Efficient Pixel-Value Differencing Based Hybrid Steganographic Method Using Modulus Function

Efficient Pixel-Value Differencing Based Hybrid Steganographic Method Using Modulus Function

Aruna Malik (Department of CSE, NIT Jalandhar, India) and Sonal Gandhi (Department of Computer Science and Engineering, G.L. Bajaj Institute of Engineering Technology, Greater Noida, India)
Copyright: © 2020 |Pages: 12
DOI: 10.4018/IJIRR.2020100104
OnDemand PDF Download:
No Current Special Offers


In the era of cloud computing and Big Data, steganographic methods are playing a pivotal role to provide security to sensitive contents. In the steganographic domain, pixel-value differencing (PVD) proposed by Wu and Tsai has been one of the most researched and popular methods as the PVD technique provides good quality stego-image along with high embedding capacity. This article extends the Wu and Tsai's work by proposing a new hybrid steganography scheme which works in two phases to increase the embedding capacity along with stego-image quality. In the first phase, the cover image is preprocessed using a segmentation table to make the image more robust for PVD method. In the second phase, the resultant image is partitioned into 2×1 pixels size blocks in a non-overlapping fashion and then modulus function based scheme is applied in reversible manner. Thus, a significant amount of secret data is embedded into the image. The experimental results prove that the proposed scheme has significantly improved in embedding capacity and quality as compared to the other related PVD-based methods.
Article Preview

1. Introduction

Today, Big Data has become a buzzword. However, the security of the contents like Log Files, which is one of the biggest chunks in Big Data Technology has been a major concern. To address security concerns, there exist Cryptography and Steganography techniques. The cryptography techniques basically change or encrypt/encode the contents of the secret message in such a form that it is unreadable/undecidable to any unauthorized user/party whereas the steganography techniques embed the secret message behind a cover media so that nobody is even aware of the existence of the hidden message. Therefore, in cryptography, the visible secret message (though in the unreadable form) makes it prone to attacks whereas the secret message in steganography is hidden. Thus, the steganography techniques are preferable for confidential/coveted communication. To assess the performance of steganography techniques, there are basically three quality matrices, namely: 1) visual/image quality refers to non-detective presence of secret message, 2) data hiding/embedding capacity refers to number of bits/bytes embedded in the cover image, and 3) robustness refers to resistibility against unauthorized modifications (Chang, Pai, Yeh, & Chan, 2010).

The steganography techniques work in all three domains i.e., transform/frequency domain, compression domain, and spatial domain. The transform/frequency domain based techniques (Chang, Pai, Yeh, & Chan, 2010) basically convert the cover media in frequency coefficients and then embed the secret data into them. In the compression domain based techniques (Kumar, Malik, Singh, Kumar, & Chand, 2016), (Malik, Sikka, & Verma, 2017a), (Malik, Sikka, & Verma, 2017b) the cover media is first compressed and then secret data is embedded into compression codes. These methods are most suitable for low bandwidth communications. The spatial domain-based data hiding (Kumar, Chand, & Singh, 2017), (Malik, Singh, & Kumar, 2017) techniques are most popular and simple in terms of computational complexity as these techniques embed the secret data by directly modifying the cover media elements.

Some of the popular and efficient spatial domain-based data hiding methods are LSB substitution (Bender, Morimoto, & Lu, 1996), (Wang, Lin, & Lin, 2000), (Chan & Cheng, 2001) pixel value differencing (Wu & Tsai, 2003), difference expansion (Tian, 2003), (Jung, 2018), (Maniriho & Ahmad, 2018) histogram shifting (Ni, Shi, Ansari, & Su, 2006), (Wang, Lee, & Chang, 2013), (Hwang, Kim, & Choi, 2006) and prediction error expansion (Kumar, & Jung, 2020), (Kumar, Kim, Lim, Jung, 2019). The LSB substitution techniques basically replace the bits of the cover media with the bits of secret data. The pixel value differencing technique (Wu & Tsai, 2003) hides the secret data into the pixels of the image by dividing the image into blocks and taking the Human Visual system into account. The difference expansion technique (Tian, 2003) also divides the image into blocks but expands the difference of the pixels to embed the secret data. Histogram shifting technique builds the histogram of cover media then embeds the secret data in high-frequency elements by shifting bins (Ni, Shi, Ansari, & Su, 2006). The prediction error expansion-based schemes (Kumar & Jung, 2020), (Kumar, Kim, Lim, & Jung, 2019) first predict the pixels values using a predictor and then embed the secret data by expanding the difference of the original pixel value and the predicted value. There have been several variants of these spatial domain-based steganography methods to improve their performance; however, one of the most successful methods has been the PVD. In this paper, we to propose a spatial domain-based steganographic scheme using the PVD method which works in two phases. In the first phase, we preprocess the cover image using a segmentation/range table before applying the original pixel value differencing method. In the second phase, the resultant image of the first phase is partitioned into 2×1 pixel size blocks and then the modulus function based scheme is applied in a reversible manner. Thus, a significant amount of secret data can be embedded in the image. Thus, we are able to achieve both good data hiding capacity and visual quality at the same time.

Complete Article List

Search this Journal:
Volume 12: 4 Issues (2022): 1 Released, 3 Forthcoming
Volume 11: 4 Issues (2021)
Volume 10: 4 Issues (2020)
Volume 9: 4 Issues (2019)
Volume 8: 4 Issues (2018)
Volume 7: 4 Issues (2017)
Volume 6: 4 Issues (2016)
Volume 5: 4 Issues (2015)
Volume 4: 4 Issues (2014)
Volume 3: 4 Issues (2013)
Volume 2: 4 Issues (2012)
Volume 1: 4 Issues (2011)
View Complete Journal Contents Listing