Text embedding/steganography in image pixels | ZeroTools

A digital image is defined as a collection of pixels arranged in a grid, with each pixel consisting of four channel components representing red, green, blue, and opacity.

This mechanism implements a lowest bit replacement steganography technology that targets the microscopic structure of spatial color information representation and hides information.

Specifically, we use a method that occupies the lowest single bit of the 8-bit gradation data of each color channel, which is difficult to detect as a color difference due to human visual characteristics, as space for storing information.

One pixel secures a 4-bit data storage area, and if a large-scale digital image file is used, it can function as a concealed communication medium with a considerable capacity.

In this processing procedure, image data is first converted into a one-dimensional array, and a binary stream of secret information is allocated bit by bit while sequentially scanning the index of the target pixel data.

The extraction algorithm works as a reverse process and is responsible for completely reconstructing the original binary stream by rescanning the color information of the target image and extracting and combining only the least significant bits.

By assuming a bitmap format or portable network graphics format that does not pass through image data irreversible compression, complete retention of the least significant bit information is guaranteed, and a robust information-hiding communication path with no data loss is established.

Simply replacing the lowest bit leaves the system vulnerable to binary analysis by a third party, so this system forcibly executes an encryption process on the secret text data before embedding.

The encryption standard adopted here is the Advanced Encryption Standard Protocol, which converts plaintext data into a pseudo-random bit stream using a symmetric key encryption method with a key length of 256 bits.

This highly entropy-enhanced encrypted data is then scattered across each bit-plane of the image data according to mathematical rules.

A pixel value represented by 8 bits can be interpreted as eight layers from the seventh bit plane at the highest level to the zeroth bit plane at the lowest level, but this mathematical model introduces a scattering algorithm that dynamically distributes data across multiple lower bit planes in order to maximize storage efficiency while minimizing visual impact.

This scattering pattern is determined by a spatial pseudo-random number sequence derived from a hash function whose inputs are the initialization vector generated during encryption and the private key defined by the user.

Data mapping based on this spatial pseudo-random number sequence causes each bit of secret information to be stored in a discontinuous and unpredictable order relative to the pixel coordinates of the entire image, making it extremely difficult to detect the presence of a message using statistical stegano analysis.

The most important evaluation criterion for information hiding technology is the invisibility of visual changes to the medium, and it is necessary to quantitatively manage the differences that occur between the original image and the image after information embedding.

This mechanism includes a monitoring system that constantly calculates the peak signal-to-noise ratio and provides it to the user as an indicator to mathematically prove this invisibility.

Since the change in the value of each pixel due to the manipulation of the lowest bit is at most one increase or decrease, the theoretical peak signal-to-noise ratio remains at a level well over 40 decibels, which means that it is completely invisible to the human eye, where no image quality degradation is perceivable.

During the extraction phase, strict recovery key validation logic is activated to allow message extraction.

The passphrase entered by the user is irreversibly converted to a 256-bit verification hash value through a secure hash algorithm, which is a cryptographic hash function, and is compared with the metadata in the header area hidden at specific coordinates within the image file.

This verification mechanism determines whether the input keyboard string is the same as the valid private key used in the encryption process at the early stage of the extraction process, and if a mismatch is detected, it functions as a safety device that immediately forcibly terminates the extraction process to prevent the generation of invalid noise data.

For information hiding systems that handle highly confidential text information and private image files, it is essential to design a system that fundamentally eliminates the risk of data leakage in the communication path.

This system is based on a complete client-side information processing architecture that operates within the framework of a web browser but does not send any data to an external processing server.

The entire computational process, from image loading to pixel array analysis to application of encryption algorithms and manipulation of the least significant bits, is completed solely within the local device's central processing unit and main memory.

By utilizing web assembly and an advanced JavaScript execution engine, we achieve high-speed encryption processing and image encoding processing equivalent to those on the server side within the sandbox environment of the web browser.

By adopting this architecture, users can use all functions even when the Internet connection is completely cut off and offline, and the possibility of information leakage due to packet interception or unauthorized access to the server can be physically and logically reduced to zero.

In addition, the image buffer and encryption key data temporarily expanded in memory are volatilized the moment the browser tab is closed, leaving no trace, so the structure also ensures a high degree of evidence destruction.

The system implements an advanced image comparison analysis module to allow users to visually and numerically confirm the reliability of steganography operations.

This module provides an analysis mode that calculates the difference between the original image and the stego image with embedded information on a pixel-by-pixel basis and highlights only the changed coordinate positions with a unique color.

This highlighting forcefully visualizes changes in the least significant bits, which are usually invisible to the human eye, and serves as an auxiliary analysis method to spatially understand how messages are distributed and stored in which areas of the image.

On the other hand, regarding the operability of the extraction process, a sophisticated user interface has been designed to minimize the cognitive burden on the user.

It is equipped with a one-click extraction mechanism where all the extraction algorithms and decryption processes are automatically chained behind the scenes by simply dropping the image file with hidden information into a predetermined area and entering the pre-agreed private key into the input field.

The extraction progress is displayed in real-time as a percentage of memory read, and after a few milliseconds of calculation, the original secret text message is instantly restored to the screen in a format that can be transferred to the clipboard.

This information hiding mechanism by replacing the least significant bit is not only a means of secret communication, but also has a wide range of practical applications as digital watermarking technology for copyright protection and authenticity certification of digital content.

For example, by embedding the creator's identification information and time stamp invisibly as a watermark in the image file of a digital certificate or confidential design drawing issued by a company, an electronic signature-like operation is realized that makes it possible to track future falsification or unauthorized copying of data.

Additionally, this tool serves as an educational platform for students in the field of information security to give them a practical understanding of the basic principles of steganography and the effectiveness of combining cryptographic elements.

The theory of information hiding and the theory of stegano analysis that detects it are two sides of the same coin, and directly manipulating and observing the behavior of embedded algorithms and scattering models provides deep insight into information hiding technology.

In practice, users can appropriately judge the tradeoff between limiting embedding capacity in favor of image quality or consuming multiple lower bit planes in favor of information content while observing fluctuations in the peak signal-to-noise ratio, allowing users to formulate guidelines for deriving parameter settings optimized for the application.