Steganography Encoding System Pixel Classification

Solution Overview

Problem

Existing steganography techniques suffer from visible borderline pixels and loss of data during transmission, leading to degraded image quality when embedding secret data in images.

Innovation Solution

A computer-implemented steganography system comprising an encoding machine and a decoding machine, which uses target pixel selection rules and gray value calculations to embed and extract secret data, minimizing error and preserving image quality by selecting target pixels based on predetermined rules and calculating the number of bits to be embedded or extracted.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If secret data is embedded in image pixels using conventional steganography techniques, then data hiding capacity is achieved, but image quality degrades with visible borderline pixels and higher error rates

Engineering Contradiction:

Improvesecret data capacityVSAvoidimage quality

Core Design Contradiction:

Quantity of substanceVSManufacturing precision

Solution Approach 1:

The patent applies local quality by classifying pixels into different types (corner, edge, smooth, textured) and applying different embedding strategies to each type. Corner pixels use 2-bit embedding, edge pixels use 1-bit embedding, and smooth/textured pixels use adaptive embedding based on local variance thresholds. This localized approach ensures that embedding operations are optimized for each region's characteristics, minimizing visible artifacts while maximizing data capacity.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The patent implements partial action by selectively embedding data only in pixels that meet specific criteria (e.g., pixels with sufficient local variance, pixels below certain intensity thresholds). The embedding process is performed partially on the image, skipping pixels that would cause visible degradation. This selective partial embedding maintains image quality while achieving adequate data hiding capacity.

Inventive Principle:

Principle #16Partial or excessive action

2Quantity of substance

If pixel values are adjusted to embed secret data, then data embedding is achieved, but data loss occurs during transmission

Engineering Contradiction:

Improveembedded dataVSAvoiddata transmission integrity

Core Design Contradiction:

Quantity of substanceVSReliability

Solution Approach 1:

The patent applies preliminary action by calculating and storing the difference values (original pixel value minus embedded pixel value) before transmission. These difference values are used as verification data to detect and correct transmission errors. The system prepares error detection and correction mechanisms in advance, allowing reliable data extraction even when transmission errors occur.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent implements feedback through a verification mechanism that uses the stored difference values to check the integrity of received pixels. When a pixel is received, the system calculates the expected value based on the difference and compares it with the received value. Discrepancies trigger error correction procedures, ensuring reliable data extraction despite transmission imperfections.

Inventive Principle:

Principle #23Feedback

3Quantity of substance

If conventional embedding methods are used to hide secret data, then data hiding is achieved, but visible borderline pixels appear degrading image quality

Engineering Contradiction:

Improvehidden secret dataVSAvoidvisible artifacts

Core Design Contradiction:

Quantity of substanceVSObject-affected harmful factors

Solution Approach 1:

The patent applies local quality by classifying pixels into different types (corner, edge, smooth, textured) and applying different embedding strategies to each type. Corner pixels use 2-bit embedding, edge pixels use 1-bit embedding, and smooth/textured pixels use adaptive embedding based on local variance thresholds. This localized approach ensures that embedding operations are optimized for each region's characteristics, minimizing visible artifacts while maximizing data capacity.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The patent implements parameter changes by dynamically adjusting the embedding strength and method based on local image characteristics. The system calculates local variance for each pixel and adapts the embedding parameters accordingly - using weaker embedding in smooth regions and stronger embedding in textured regions. This adaptive parameter adjustment prevents visible borderline pixels by matching the embedding intensity to the local image structure.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS10489874B2Computer implemented system and method for steganography

Publication Date: 2019.11.26 ZENSAR TECHNOLOGIES

AI summary

A computer implemented system for steganography comprises an encoding machine for encoding secret data into a cover image to generate the steganography image, and a decoding machine for extracting the secret data from the steganography image. The encoding machine comprises a first memory, a first processor, an input module and an encoder module. The decoding machine comprises a second memory, a second processor, a receiver module, and a decoder module. The system has a comparatively lower error rate (approx. 19.5-21.5 percent) than the conventional steganography method Optimal Pixel Adjustment Procedure (OPAP).