Enhancing Data Protection in Digital Communication: A Novel Method of Combining Steganography and Encryption

Abstract

In today's highly digitized landscape, securing digital communication is paramount due to threats like hacking, unauthorized data access, and network policy violations. The response to these challenges has been the development of cryptography applications, though many existing techniques face issues of complexity, efficiency, and limitations. Notably, sophisticated intruders can easily discern encrypted data during transmission, casting doubt on overall security. In contrast to encryption, steganography offers the unique advantage of concealing data without easy detection, although it, too, grapples with challenges. The primary hurdles in image steganography revolve around the quality and payload capacity of the cover image, which are persistently compromised. This article introduces a pioneering approach that integrates image steganography and encryption, presenting the BitPatternStego method. This novel technique addresses prevalent issues in image steganography, such as stego-image quality and payload, by concealing secret data within image pixels with identical bit patterns as their characters. Consequently, concerns regarding the quality and payload capacity of steganographic images become obsolete. Moreover, the BitPatternStego method boasts the capability to generate millions of keys for the same secret message, offering a robust and versatile solution to the evolving landscape of digital security challenges.

1. Introduction

Ensuring information security has emerged as a paramount concern in the realm of information technology and communication, especially with the advent of the Internet. The expansion of digital communication networks has heightened the public's expectations for privacy and security in data transmission. Various strategies have been implemented to protect data integrity and facilitate secure data transmission processes, mitigating the risks of unauthorized access or data leakage due to irresponsible actions.

For centuries, techniques in cryptography and steganography have been employed to safeguard sensitive data. Cryptography involves transforming confidential information into ambiguous data through mathematical algorithms and keys. While still identifiable, this obscured data can attract the attention of unintended third parties, leading to potentially dire consequences. The primary objective of cryptography is to ensure that even if intercepted, the encrypted message remains confidential and incomprehensible to unauthorized individuals.

In contrast, steganography focuses on the concealment of sensitive data to evade detection [1]. It revolves around hiding the communication itself within another medium, such as an image, audio file, or video, in a manner imperceptible to human senses [2]. The most popular carrier format on the Internet is images because of their high frequency on the Internet. There are many existing steganography techniques for hiding secret messages in images with their respective pros and cons [3]. Obviously, techniques with modest strengths would increase data vulnerability to discovery, extraction, and attack.

Image steganography has many applications. In addition to sending sensitive information embedded in an image through channels such as email and social media, it embeds a digital watermark in the image as well [4]. These watermarks can be used for copyright protection or image authentication. Moreover, image steganography can be employed to store personal or sensitive information, such as passwords or financial data, within an image. Therefore, image steganography is an effective technique for preventing unauthorized parties from accessing sensitive information [3].

The accuracy in identifying the pixels essential for concealing the secret message is a crucial aspect of steganography embedding methods. Furthermore, it is insufficient merely to identify the current embedding location; it is equally vital to establish a sequence of locations. Thus, there is a need for an effective and adaptable technique to define the entire path of embedded locations.

Moreover, an effective steganography technique should exhibit high visual quality with minimal distortion and provide a sufficient payload [5]. While certain steganography approaches prioritize improved stego-image quality, they often come with inherent payload constraints. The trade-off between hiding payload and the quality of the stego-image is obvious, as reducing payload enhances un-detectability. Striking a balance among various steganography requirements poses a challenge for steganographers [6][7].

This study introduces a novel method known as BitPatternStego, aimed at concealing desired text bits within the blue channel of an RGB image. The approach involves utilizing an identical bit pattern shared between the text and the blue channel, ensuring a seamless integration without altering the cover image. By maintaining this consistency, the method successfully avoids any distortion, enhancing security through undetectability. Moreover, it concurrently augments the image's data storage capacity (payload), enabling it to accommodate a substantial volume of data, potentially unlimited.

2. Related work and Background

The development of image steganography has gathered extensive research attention as a result of its superiority to some limitations of cryptographic methods whose computational complexity is enormous as well as their ability to attract the attention of attackers [8]. Consequently, image steganography is used in many practical applications, including secure mobile computing [9], secure online voting systems [10], and secure communication between two parties [11].

Implementing image steganography methods is generally divided into two domains: the frequency domain and the spatial domain [12]. In a frequency-domain method, it manipulates the image's orthogonal transformation, which involves magnitude and phase, which represent frequency and space respectively. This method consists of an algorithm plus the transformed image. Although it is more flexible against image processing attacks, the frequency domain has a limited payload [13] and is not suitable for many real-time applications [14]. Conversely, spatial domain methods [15][16] act directly on the image's pixels, replacing the least significant bit (LSB) with the embedded secret message. In this regard, spatial domain algorithms are simpler, faster, more powerful, and more resistant to attacks [17]. In addition, despite having a good payload capacity, it lacks flexibility against statistical attacks and causes slight changes to the cover media [18][13].

The LSB method [12] is one of the spatial domain techniques based on the RGB (Red Green Blue) color model. The pixel's LSB is replaced with a secret message bit based on a secret key. It begins with one bit per pixel (BPP) and proceeds to two or three LSB or even more bits per pixel. The more substituted LSBs there are, the more obvious the image distortion [19].

Numerous efforts have been made to mitigate image distortion, as evidenced in works like [19][6]. Tsai et al. [20] researched into the concept of pixel relationships, a crucial consideration when concealing secret data within an image's pixels. Failing to account for this can lead to diminished visual quality and payload, impacting both smooth and edgy areas of the input image. To enhance payload and visual quality, [20] proposes a strategy of concentrating data hiding in edgy pixels while reducing it in smooth areas. This idea has inspired several researchers, resulting in various improved versions of Tsai's approach, as seen in [21][22] and [23]. Additionally, efforts have been directed towards enhancing visual quality. For instance, Chang et al. [24] devised a strategy to specify pixel adjustments for embedding data, further refined by Tsai et al. through pixel value differencing. Despite improvements in visual quality and payload capacity in both approaches, suitable for a wide range of applications, the original cover image still struggles with distortion and payload challenges.

In [25], the Sequential Color Cycle (SCC) technique serves as an illustration of a steganography method overwhelmed by distortion and payload issues. This approach involves cyclic modifications to the color channels (RGB) of each pixel, following a specific pattern. This alteration facilitates the concealment of one to four bits within the least significant bits (LSBs). For instance, in the case of the one-LSB pattern, secret message bits are distributed among the LSBs of the red, green, and blue channels. While SCC offers enhanced security and an increased payload compared to the original LSB method, it comes at the cost of declined image quality. This degradation may result in challenges related to uncovering hidden information and detecting the employed cycling pattern. Furthermore, SCC still exhibits a limited payload capacity, particularly when dealing with large hiding text.

In [26], there's another instance of limitations in cover image payload and visual quality. Despite the authors' efforts to enhance security through a simple randomization technique for embedding and extracting data from the stego-image, a significant portion of the cover image payload is wasted. The algorithm outlined in the paper, exemplified by the pattern (10), involves using one pixel to conceal secret bits and skipping the next pixel, resulting in a 50% payload waste. Similarly, the key pattern (1011) dictates that three out of four pixels will be utilized for concealing hidden text, leading to a 25% payload wastage, along with a reduction in stego-image quality. Despite the authors' attempts to minimize wasted payload capacity, it unintentionally affects the quality of the stego-image.

As a result, numerous steganography techniques strive to enhance security levels, while others focus on improving the quality of the stego-image [27][28]. However, these approaches often come with inherent capacity limitations. A proficient steganography technique should aim to strike a harmonious balance among various steganography requirements [6][7].

3. Preliminaries

3.1 Bits for Digital Text and Colors.

Bits represent both text data and colors (in digital forms) through encoding. As known, encoding is the process of converting information into a format that can be stored or transmitted digitally. In the case of text data, the ASCII code is a common encoding system that uses 8 bits to encode each character or symbol. Therefore, 2^8 = 256 different combinations of 0s and 1s that can be used to represent 256 different characters. For example, the letter "A" is represented by the binary code 01000001, see ASCII code table.

On the other side, the colors (in digital images) in each pixel are encoded by a combination of bits that indicate the intensity of the red, green, and blue components of the color. This is known as RGB (red, green, blue) encoding. Each component is usually represented by 8 bits, which means that there are 256 possible values for each component, ranging from 0 to 255 (the 0 value represents no color, while 255 represents the maximum amount of color). Therefore, each pixel can be represented by a combination of 24 bits (8 bits for red, 8 bits for green, and 8 bits for blue) (see Fig. 1), allowing for a total of 16,777,216 (2^24) different colors.

Fig. 1. Pixel structure at RGB images

3.2 LSB Steganography: A Basic Overview

In the LSB steganography technique, the last bit of each pixel is replaced (changed) with one bit of the secret text. For example, if a secret text of 100 characters is to be hidden in a cover image, 800 pixels are needed. This is because 100 letters require 800 bits (ASCII Code system) and for each bit, one pixel is needed. In the embedding process, there are two possibilities. The first one happens when changing the bit with the same bit type, 0 by 0, or 1 by 1. While the second possibility is to replace the bit with a different bit type, for example, 0 by 1 or 1 by 0, see Fig. 2.

Fig. 2. illustrates image distortion in scenario (b) and the absence of distortion in scenario (a).

As for distortion of the stego-image, the first possibility produces no distortion since the original pixel bit remains unchanged. In the second possibility, distortion occurs in the original image due to a change in bit type. Other LSB techniques would require more LSBs for the embedding process [19], for instance, 2 LSBs, 3 LSBs, or even 4 LSBs, see Fig. 3. With more LSBs, the cover image payload would be higher, but distortion would be higher as well. This would result in poor quality and easy detection of the secret message.

Fig. 3. Illustrates LSB techniques necessitating a greater number of LSBs for the embedding process, such as 2 LSBs, 3 LSBs, or even 4 LSBs, leading to increased image distortion.

In summary, with dimensions of 256x256 pixels encompasses a total of 65,536 pixels, allowing for the concealment of around 8,192 characters if only one least significant bit (LSB) is utilized, and twice this size if two LSBs are employed. Earlier methods involved embedding through sequential or random pixel selection, as seen in [29][25] and [26]. However, these approaches ultimately pose limitations on payload capacity and the risk of compromising visual quality. The proposed technique in this research, BitPatternStego, addresses the crucial concerns of both image quality and payload capacity.

4. BitPatternStego Methodology

This study proposes to map the corresponding (identical) bit patterns in both the secret text and the cover image, called BitPatternStego. In the literature, selecting the cover image pixels for embedding the secret text is done sequentially or randomly.

BitPatternStego systematically scans the pixel channels of a cover image to determine the bit pattern corresponding to the alphabet (a-z) and some commonly used characters (!, ?, space, etc). Initially, the focus is solely on the blue channel, while potential utilization of other channels (red and green) to triple the payload capacity is reserved for future work. However, for the purposes of this study, the collected data from the blue channel proves sufficient.

After analyzing numerous commonly used images on the internet, it is observed that each character may have thousands of identical blue channels corresponding to its bit pattern (refer to appendices B, C, D). Consequently, the BitPatternStego technique compiles all indexes corresponding to each character into a separate array.

During the embedding process, a random selection of indexes is made for each character, enhancing security compared to sequential embedding. The result is an unchanged output image quality, as no alterations are made to the bit types (refer to Fig. 2). Furthermore, the cover image's payload becomes virtually limitless, accommodating a substantial volume of number of characters, as their corresponding indexes are subject to repetitive selection as needed.

The embedding and extraction processes of BitPatternStego are inherently reversible and straightforward. During the embedding process, as outlined earlier, the cover image is scanned to identify pixels with bit patterns corresponding to alphabet characters and symbols. Subsequently, secret text characters are assigned to the respective pixels' indexes, resulting in an array of indexes for the secret text characters (see Fig. 4).

Fig. 4. displays an array where each number corresponds to a specific pixel index in the stego-image, containing an identical bit pattern to the secret message characters in sequential order.

For the extraction process, it is imperative to determine the indexes array, stego-image name, and size. Utilizing the ASCII code system, the recipient must read the bits of the pixel channel (specifically the blue channel) indicated by the indexes array and convert them into characters, thereby reconstructing the complete secret text, as will be explained in detail in section 5.3.3, "Extraction Stage."

4.1 BitPatternStego Stage 1: The cover image.

1. Choose a JPEG file for the cover image. (which supports 24-bits of RGB color per pixel).

2. Flattening the cover image indexes. which involves copying the 24 bits of each pixel into the cells of one dimensional array (1D array). In this process, each cell within the new array accommodates 24 bits of each pixel, as illustrated in Fig. 5. It's noteworthy that this flattening step is crucial for minimizing the size of the resulting indexes array; Instead of having two coordinate values (row and column) for each pixel, it produces only one single coordinate value for each pixel, which will be referred to as the index.

Fig. 5. Flatten the coordinates of the cover image (represented by two coordinates) into a one-dimensional array (indicated by a single coordinate index).

However, the index can be easily calculated based on the two coordinate values (row, column) of the pixel and the image’s width (number of columns). The formula for this computation is:

Index = Row × I_width + Column

As illustrated in Fig. 5, the index (2,3) can be substituted with the single index (11) through the calculation 2 * 4 + 3.

3. Scan (read) the 8-bits of the blue channel at each cell in the 1D array.

4. Create individual arrays to store corresponding indexes for each character, where, the 8-bit pattern of the blue channel corresponds with the alphabet and commonly used characters (refer to Appendices B, C, D). see sample Table 1.

Table 1. depicts the single-coordinate indexes where the 8-bit pattern of the blue channel corresponds to each alphabet character

Here is the pseudocode for stage 1 of BitPatternStego:

4.2 BitPatternStego Stage 2: The secret message.

1. Select the text that needs to be embedded.

2. Convert text characters to lowercase. (e.g. 'A' to 'a').

3. For every character in the secret text, randomly assign one index from the array of indexes (where the 8-bit pattern of the Blue-channel corresponds to the given character). (See Fig. 6).

Fig. 6. Assign randomly one coordinate index of the cover image to each character of the secret text “help me”, where the 8-bits of blue channel corresponds.​​​​​​​

4. Establish an array to store the selected indexes corresponding to characters in the secret text. Consequently, a string of 10 characters requires an array of size 10, each holding a unique index (refer to Fig. 6).

Here is the pseudocode for stage 2 of BitPatternStego:

5. BitPatternStego Implementation

5.1 Underlining Technologies

The Java Programming Language has been utilized to implement BitPatternStego using the NetBeans IDE. Java is a powerful programming language with several classes offering a variety of image manipulation tools and techniques. In terms of image processing using Java, a certain type of object with the name 'BufferedImage' needs to be implemented and created. Through this object, all methods concerning image editing can be accessed. Nevertheless, it is worth mentioning here that Python could be another option for implementation as it is one of the easiest programming languages to learn and utilize since it has a wide array of built-in libraries designed to help with different aspects of programming. Despite that, using Java wasn't difficult as we had mastered it.

5.2 BitPatternStego Dataset

In this study, BitPatternStego is implemented using a dataset consists of 18 standard test images, which vary in sizes such as 512x512, 256x256, or 2560x1150 pixels. Refer to Appendix A for details. The images were in JPEG format which is ideal for RGB files as it is a reasonable middle-ground between file size and quality, and it can be read almost anywhere. Other formats can be used as well, but they are not ideal for RGB like PNG and GIF. Other formats, such as TIFF, EPS, PDF, and BMP, should be avoided for RGB purposes. These formats are not compatible with most software and can be unnecessarily large in terms of data [30].

As mentioned earlier, that RGB images are made up of red, green, and blue color channels. And each pixel in the image is represented by a combination of these three primary colors. These colors (red, green, or blue) are typically represented using 8 bits per color channel (3 X 8 = 24 bits per pixel), allowing for 16.7 million possible color combinations.

In an RGB image, the color of each pixel is determined by the intensity of each color channel. RGB images are widely used in digital photography, computer graphics, and other applications where color accuracy is paramount. They are also the standard format for displaying images on computer monitors and other electronic devices [31].

5.3 BitPatternStego Implementation Example.

5.3.1 Embedding stage

Consider the following inputs for embedding secret text into a cover image:

• Secret text: “tic tac toe”

• Cover Image: mandril_color.jpg (see Fig. 7.a).

After scanning the cover image, mandril_color.jpg, Fig. 7.b displays the indexes corresponding to the most frequently used characters and symbols.

Fig. 7.a. 512 X 512 cover image size in jpg format. b. Upon scanning the cover image, the amount of bit patterns corresponds to each character/symbol is calculated.​​​​​​​

For instance, the letter 'a' can be represented by 1579 indexes of identical bit pattern in the cover image. The 1579 indexes are stored in a separated array for each character. See Table 1.

As depicted in Fig. 7.b, numerous pixels (indexes) exhibit identical bit patterns for each character. This implies the potential to generate millions of distinct encryption keys for a single secret message. Presented below are just five keys for the provided secret text (refer to Fig. 8).

Fig. 8. Illustrate five district keys for the given secret message “tic tac toe”​​​​​​​

Significantly, in the provided example, the alphabet character 't' is observed with 15 different indexes among the five generated keys, each sharing an identical bit pattern. These 15 indexes represent only a fraction of the total number of indexes for the letter 't,' which amounts to 1186, as illustrated in Fig. 7.b. The BitPatternStego algorithm employs a random selection of indexes for each character, potentially resulting in the repetition of index selection, further elaboration on this aspect will be provided in the subsequent section.

Various randomization techniques, recognized for their efficacy in enhancing encryption security [26][32] and [33], have been widely applied in steganography. Consequently, the random selection of indexes serves to mitigate the vulnerabilities associated with breaking ciphers through methods that rely on letter frequency analysis, such as those employed in mono-alphabetic substitution ciphers, Caesar shift ciphers, and Vatsyayana cipher

The proposed technique, BitPatternStego, incorporates various intermediate encryption methods to enhance security. This includes mapping the ASCII code of the text to the image RGB code and employing a randomization technique for selecting embedding pixels. Additionally, the encrypted text is structured as a set of single coordinate values instead of two coordinates for the indexes. While the resulting encrypted text can undergo established encryption techniques like AES, Homomorphic Encryption, Paillier Cryptosystem, or Lattice-based Encryption for heightened complexity and security, our research intentionally maintains simplicity. This decision leaves room for future enhancements and improvements in the ongoing exploration of these concepts.

5.3.2 Index Repetition and Cover Image Suitability

BitPatternStego algorithm employs a random selection of indexes for each character, which can lead to instances of index repetition. Such repetition may occur coincidentally or when the hidden message is extensive and the cover image is small, resulting in a limited number of indexes for certain letters. Therefore, it is recommended to utilize appropriately sized cover images that match the text size, along with rich cover images containing a sufficient source of data bits.

For illustration, in Fig. 9, a secret message containing 100 words (618 letters) demonstrates differing repetition patterns across two cover images of identical dimensions. In the first cover image, labeled 'House', all letters, except for two instances of index repetition, do not repeat. Similarly, in the 'mandril_colore' cover image, the majority of letters do not repeat, with only six exceptions. It's important to highlight that even repeated trials with the same cover image can yield varying numbers of repetitions, hence concealing secret text in various ways. This variability adds an additional layer of security and complexity to the concealment process, making it more robust against potential decryption attempts.

Fig. 9. The number of repeating indexes in two images for a 100-word secret message differs slightly. The House image (black) exhibits fewer repetitions compared to the mandril_color image (grey), indicating that it contains a greater abundance of indexes for each letter.​​​​​​​

When managing longer secret messages, such as a 5000-word (27144-character) composition, within images of two different sizes (512 x 512 and 2560 x 1150 pixels), a significant increase in the occurrence of repeated indexes can be observed in the smaller size image. Conversely, the larger size image exhibits far fewer repetitions, as well as many letters still show no repetitions, as demonstrated in Fig. 10.

Fig. 10. illustrates the frequencies of index repetition observed in two cover images of different sizes (512x512 in grey and 2560x1150 in black), each concealing a 5000-word message.​​​​​​​

The concept is that as the data volume increases, so does the frequency of indexes, regardless of the quantity of data. Yet, the repetition of indexes does not yield evident outcomes, complicating the task of letter prediction, especially with the inclusion of additional characters and symbols in the secret message (e.g., spaces, dots, numbers, etc.).

Accordingly, it is fundamental to acknowledge that factors such as image size, pixel density, and available bit patterns play a significant role in dispersing index frequencies. Careful consideration of these factors is essential to mitigate the risk of statistical analysis or detection.

When considering the suitability of a cover image, it's crucial to recognize that certain images may lack the necessary bit patterns in their blue channels to accommodate specific letters, rendering them incompatible with the proposed method. For example, in Appendix D, images 5 ("woman_darkhair") and 16 ("lena_color_512") demonstrate a deficiency of indexes for the "space" character, indicated by the absence of any indexes. Consequently, it is prudent to confirm the availability of corresponding bit patterns for all characters within the cover image prior to employing the BitPatternStego method.

5.3.3 Extraction stage

The decryption process for the BitPatternStego method relies on several key elements: the indexes array, which indicates the pixel locations in the image corresponding to characters of the secret message, the agreed-upon image name, and its size. Using these components, the recipient retrieves the bits from the blue channel of the specified indexes (pixels) and converts them into text data using the ASCII code system to reconstruct the original message.

For example, if key-1 in Fig. 8, along with the image name and size (mandril_color.jpg, 512x512), is provided to the recipient, they can extract the corresponding bit patterns to the characters of the secret message. This can be achieved by either flattening the stego-image, as described in point 2 of section 4.1, or converting each value of key-1 (indexes array) into two coordinate values (row, column) for the image.

To clarify, according to key-1, the first character of the secret message is located at index 207365 of the stego-image. If the image has been flattened, the recipient can directly extract the bits from the specified index. Otherwise, the recipient needs to convert each index into two coordinate values (row, column) of the image. This conversion involves a straightforward calculation:

Row = index Div I_width,

Column = index mod I_width

Where, div is the integer division, mod is the modulus operation, and I_width is the Image width (number of image’s columns).

Therefore,

Row = 207365 Div 512 = 405,

Column = 207365 mod 512 = 5

This suggests that index 207365 in the indexes array corresponds to row 405 and column 5 of the stego-image. Consequently, the recipient must extract the blue channel bits at this specific pixel location, which will represent the character 't' according to the ASCII code (refer to Fig. 11).

Fig. 11. Illustrates the procedure for extracting a single character of the secret message from the image.​​​​​​​

5.4 Evaluating BitPatternStego

In the assessment of image steganography and other image processing methods, the effectiveness of steganographic algorithms is commonly evaluated using metrics such as Mean Squared Error (MSE) and Peak Signal-to-Noise Ratio (PSNR). This evaluation involves comparing the original image with the distorted images (stego-image). MSE calculates the average squared difference between each pixel in the original and distorted images, as shown in equation (1).

\(\begin{align}M S E=\frac{\sum(\text { I_original }- \text { I_distorted })^{2}}{I_{\text {width }} * I_{\text {height }}}\end{align}\)       (1)

PSNR is determined by employing the MSE and the maximum pixel value, as illustrated in equation (2)

\(\begin{align}P S N R=10 * \log _{10} \frac{(\text { max pixel value })^{2}}{M S E}\end{align}\)       (2)

The result is usually expressed in decibels (dB). A lower MSE value indicates greater similarity between the stego-image and the original image, which is a desirable outcome in steganography. However, relying solely on MSE may not be adequate for assessing stego-image quality, as it overlooks visual considerations. PSNR, in contrast, offers a more comprehensive evaluation by considering both visual quality and the differences between the original and stego-images. A higher PSNR value suggests that the stego-image closely resembles the original image and exhibits superior visual quality.

Accordingly, The BitPatternStego method effectively conceals information without altering any bits within the cover image. As it scans the cover image to locate bit patterns aligning with those of the secret message. Consequently, there is no meaningful deviation on the calculation of MSE or PSNR, as the cover image remains unchanged.

However, when calculating the MSE and PSNR for two images, the “mandril_color” and the “lena_gray_512”, before and after implementing the BitPatternStego process, the resulting MSE is 0, and the PSNR is Infinity decibels (dB) for both images. (refer to Table 2). This scenario arises when the stego-image precisely matches the original, resulting in no difference between them.

Table 2. Calculating MSE and PSNR for two images before and after extracting the corresponding bits pattern​​​​​​​

Consequently, when comparing the MSE and PSNR outcomes of the proposed method to those of most, if not all, LSB methods, our results exhibit superiority. Furthermore, in terms of cover image payload, which constitutes one of the primary concerns for the majority of LSB techniques, BitPatternStego has demonstrated its capability to handle a substantial volume of data, potentially unlimited, as discussed in section 5.3.2. This once again underscores its superiority in this aspect as well.

The BitPatternStego method's robustness stems from its capability to conceal data without altering any bits of the cover image, thus preserving its visual fidelity and evading detection. However, the susceptibility to attacks may fluctuate depending on factors like the randomness of index selection and the diversity of embedding symbols, of the secret message, employed in the process. Furthermore, possessing knowledge of the image name and size, along with the indexing array, significantly bolsters the algorithm's resistance to attacks.

The algorithm's complexity is moderate, involving tasks such as scanning the cover image, mapping bit patterns, selecting random indexes, and embedding text. Nevertheless, it can become computationally intensive when dealing with large images or extensive text.

6. Conclusion

Securing digital data transmission is paramount in contemporary times, prompting the implementation of diverse techniques to frustrate unauthorized access. Among these, steganography stands out as a potent method for safeguarding transmitted digital data. Despite its efficacy, certain steganography techniques face limitations, such as constraints on cover-image payload and potential compromises to visual quality, influencing overall security. Striking a delicate balance between concealing data and preserving carrier integrity underscores the need for ongoing refinement in the realm of digital security measures.

In this paper, scholars introduce a steganography method suggesting the alignment of identical 8-bit patterns between the characters of a secret message and the blue channel of the pixels in a cover image. This approach ensures that the bit-type of the stego-image remains unchanged, maintaining the quality of the stego-image identical to the original. Furthermore, the payload capacity of the cover image is significantly enhanced, becoming virtually limitless, as character indexes can be repetitively utilized on random bases. This technique ensures covert data transmission without raising suspicion.

Appendix