16-Bit Data Decompression Using Predictors for Format Versatility
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional data compression techniques do not support 16-bit unsigned, signed, or floating-point data formats, limiting their applicability in representing and processing various data types.
Innovation Solution
A system and method for compressing and decompressing 16-bit data in floating-point or fixed-point formats, using a fixed-size block approach with extraction, unquantization, and reverse transformation units to produce 16-bit values, enabling support for unsigned and signed fixed-point, as well as floating-point formats.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional compression techniques (e.g., DXTC) are used, then compression is achieved for 3-channel 16-bit unsigned data, but support for 16-bit signed data and floating-point format data is lost
Solution Approach 1:
The compression system is designed to handle multiple data formats (unsigned 16-bit, signed 16-bit, and floating-point 16-bit) through a unified compression architecture. The encoder and decoder are configured to process different data types using the same compression algorithm, enabling universal support across various formats while maintaining compression efficiency.
2Quantity of substance
If data is compressed into a fixed length format, then storage efficiency is improved, but exact reconstruction of original 16-bit values becomes difficult
Solution Approach 1:
The 16-bit data values are segmented into multiple components during compression: a base value (BV) representing the common portion and difference values (DV) representing the deviations. During decompression, these segments are recombined to reconstruct the original values. This segmentation allows efficient storage while enabling accurate reconstruction of the original 16-bit data.
Solution Approach 2:
The compression algorithm transforms the original 16-bit values by calculating base values and difference values, changing the parameter representation from direct 16-bit storage to a two-component representation (BV + DV). This parameter transformation enables more efficient storage while preserving the ability to exactly reconstruct original values through reverse transformation.
3Productivity
If quantization is applied during compression, then compression ratio is improved, but loss of precision occurs in the decompressed data
Solution Approach 1:
The algorithm applies quantization selectively only to the difference values (DV) rather than to the entire 16-bit data or to base values. By limiting quantization to only the correction component and using higher precision for the base values, the system achieves compression while minimizing precision loss. This partial application of quantization maintains data quality while improving compression efficiency.
Data Source
AI summary
Systems and methods compress and decompress 16 bit data. The 16 bit data may be signed or unsigned and represented in a fixed point or floating point format. A fixed block size of data is compressed into a fixed length format. Data compressed using a medium quality compression scheme may be efficiently decompressed in hardware. Data may be efficiently compressed and decompressed in hardware using a high quality compression scheme. The high quality compression scheme has a lower compression ratio compared with the medium quality compression scheme, but is near lossless in terms of quality.