Fixed-Point Approximation for Display Stream Compression Rate Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video compression techniques, such as display stream compression (DSC), face challenges in hardware implementation due to the complexity of division operations required for calculating rate control parameters, which increases costs and difficulties in achieving visually lossless compression for high-resolution displays.
Innovation Solution
The use of fixed-point approximation methods to determine rate control parameters like target rate, flatness QP, buffer fullness, and Lagrangian parameters, reducing the need for division operations and simplifying hardware implementation by employing multiplications, additions, and bit shifting, along with the use of lookup tables to precompute values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If division operations are used to calculate rate control parameters in DSC, then compression accuracy is improved, but hardware implementation complexity increases
Solution Approach 1:
The patent transforms the calculation parameters from requiring division operations to using only multiplication, addition, and bit-shifting operations. Specifically, rate control parameters such as target bitrate and buffer fullness are recalculated using fixed-point arithmetic and lookup tables, changing the mathematical parameters from floating-point division to integer-based operations that are hardware-friendly while preserving compression accuracy.
Solution Approach 2:
The patent replaces the mechanical division operation with an equivalent computational system using multiplication and bit-shifting. Instead of implementing physical division circuits which are complex and slow, the invention uses a substitution approach where division by constant values is replaced with pre-computed multiplication factors and bitwise operations, significantly simplifying the hardware architecture.
2Quantity of substance
If conventional image compression is used on pixel data, then bandwidth requirements are reduced, but implementation cost and difficulty increase
Solution Approach 1:
The patent enables display devices to perform self-service compression using simplified algorithms that can be implemented in standard display hardware without requiring external compression devices. The rate control mechanism using fixed-point arithmetic and lookup tables allows the display device itself to compress video data efficiently, eliminating the need for expensive external compression equipment while reducing bandwidth requirements.
Solution Approach 2:
The patent employs computationally inexpensive algorithms that can be implemented using simple, cost-effective hardware components. By using fixed-point arithmetic and pre-computed lookup tables instead of complex floating-point processors, the invention achieves compression functionality using cheap, readily available hardware that can be manufactured at low cost and integrated into conventional display devices.
3Manufacturing precision
If high-resolution video is transmitted without compression, then picture quality is maintained, but bandwidth requirements exceed display link capacity
Solution Approach 1:
The patent applies partial compression selectively to different regions and blocks of video data based on their complexity. By using rate control parameters calculated through fixed-point arithmetic, the system applies compression only where necessary while maintaining higher quality in critical regions, achieving a balance between bandwidth reduction and picture quality preservation that allows high-resolution video transmission over limited bandwidth links.
Data Source
Figure 1A
Figure 1B
Figure 2A
AI summary
Provided are systems and methods for used fixed-point instead of floating point techniques in order to calculate various parameters for coding video data, including target rate, QP adjustment, buffer fullness, a Lagrangian parameters for a bitrate, and/or a Lagrangian parameter for the fullness of the buffer. By determining one or more of the parameters using fixed-point, hardware implementation costs may be decreased.