# US 7,412,104 B2 · Optimized lossless data compression methods

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 7,412,104 B2
- Document ID: US7412104B2
- Status: Granted
- Kind: B2
- Application number: 11/343,908
- Filed: 2006-01-30
- Published: 2007-03-08
- Granted: 2008-08-12
- Earliest claimed: 2006-01-30
- Inventors: Andrei Talapov; Daniel Kilbank; John Quagliano; Thomas Lyon Gideon; Ted Malaska; Janek Claus; Hadon Nash
- Web page: https://patentgrep.com/patent/US7412104B2

## CPC

- H04N 19/94: Vector quantisation
- H04N 19/50: using predictive coding (H04N19/61 takes precedence)
- H04N 19/60: using transform coding
- H04N 19/85: using pre-processing or post-processing specially adapted for video compression
- H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION

## Abstract

A method for losslessly encoding and compressing and decoding original, raw, raster data, and other files by optimizing the results of preprocessing and transformation techniques. Four stages are involved: pre-processing during which the data is replaced by predictive values and the deviations from the predictive value; mapping each block through a transform sequence using a lossy reversible mapping; minimizing the joint entropy of the transformed data and deviation data by varying parameters for the predictive and transform step; and encoding the transformed sequence. At each stage of the compression of the data, different techniques are tried and compared and an optimal technique used to carry out that state.

## Claims

### Claim 1

A method for lossless compression of data by a digital processor employing a lossy transformation defined as a function of parameters, comprising
compressing an initial set of data yinit by a parametrized lossy transformation into a set of data ytrans,
reversing the transformation to provide a lossy version of the initial data, ylossy,
determining a difference between the initial data and the reverse transformed lossy version of the data by calculating, ydifference=yinitial−ylossy.
minimizing the sum of entropies of ydifference and ytrans as a function of the transformation parameters to determine a value of the parameters for optimal compression,
using the values of the parameters that maximize compression to perform a compression.

### Claim 2 (depends on 1)

The method of claim 1 further comprising decompressing the data by reversing the compression transformation.