ITERATIVE BACKWARD REFERENCE SELECTION WITH REDUCED ENTROPY FOR IMAGE COMPRESSION

Patent №

US 8,175,403

Granted

2012-05-08

Filed 2008

Owner

GOOGLE INC.

AI components

1

planning

Assignment

Recorded

Dataset

AIPD

2023_r1 edition

Application

12134182

A system (and a method) are disclosed for optimizing lossless data compression in digital images systems iteratively. The system includes a backward reference module and an entropy encoder. The backward reference module comprises three sub-modules: a backward reference selection module, a backward reference statistics measurement module and a backward reference selection control module. The backward reference selection module initially selects backward references that have shorter backward distances. The statistics measurement module analyzes the backward references of input data. The backward reference selection module updates the backward references based on the statistics of previously generated backward references at subsequent selection iterations. The backward reference selection control module controls the number of selection iterations needed for optimized data compression performance.

PlanningH03M 7/3097H04N 19/64H04N 19/91

AI classification

Planning0.72
Machine learning0.33
Evolutionary computation0.12
Vision0.01
AI hardware0.00
Natural language0.00
Knowledge representation0.00
Speech0.00

Ownership

GOOGLE INC.

assignment · 210590391

Assignors

ALAKUIJALA, JYRKI

On an employer assignment, the assignors are typically the inventors.

© 2026 NYSGPT2525 LLC