Image Data Decompression Using Multi-Level Difference Tables
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing lossy compression techniques for image data in GPUs and CPUs, especially in mobile devices, face challenges in achieving a guaranteed compression ratio without significant quality loss, often requiring complex operations and causing visually perceptible artifacts.
Innovation Solution
A lossy compression method using a multi-level difference table to compress image data, involving simple integer operations like addition and subtraction, ensuring a fixed compression ratio without noticeable quality degradation, suitable for low-power and low-cost systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If lossy compression techniques are used to achieve a guaranteed compression ratio, then the compression ratio is improved, but image quality deteriorates with visually perceptible artifacts
Solution Approach 1:
The patent applies parameter changes by using a multi-level difference table where each level represents a different quantization precision. The compression algorithm selects appropriate levels for different image regions, changing the precision parameter dynamically to achieve both high compression ratios and maintain image quality. This allows guaranteed compression ratios while minimizing visually perceptible artifacts through adaptive precision control.
2Manufacturing precision
If complex compression operations are used to maintain image quality, then image quality is improved, but computational complexity increases
Solution Approach 1:
The patent segments the image data into blocks and applies a multi-level difference table approach where each block can use different compression levels. This segmentation allows simple integer operations to be applied locally while achieving global image quality through selective precision application across different regions, avoiding the need for complex global operations.
Solution Approach 2:
By changing the precision parameter through multi-level difference tables, the patent enables simple integer operations to achieve quality results. The different levels in the difference table represent different precision settings, allowing the system to switch between simplicity and quality as needed without requiring complex computational operations.
3Productivity
If high compression ratios are achieved through complex algorithms, then compression ratio is improved, but power consumption increases
Solution Approach 1:
The patent uses parameter changes through multi-level difference tables to achieve high compression ratios with simple integer operations. By pre-computing difference tables at multiple precision levels and selecting appropriate levels during compression, the system avoids complex real-time computations that would consume excessive power, enabling efficient compression suitable for mobile and embedded devices.
Data Source
Figure 1~2
Figure 3a~3b
Figure 4
AI summary
There is provided a computer-implemented method and a decompression unit for decompressing a compressed block of data in accordance with a multi-level difference table. The compressed block of data represents a block of image data comprising a plurality of image element values, wherein each level of the multi-level difference table comprises a plurality of entries. An origin value for the block of image data is determined using data representing the origin value from the compressed block of data. A level within the multi-level difference table for the block of image data is identified using an indication of the level from the compressed block of data. For each image element value in the block of image data, a respective entry indication from the compressed block of data is used to identify one of the entries at the identified level within the multi-level difference table, and the image element value is determined using: (i) the determined origin value for the block of image data, and (ii) the identified entry at the identified level within the multi-level difference table for the image element value.