63
IMAGE DATA HIDING SCHEME WITH OPTIMAL DATA
RECOVERY FEATURES
Mr. N. Vipin Raj1, Dr. T. Senthil Prakash2, Mr. R. Senthilkumar3,Ms. N. Anusha4 II Year M.E (CSE)1, Professor & HOD2, Associate Professor3, II Year M.E(CSE)4
Shree Venkateshwara Hi-Tech Engg. College, Gobi, Tamilnadu, India , 1,2,3,4 [email protected], [email protected], [email protected]
Abstract
Reversible data hiding (RDH) is used to embed secret message into a cover image by slightly modifying its pixel values. Embedded message and the cover image are completely recovered from the marked content. RDH supports information hiding with the lossless compressibility of natural images. Lossless compression, difference expansion, histogram modification, prediction-error expansion and integer transform techniques are used for RDH process. Histogram based RDH method is divided into two steps histogram generation and histogram modification. Histogram construction is performed with the pixel pairs sequences and their different values. Histogram modification is carried out to embed data into the cover image. The un-hiding process recovers the message and also the cover image. Prediction-error expansion (PEE) technique is applied for Reversible Data Hiding (RDH) process. One-or two-dimensional Prediction-error Histogram (PEH) are used in the PEE techniques. The two-dimensional PEH-based methods perform better than one dimensional PEH. PEH modification is fixed and independent of image content. Multiple histograms based PEE method is adopted to improve reversible data hiding (RDH) process. Multiple Histograms Modification (MHM) method uses a sequence of histograms for the hiding process. A complexity measurement is computed for each pixel with reference to its context. Prediction-Error Histogram (PEH) is generated using the pixels with the complexity value. A sequence of histograms can be generated by varying the complexity to cover the whole image. Two expansion bins are selected in each generated histogram. The expansion bins are selected with reference to the image content. Data embedding is carried out on Multiple Histograms Modification (MHM).
1. Introduction
Steganography is the art or practice of concealing a message, image, or file within another
message, image, or file. The
word steganography combines the Ancient
Greekwords steganos, meaning "covered, concealed, or protected" and graphei meaning "writing". The first recorded use of the term was in
1499 by Johannes Trithemius in
his Steganographia, a treatise on cryptography and steganography, disguised as a book on magic [11]. The hidden messages will appear to be something else: images, articles, shopping lists, or some other cover text. For example, the hidden message may be in invisible ink between the visible lines of
a private letter. Some implementations of steganography which lack a shared secret are forms of security through obscurity, whereas key-dependent steganographic schemes adhere to Kerckhoffs's principle\
The advantage of steganography
64 fact that a secret message is being sent, as well as concealing the contents of the message. Steganography includes the concealment of information within computer files. In digital steganography, electronic communications may include steganographic coding inside of a transport layer, such as a document file, image file, program or protocol. Media files are ideal for steganographic transmission because of their large size. For example, a sender might start with an innocuous image file and adjust the color of every 100th pixel to correspond to a letter in the alphabet, a change so subtle that someone not specifically looking for it is unlikely to notice it.
In general, terminology analogous to more
conventional radio and communications
technology is used; a brief description of some terms which show up in software specifically and are easily confused, is appropriate. These are most relevant to digital stegano graphic systems.
The payload is the data to be covertly
communicated [12]. The carrier is the signal, stream, or data file into which the payload is hidden; which differs from the "channel". The resulting signal, stream, or data file which has the payload encoded into it is sometimes referred to as the package, stego file, or covert message. The percentage of bytes, samples, or other signal elements which are modified to encode the payload is referred to as the encoding density and is typically expressed as a number between 0 and 1. In a set of files, those files considered likely to contain a payload are called suspects. If the suspect was identified through some type of statistical analysis, it might be referred to as a candidate.
Detection of physical steganography
requires careful physical examination, including the use of magnification, developer chemicals and ultraviolet light. It is a time-consuming process with obvious resource implications, even in countries where large numbers of people are employed to spy on their fellow nationals. It is
feasible to screen mail of certain suspected individuals or institutions, such as prisons or prisoner-of-war (POW) camps. During World War II, a technology used to ease monitoring of POW mail was specially treated paper that would reveal invisible ink. An article in the 24 June 1948 issue of Paper Trade Journal by the Technical Director of the United States Government Printing Office, Morris S. Kantrowitz, describes in general terms the development of this system, three prototypes of which were named Sensicoat, Anilith
and Coatalith paper. These were for the
manufacture of post cards and stationery to be given to German prisoners of war in the US and Canada. If POWs tried to write a hidden message the special paper would render it visible. At least two US patents were granted related to this technology, one to Mr. Kantrowitz, U.S. Patent 2,515,232, "Detecting paper and Water-Detecting Coating Composition Therefor", patented 18 July 1950 and an earlier one, "Moisture-Sensitive Paper and the Manufacture Thereof", U.S. Patent 2,445,586, patented 20 July 1948. A similar strategy is to issue prisoners with writing paper ruled with a water-soluble ink that "runs" when in contact with a water-based invisible ink.
In computing, detection of stegano
graphically encoded packages is
65 payload in, raising the encoding density and facilitating easier detection.
2. Related work
Ni et al. proposed a reversible data hiding method based on histogram modification. In the scheme, part of the cover image histogram is shifted rightward or leftward to produce redundancy for data embedding. First, the peak and zero point bins of the original histogram are found denoted as b(P) and b(Z), respectively. Then all the bins belonging to b(P) and b(Z) are shifted rightward one level. In this way, the bin of b(P) is emptied and b(P + 1) becomes the new peak point. Next, the confidential data can be embedded by modulating the pixel values equaling P + 1. That is, if encounter a pixel with value equaling P + 1, then one bit confidential data can be hidden. For example, if the current processing confidential bit is “0”, we modify the pixel value as P; whereas if the current processing confidential bit is “1”, the pixel with value P + 1 is kept no changed. In decoder, the data extraction and image recovery is the inverse processing of data embedding.
In [3], Li et al. proposed a reversible data hiding method named adjacent pixel difference (APD) based on the neighbor pixel differences modification. In this method, an inverse “S” order is adopted to scan the image pixels. A 3 × 3 image block is used to illustrate this principle. The scan direction is marked as the blue line and the block can be rearranged into a pixel sequence as p1, p2, . . ., p9.
Suppose the host image I is an 8-bit gray level image sized as M × N. Then a pixel sequence p1, p2, . . ., pM×N are obtained via the inverse “S” order scan. The differences of adjacent pixels are computed as:
(1)
Considering the pixel values similarity between pi−1 and pi, a large quantity of di (2 ≤ i≤ M
× N) is equal or close to 0. The difference histogram is constructed based on these M × N − 1 difference statistics. Suppose the histogram bins from left to right are denoted by b(−255), b(−254), . . ., b(−1), b(0), b(1), . . ., b(254), b(255). The 512 × 512 Lena image’s difference histogram. Obviously most differences are concentrated around b(0). When the curve spreads away t both sides, it drops dramatically and no differences fall into those bins far from b(0).
Basically, APD selects one pair of bins b(p1) and b(z1) (suppose p1 < z1) where b(p1) and b(z1) denote the peak point and zero point, respectively. Then the bins between [b(p1 + 1), b(z1 − 1)] are shifted rightward one level. Thus b(p1 1) are emptied for data embedding. That is, if a secret bit “1” is embedded, the differences equaling p1 are added by 1. If “0” is embedded, they are not changed. To enhance the capacity, APD can also select two pairs of peakzero points, e.g. [b(p1), b(z1)] and [b(z2), b(p2)] (suppose p1 < z1 and z2 < p2). Then the bins between [b(p1 + 1), b(z1 − 1)] are shifted rightward one level and those between [b(z2 + 1), b(p2 − 1)] are shifted leftward one level. Thus b(p1 + 1) and b(p2 − 1) are emptied for data embedding [4]. The secret bits modulation is similar as that in one pair of peak-zero points embedding. Note the ranges of [b(p1), b(z1)] and [b(z2), b(p2)] must not be overlapped.
3. Reversible Data Hiding Based on Multiple Histograms Modification
66 in the cloud [10], etc. RDH is a fragile technique and the marked image cannot undergo any degradation. RDH method is usually evaluated by its capacity-distortion performance, i.e., for a given embedding capacity (EC), one expects to minimize the embedding distortion measured by PSNR of the marked image versus the original one.
Early RDH methods are mainly based on lossless compression. The idea behind these methods is to losslessly compress a feature set of cover image and utilize the saved space for reversible embedding. Fridrich et al. proposed to compress a proper bit-plane with the minimum redundancy. Celik et al. proposed a generalized least significant bit (LSB) compression method to improve the compression efficiency by using unaltered bit-planes as side information. The lossless compression-based methods cannot yield satisfactory performance, since the correlation within a bit-plane is too weak to provide a high EC. As EC increases, one needs to compress more bit-planes, thus the distortion increases dramatically.
Later on, more efficient RDH methods based on histogram modification and expansion technique have been devised. The histogram-modification-based method is firstly proposed by Ni et al. This method focuses on high visual quality with quite limited EC, in which the peak point of image histogram is utilized for data embedding. In this method, each pixel value is modified at most by 1 and thus the marked image quality is well guaranteed. Ni et al.’s method is
improved by Lee et al. using the histogram of difference image. The spatial correlation of natural images is exploited considering the difference of adjacent pixels. Thus, a regular-shaped histogram is utilized in Lee et al.’s method. This histogram is
centered at origin and has rapid two-sided decay which is more suitable for RDH.
The expansion technique is firstly
proposed by Tian. This method is performed on pixel pairs and one data bit is embedded into each
selected pixel pair by expanding its difference. Compared with the lossless-compression based RDH, Tian’s difference expansion (DE) based method can provide a higher EC with an improved PSNR. The DE approach has attracted considerable attention and it makes an important progress in RDH. Afterwards, the expansion technique has been widely investigated and developed, mainly in the aspects of integer-to-integer transformation [1], location map reduction [9] and prediction-error expansion (PEE) [5]. Besides the histogram modification and the expansion technique, the analysis about theoretical capacity limit subjected to admissible distortion has also been studied in some recent works.
The most effective and extensively
exploited RDH technique is the PEE technique which is firstly proposed by Thodi and Rodriguez. Instead of the difference value in DE, the prediction-error is utilized in PEE for expansion embedding. Thus, unlike DE where only the correlation of two adjacent pixels is considered, the local correlation of a larger neighborhood is exploited in PEE. As a result, compared with DE, better performance can be derived by PEE. Following Thodi and Rodriguez’s work, many RDH techniques related to PEE have been proposed in recent years, for example, double-layered embedding, adaptive embedding, context modification, optimal expansion bins selection and two-dimensional histogram modification, etc. Some PEE-based methods exploit advanced prediction techniques to generate a more sharply distributed prediction-error histogram (PEH) and this is also helpful for enhancing the embedding performance.
Most previous PEE-based methods are
based on one- or two-dimensional PEH
67 We focus on PEE and propose a new RDH method based on PEE for multiple histograms. Unlike the previous methods, we consider here a sequence of histograms and devise a new embedding
mechanism based on multiple histograms
modification (MHM). By MHM, the embedding performance can be optimized by adaptively selecting expansion bins in each histogram considering the image content. Its prediction value and complexity measurement are computed according to its context and multiple histograms are generated for different complexity levels. The pixels with a given complexity are collected together to generate a PEH and by varying the complexity measurement to cover the whole image, a sequence of histograms can be derived. Two expansion bins are selected in each generated histogram and data embedding is realized based on MHM. Based on an estimation of embedding distortion, the expansion bins can be effectively determined such that the distortion is
minimized. The proposed method is a
generalization of some existing methods and it can well exploit image redundancy to achieve improved embedding performance. Experimental results show that the proposed method outperforms
the conventional PEE (C-PEE) and its
miscellaneous extensions including both one- or two-dimensional PEH based ones. Our advantages mainly lie in the MHM-based embedding mechanism and the selection of optimal expansion bins.
4. Problem Statement
Prediction-error expansion (PEE)
technique is applied for Reversible Data Hiding
(RDH) process. One-or two-dimensional
Prediction-error Histogram (PEH) are used in the PEE techniques. The two-dimensional PEH-based methods perform better than one dimensional PEH. PEH modification is fixed and independent of image content. Multiple histograms based PEE method is adopted to improve reversible data
hiding (RDH) process. Multiple Histograms Modification (MHM) method uses a sequence of histograms for the hiding process. A complexity measurement is computed for each pixel with reference to its context. Prediction-Error Histogram (PEH) is generated using the pixels with the complexity value. A sequence of histograms can be generated by varying the complexity to cover the whole image. Two expansion bins are selected in each generated histogram. The expansion bins are selected with reference to the image content. Data embedding is carried out on Multiple Histograms Modification (MHM). The following problems are identified from the existing system.
• Embedding capacity (EC) is low • Predictor selection is not optimized • Limited embedding performance
• Complexity measure selection is not optimized
• Limited secret data security
5. Image Data Hiding Scheme with Optimal Data Recovery
The reversible data hiding scheme is improved with security features. The data values are hided with reference to the histogram information. The RSA algorithm is applied to secure the secret data value. The system is divided into four major modules. They are Sender, Data security, Receiver and Data extraction.
The sender module is designed to send the image data values. Data security is designed to perform hiding and encryption process. Data receiver module collects data from the sender. Data extraction module is designed to perform unhide operations.
5.1. Sender
68 applied with Prediction Error Histogram (PEH) with different complexity levels. Histogram construction is carried out using the cover data image.
5.2. Data Security
Data security process is used to hide secret data values. Secret data is converted into bits. RSA algorithm is used to encrypt the data values. The sender collects the public key from the receiver node for the encryption process. Encrypted data values are hided in the cover image.
5.3. Receiver
Data receiver collects data from the sender node. Received data values are updated into the local memory. The received data value is passed to unhide and decrypt process. The receiver node maintains the secret key for decryption process.
5.4. Data Retrieval
Secret data is separated from the received data values. Cover data is also separated from the received data values. Decryption process is carried out to fetch the secret data. Cover data quality is analyzed with image quality measures.
6. Conclusion
Reversible Data Hiding (RDH) techniques are used to support data hiding with message and cover image retrieval mechanism. Multiple Histogram Modification (MHM) scheme is employed for the data hiding process. The system is enhanced to improve the embedding capacity with optimized predictor selection approach. RSA algorithm is adapted to ensure the security level of secret data values. The Multiple Histogram Modification (MHM) scheme is enhanced to improve the embedding capacity. Embedding performance is improved by the system. The system supports efficient coverage image retrieval process. The system reduces the process time in hiding and un-hiding process.
REFERENCES
[1] F. Peng, X. Li and B. Yang, “Adaptive Reversible Data Hiding Scheme Based On Integer Transform,” Signal Process., vol. 92, no. 1, pp. 54–62, Jan. 2012.
[2] R. Caldelli, F. Filippini and R. Becarelli, “Reversible Watermarking Techniques: An Overview And A Classification,” EURASIP J. Inf. Security, vol. 2010, Jun. 2010, Art. ID 134546.
[3] Li YC, Yeh CM, Chang CC. Data hiding based on the similarity between neighboring pixels with reversibility. Digital Signal Processing 2009.
[4] Zhenfei Zhao, Hao Luo, Zhe-Ming Lu and Jeng-Shyang Pan, “Reversible data hiding based
on multilevel histogram modification and sequential recovery”, Elsever, 2011
[5] D. Coltuc, “Improved Embedding For Prediction-Based Reversible Watermarking,” IEEE Trans. Inf. Forensics Security, vol. 6, no. 3, pp. 873–882, Sep. 2011.
[6] M. Fontani et al., “Reversible Watermarking For Image Integrity Verification In Hierarchical PACS,” in Proc. ACM MM&SEC, 2010, pp. 161– 168.
[7] F. Battisti, M. Carli and A. Neri, “Secure Annotation For Medical Images Based On Reversible Watermarking In The Integer Fibonacci–Haar Transform Domain,” Proc. SPIE, vol. 7870, p. 78700G, Feb. 2011.
69
[9] D. Coltuc, “Low Distortion Transform For Reversible Watermarking,” IEEE Trans. Image Process., vol. 21, no. 1, pp. 412–417, Jan. 2012.
[10] K. Hwang and D. Li, “Trusted Cloud Computing With Secure Resources And Data Coloring,” IEEE Internet Comput., vol. 14, no. 5, pp. 14–22, Sep./Oct. 2010.
[11] Yanzhen Ren, Tingting Cai, Ming Tang and Lina Wang, “AMR Steganalysis Based on the Probability of Same Pulse Position”, IEEE Transactions On Information Forensics And Security, Vol. 10, No. 9, September 2015
