Variable Length Encoding Storage Size Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional software applications for compressing digital images using variable length encoding do not predict storage size effectively, requiring users to generate and measure compressed representations multiple times to find suitable quality parameters, leading to inefficiencies and delays, especially when dealing with large datasets.
Innovation Solution
A method and apparatus for predicting the storage size of digital data in representations using variable length encoding by generating bit width distributions based on the frequency of occurrence of different bit widths, allowing for the estimation of storage size without generating the output representation, and using this estimation to determine compression parameters that satisfy global constraints such as maximum storage size.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional software applications generate and measure compressed representations multiple times to find suitable quality parameters, then storage size prediction accuracy improves, but processing time and productivity deteriorate
Solution Approach 1:
The patent applies preliminary action by analyzing the input representation and generating bit width distributions before actual compression occurs. The system predicts storage size by examining code value patterns, bit width frequencies, and representation component characteristics in advance, eliminating the need for multiple trial compressions and measurements to determine suitable quality parameters.
2Manufacturing precision
If multiple compressed representations are generated to determine quality parameters, then compression parameter optimization improves, but loss of time increases
Solution Approach 1:
The system performs preliminary analysis of the input representation to predict how different quality parameters will affect compression results. By generating bit width distributions and analyzing code value patterns beforehand, the system can determine optimal compression parameters without actually generating multiple compressed representations, thus avoiding time loss.
Solution Approach 2:
Instead of generating multiple actual compressed representations, the system creates predictive models and bit width distributions that copy the essential characteristics of what the compressed data would look like. This allows parameter optimization based on predicted outcomes rather than actual multiple generations.
3Quantity of substance
If variable length encoding is used for compression, then storage size reduction improves, but difficulty of detecting and measuring storage size increases
Solution Approach 1:
The patent segments the compression analysis into distinct components: identifying code values, determining bit widths, generating bit width distributions, and calculating predicted storage size. This segmentation makes the measurement process systematic and manageable despite the complexity of variable length encoding, allowing accurate storage size prediction without generating compressed data.
Data Source
AI summary
Methods, systems and apparatus, including computer program products, for processing digital data. An approximate storage size is predicted for an output representation that uses variable length encoding. The approximate storage size can be used to determine one or more compression parameters to satisfy a global constraint, such as a maximum storage size for a compressed representation of the digital data. In a user interface, storage sizes can be graphically represented for multiple images. In the graphical representation, the storage size is represented for each image by a corresponding graphics object that includes a visual representation of the image and has a linear size that is proportional to the storage size of that image.