Encoding Data Arrays Using 1D Marginal Approximation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data encoding methods for streams of data arrays, such as video frames, face challenges in optimizing compression while maintaining quality, particularly in resource-constrained devices like portable devices, where memory and processing resources are limited.
Innovation Solution
The method involves generating two one-dimensional arrays from a two-dimensional array of data values representing properties of data elements, which are then used to approximate data values for encoding, reducing memory storage needs and processing resources while maintaining accurate encoding decisions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If statistical information is collected and stored in a two-dimensional array during encoding, then encoding quality and compression optimization are improved, but memory consumption and processing resources increase
Solution Approach 1:
The patent extracts only the essential statistical information from the complete two-dimensional array by computing one-dimensional marginal distributions (row sums and column sums). This extraction retains the critical statistical properties needed for encoding decisions while discarding redundant data, thereby reducing memory requirements while maintaining encoding quality.
Solution Approach 2:
The patent segments the two-dimensional statistical array into two separate one-dimensional arrays representing row and column marginal distributions. This segmentation allows the system to store and process statistical information more efficiently by treating dimensions independently, reducing overall memory footprint while preserving the statistical characteristics needed for optimal encoding.
2Productivity
If detailed statistical analysis of data element properties is performed, then encoding optimization is improved, but processing time and computational resources increase
Solution Approach 1:
The patent extracts only the necessary statistical summaries (marginal distributions) from complete data analysis. By computing only the row and column sums rather than processing every individual data element in detail, the system achieves encoding optimization while significantly reducing processing time and computational burden.
Solution Approach 2:
The patent applies partial action by performing statistical analysis only on aggregated data (marginal distributions) rather than on every individual element. This partial analysis provides sufficient statistical information for encoding decisions without the excessive processing required for complete element-by-element analysis, thus balancing optimization needs with processing time constraints.
Data Source
AI summary
An apparatus for encoding arrays of data elements of a stream of arrays of data elements. The apparatus includes encoding circuitry for encoding an array of data elements. The apparatus also includes processing circuitry for producing two one-dimensional arrays of data values from a two-dimensional array of data values. The data values of the two-dimensional array are representative of the properties of the array of data elements. The data values of the two one-dimensional arrays are representative of the data values along the two dimensions of the two-dimensional array respectively. The processing circuitry is operable to use the two one-dimensional arrays to approximate data values in the two-dimensional array, and to use the approximated data values of the two-dimensional array when encoding arrays of data elements in the stream of arrays to be encoded.