Data Compression Using Color Grouping and Bit Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data compression methods in the spatial frequency domain are computationally intensive and increase hardware circuit scale, while tone compression reduces image quality by decreasing the number of bits, leading to noticeable deterioration.
Innovation Solution
A data processing apparatus and method that sorts input data into groups by type, extracts maximum values, and determines common bit reduction positions to minimize bit reduction while maintaining image quality, thereby reducing processing time and circuit scale.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If raw data is compressed by variable-length-encoding in spatial frequency domain, then compression is achieved, but computation volume increases and processing time becomes long
Solution Approach 1:
The patent segments the raw data into multiple groups based on color information data type (e.g., red, green, blue channels), then processes each group separately by extracting maximum values and determining bit reduction positions. This segmentation allows simpler, faster operations on smaller data portions rather than processing all data uniformly in the spatial frequency domain.
Solution Approach 2:
The patent changes the compression approach from using spatial frequency domain parameters to using color information data type parameters. By grouping data according to color information and determining bit reduction positions based on maximum values within each group, the system achieves compression through parameter-based organization rather than complex frequency transformations.
2Loss of energy
If raw data is compressed by variable-length-encoding in spatial frequency domain, then compression is achieved, but hardware circuit scale increases
Solution Approach 1:
The patent divides the data processing into separate blocks: a data sort block for grouping, a maximum value extraction block for finding max values in each group, and a bit position decision and bit reduction block for compression. This modular segmentation reduces the complexity of any single circuit component compared to implementing full spatial frequency domain encoding hardware.
Solution Approach 2:
The patent replaces complex hardware-based spatial frequency domain transformation circuits with a more straightforward algorithmic approach that processes data in the original domain by sorting and grouping based on color information, thereby reducing the required hardware circuit scale.
3Loss of energy
If tone compression is applied to decrease number of bits, then compression is achieved, but image deterioration becomes conspicuous
Solution Approach 1:
The patent applies different bit reduction strategies to different groups of data based on their maximum values. Instead of uniformly reducing bits across all data, it determines optimal reduction positions for each group (e.g., red, green, blue channels) based on their specific maximum values, thereby preserving image quality while achieving compression in each local region.
Solution Approach 2:
The patent changes the compression parameter from a uniform bit reduction approach to a variable bit reduction approach where the reduction position is determined by the maximum value of each color group. This parameter adaptation allows maintaining more bits in regions where maximum values indicate higher detail requirements, thereby reducing image deterioration.
Data Source
AI summary
Disclosed herein is a data processing apparatus including: a data sort block configured to sort, into multiple groups by information data of a same type, input data including information data of multiple types formed by multiple bits; a maximum value extraction block configured to extract a maximum value of the information data in each of the groups sorted by the data sort block; and a bit position decision and bit reduction block configured to determine a reduction position of bits common to the information data of the same type in each group on the basis of the maximum value extracted by the maximum value extraction block and reduce the common bits in accordance with the determination.