Image Data Compression via Linear Prediction and Pixel Grouping

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing image compression methods are inefficient, leading to high computational requirements and the generation of artifacts in decompressed images, especially in high-resolution images from digital cameras.

Innovation Solution

A compression process that groups pixels into predetermined groups based on their arrangement in rows, estimates pixel values using weighted combinations of neighboring pixels, and encodes the differential values using an efficient coding procedure.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of substance

If transformation into frequency domain is used for compression, then compression factor is improved, but computational effort and technical complexity increase significantly

Engineering Contradiction:

Improveimage data volumeVSAvoidcomputational effort

Core Design Contradiction:

Loss of substanceVSDevice complexity

Solution Approach 1:

The image is divided into multiple blocks of pixels, and each block is processed independently through linear prediction. This segmentation allows parallel processing and reduces computational complexity compared to global frequency transformation methods.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent replaces complex frequency domain transformation (mathematical mechanism) with simple linear prediction using neighboring pixel values (arithmetic operation). This substitution dramatically reduces computational effort while maintaining compression effectiveness.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Loss of substance

If frequency domain transformation is used for compression, then compression factor is improved, but image quality deteriorates due to artifact formation

Engineering Contradiction:

Improveimage data volumeVSAvoidimage artifacts

Core Design Contradiction:

Loss of substanceVSObject-affected harmful factors

Solution Approach 1:

The patent replaces frequency domain transformation with linear prediction based on spatial relationships between neighboring pixels. This approach preserves image quality by avoiding the high-frequency coefficient quantization that causes artifacts, while still achieving effective compression through entropy coding of prediction errors.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

Each pixel's value is predicted using only its neighboring pixels within the same block, making the compression process self-contained and avoiding introduction of external artifacts. The prediction naturally adapts to local image characteristics without generating the washed-out appearance associated with frequency domain methods.

Inventive Principle:

Principle #25Self-service

3Loss of substance

If conventional compression methods are used, then compression is achieved, but processing speed is insufficient for real-time applications

Engineering Contradiction:

Improveimage data volumeVSAvoidcompression speed

Core Design Contradiction:

Loss of substanceVSProductivity

Solution Approach 1:

By dividing the image into small independent blocks and applying simple linear prediction to each, the patent enables highly parallel processing. This segmentation strategy dramatically increases compression speed compared to sequential frequency domain transformations, making real-time compression feasible.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The replacement of computationally intensive frequency transformation with simple arithmetic operations (addition, subtraction, multiplication by small integers) dramatically reduces processing time per pixel, thereby increasing overall compression productivity.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentEP3624455B1Method, computer program product and storage medium for compressing image data

Publication Date: 2025.04.16 ARNOLD & RICHTER CINE TECHNIK GMBH & CO BETRIEBS KG

Figure 1
Figure 2
Figure 3

AI summary

A method for compressing image data comprising a multitude of pixel values assigned to a given pixel is described, wherein, for at least some pixels, the pixel values are compressed according to the following steps: assigning the pixel in question to one of several pixel groups according to a predetermined grouping scheme, wherein each of the several pixel groups is defined by the pixels of several predetermined pixel rows or by a subset of pixels from several predetermined pixel rows; determining an estimated pixel value of the pixel in question as a function of the pixel value of at least one other predetermined pixel using at least one estimation rule, wherein the at least one other predetermined pixel is chosen as a function of the pixel group of the pixel in question;Calculating a difference value for the pixel in question, corresponding to a predetermined relationship between the pixel value and the estimated pixel value of the pixel in question; and encoding the difference value using a coding procedure.