent introduces dynamic quantization parameter adjustment at the CTU level, allowing the encoder to adaptively modify compression strength based on local image complexity. This dynamic approach enables constant bitrate transmission while preserving image quality in high-complexity regions by applying less aggressive quantization where needed.
Solution Approach 2:
The patent applies different quantization parameters to different CTUs (Coding Tree Units) within the same frame based on their individual complexity characteristics. This local adaptation allows the encoder to maintain overall bitrate constraints while ensuring high-quality encoding for complex regions and accepting lower quality for simple regions.
3Manufacturing precision
If VBR encoding is used, then image quality is maintained, but the output bitrate cannot be determined for subsequent processing
Solution Approach 1:
The patent implements a feedback mechanism where the quantization parameter calculation incorporates information about the actual bitrate consumption and image complexity. This feedback loop allows the encoder to adaptively adjust parameters to achieve both quality preservation and predictable bitrate output, resolving the uncertainty issue of pure VBR encoding.
Data Source
AI summary
An encoder includes a frame level processing circuit, a coding tree unit (CTU) level processing circuit and an encoding circuit. The frame level processing circuit is arranged to calculate a bit number of a current frame according a target bitrate and a frame rate, and the frame level processing circuit is further arranged to calculate a quantization parameter of the current frame according to the bit number of the current frame and at least one parameter. The CTU level processing circuit is arranged to use an adaptive quantization mode to adjust the quantization parameter to generate an adjusted quantization parameter. The encoding circuit is arranged to encode the current frame to generate output data according to the adjusted quantization parameter.