# Hybrid Data Hiding Scheme Using Right-Most Digit Replacement and Adaptive Least Significant Bit for Digital Images

^{1}

^{2}

^{3}

^{4}

^{*}

## Abstract

**:**

## 1. Introduction

## 2. Proposed Hybrid Data Embedding Method

_{1}[0, 31] to R

_{4}[128, 255]. On the other hand, the second row of Table 1 shows the pre-estimated number of secret bits for embedding in each pixel. For example, the R

_{1}range exists under region-1 and 3-bits secret data are used to embed in each pixel. Furthermore, these ranges can be dynamically generated depending on the steganographic requirement. For experiments, we propose the following regions and ranges of Table 1, which meet the proposed method goals, i.e., high payload and acceptable visual quality.

#### 2.1. RMDR Embedding Method

_{0}(b) and SRD

_{1}(b), where b is a three-bit secret decimal data, and SRD

_{0}(b) and SRD

_{1}(b) are the stegoRDs (mapped RDs) against b. In our experiments, the generation of mapping table (Table 2) is as follows: the three-bit secret digit, 2

^{3}range is [0, 7], and a pixel RD range is [0, 9]. The pixel RD range has two extra digits as {8, 9} that can be reused with three-bit secret digits as replacing with {3, 4} digit value. These two extra digits aim at minimizing the difference between cover and stego-pixels RDs. Alternatively, it can be adaptively generated by considering the frequency of RDs in cover images.

#### 2.2. Hybrid Embedding Method

#### 2.3. RMDR Extracting Method

#### 2.4. Hybrid Extraction Method

## 3. Experimental Results and Discussion

#### 3.1. Embedding Capacity and Visual Quality Analysis

#### 3.2. Universal Quality Index Analysis

#### 3.3. Embedding Capacity versus PSNR

#### 3.4. Security against Statistical RS-Steganalysis

#### 3.5. Pixel Difference Histogram Analysis

#### 3.6. Undetectability under SPAM Analysis Using Ensemble Classifier

## 4. Conclusions

## Acknowledgments

## Author Contributions

## Conflicts of Interest

## References

- Fan, C.-I.; Tseng, Y.-F. Anonymous Multi-Receiver Identity-Based Authenticated Encryption with CCA Security. Symmetry
**2015**, 7, 1856–1881. [Google Scholar] [CrossRef] - Shahzad, A.; Lee, Y.-K.; Kim, S.; Xiong, N.; Choi, J.-Y.; Cho, Y. Real Time MODBUS Transmissions and Cryptography Security Designs and Enhancements of Protocol Sensitive Information. Symmetry
**2015**, 7, 1176–1210. [Google Scholar] [CrossRef] - Katzenbeisser, S.; Petitcolas, F. Information Hiding Techniques for Steganography and Digital Watermarking; Artech House: Boston, MA, USA, 2000. [Google Scholar]
- Subhedar, M.S.; Mankar, V.H. Current status and key issues in image steganography: A survey. Comput. Sci. Rev.
**2014**, 13, 95–113. [Google Scholar] [CrossRef] - Kougianos, E.; Mohanty, S.P.; Mahapatra, R.N. Hardware assisted watermarking for multimedia. Comput. Electr. Eng.
**2009**, 35, 339–358. [Google Scholar] [CrossRef] - Kuo, W.-C.; Kuo, S.-H.; Wuu, L.-C. Multi-Bit Data Hiding Scheme for Compressing Secret Messages. Appl. Sci.
**2015**, 5, 1033–1049. [Google Scholar] [CrossRef] - Lee, C.F.; Weng, C.Y.; Sharma, A. Steganographic access control in data hiding using run-length encoding and modulo-operations. Secur. Commun. Netw.
**2016**, 9, 139–148. [Google Scholar] [CrossRef] - Collins, J.C.; Agaian, S.S. Taxonomy for spatial domain LSB steganography techniques. In SPIE Sensing Technology+ Applications; International Society for Optics and Photonics: Baltimore, MD, USA, 2014; Volume 9120, p. 912006. [Google Scholar]
- Chan, C.-K.; Cheng, L.-M. Hiding data in images by simple LSB substitution. Pattern Recognit.
**2004**, 37, 469–474. [Google Scholar] [CrossRef] - Fridrich, J.; Goljan, M.; Du, R. Reliable detection of LSB steganography in color and grayscale images. In Proceedings of the 2001 Workshop on Multimedia and Security: New Challenges, New York, NY, USA, 5 October 2001; pp. 27–30.
- Fillatre, L. Adaptive steganalysis of least significant bit replacement in grayscale natural images. IEEE Trans. Signal Process.
**2012**, 60, 556–569. [Google Scholar] [CrossRef] - Zhang, J.; Cox, I.J.; Doërr, G. Steganalysis for LSB matching in images with high-frequency noise. In Proceedings of the IEEE 9th Workshop on Multimedia Signal Processing, Crete, Greece, 1–3 October 2007; pp. 385–388.
- Al-Dmour, H.; Al-Ani, A. Quality optimized medical image information hiding algorithm that employs edge detection and data coding. Comput. Methods Programs Biomed.
**2016**, 127, 24–43. [Google Scholar] [CrossRef] [PubMed] - Pevný, T.; Filler, T.; Bas, P. Using high-dimensional image models to perform highly undetectable steganography. In Information Hiding; Springer: Berlin/Heidelberg, Germany, 2010; pp. 161–177. [Google Scholar]
- Holub, V.; Fridrich, J. Designing steganographic distortion using directional filters. In Proceedings of the 2012 IEEE International Workshop on Information Forensics and Security (WIFS), Tenerife, Spain, 2–5 December 2012; pp. 234–239.
- Holub, V.; Fridrich, J.; Denemark, T. Universal distortion function for steganography in an arbitrary domain. EURASIP J. Inf. Secur.
**2014**, 2014, 1–13. [Google Scholar] [CrossRef] - Li, B.; Wang, M.; Huang, J.; Li, X. A new cost function for spatial image steganography. In Proceedings of the 2014 IEEE International Conference on Image Processing (ICIP), Paris, France, 27–30 October 2014; pp. 4206–4210.
- Sedighi, V.; Fridrich, J.; Cogranne, R. Content-adaptive pentary steganography using the multivariate generalized Gaussian cover model. In IS&T/SPIE Electronic Imaging; International Society for Optics and Photonics: San Francisco, CA, USA, 2015. [Google Scholar]
- Wu, D.-C.; Tsai, W.-H. A steganographic method for images by pixel-value differencing. Pattern Recognit. Lett.
**2003**, 24, 1613–1626. [Google Scholar] [CrossRef] - Lee, Y.-P.; Lee, J.-C.; Chen, W.-K.; Chang, K.-C.; Su, J.; Chang, C.-P. High-payload image hiding with quality recovery using tri-way pixel-value differencing. Inf. Sci.
**2012**, 191, 214–225. [Google Scholar] [CrossRef] - Balasubramanian, C.; Selvakumar, S.; Geetha, S. High payload image steganography with reduced distortion using octonary pixel pairing scheme. Multimedia Tools Appli.
**2014**, 73, 2223–2245. [Google Scholar] [CrossRef] - Zhang, X.; Wang, S. Efficient steganographic embedding by exploiting modification direction. IEEE Commun. Lett.
**2006**, 10, 781–783. [Google Scholar] [CrossRef] - Wang, X.-T.; Chang, C.-C.; Lin, C.-C.; Li, M.-C. A novel multi-group exploiting modification direction method based on switch map. Signal Process.
**2012**, 92, 1525–1535. [Google Scholar] [CrossRef] - Kim, C. Data hiding by an improved exploiting modification direction. Multimedia Tools Appl.
**2014**, 69, 569–584. [Google Scholar] [CrossRef] - Wu, K.-S.; Liao, W.-D.; Lin, C.-N.; Chen, T.-S. A high payload hybrid data hiding scheme with LSB, EMD and MPE. Imag. Sci. J.
**2014**, 63, 174–181. [Google Scholar] [CrossRef] - Shen, S.-Y.; Huang, L.-H. A data hiding scheme using pixel value differencing and improving exploiting modification directions. Comput. Secur.
**2015**, 48, 131–141. [Google Scholar] [CrossRef] - Wu, H.-C.; Wu, N.-I.; Tsai, C.-S.; Hwang, M.-S. Image steganographic scheme based on pixel-value differencing and LSB replacement methods. IEEE Proc. Vis. Image Signal Process.
**2005**, 152, 611–615. [Google Scholar] [CrossRef] - Khodaei, M.; Faez, K. New adaptive steganographic method using least-significant-bit substitution and pixel-value differencing. IET Image Process.
**2012**, 6, 677–686. [Google Scholar] [CrossRef] - Jung, K.-H. High-capacity steganographic method based on pixel-value differencing and LSB replacement methods. Imag. Sci. J.
**2010**, 58, 213–221. [Google Scholar] [CrossRef] - Tsai, Y.-Y.; Chen, J.-T.; Chan, C.-S. Exploring LSB substitution and pixel-value differencing for block-based adaptive data hiding. Int. J. Netw. Secur.
**2014**, 16, 359–364. [Google Scholar] - Jung, K.-H.; Ha, K.-J.; Yoo, K.-Y. Image data hiding method based on multi-pixel differencing and LSB substitution methods. In Proceedings of the International Conference on Convergence and Hybrid Information Technology, Daejeon, South Korea, 28–30 August 2008.
- Zaker, N.; Hamzeh, A. A novel steganalysis for TPVD steganographic method based on differences of pixel difference histogram. Multimedia Tools Appl.
**2012**, 58, 147–166. [Google Scholar] [CrossRef] - Sajedi, H.; Jamzad, M. HYSA: hybrid steganographic approach using multiple steganography methods. Secur. Comm. Netw.
**2011**, 4, 1173–1184. [Google Scholar] [CrossRef] - Yang, C.-H.; Weng, C.-Y.; Wang, S.-J.; Sun, H.-M. Varied PVD + LSB evading detection programs to spatial domain in data embedding systems. J. Syst. Softw.
**2010**, 83, 1635–1643. [Google Scholar] [CrossRef] - Huang, H.-S. A combined image steganographic method using multi-way pixel-value differencing. In Proceedings of the Sixth International Conference on Graphic and Image Processing (ICGIP 2014), Beijing, China, 24–25 October 2015.
- Shen, S.; Huang, L.; Tian, Q. A novel data hiding for color images based on pixel value difference and modulus function. Multimedia Tools Appl.
**2015**, 74, 707–728. [Google Scholar] [CrossRef] - Petitcolas, F.A.; Anderson, R.J. Evaluation of copyright marking systems. In Proceedings of the IEEE International Conference on Multimedia Computing and Systems, Florence, Italy, 7–11 June 1999.
- Implementation of the Ensemble Classifier (Using Weka) and the SPAM Features + HUGO Algorithm. Available online: http://dde.binghamton.edu/download/ (accessed on 20 April 2016).
- Schaefer, G.; Stich, M. UCID: An uncompressed color image database. In Proceedings of the SPIE, Storage and Retrieval Methods and Applications for Multimedia 2004, San Jose, CA, USA, 18 January 2004.
- The USC-SIPI Image Database. Available online: http://sipi.usc.edu/database/ (accessed on 1 September 2015).
- Yang, H.; Sun, X.; Sun, G. A high-capacity image data hiding scheme using adaptive LSB substitution. Radioengineering
**2009**, 18, 509. [Google Scholar] - Wang, Z.; Bovik, A.C. A universal image quality index. IEEE Signal Process. Lett.
**2002**, 9, 81–84. [Google Scholar] [CrossRef] - Pevny, T.; Bas, P.; Fridrich, J. Steganalysis by subtractive pixel adjacency matrix. IEEE Trans. Inf. Forensics Secur.
**2010**, 5, 215–224. [Google Scholar] [CrossRef]

**Figure 5.**Comparison graphs for various embedding capacities vs. PSNR: (

**a**) Lena; (

**b**) Baboon; (

**c**) Pepper; (

**d**) Jet.

**Figure 6.**RS-analysis graphs for 3-bit LSB vs. RMDR of stego-images. (

**a**) Lena 3-bit LSB; (

**b**) Lena RMDR; (

**c**) Baboon 3-bit LSB; (

**d**) Baboon RMDR.

Let | ${\mathit{g}}_{\mathit{i}}$ and ${\mathit{g}}_{\mathit{i}\mathbf{+}\mathbf{1}}$ are the ith block pixels values of C, M be the secret message bits streams. In Table 2, b refers to both ${\mathit{b}}_{\mathit{i}}$ and ${\mathit{b}}_{\mathit{i}\mathbf{+}\mathbf{1}}$ For $\mathit{S}\mathit{R}{\mathit{D}}_{\mathbf{0}}$(b) and $\mathit{S}\mathit{R}{\mathit{D}}_{\mathbf{1}}$(b) are stego-right-digits (stegoRD), refers to $\mathit{S}\mathit{R}{\mathit{D}}_{\mathbf{0}\mathbf{\left(}{\mathit{b}}_{\mathit{i}}\mathbf{\right)}}$, $\mathit{S}\mathit{R}{\mathit{D}}_{\mathbf{0}\mathbf{\left(}{\mathit{b}}_{\mathit{i}\mathbf{+}\mathbf{1}}\mathbf{\right)}}$ and $\mathit{S}\mathit{R}{\mathit{D}}_{\mathbf{1}\mathbf{\left(}{\mathit{b}}_{\mathit{i}}\mathbf{\right)}}$, $\mathit{S}\mathit{R}{\mathit{D}}_{\mathbf{1}\mathbf{\left(}{\mathit{b}}_{\mathit{i}\mathbf{+}\mathbf{1}}\mathbf{\right)}}$ respectively. | |

Begin: | ||

Step 1 | Read n = 6 ( ${n}_{5},{n}_{4},{n}_{3},{n}_{2},{n}_{1},{n}_{0}$ ) bits from M, generate two decimals as ${b}_{i}=\left({n}_{5}{n}_{4}{n}_{3}\right)$_{10} and ${b}_{i+1}=\left({n}_{2}{n}_{1}{n}_{0}\right)$_{10}. | |

Step 2 | Find the respective $SR{D}_{0\left({b}_{i}\right)}$, $SR{D}_{1\left({b}_{i}\right)}$ against ${b}_{i}$ and $SR{D}_{0\left({b}_{i+1}\right)}$, $SR{D}_{1\left({b}_{i+1}\right)}$ for ${b}_{i+1}$ (from Table 2). | |

Step 3 | Discard the $SR{D}_{1\left({b}_{i}\right)}$ and $SR{D}_{1\left({b}_{i+1}\right)}$ in case of −1. | |

Step 4 | Generate the nearest pixels (high, medium and low values) against ${g}_{i}$ using Equation (2) where its RDs must be matched either with $SR{D}_{0\left({b}_{i}\right)}$ or $SR{D}_{1\left({b}_{i}\right)}$ for ${b}_{i}$ case, denoted as ${S}_{0}{g}_{iH}$, ${S}_{0}{g}_{iM}$, ${S}_{0}{g}_{iL}$, ${S}_{1}{g}_{iH}$, ${S}_{1}{g}_{iM}$, ${S}_{1}{g}_{iL}$ and ∈ [0, 255].
$$\begin{array}{lll}{S}_{0}{g}_{iL}=NearPixFun\left({g}_{i}-10,SR{D}_{0\left({b}_{i}\right)}\right)& & {S}_{1}{g}_{iL}=NearPixFun\left({g}_{i}-10,SR{D}_{1\left({b}_{i}\right)}\right)\\ {S}_{0}{g}_{iM}=NearPixFun\left({g}_{i},SR{D}_{0\left({b}_{i}\right)}\right)& & {S}_{1}{g}_{iM}=NearPixFun\left({g}_{i},SR{D}_{1\left({b}_{i}\right)}\right)\\ {S}_{0}{g}_{iH}=NearPixFun\left({g}_{i}+10,SR{D}_{0\left({b}_{i}\right)}\right),& & {S}_{1}{g}_{iH}=NearPixFun\left({g}_{i}+10,SR{D}_{1\left({b}_{i}\right)}\right)\end{array}$$
$$NearPixFun\left(arg1,arg2\right)=\left(\left(floor\left(\frac{arg1}{10}\right)\times 10\right)+arg2\right)$$
| |

Step 5 | Repeat the step 4 for ${g}_{i+1}$ pixel and compute its nearest values ${S}_{0}{g}_{i+1H}$, ${S}_{0}{g}_{i+1M}$, ${S}_{0}{g}_{i+1L}$, ${S}_{1}{g}_{i+1H}$, ${S}_{1}{g}_{i+1M}$ and ${S}_{1}{g}_{i+1L}$. | |

Step 6 | Choose the best (minimum difference) stego-pixels ($g{\prime}_{i},g{\prime}_{i+1}$), for $g{\prime}_{i}$ value with ${S}_{0}{g}_{iH}$, ${S}_{0}{g}_{iM}$, ${S}_{0}{g}_{iL}$, ${S}_{1}{g}_{iH}$, ${S}_{1}{g}_{iM}$ and ${S}_{1}{g}_{iL}$ using Equation (3).
$$g{\prime}_{i}=argmi{n}_{\left\{xinCPVj\right\}}\left\{\right|x-{g}_{i}\left|\right\}$$
Repeat this step for $g{\prime}_{i+1}$ value with ${S}_{0}{g}_{i+1H}$, ${S}_{0}{g}_{i+1M}$, ${S}_{0}{g}_{i+1L}$, ${S}_{1}{g}_{i+1H}$, ${S}_{1}{g}_{i+1M}$ and ${S}_{1}{g}_{i+1L}$. | |

Step 7 | If the $g{\prime}_{i}$ and $g{\prime}_{i+1}$ pixel values ∈ [0, 255] and the new difference $d{\prime}_{i}=|g{\prime}_{i}$ − $g{\prime}_{i+1}|$ belongs to region-1 level (of Table 1) return/stop otherwise go to step 4. | |

End |

Let | ${g}_{i}$ = 74 and ${g}_{i+1}$ = 99 are the ith block pixels values of C, the secret message bits M = (10101101010…)_{2} | |

Begin: | ||

Step 1 | Read n = 6 = (1 0 1 0 1 1 )_{2} bits from M, generated decimals are ${b}_{i}$ = (1 0 1 )_{2} =(5)_{10} and ${b}_{i+1}$ = (0 1 1)_{2} = (3)_{10}. | |

Step 2 | Found respective $SR{D}_{0\left(5\right)}$ = 5, $SR{D}_{1\left(5\right)}$ = −1 against ${b}_{i}$ = 5 and $SR{D}_{0\left(3\right)}$ = 3, $SR{D}_{1\left(3\right)}$ = 8 against ${b}_{i+1}$ = 3 from Table 2. | |

Step-3 | Discarded the $SR{D}_{1\left(5\right)}$ = −1, because pixel right digit should be ∈ [0, 9]. | |

Step 4 | Generated nearest pixels for ${g}_{i}$ = 74 as ${S}_{0}{g}_{iH}$ = 85, ${S}_{0}{g}_{iM}$ = 75 and ${S}_{0}{g}_{iL}$ = 65 with $SR{D}_{0\left(5\right)}$ = 5 and ∈ [0, 255].
$$\begin{array}{l}{S}_{0}{g}_{iL}=65=NearPixFun\left(\left(74-10\right),5\right)\\ {S}_{0}{g}_{iM}=75=NearPixFun\left(74,5\right)\\ {S}_{0}{g}_{iH}=85=NearPixFun\left(\left(74+10\right),5\right)\end{array}$$
| |

Step 5 | Repeated step 4 for ${g}_{i+1}$ = 99 pixel and its nearest values with $SR{D}_{0\left(3\right)}$ = 3 are ${S}_{0}{g}_{i+1H}$ = 103, ${S}_{0}{g}_{i+1M}$ = 93 and ${S}_{0}{g}_{i+1L}$ = 83. On the other hand, the nearest values for ${g}_{i+1}$ = 99 with $SR{D}_{1\left(3\right)}$ = 8 are ${S}_{1}{g}_{i+1H}$ = 108, ${S}_{1}{g}_{i+1M}$ = 98 and ${S}_{1}{g}_{i+1L}$ = 88.
$$\begin{array}{lll}{S}_{0}{g}_{i+1L}=83=NearPixFun\left(\left(99-10\right),3\right)& & {S}_{1}{g}_{i+1L}=88=NearPixFun\left(\left(99-10\right),8\right)\\ {S}_{0}{g}_{i+1M}=93=NearPixFun\left(99,3\right)& & {S}_{1}{g}_{i+1M}=98=NearPixFun\left(99,8\right)\\ {S}_{0}{g}_{i+1H}=103=NearPixFun\left(\left(99+10\right),3\right)& & {S}_{1}{g}_{i+1H}=108=NearPixFun\left(\left(99+10\right),8\right)\end{array}$$
| |

Step 6 | Selected the best closest $g{\prime}_{i}$ = 75 and $g{\prime}_{i+1}$ = 98 values from $g{\prime}_{i}$ = 75 = $argmi{n}_{\left[85,75,65\right]}${|74 − [85, 75, 65]|}, $g{\prime}_{i+1}$ = 98 = $argmi{n}_{\left[83,88,93,98,103,108\right]}${|99 − [83, 88, 93, 98, 103, 108]|} | |

Step 7 | Both $g{\prime}_{i}$ = 75 and $g{\prime}_{i+1}$ = 98 pixel values ∈ [0, 255] and the new difference $23=\left|75-98\right|$ exists in region-1 level (of Table 1). | |

End |

Let | ${\mathit{g}}_{\mathit{i}}$ and ${\mathit{g}}_{\mathit{i}\mathbf{+}\mathbf{1}}$ are the ith block pixels values of cover-image C, M be the secret message bits streams, k indicates the number of least bits for LSB embedding, Table 1 is used to identify the level of region-1 and region-2 difference range of pixel blocks. | |

Begin: | ||

Step 1 | Partitioned the C into two consecutive pixels with i no. of blocks in raster scan order, $bloc{k}_{i}=({g}_{i}$, ${g}_{i+1})$. | |

Step 2 | Calculate the difference ${d}_{i}=\left({g}_{i+1}-{g}_{i}\right)$. | |

Step 3 | If the $\left|{d}_{i}\right|$ belongs to region-1 level of Table 1, apply RMDR embedding method (Section 2.1) with M to satisfy the following condition.- -
- Stego-pixels values $g{\prime}_{i}$ and $g{\prime}_{i+1}$ ∈ [0, 255].
- -
| |

Step 4 | If the $\left|{d}_{i}\right|$ belongs to region-2 level of Table 1, apply the k-bit ALSB OPAP with region-2 k secret bits (M) embedding method to satisfy the following condition.- -
**If**new difference $\left|d{\prime}_{i}\right|\in $ region-1 level of Table 1.- Then the readjustment process is applied on $g{\prime}_{i}$ and $g{\prime}_{i+1}$ using Equation (4).
| |

Step 5 | Repeat Steps 1–5 until all M are embedded; if all cover blocks are traversed while M has not been embedded completely, restart Step 1 with new C. | |

End |

Let | Stego-pixel $\mathit{g}{\mathbf{\prime}}_{\mathit{i}}$ and $\mathit{g}{\mathbf{\prime}}_{\mathit{i}\mathbf{+}\mathbf{1}}$ are the ith block pixels values of S.
Table 3
, ExRD is consider as $ExR{D}_{i}$ and $ExR{D}_{i+1}$ and b is treated as ${b}_{i}$ and ${b}_{i+1}$. | |

Begin: | ||

Step 1 | Read $g{\prime}_{i}$ and $g{\prime}_{i+1}$ pixels of ith block of S. | |

Step 2 | Extract the RDs as $ExR{D}_{i}$ and $ExR{D}_{i+1}$ from $g{\prime}_{i}$ and $g{\prime}_{i+1}$, respectively, i.e., $ExR{D}_{i}$ = Mod ($g{\prime}_{i}$, 10). | |

Step 3 | Find the ${b}_{i}$ and ${b}_{i+1}$ equivalent values against $ExR{D}_{i}$ and $ExR{D}_{i+1}$ from Table 3. | |

Step 4 | Convert the decimal values of ${b}_{i}$ and ${b}_{i+1}$ into binary and concatenate the n = 6 (${n}_{5},{n}_{4},{n}_{3},{n}_{2},{n}_{1},{n}_{0}$) bits as a recovered M bit stream. | |

End |

Let | Stego-pixel $g{\prime}_{i}$, $g{\prime}_{i+1}$ = (75, 98) are the ith block pixel values of S. | |

Begin: | ||

Step 1 | Read $g{\prime}_{i}$ = 75 and $g{\prime}_{i+1}$ = 98 pixels. | |

Step 2 | Extracted RD’s are $ExR{D}_{i}$ = 5 and $ExR{D}_{i+1}$ = 8. e.g., 5 = Mod (75, 10). | |

Step 3 | Found equivalent of $ExR{D}_{i}$ = 5 and $ExR{D}_{i+1}$ = 8 are ${b}_{i}$ = 5 and ${b}_{i+1}$ = 3 from Table 3. | |

Step 4 | Converted binary and concatenated as ${b}_{i}$ = (5)_{10} = (101)_{2} and ${b}_{i+1}$ = (3)_{10} = (011)_{2} and (101011)_{2} bits as a recovered six-bit stream. | |

End |

Let | $\mathit{g}{\mathbf{\prime}}_{\mathit{i}}$ and $\mathit{g}{\mathbf{\prime}}_{\mathit{i}\mathbf{+}\mathbf{1}}$ are the ith block pixels values of stego-image S. | |

Begin: | ||

Step 1 | Partitioned S into two consecutive pixels with i no. of blocks in raster scan order, $bloc{k}_{i}=\left({{g}^{\prime}}_{i},g{\prime}_{i+1}\right)$. | |

Step 2 | Compute the difference $d{\prime}_{i}=\left(g{\prime}_{i+1}-g{\prime}_{i}\right)$. | |

Step 3 | If the $\left|d{\prime}_{i}\right|$ belongs to region-1 level of Table 1, apply RMDR extraction method (Section 2.3); otherwise apply ALSB extraction using Table 1 with k-bit secret bits process. | |

Step 4 | Repeat Steps 1–4 until all M is extracted from S. | |

End |

**Table 1.**Proposed hybrid embedding method range table divisions as region-1 and region-2 levels, where k denotes the least bits for embedding.

Regions | Region-1 Level | Region-2 Level | ||
---|---|---|---|---|

Lower-Upper bound of R_{n} | ${R}_{1}\text{}\in \left[0,31\right]$ | ${R}_{2}\text{}\in \text{}\left[32,63\right]$ | ${R}_{3}\text{}\in \left[64,127\right]$ | ${R}_{4}\in \left[128,255\right]$ |

Secret bits | 3 | k = 4 = ${\text{log}}_{2}\left(63-32\right)-1$ | k = 5 = ${\text{log}}_{2}\left(127-64\right)-1$ | k = 6 = ${\text{log}}_{2}\left(225-128\right)-1$ |

b | SRD_{0}(b) | SRD_{1}(b) |
---|---|---|

0 | 0 | −1 |

1 | 1 | −1 |

2 | 2 | −1 |

3 | 3 | 8 |

4 | 4 | 9 |

5 | 5 | −1 |

6 | 6 | −1 |

7 | 7 | −1 |

ExRD | b |
---|---|

0 | 0 |

1 | 1 |

2 | 2 |

3 | 3 |

4 | 4 |

5 | 5 |

6 | 6 |

7 | 7 |

8 | 3 |

9 | 4 |

**Table 4.**Performance comparison between the proposed hybrid embedding method and singular steganographic methods.

Parameters | Methods | Lena | Baboon | Pepper | Jet | Barbara | Zelda | Tiffany | Elaine | Average |
---|---|---|---|---|---|---|---|---|---|---|

PSNR (dB) | Proposed | 39.19 | 38.03 | 39.34 | 39.18 | 37.92 | 39.56 | 39.24 | 39.50 | 39.00 |

3-bit LSB | 37.90 | 37.90 | 37.91 | 37.95 | 37.92 | 37.92 | 37.89 | 37.92 | 37.91 | |

Wu and Tsai [19] | 41.10 | 36.98 | 41.55 | 40.42 | 36.33 | 42.94 | 41.48 | 41.88 | 40.34 | |

Yang et al. [41] | 39.31 | 39.16 | 39.06 | 39.55 | 39.16 | 39.00 | 39.12 | 39.01 | 39.17 | |

Capacity (bits) | Proposed | 793,810 | 820,776 | 792,384 | 795,728 | 824,084 | 788,516 | 792,864 | 791,994 | 800,019 |

3-bit LSB | 786,432 | 786,432 | 786,432 | 786,432 | 786,432 | 786,432 | 786,432 | 786,432 | 786,432 | |

Wu and Tsai [19] | 409,776 | 456,952 | 405,424 | 409,531 | 450,650 | 339,918 | 398,980 | 408,582 | 417,447 | |

Yang et al. [41] | 757,332 | 785,572 | 786,014 | 735,236 | 786,012 | 778,014 | 777,888 | 760,016 | 770,761 |

**Table 5.**Performance comparison of existing hybrid LSB-based (PVD, EMD, and MPE) methods against the proposed method.

Parameters | Methods | Lena | Baboon | Pepper | Jet | Barbara | Zelda | Tiffany | Elaine | Average |
---|---|---|---|---|---|---|---|---|---|---|

PSNR (dB) | Proposed | 39.19 | 38.03 | 39.34 | 39.18 | 37.92 | 39.56 | 39.24 | 39.50 | 39.00 |

Wu et al. [27] | 37.12 | 35.30 | 37.20 | 36.98 | 34.79 | 37.63 | 37.25 | 37.28 | 36.70 | |

Jung et al. [29] | 36.28 | 35.94 | 36.10 | 36.20 | 36.02 | 36.05 | 35.12 | 35.92 | 35.95 | |

Yang et al. [34] | 38.71 | 36.19 | 38.92 | 38.56 | 35.57 | 39.78 | 39.11 | 39.26 | 38.26 | |

Khodaei et al. [28] | 37.56 | 34.85 | 35.88 | 36.29 | 32.91 | 38.49 | 37.78 | 38.17 | 36.49 | |

Wu et al. [25] | 35.10 | 35.10 | 35.10 | 35.10 | 35.10 | 35.10 | 35.10 | 35.10 | 35.10 | |

Capacity (bits) | Proposed | 793,810 | 820,776 | 792,384 | 795,728 | 824,084 | 788,516 | 792,864 | 791,994 | 800,019 |

Wu et al. [27] | 765,969 | 717,749 | 768,455 | 770,176 | 740,147 | 776,196 | 766,663 | 760,170 | 758,191 | |

Jung et al. [29] | 786,432 | 786,441 | 786,586 | 786,475 | 786,406 | 786,647 | 786,559 | 786,440 | 786,498 | |

Yang et al. [34] | 765,969 | 717,749 | 768,455 | 770,176 | 740,147 | 776,196 | 766,663 | 760,170 | 758,191 | |

Khodaei et al. [28] | 791,443 | 809,435 | 790,299 | 792,443 | 811,747 | 787,887 | 790,503 | 788,356 | 795,264 | |

Wu et al. [25] | 639,761 | 603,894 | 620,920 | 650,362 | 626,994 | 641,866 | 643,305 | 615,116 | 630,227 |

Methods | Average Capacity | Average PSNR | Average Q | |||
---|---|---|---|---|---|---|

UCID [39] | USC-SIPI [40] | UCID [39] | USC-SIPI [40] | UCID [39] | USC-SIPI [40] | |

Proposed | 554,156 | 1,636,102 | 38.88 | 38.32 | 0.9988 | 0.9952 |

Wu et al. [27] | 514,375 | 150,3772 | 36.67 | 35.99 | 0.9979 | 0.9919 |

Yang et al. [34] | 514,375 | 150,3772 | 37.91 | 37.19 | 0.9984 | 0.9940 |

Jung et al. [29] | 540,682 | 1,591,207 | 34.48 | 34.12 | 0.9951 | 0.9862 |

Khodaei et al. [28] | 549,448 | 1,623,681 | 35.02 | 34.72 | 0.9955 | 0.9873 |

Wu et al. [25] | 481,140 | 141,6278 | 35.10 | 35.10 | 0.9972 | 0.9890 |

**Table 7.**Universal quality index (Q) of proposed and existing LSB-based methods by Wang and Bovik [42].

Methods | Lena | Baboon | Pepper | Jet | Barbara | Zelda | Tiffany | Elaine | Average |
---|---|---|---|---|---|---|---|---|---|

Proposed | 0.9983 | 0.9979 | 0.9987 | 0.9982 | 0.9972 | 0.9988 | 0.9985 | 0.9983 | 0.9982 |

3-bit LSB | 0.9977 | 0.9971 | 0.9982 | 0.9976 | 0.9983 | 0.9968 | 0.9939 | 0.9975 | 0.9971 |

Yang et al. [34] | 0.9981 | 0.9955 | 0.9985 | 0.9979 | 0.9969 | 0.9979 | 0.9952 | 0.9980 | 0.9973 |

Khodaei et al. [28] | 0.9975 | 0.9940 | 0.9971 | 0.9965 | 0.9944 | 0.9972 | 0.9937 | 0.9977 | 0.9960 |

Wu et al. [25] | 0.9959 | 0.9948 | 0.9967 | 0.9957 | 0.9968 | 0.9943 | 0.9893 | 0.9956 | 0.9949 |

**Table 8.**Undetectabilityperformance under SPAM steganalysis with ensemble classifier for the proposed method as compared with the LSB matching and HUGO embedding methods.

Bit Rate (bpp) | Method | TP | FP | TN | FN | Error rate |
---|---|---|---|---|---|---|

0.2 | Proposed | 228 | 249 | 251 | 272 | 52.10% |

LSBM (±1) | 421 | 174 | 326 | 79 | 25.30% | |

HUGO | 234 | 252 | 248 | 266 | 51.80% | |

0.4 | Proposed | 276 | 267 | 233 | 224 | 49.10% |

LSBM (±1) | 443 | 102 | 398 | 57 | 15.90% | |

HUGO | 279 | 259 | 241 | 221 | 48.00% | |

0.6 | Proposed | 301 | 228 | 272 | 199 | 42.70% |

LSBM (±1) | 453 | 55 | 445 | 47 | 10.20% | |

HUGO | 298 | 210 | 290 | 202 | 41.20% | |

0.8 | Proposed | 336 | 155 | 345 | 164 | 31.90% |

LSBM (±1) | 478 | 38 | 462 | 22 | 06.00% | |

HUGO | 339 | 149 | 351 | 161 | 31.00% | |

1.0 | Proposed | 441 | 103 | 397 | 59 | 16.20% |

LSBM (±1) | 497 | 14 | 486 | 3 | 01.70% | |

HUGO | 446 | 106 | 394 | 54 | 16.00% |

© 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Hussain, M.; Abdul Wahab, A.W.; Javed, N.; Jung, K.-H.
Hybrid Data Hiding Scheme Using Right-Most Digit Replacement and Adaptive Least Significant Bit for Digital Images. *Symmetry* **2016**, *8*, 41.
https://doi.org/10.3390/sym8060041

**AMA Style**

Hussain M, Abdul Wahab AW, Javed N, Jung K-H.
Hybrid Data Hiding Scheme Using Right-Most Digit Replacement and Adaptive Least Significant Bit for Digital Images. *Symmetry*. 2016; 8(6):41.
https://doi.org/10.3390/sym8060041

**Chicago/Turabian Style**

Hussain, Mehdi, Ainuddin Wahid Abdul Wahab, Noman Javed, and Ki-Hyun Jung.
2016. "Hybrid Data Hiding Scheme Using Right-Most Digit Replacement and Adaptive Least Significant Bit for Digital Images" *Symmetry* 8, no. 6: 41.
https://doi.org/10.3390/sym8060041