GPU Image Decompression Using Channel Decorrelation Modes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data compression techniques for image data in graphics processing units (GPUs) struggle to achieve a guaranteed compression ratio while maintaining data integrity, particularly in low-power mobile devices, where high-quality rendering and reduced latency are essential.
Innovation Solution
A method and apparatus for image data decompression that utilizes a channel decorrelating mode and a non-channel decorrelating mode to determine decompressed data values, ensuring a guaranteed compression ratio by selecting the appropriate mode for each non-reference channel, allowing for lossless or lossy compression based on the compression ratio requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If data compression is applied to image data in GPU memory, then memory bandwidth is reduced and power consumption decreases, but compression ratio guarantees and data integrity become difficult to maintain
Solution Approach 1:
The image data is divided into multiple blocks, and each block is further divided into channels (e.g., R, G, B channels). The compression is applied independently to each channel within each block, allowing selective application of different compression techniques (lossless or lossy) to different channels based on their importance and correlation characteristics.
Solution Approach 2:
The patent changes the parameter representation by transforming channel data into difference values (delta encoding) where each pixel's channel value is represented as the difference from a reference pixel's channel value. This transformation enables better compression ratios while maintaining the ability to guarantee minimum quality thresholds through controlled parameter selection.
2Quantity of substance
If compression is applied to reduce memory footprint, then data transfer amount decreases, but achieving guaranteed compression ratio while maintaining data integrity becomes problematic
Solution Approach 1:
The patent introduces difference values as an intermediary representation between the original channel data and the compressed storage format. By storing the difference between a pixel's channel value and a reference pixel's channel value, the system achieves better compression while maintaining the ability to reconstruct the original data with controlled precision.
Solution Approach 2:
The compression approach is made dynamic by allowing the system to switch between lossless and lossy compression modes for different channels based on their characteristics. Reference channels (with high correlation to non-reference channels) use lossy compression with difference values, while non-reference channels use lossless compression, optimizing the balance between compression ratio and data integrity.
3Productivity
If channel decorrelating mode is used for compression, then compression ratio improves, but complexity of determining decompressed values increases
Solution Approach 1:
The decompression process is segmented into distinct stages: first reading the reference channel values, then using those to compute non-reference channel values through the stored difference values. This segmentation simplifies the overall complexity by breaking down the decorrelation process into manageable, sequential steps that can be efficiently implemented in hardware.
Data Source
AI summary
Image element values are determined from a compressed block of image data relating to a reference channel and non-reference channels. Compressed channel data is used to determine an initial data value relating to a channel. A decompressed data value is determined for the non-reference channels by: (i) reading an indication of a compression mode, the compression mode being either a channel decorrelating mode or a non-channel decorrelating mode, (ii) if the compression mode is non-channel decorrelating mode, determining the decompressed data value for the non-reference channel to be the determined initial data value relating to the non-reference channel for the image element value; (iii) if the compression mode is a channel decorrelating mode, determining the decompressed data value for the non-reference channel to be a function of the determined initial data value relating to the non-reference channel and the determined initial data value relating to one of the reference channels.