Concealing a Secret Message in a Colour Image Using an Electronic Workbench

Steganography is the art of concealing security data in media, such as pictures, audio, video, text, and protocols. The objective of this paper is hiding a secret message in a colour image to prevent an attacker from accessing the message. This is important because more people use the Internet all the time and network connections are spread around the world. The hidden secret message uses two general algorithms that are embedded and extracted. This paper proposes a new algorithm to conceal a secret message in a colour image in LSB. This algorithm includes three phases: 1) dividing the colour image into a number of blocks, 2) concealing the secret message, and 3) transmitting the stego-image from the sender in a multiplexer network and receiving it through a demultiplexer network using an electronic workbench. The outcome of the new algorithm demonstrates good efficiency, high security, and robustness and is executed quickly. The system is evaluated through the measurements of mean square error, peak signal-to-noise ratio, correlation, histogram, and capacity.


Khodher and Alabaichi
Iraqi Journal of Science, 2021, Vol. 62, No. 12, pp: 4964-4977 5:76 Digital image handling deals with the processing of other images through a digital computer. Image treatment is the application of a signal treatment method, such as two-dimensional (2D) signals, in photographs or videos. Image treatment typically includes filtering or enhancing a picture using different kinds of functions in addition to other methods to extract data from the image [1]. Image treatment via digital means comprises many branches involving picture recognition, segmentation, compression, and so on. It is essential in many applications, such as pattern recognition and goal identification. Picture treatment normally means digital image treatment [2]. Steganography is a method of hiding data in a carrier file so that it is imperceptible to prohibited parties [3]. Data concealment methods have recently become substantial in sundry application fields. Most issues of the information-hiding method stem from the unsecure carriage medium. The information-hiding method is an innovative kind of mystery communication technology. It comprises embedding data into digital media (image, video, audio, text, and protocol), with a minimum appreciable declination in the steward signal, for the objectives of identification, controlling access to digital media, and copyright [4]. A computer network is a framework where two or more computers and/or suitable calculation devices are interconnected to a portion of interchange data. The computers and other suitable devices are joined by telecommunication devices that can also recall and transfer media, causing the information or data to be transmitted within the network through these devices [5]. Facciolo et al. (2014) suggested integrating image impersonation, which is a wonderful idea that allows one to evaluate the sum of picture values in rectangular areas of the image for four processes, regardless of the volume of the areas. It was first suggested under the name of the "summed region table" in the Computer Graphics Society to efficiently change textile maps. It was popularised in the computer community with its use in their real-time topic discovery framework. In this work, it was suggested to depict the integrated image algorithm and study its implementation in the context of block matching. It was used to test trade-offs and the limit of the implementation success with respect to exhaustive block matching [6]. Khodher (2015) suggested an image as a covering to conceal the message using a secret random key. The secret key is a 9×9 array, and random items are selected for the reverse matrix. The reverse matrix is multiplied by a one-dimensional matrix, and the outcome is in a hidden image. This algorithm keeps the image secure through a transformation via a network. Such an operation is defined as a random picture transformation and is regarded as a credible means of hiding. The outcomes obtained from the proposed algorithm rely on the covering image. A secret key is created so that it can be retrieved from the original image after receiving it, without losing any concealed data by the recipient in the network [7]. Pandey et al. (2015) suggested image compression in DCT. The quantisation encoding manner of covert coding is widely used in picture treatment methods, but in these, the 2D pictures are split into sub-blocks, and every block is converted separately into prime frequency compositions. The frequency compositions (DC and AC) are decreased to zero over the operation of quantisation, which is a lossy operation. The result of the proposed manner applies to all the images, and its performance is analysed in terms of lowering the picture volume. The difference in picture quality between the original picture and the reconstructed picture is measured using the PSNR value with different quantisation matrices [8]. Barapatre et al. (2017) suggested a survey of three rising velocity technologies, which are X.25, frame relay, and ATM and compared them. Frame relay and ATM are variations of the requisite X.25 technologies. Based on the different execution metrics, the comparison survey explained that ATM has a lower retard contrast as compared to the X.25 and frame relay, and Khodher and Alabaichi Iraqi Journal of Science, 2021, Vol. 62, No. 12, pp: 4964-4977 5:77 therefore is effective for carrying real-time data. The result of ATM uses constant volume packets (53 bytes) for telecommunication. The frame relay uses changing packet volumes that rely on the kinds of data to be transmitted. The frame relay is used to join local area networks (LAN), and it is not executed in a single area network, whereas the data in ATM are within a single LAN. In addition, ATM is designed to be suitable for hardware execution and thus, the cost is higher compared to frame relay, which is software controlled [9]. Shirur (2018) suggested a wireless criterion frame status that is executed for time-critical signals (voice and video) and datum signals based on user priority in version 802.11. This proposed action is designed in such a way that forthcoming data are prioritised and integrated into Ethernet frames by tagging. Then, they are mapped to access categories. Based on user priorities, various queues are formative for every access category and difference for the channel differentiation. To support service quality, an enhanced distributed channel access is used, which was based on 802.11e [10]. Bandyopadhyay (2018) stated that the hiding of all information is a challenge in the field of safety and that security methods try to preserve privacy, safety, and availability. A steganography system includes two functions, namely embedding and extraction. The goal of the proposed action is to design robust algorithms that generate stego media and can load a large capacity of secure data without lowering the imperceptibility [11].

Electronic Workbench
The Electronic Workbench (EWB) is a designing instrument that supplies the user with all the components and tools to conduct board-level designing on a personal computer [12]. The EWB includes two kinds of data connections: 1) multiplexer (MUX) and 2) demultiplexer (DEMUX). A multiplexer is a circuit that accepts numerous inputs but gives only one output. A demultiplexer task is precisely the inverse of a multiplexer; that is, a demultiplexer accepts only one input and gives numerous outputs. In general, a multiplexer and demultiplexer are used jointly because the connection systems are bi-directional [13], as shown in Figure 1 a, b.

3.Steganography
In information hiding, i.e. steganography, the data are concealed in the covering media. The covering medium can be in the form of images, script, video, or auditory files. Information hiding is defined as the science or technique of concealing the message into some covering medium. The word "steganography" comprises two words of old Greek origin: steganos, which means "covering, concealed, or protected" and graphical, meaning "lettering" [14]. Khodher  Information hiding is a secure message embedded into the points of the load content, either by changing or replacing data [15]. It realises good imperceptibility and full capacity that can be realised comparatively faster than in the frequency domain, but with poor robustness. Alternatively, in frequency domain methods, the secret message is embedded inside the features of the load media after a certain transformation; however, this offers high robustness at the price of complexity and treatment time [16]. Information hiding uses two main algorithms, namely the embedded and extracted algorithms between the sender and receiver [17], as shown in Figure 2.

4.Evaluation System Performance
We applied the mean square error (MSE), peak signal-to-noise ratio (PSNR), correlation, and histogram [18] for the evaluation of the system performance.

Mean Square Error
MSE is calculated by comparing the bytes of two images. A pixel comprises 8 bits, and thus, 256 levels are available to represent various gray levels. MSEs are valuable when the bytes of an image are compared with the corresponding bytes of another image. Equation (1) is used to compute MSE [18,19].
(1) where M and N represent row and column, respectively, in Figures 1 and 2 to obtain the value of MSE.

Peak Signal to Noise Ratio
PSNR is a parameter used to measure the amount of imperceptibility in decibels. It measures the quality between the two images. A large PSNR value indicates that a small difference exists between two images. By contrast, a small PSNR value indicates a huge distortion between two images. Equation (2) is used to compute PSNR [18,19].
(2) MSE where R is the maximum number of fluctuation in the input image data type.

Correlation Coefficient
The correlation coefficient r is the measurement of the range and trend of the linear group of two random variables. If two variables are closely related, the correlation coefficient is close to the value of 1. If the coefficient is close to 0, the two variables are not related. The coefficient r can be calculated using equation (3) [20] . where X i is the pixel intensity of the original image, X m is the mean value of the original image intensity, Y i is the pixel intensity of the stego-image, and Y m is the mean value of the stego-image intensity [21,22].

Histogram
An image histogram is a diagram displaying how many pixels there are at every scale or at every index for the indexed colour image. The histogram includes data that are necessary for image equalisation, where the image pixels are stretched to give a sensible dissimilarity [22,23]. With the histogram, the equalisation method can evolve. Equalization stretches the scale range of the pixel level to the full range to improve the contrast of the given image. To use this technique, the equalised new pixel value is redefined using equation (4) [23].
where m and n represent row and column, respectively in image p. which are applied in equation (4) to obtain the histogram.

The Proposed System
The proposed system uses colour images to conceal a secret message inside an image in three phases: 1) image blocks, 2) secret message, and 3) image transmission, as shown in the block diagram of the proposed system in Figure 3.

First Phase: Image Blocks
This phase divides a colour image of any format into a number of blocks, as shown in the original image in Figure 4. The size of the image is 96×96 for 16 blocks. Figure 5 starts from 0 to 23 for block 1, from 24 to 47 for block 2, from 48 to 71 for block 3, and from 72 to 96 for

Second Phase: Hide the Secret Message
The secret message comprises 20 characters "This computer is very fast", and each character is represented as ASCII in 8 bits. For the secret message, every 10 bits hide in one block. For 20 characters times 8 bits in ASCII, that is equal to 160 bits hiding in the image of 96×96 pixels in 16 blocks. This system can hide more than 20 characters. As shown in Figure 6a, there are 16 blocks in the image, and Figure 6b shows the readout for block 1 in the block image. The pixels hide one bit in LSB after converting this pixel to binary in 137, 89, 76, 108, 208, 219, 225, 209, 215, and 223, and it resets LSB in each pixel, adding one bit from the secret message. In each of the 50 locations, one bit hides from 0 to 49 and continues in 50 to 99 to n blocks, hiding 10 bits in each block. In all, the image hides 160 bits. The pixels become 136, 88, 77,108, 209, 218, 225, 208, 215, and 222. There is no effect on the colour image. The operation is repeated in ten locations from 137 to 223 in block 1 and is repeated in all blocks to hide all parts of the secret message in all image blocks before sending the image (Figure 7). Each block can hide a character and two bits from the next character.
After embedding the secret message as shown in Figure 8, we have the stego-image.

Implementation Embedding and Extraction Algorithms
Step 2: Divide the original image into 16 blocks; each block has 576 pixels in B.
Step 3: Find the location in 16 blocks considering the secret key, select each of the 50 pixels Khodher and Alabaichi Iraqi Journal of Science, 2021, Vol. 62, No. 12, pp: 4964-4977 5:84 by 50 pixels in C.
Step 4: Hide 10bitsin binary in LSB in each location in each block using the secret key: 160 bits in D.
Step 5: Divide each block into 72 frames, where each frame has 8 slots in E.
Step 6: Put the result (stego-image) in F and transmit it using multiplexers in one channel. Step 1: Load all frames from one channel to DEMUX in A.

Extraction Process
Step 2: Sum all frames and rearrange to 16 blocks, each block with 576 pixels, and each block includes72 frames in B.
Step 3: Load stego-image in C.
Step 4: Find the secret key to select the location (every 50 pixels) in all blocks in D.
Step 5: Retrieve 10 bits in binary from LSB in all blocks using the secret key in E.
Step 6: Sum the 160 bits in binary from the 16 blocks in E.
Step 7: Put the results (secret message) in F.

Testing the Result
This paper indicates that the outcome in the proposed system is good; it hides the secret message in blocks in an image. The steganography system provides high security and is efficient and robust because it cannot be viewed and changed by attackers. Through the test measurements using PSNR MSE, correlation, and histograms, the outcomes are very good. Figure 11 explains the testing results among the original image, image block, and stegoimage.
Original Image Cut Block Image Cut Stego-Image Cut Original Image House Block Image House Stego-Image House Original Image Nature Block Image Nature Stego-Image Nature Original Image River Block Image River Stego-Image River Original Image Tree Block Image Tree Stego-Image Tree Figure 11-The original, block, and stego-image

System Analysis
The proposed system is analysed by evaluating the outcome using MSE, PSNR, correlation, and histogram to explain the robustness, efficiency, and security. The system is very powerful in analogous between original image, block image, and stego-image, when concealing the secret message. Table 1 indicates the measurements of MSE, PSNR, and correlation, using equations (1), (2), and (3), respectively, shown in the section on the performance evaluation. Table 2 indicates the histogram of the original image, which varies in the block image and stego-image, where the histogram between the block image and stego-image relies on equation (4) shown in the section on the performance evaluation. This indicates that the system is powerful for concealment. When MSE, PSNR, and the correlation are run in five tests, the range of MSE is the reverse of the range of PSNR. The range of MSE in the block image and stego-image increases compared with the original image, and the range of the PSNR in the block image and stegoimage decreases compared with the original image. When the correlation in the block image and stego-image is relatively equal, as shown in Table 1, this indicates that the proposed system is good for concealing a secret message and can transfer the stego-image using EWB. Correlation=0.9932 Correlation=0.9250

Histogram
The histogram displays the accurate appearance of all pixels in the image. Table 2 indicates the histogram of the original image, block image, and stego-image. The histogram shows variation in the proposed system among the original image and stego-image, but also shows the similarity between the block image and stego-image. This indicates that the proposed system is good for hiding a secret message. The similarity between block and stego-images indicates that it prevents the detection of the secret message by attackers.

High Capacity
The hiding capacity determines the maximum number of bits that can be hidden in the original image (96x96) after dividing 16 blocks for an acceptable quality of the resultant stego-image. The steganography system has better performance if it has a large hiding capacity, equal to 160 bits. In the proposed algorithm, one bit is embedded in each of the 50 pixels of each block. Therefore, the capacity of the hiding rate in the proposed algorithm is equal to the number of characters/size of the image (number of pixels). For 20 characters, it is 160 bit/9216 pixels, equal 0.017361, and for 80 characters, it is 640 / 9216 pixels, equal 0.0694 capacity for hiding data in a colour image.

Conclusions
This paper describes a proposed system to hide a secret message inside a block image using LSB. More than 20 characters can be sent in a secret message in this system. Each block image is transformed via EWB, using a circuit of multiplexers and demultiplexers (MUX and DEMUX) and TDM. This method is used to send a secret message across a network Internet. The outcomes of this system show that it is executed quickly, efficiently, robustly, and securely, through the use of TDM and a number of tests shown in table 1 and table 2. It can be used as a secret key by selecting the locations of all 50 pixels to hide one bit in each block image, where ten bits are used without revealing sensitive information to an unauthorised attacker when exchanging a message.