Prediction-error expansion (PEE) is an efficient technique for reversible data hiding (RDH). Instead of expanding the highest histogram bins in conventional PEE, in this paper, to better utilize the image redundancy, we propose a new PEE-based RDH scheme with an advisable expansion strategy utilizing referential prediction-errors. For each pixel, we first calculate its prediction-error and use its neighbor prediction-error as a reference. The correlation of the prediction-error and its reference is exploited to adaptively select bins for expansion embedding. In addition, to further enhance the reversible embedding performance, we apply the pixel selection technique in our scheme such that the pixels located in smooth image areas are priorly embedded. Experimental results show that the proposed scheme outperforms conventional PEE and it is better than some state-of-the-art RDH works as well.

In this paper, we propose a scheme to detect the copy-move forgery in an image, mainly by extracting the keypoints for comparison. The main difference to the traditional methods is that the proposed scheme first segments the test image into semantically independent patches prior to keypoint extraction. As a result, the copy-move regions can be detected by matching between these patches. The matching process consists of two stages. In the first stage, we find the suspicious pairs of patches that may contain copy-move forgery regions, and we roughly estimate an affine transform matrix. In the second stage, an Expectation-Maximization-based algorithm is designed to refine the estimated matrix and to confirm the existence of copy-move forgery. Experimental results prove the good performance of the proposed scheme via comparing it with the state-of-the-art schemes on the public databases.

Prediction-error expansion (PEE) is the most successful reversible data hiding (RDH) technique, and existing PEE-based RDH methods are mainly based on the modification of one- or two-dimensional prediction-error histogram (PEH). The two-dimensional PEH-based methods perform generally better than those based on one-dimensional PEH; however, their performance is still unsatisfactory since the PEH modification manner is fixed and independent of image content. In this paper, we propose a new RDH method based on PEE for multiple histograms. Unlike the previous methods, we consider in this paper a sequence of histograms and devise a new embedding mechanism based on multiple histograms modification (MHM). A complexity measurement is computed for each pixel according to its context, and the pixels with a given complexity are collected together to generate a PEH. By varying the complexity to cover the whole image, a sequence of histograms can be generated. Then, two expansion bins are selected in each generated histogram and data embedding is realized based on MHM. Here, the expansion bins are adaptively selected considering the image content such that the embedding distortion is minimized. With such selected expansion bins, the proposed MHM-based RDH method works well. Experimental results show that the proposed method outperforms the conventional PEE and its miscellaneous extensions including both one- or two-dimensional PEH-based ones.

In this paper, referred to the general framework of histogram-shifting-based reversible data hiding, some valuable related works based on two-dimensional pixel-intensity-histogram modification are first introduced. In these schemes, all pixels are classified into two categories, which are embedded with data or shifted for reversibility, respectively. From this point of view, novel embedding schemes are proposed with more applicable pixel partition and more redundant information utilized. Moreover, pixel selection is conducted such that smooth pixels are priorly embedded to further exploit the image redundancy. The experimental results show the superiority of the proposed methods over some state-of-the-art works in terms of capacity-distortion performance.

Compression of encrypted data has attracted considerable research interests nowadays due to distributed processing and cloud computing. In this work, we propose a novel lossy compression scheme for encrypted gray-scale images. The original image is first divided into non-overlapping blocks. Then, it is encrypted by a modulo-256 addition and block permutation. In compression phase, the spatial correlation and quantization are exploited to reduce the compression ratio. At the decoder side, context-adaptive interpolation with an image-dependent threshold is used to make image reconstruction precise. Experimental results show that the proposed scheme achieves better performance compared to the previous work.

In this paper, based on two-dimensional difference- histogram modification, a novel reversible data hiding (RDH) scheme is proposed by using difference-pair-mapping (DPM). First, by considering each pixel-pair and its context, a sequence consisting of pairs of difference values is computed. Then, a two-dimensional difference-histogram is generated by counting the frequency of the resulting difference-pairs. Finally, reversible data embedding is implemented according to a specifically designed DPM. Here, the DPM is an injective mapping defined on difference-pairs. It is a natural extension of expansion embedding and shifting techniques used in current histogram-based RDH methods. By the proposed approach, compared with the conventional one-dimensional difference-histogram and one-dimensional prediction-error-histogram-based RDH methods, the image redundancy can be better exploited and an improved embedding performance is achieved. Moreover, a pixel-pair-selection strategy is also adopted to priorly use the pixel-pairs located in smooth image regions to embed data. This can further enhance the embedding performance. Experimental results demonstrate that the proposed scheme outperforms some state-of-the-art RDH works.

Histogram shifting (HS) is a useful technique of reversible data hiding (RDH). With HS-based RDH, high capacity and low distortion can be achieved efficiently. In this paper, we revisit the HS technique and present a general framework to construct HS-based RDH. By the proposed framework, one can get a RDH algorithm by simply designing the so-called shifting and embedding functions. Moreover, by taking specific shifting and embedding functions, we show that several RDH algorithms reported in the literature are special cases of this general construction. In addition, two novel and efficient RDH algorithms are also introduced to further demonstrate the universality and applicability of our framework. It is expected that more efficient RDH algorithms can be devised according to the proposed framework by carefully designing the shifting and embedding functions.

Least significant bit (LSB) matching is a well-known steganographic method with advantages of high payload, good visual/statistical imperceptibility and extreme ease of implementation. However, by utilizing the distortion of one-dimensional histogram or the generated additive embedding noise, some steganalyzers can perceive the existence of covert communication to some extent. Due to this, we extend the LSB matching steganography by minimizing the distortion of first order statistics (i.e., one-dimensional histogram) and adaptively embedding data into noise regions. With these extensions, our method significantly improves the stego-security. The experimental results also prove its superiority over some state-of-the-art steganographic methods against various steganalyzers.

The steganalysis of least significant bit (LSB) matching steganography has been well realized by many highly sensitive detectors so far, but few of them can analyze further to higher levels. Most of current steganalysis methods still stay in the low-level stage that can only detect the existence of secret message. In this paper, we focus on higher level steganalysis and an improved method for payload location of LSB matching is proposed. Our work is based on the combination and improvement of previous techniques including message length estimation and payload location. The proposed method first solves a regression problem to estimate the embedding rate utilizing steganalysis features. Then, the embedding payload can be located via optimal cover estimation and improved residuals calculation, in which the embedding rate achieved above is a crucial parameter. By this approach, the payload location method can be applied to practical use needless of any prior knowledge. Experimental results show that the proposed method outperforms the state-of-the-art work with a higher accuracy for payload location of LSB matching.

This paper presents a new reversible watermarking algorithm for color image. The traditional technique is to embed data into each color channel independently. Considering that the color channels correlate with each other, we propose a reversible watermarking algorithm based on prediction-error expansion that can enhance the prediction accuracy in one color channel through exploiting the gradient information from the other channels. In doing so, the magnitude of the prediction-error is decreased generally, and consequently, the distortion to the host image is diminished. Experimental results demonstrate that the proposed algorithm outperforms the traditional methods that independently embed watermark into each channel.

Discriminating computer generated graphics from photographic images is a challenging problem of digital forensics. An important approach to this issue is to explore usual image statistics. In this way, when the statistical distributions (i.e., histograms) of some types of residual images are established, previous works usually apply operations on these histograms or compute statistical quantities to extract features. However, as the histograms are fundamental resources and can present most image information, the histograms themselves can be directly used as features and we do not need further manipulations on them. Based on this consideration, we simply take several highest histogram bins of the difference images as features to carry out classification, and these simple histogram features work well in terms of both detection accuracy and computational complexity. Actually, experimental results demonstrate that, with only 112 features, the proposed method outperforms some state-of-the-art works.

Many highly sensitive detectors for least significant bit (LSB) matching steganography have been proposed so far, but few of them can locate the embedding payload. Assuming a number of stego images with payload placed in the same locations are provided, Ker presented a residual-based method to identify the payload-carrying locations for LSB matching. In this paper, we put forward two novel residuals and theoretically prove their reliability for LSB matching detection. Experimental results show that the proposed weighted l1and l2residuals significantly outperform the previous work in terms of both detection performance and computational complexity. Even very small payload can be perfectly detected and precisely located as long as adequate stego images are available.

Prediction-error expansion (PEE) develops Tian's difference expansion technique by performing reversible data embedding in the prediction-error histogram. As the prediction accuracy greatly affects the embedding performance, the PEE-based algorithms depend heavily on image's spatial redundancy. However, one drawback of these algorithms is that the modification of pixels may distort the high correlation in their local regions, and thus affect the prediction accuracy. To deal with this problem, we propose an adjustment process named "compensation" to compensate for the already modified pixels, before using them for prediction. Combining PEE and compensation together, a new reversible image watermarking algorithm is then presented. Comparisons with state-of-the-art works indicate that the compensation technique may benefit reversible watermarking in term of capacity-distortion behavior.

Prediction-error expansion (PEE) is an important technique of reversible watermarking which can embed large payloads into digital images with low distortion. In this paper, the PEE technique is further investigated and an efficient reversible watermarking scheme is proposed, by incorporating in PEE two new strategies, namely, adaptive embedding and pixel selection. Unlike conventional PEE which embeds data uniformly, we propose to adaptively embed 1 or 2 bits into expandable pixel according to the local complexity. This avoids expanding pixels with large prediction-errors, and thus, it reduces embedding impact by decreasing the maximum modification to pixel values. Meanwhile, adaptive PEE allows very large payload in a single embedding pass, and it improves the capacity limit of conventional PEE. We also propose to select pixels of smooth area for data embedding and leave rough pixels unchanged. In this way, compared with conventional PEE, a more sharply distributed prediction-error histogram is obtained and a better visual quality of watermarked image is observed. With these improvements, our method outperforms conventional PEE. Its superiority over other state-of-the-art methods is also demonstrated experimentally.

In this paper, a recently proposed reversible image watermarking algorithm based on reversible contrast mapping (RCM) is further developed. The integer transform, RCM, which is originally defined on a pair of integers, is extended to integer array of arbitrary length. Based on this generalization, the embedding capacity can be significantly increased. Meanwhile, the embedding distortion is well controlled by a suitable selection of embeddable pixel blocks. In fact, the hidden data is priorly embedded into the blocks that introduce less distortion. Furthermore, the proposed scheme does not need additional data compression, and it has low computational complexity. The experiments verify that the novel scheme outperforms the original RCM-based method and some state-of-the-art reversible watermarking algorithms, especially the ones based on integer transform.

In this paper, a content-adaptive steganographic scheme is proposed. The novel scheme can be viewed as an improvement of the conventional LSB matching. In this scheme, we take advantage of embedding redundancy in LSB matching to select modification direction (i.e., increasing or decreasing the pixel value by 1), and the dependency of neighboring pixels is taken into consideration. More specifically, if the secret message bit does not match the LSB of the corresponding cover pixel value, the choice of modification direction is not random but a specific selection, in order to hold the correlation of neighboring pixels as far as possible. The resulting stego image looks more like a natural one, and smooth to some extent. Comparing with LSB matching and other state-of-the-art steganography, higher level security of the proposed scheme is experimentally verified. In addition, the proposed approach can be also applied in LSB-based steganography to enhance the security.

This paper proposes a novel steganalyzer for detecting one of the most popular steganography, LSB matching (also known as “±1 embedding”). The histogram of difference image (the differences of adjacent pixels), which is usually a generalized Gaussian distribution centered at 0, is exploited for deriving statistical features. We have proved theoretically that the peak-value of the histogram would decrease after LSB matching embedding, while the renormalized histogram (the ratio of the histogram to the peak-value) would increase. Then we take the peak-value and the renormalized histogram as features for classification. Extensive experimental results show that the proposed steganalytic method outperforms some previous ones.

Many reversible watermarking algorithms have been proposed in the literature, and all of these solutions put emphasis on high embedding capacity with low distortion. In this paper, a novel reversible image watermarking method based on integer transform is presented. Through taking pixel block of arbitrary size as embedding unit, the method can provide a very high embedding capacity. For instance, it can embed as large as 1.85 bits per pixel in “Lena”. Besides, by pre-estimating the embedding distortion, one can suitably select embeddable blocks so that the visual quality of the watermarked image is well guaranteed. Furthermore, extensive experiments show that the novel method performs better than some state-of-the-art algorithms.

Reversible watermarking is a special watermarking technique which allows one to extract both the hidden data and the exact original signal from the watermarked content. In this paper, a recently introduced reversible image watermarking method based on prediction-error expansion is further investigated and improved. Instead of taking the pixels with small prediction-error as embedding pixels (i.e., the pixels that carry watermark bits), we propose to select these pixels in a dynamical way. In fact, we can pre-calculate the embedding distortion for each possible choice of embedding pixels, and determine the one with minimal distortion. We see that, with this choice of embedding pixels, the distortion is reduced comparing with the original method, and thus, the proposed approach has a better performance. In addition, experimental results show that the novel method outperforms some state-of-the-art algorithms.

In this letter, an efficient integer transform based reversible watermarking is proposed. We first show that Tian's difference expansion (DE) technique can be reformulated as an integer transform. Then, a generalized integer transform and a payload-dependent location map are constructed to extend the DE technique to the pixel blocks of arbitrary length. Meanwhile, the distortion can be controlled by preferentially selecting embeddable blocks that introduce less distortion. Finally, the superiority of the proposed method is experimental verified by comparing with other existing schemes.

In this paper, we present an improved method for detecting LSB matching steganography in gray-scale image. Our improvements focus on three aspects: (1) instead of using the amplitude of local extrema of the image's histogram in the previous work, we turn to considering the sum of the amplitude of each point in the histogram; (2) incorporating the calibration (downsample) technique with the current method; (3) the sum/difference image (which is defined as the sum or difference of two adjacent pixels in the original image) is taken into consideration to provide additional statistical features. Extensive experimental results show that the novel steganalyzer out-performs the previous ones.

Matrix embedding is proved an effective way to improve embedding efficiency of steganography, and usually, higher dimensional matrix will provide better embedding efficiency. However, the sender might suffer huge computational complexity when employing high dimensional matrix. In this paper, a variation of matrix embedding is proposed. Instead of finding a coset leader as the modification of the cover image (which is the most time consuming step in matrix embedding), we turn to finding a vector in the coset which has relatively small Hamming weight. Such a vector can be found at reduced computational cost. Consequently, higher dimensional matrix can be used for practically implementable matrix embedding. The parity check matrix of random linear code is used in the experiments. It has shown that the proposed scheme can enhance the practicability and efficiency of the conventional matrix embedding.

In order to enhance the steganographic security, higher embedding efficiency (average number of secret data bits embedded per one embedding change) is desired. Matrix embedding is a well-known steganographic scheme that can improve the embedding efficiency. However, the embedding cost (running time) is expensive when employing high dimensional matrix, while higher dimensional matrix usually provides better embedding efficiency. In this paper, we construct a specific matrix for matrix embedding, in which data embedding procedure can be implemented with linear computational complexity. The novel scheme is computationally efficient, and meanwhile it leads to a relatively good embedding efficiency.

Embedding efficiency, which is defined as the average number of secret data bits embedded per one embedding change, is an important attribute directly influencing the security of steganography. Higher embedding efficiency is usually expected to get more secure steganographic method. In this paper, by incorporating two previously introduced techniques: the ldquosum and difference covering setrdquo (SDCS) and the ldquowet paper coderdquo (WPC), a novel double-layered embedding method is presented. As compared with the original SDCS-based steganography, the novel method can embed more bits without introducing additional embedding distortion, therefore the embedding efficiency is improved. In summary, the proposed method provides some steganographic schemes with diverse embedding rates and good embedding efficiency.

Recently, a significant improvement of the well-known least significant bit (LSB) matching steganography has been proposed, reducing the changes to the cover image for the same amount of embedded secret data. When the embedding rate is 1, this method decreases the expected number of modification per pixel (ENMPP) from 0.5 to 0.375. In this letter, we propose the so-called generalized LSB matching (G-LSB-M) scheme, which generalizes this method and LSB matching. The lower bound of ENMPP for G-LSB-M is investigated, and a construction of G-LSB-M is presented by using the sum and difference covering set of finite cyclic group. Compared with the previous works, we show that the suitable G-LSB-M can further reduce the ENMPP and lead to more secure steganographic schemes. Experimental results illustrate clearly the better resistance to steganalysis of G-LSB-M.

In this paper, based on a careful investigation on the calibration (downsample) technique, we improve two detectors for detecting LSB matching: calibrated HCF COM and calibrated adjacency HCF COM. Instead of using the COM (center of mass) of the HCF (histogram characteristic function), we consider the ratio of the histogram's DFT coefficients of the image to the corresponding coefficients of the down-sampled image. Moreover, we propose to down-sample only for non-oscillating pixels. With a same level of computational complexity, the new detectors thus obtained are better than the old ones, especially for uncompressed images.

As an important attribute directly influencing the steganographic scheme, the embedding efficiency is defined as the average number of random data bits per one embedding change. In this paper, we propose a novel approach to improve the embedding efficiency of LSB matching based on the sum and difference covering set (SDCS) of finite cyclic group. We show that the suitable choice of SDCS will lead to a new steganographic scheme which is more efficient than LSB matching. Then we illustrate that the new scheme keeps the statistical imperceptibility of LSB matching. The detailed constructions of SDCS are given and some related problems are also discussed.