Page 51 - 2024-Vol20-Issue2
P. 51
47 | Alobaidi & Mikhael
The size of the Huffman encoded bit stream and Huffman within the time domain of a digital signal, such as an image
Table were also embedded inside the cover image, making or audio file. This technique takes advantage of the fact that
the Stego-Image a standalone information for the receiver. changing the least significant bit of a pixel or a sample in an
The experimental results showed that the algorithm had a audio signal has minimal impact on the overall perception
high capacity and good invisibility. Furthermore, the Peak of the signal. In LSB insertion, the binary representation
Signal to Noise Ratio (PSNR) of the stego image with the of the secret message is embedded by replacing the least
cover image yielded better results compared to other existing significant bit of selected pixels or audio samples with the
steganography approaches. Additionally, satisfactory security corresponding bits from the message. The LSBs are typically
was maintained since the secret message/image could not be modified because they have the least impact on the visual or
extracted without knowing the decoding rules and Huffman auditory quality of the carrier signal. For example, in image
table. steganography using LSB insertion, the pixels of an image
In this work, a new image Steganogrphy insertion approach are represented by three color channels: red, green, and blue
is presented. The approach starts with dividing the cover im- (RGB). Each color channel consists of 8 bits per pixel, rang-
age into predefined number of non-overlapping blocks. Then, ing from 0 to 255. The LSB of each color channel can be
cover image blocks are transformed to the Cosine domain modified to store a single bit of the secret message, effectively
by utilizing the 2D DCT. The adaptive algorithm explained hiding the information. The process involves the following
in [17] is employed to find the weights of each coefficient, for steps:
each block individually, in the Cosine domain. Blocks with
coefficients that have low, compared to the rest of the coeffi- 1. Convert the secret message into binary representation;
cients, total weights are chosen. The coefficients in that block
is converted to binary representation, refereed to as ”cover in 2. Iterate through the pixels of the image or audio samples;
binary”. The secret data, or more precisely its binary form, is
embedded in the LSB bit of ”cover in binary”. The block is 3. Modify the LSB of each selected pixel or sample ac-
converted back to the decimal representation. Then, 2D IDCT cording to the corresponding bit of the secret message;
is applied to obtain the stego image.
The metrics utilized to evaluate the performance of the pro- 4. Repeat the process until all bits of the secret message
posed technique are RMSE, and PSNR. In addition, human are embedded.
visual inspection is also considered. The proposed system is
compared with three other techniques. The first one is the By modifying only the LSB, the changes introduced to the car-
traditional Spatial LSB, the second is energy based DCT in- rier signal are generally imperceptible to the human eye or ear.
sertion (in which total block energy is used as the selection However, it is essential to consider the capacity of the carrier
parameter), and the third is the most recent reported work signal and ensure that the secret message can be embedded
in [12]. The effect of size of cover image blocks is also exam- without causing significant distortion or noticeable artifacts.
ined. As shown in the results, the proposed approach performs In conclusion, LSB insertion in the time domain provides a
better than the other techniques when tested with 10 samples simple and straightforward method for steganographic data
from BossBase [18], and a custom-built databases. hiding, but it may be susceptible to detection by steganaly-
The rest of the paper is organized as follows. Details about sis techniques that analyze statistical properties or deviations
Stegnaogrpahy techniques are presented in Section II. In Sec- from expected patterns in the carrier signal. Therefore, ad-
tion III. , the proposed technique is explained. The results are ditional techniques such as encryption and more advanced
presented in Section IV. Section V. contains the discussion. steganographic methods may be employed to enhance the
The conclusions are shown in Section VI. security and robustness of the hidden information.
II. STEGANOGRAPHY TECHNIQUES B. Cosine Domain Insertion
Discrete Cosine Transform (DCT) is a mathematical technique
In this section, details about Steganography insertion tech- commonly used in signal processing and data compression.
niques [19] are presented. These techniques are in Spatial It is also utilized in certain forms of steganography to hide
domain and in Discrete Cosine Transform (DCT) domain. information within digital media such as images or videos.
The equations, forward and inverse, for calculating such coef-
A. Spatial Domain LSB Insertion
The LSB (Least Significant Bit) insertion technique is a com-
monly used method in steganography for hiding information
