Xuefeng Tong

Optimum Histogram Pair Based Image Lossless Data Embedding

This paper presents an optimum histogram pair based image reversible data hiding scheme using integer wavelet transform and adaptive histogram modification. This new scheme is characterized by (1) the selection of best threshold T , which leads to the highest PSNR of marked image for a given payload; (2) the adaptive histogram modification, which aims at avoiding underflow and/or overflow, is carried out only when it is necessary, and treats the left side and right side of histogram individually, seeking a minimum amount of histogram modification; and (3) the selection of most suitable embedding region, which attempts to further improve the PSNR of marked image in particular when the payload is low. Consequently, to our best knowledge, it can achieve the highest visual quality of marked image for a given payload as compared with the prior arts of image reversible data hiding. The experimental results have been presented to confirm the claimed superior performance.

Reversible data hiding for JPEG images based on histogram pairs

This paper proposes a lossless data hiding technique for JPEG images based on histogram pairs. It embeds data into the JPEG quantized 8x8 block DCT coefficients and can achieve good performance in terms of PSNR versus payload through manipulating histogram pairs with optimum threshold and optimum region of the JPEG DCT coefficients. It can obtain higher payload than the prior arts. In addition, the increase of JPEG file size after data embedding remains unnoticeable. These have been verified by our extensive experiments.

Optimal histogram-pair and prediction-error based image reversible data hiding

This proposed scheme reversibly embeds data into image prediction-errors by using histogram-pair method with the following four thresholds for optimal performance: embedding threshold, fluctuation threshold, left- and right-histogram shrinking thresholds. The embedding threshold is used to select only those prediction-errors, whose magnitude does not exceed this threshold, for possible reversible data hiding. The fluctuation threshold is used to select only those prediction-errors, whose associated neighbor fluctuation does not exceed this threshold, for possible reversible data hiding. The left- and right-histogram shrinking thresholds are used to possibly shrink histogram from the left and right, respectively, by a certain amount for reversible data hiding. Only when all of four thresholds are satisfied the reversible data hiding is carried out. Different from our previous work, the image gray level histogram shrinking towards the center is not only for avoiding underflow and/or overflow but also for optimum performance. The required bookkeeping data are embedded together with pure payload for original image recovery. The experimental results on four popularly utilized test images (Lena, Barbara, Baboon, Airplane) and one of the JPEG2000 test image (Woman, whose histogram does not have zero points in the whole range of gray levels, and has peaks at its both ends) have demonstrated that the proposed scheme outperforms recently published reversible image data hiding schemes in terms of the highest PSNR of marked image verses original image at given pure payloads.

Reversible binary image data hiding by run-length histogram modification

A novel reversible binary image data hiding scheme using run-length (RL) histogram modification is presented in this paper. The binary image is scanned from left to right and from top to bottom to form a sequence of alternative black RL and white RL. Combining one black RL and its immediate next white RL, we form one RL couple, thus generating a sequence of RL couples. The length of each couple is fixed during data embedding in order not to fail the reversibility. Two procedures are adopted to achieve reversibility: (1) only involve those RL couples in data embedding in which the length of couple is not shorter than threshold T1; (2) increase white RL of isolated white pixels from one to two. Another parameter T indicates where to embed data in black RL histogram. Adjusting T1 and T may result in optimum performance of pure embedding rate versus visual quality of marked image. The proposed scheme works for text, graphics, and their mixture, both halftone and nonhalftone binary images. Experimental works have shown its superior performs over the prior-arts.

Double-threshold reversible data hiding

This proposed scheme reversibly embeds data into image prediction-errors by using histogram-pair method with double thresholds (embedding threshold and fluctuation threshold). The embedding threshold is used to select only those prediction-errors, whose magnitude does not exceed this threshold, for possible reversible data hiding. The fluctuation threshold is used to select only those prediction-errors, whose associated neighbor fluctuation does not exceed this threshold, for possible reversible data hiding. Only when both thresholds are satisfied the reversible data hiding is carried out. Image gray level histogram modification is conducted to shrink the image histogram towards the center to avoid underflow and/or overflow only when this is necessary. The required bookkeeping data are embedded together with pure payload for original image recovery late. The experimental results have demonstrated that the proposed scheme outperforms recently published reversible image data hiding schemes in terms of the highest PSNR of marked image vs. original image at given pure payloads.

Stereo Image Coding with Histogram-Pair Based Reversible Data Hiding

This paper presents a stereo image coding method using reversible data hiding technique so that the right frame can be recovered losslessly and the left frame can be reconstructed with high visual quality. Utilizing the similarity between two frames in a stereo image pair the required size of storage and transmission bandwidth for the stereo image pair can be reduced to 50 %. A residual error matrix with a dynamic range of [−255, 255] is obtained by applying a frame-wise disparity algorithm which first shifts the left frame horizontally by a certain amount and then computes its difference to the right frame. Next, thus the generated residual error image with gray levels [0, 255] is obtained losslessly by a proposed labeling scheme. JPEG2000 lossy compression is then applied to the residual error image. The histogram-pair based reversible data hiding scheme is then utilized to embed the JPEG2000 lossy compressed data into the right frame. Compared with the prior art, which uses a block-based disparity estimation algorithm and a location map based reversible data hiding, the proposed method has demonstrated that the stereo image can be reconstructed with higher visual quality and with faster processing speed. Specifically, the experiments have demonstrated that both the PSNR and visual quality of the reconstructed stereo image pair are higher than those achieved by the prior arts.

JPEG Steganalysis Based on Classwise Non-Principal Components Analysis and Multi-Directional Markov Model

This paper presents a new steganalysis scheme to attack JPEG steganography. The 360 dimensional feature vectors sensitive to data embedding process are derived from multidirectional Markov models in the JPEG coefficients domain. The class-wise non-principal components analysis (CNPCA) is proposed to classify steganograpghy in the high-dimensional feature vector space. The experimental results have demonstrated that the proposed scheme outperforms the existing steganalysis techniques in attacking modern JPEG steganographic schemes-F5, Outguess, MB1 and MB2.

Optimal Histogram-Pair and Prediction-Error Based Reversible Data Hiding for Medical Images

In recent years, with the development of application research on medical images and medical documents, it is urgent to embed data, such as patient’s personal information, diagnostic information and verification information into medical images. Reversible data hiding for medical images is the technique of embedding medical data into medical images. However, most existed schemes of reversible data hiding for medical images could not achieve high performance and high payloads. This paper presents a reversible data hiding scheme for medical images based on histogram-pair and prediction-error. As the prediction-error histogram of medical images, compared with the gray level histogram of medical images, is more in line with quasi-Laplace distribution, histogram-pair and prediction-error based method could achieve high performance. We adjust the following four thresholds for optimal performance: embedding threshold, fluctuation threshold, left- and right-histogram shrinking thresholds. The left- and right-histogram shrinking thresholds are used not only to avoid underflow and/or overflow but also to achieve optimum performance. Compared to previous works, the proposed scheme has significant improvement in embedding capacity and marked image quality for medical images.

Using RZL Coding to Enhance Histogram-Pair Based Image Reversible Data Hiding

An improvement of histogram-pair based image reversible data hiding by using RZL (Reverse Zero-run Length) coding is proposed in this paper. The pre-processing to compress data to a shortest one is usually adopted for raising the PSNR (Peak Signal to Noise Ratio) in data hiding. Recently, the disagreements appear that we can get better PSNR by using RZL coding after compression. We proved that our histogram-pair based image reversible data hiding is suitable to use RZL to improve the performance. The PSNR can be raised by using different RZL methods, different parameters, different embedded capacity and different images. It is hard to apply RZL to given original data with different lengths. We proposed a method to solve that by adding some 0 s to the original data to form a complete block, and the RZL needs an attached mark for lossless recovery. In our experiments it has been shown that the PSNR of image with histogram-pair based reversible data hiding by using RZL is higher than that without using RZL as the embedding data rate is not high. Zhang et al.’s RZL is better than Wong et al.’s in most cases. The average PSNR gain is about 1 dB for five test images at different payloads with the RZL used in this paper.