Steganography based on LSB Techniques
In the paper, a review of different approaches is done. Some of the stenographic algorithms discussed include the Least Significant Bit embedding (LSB), Random Least Significant Bit embedding (RLSB) as well as the Edge Least Significant Bit embedding (ELSB). Furthermore, a comparison is made to substantiate the workability of the approaches, advantages, disadvantages as well determining the most appropriate technology. The focus of the discussion is to identify of the digital images that are predominantly used as covers within the spatial domain. Furthermore, understanding of the adaptive mechanisms of the approaches and the appropriate data securing mechanism is delineated. This is because the information can be hidden in the cover data (Mehndiratta, 2015). In most instances, the term cover is used to mean the original message that can be informed of audio, written work, video among others.
One aspect to note when hiding the message is that it can involve the use of watermarks, copyrights in order to protect the rights of the owner of the message (Smitha, & Baburaj, 2016). As a result, the appearance of the message cannot be identified. Most individuals confess that they just look at the cover image with little consideration to what might be held inside the image. The spatial domain approach matches the message with the least significant pixel bits (LSB). The frequency approach matches the message bits using mathematical coefficients. The LSB approach involves increasing or decreasing the pixel value to match the message bit that is meant for concealment. Nikseresht, Dezfouli, & Alavi, (2017) researched on the possibilities of improving the "Least Significant Matching Revisited" (LSBMR) that was proposed by Mielikainen.
LSBMR aims at improving the quality of stego images through making fewer changes in the generation of greyscale images. Jung & Yoo, (2015) focused their research on the reversibility of data in steganography. They considered an approach that challenged a reversible data approach. A reversible method can easily extract the original image from a stego-image with minimal or no distortion. Jung and Yoo discarded the reversible data approach and attempted the semi-reversible data approach. Their technique employed interpolation and the Least Significant substitution method in the creation of a semi-reversible stego-image. High embedding rates increase the chances of detection and low embedding rates minimized the chances of detection. Afrakhteh Moon and Lee worked on improving the steganography scheme through the use of a spatial domain.
The spatial domain was to incorporate the use of error images that originate from the application of image quality factor to identify pixels for data embedding. The approach was identified o be less detectable than traditional approached. The method had a high embedding rate of 4bpp while maintaining a high PSNR value. Their method managed to embed data at 4pp with high efficiency. Their technique was also capable of overcoming first order statistical test. The weakness of the approach is the inability to make large amounts of data due to difficulties in storing large statistical features. Paul et al also propose that their method can be used beyond the spatial domain scheme. Sur et al, (2014) explored the use of an algorithm based spatial desynchronization in concealing an image.
The concealed information is split into four blocks. An MLE algorithm is used to encrypt the concealed information. The rotated sub-images embed the sub-blocks of the messages. The embedding of the messages is performed through a specific pattern obtained through magic LSB substitution. The results showed that the magic LSB model provided sufficient security and imperceptibility for color images. The high payload±1 uses two binary functions that employ the information of least two significant bit planes for extraction and embedding purposes. The method selects the correct modification component to increase the embedding efficiencies. In the approach, a specific number of data bits are embedded in a similar number of pixels in the cover image through the causation of a single unit alteration in one-third of the pixels.
The results showed an embedding rate of 1bpp and embedding efficiency of 3. 019bpc. A huge amount of the text material can be concealed through a small image (Patil et al. , 2017) Furthermore, the image is not distorted since it only acts as a reference image. Least Significant Bit Embedding (LSB) Least significant Bit (LSB) is a strategic implementation that employs steganographic methods. In this method, data is embedded into the cover which makes it difficult to be detected by a casual observer (Islam, Modi, & Gupta 2014). It works through replacement of information in a pixel with data from the image. To be succinct; the LSB steganography involves the LSB of the image to replace it with the bits from the data. The LSB has been known to be vulnerable to steganalysis, as a result, it requires encryption of the raw data before it is embedded in the image.
Despite the encryption process being time-consuming, it is known to be quite secure. In addition to that, the process is simple as it requires the replacement of the bytes inside the image to bits of the secret message. The approach is considered vital due to the ability to hide messages that are contained within the multimedia carrier data. The results of the PoV are the frequency of two numbers having the 1S and 0S, and a regular equality pattern (Pavani, Naganjaneyulu, & Nagaraju, 2013). Sample values can be incremented or decremented based on the predefined threshold series that can be generated by the user. In cases where we have specified stego-key, the new sample value is not dependent on the pseudorandom and the original sample value instead focus is put on the generation of values using the RLSB as compared to the traditional LSB embedding techniques due to the advanced security provided by the technique.
Edge Least Significant Bit Embedding (ELSB) The LSB method uses all the edge pixels present in the image. It involves the calculation of the masked image through the masking of two LSB available on the cover image. The pseudorandom generator does not have any considerations towards the image content and the secret image as a result of the dynamic embedding positions. Researchers assert that regions within the cover images get contaminated at data hiding because of the low embedding rate. The resultant effects of the data contamination are low visual and low security with the most affected cover images being those that have smooth regions. To determine low embedding rates, sharper edge regions are considered. Embedding rates affect the parameters for data hiding and edge regions (Hussain et al, 2018).
The data that is identified in most instance is hidden using the Canny Edge Detection Method in the edge pixels. advantages It is a simple process as it involves the replacement of the bytes inside the image to bits of the secret message. It is flexible as the sample values that are formed during incrimination and decrementation can be defined as it has a predefined threshold series that is determined by the user. The information that is transferred between the sender and the recipient cannot be easily detected by the attackers as it has the special pixels that hide the cover image and any data that needs to be shared. The transmission rate is very fast, enabling faster communication between the recipient and the sender.
Through the use of the edge least significant bit, the solution to image privacy is solved. Furthermore, communication through the use of images can be regarded efficient since it is a digitalized form of communication. However, further research should be done on the ELSB to ensure that the pixel has the ability to hold more secret data bits without revealing any ramifications. This can be done by breaking down the image into both smooth areas and edge areas to hide different forms of data. It is worth noting that all LSB algorithms concentrate in concealing information both in bits and the pixels. W. A. , Idris, Y. I. B. Edge-based image steganography. EURASIP Journal on Information Security, 2014(1), 8. Jindal, S. , & Kaur, N. Digital image steganography survey and analysis of current methods.
From $10 to earn access
Only on Studyloop