Hao-Tian Wu

Reversible image watermarking on prediction errors by efficient histogram modification

A reversible data hiding algorithm is proposed, in which the efficiency of modifying a pair of histogram bins is considered. Multiple pairs of histogram bins can be further selected for data embedding in sequence, while pre-process of pixel values is performed to prevent the possible overflow and underflow. Embedding with the prediction errors is investigated with a new prediction scheme. In each of the four prediction modes, a large amount of prediction errors can be produced from the host image. Moreover, all combinations of the four modes to generate a number of histogram pairs are enumerated to obtain the best performance. Blind extraction and recovery are enabled by embedding a pre-computed location map and other overhead information into the watermarked image. Promising experimental results are obtained on a variety of test images. Compared with the existing algorithms, the image content is better preserved in high payload data hiding.

Reversible Data Hiding: Advances in the Past Two Decades

In the past two decades, reversible data hiding (RDH), also referred to as lossless or invertible data hiding, has gradually become a very active research area in the field of data hiding. This has been verified by more and more papers on increasingly wide-spread subjects in the field of RDH research that have been published these days. In this paper, the various RDH algorithms and researches have been classified into the following six categories: 1) RDH into image spatial domain; 2) RDH into image compressed domain (e.g., JPEG); 3) RDH suitable for image semi-fragile authentication; 4) RDH with image contrast enhancement; 5) RDH into encrypted images, which is expected to have wide application in the cloud computation; and 6) RDH into video and into audio. For each of these six categories, the history of technical developments, the current state of the arts, and the possible future researches are presented and discussed. It is expected that the RDH technology and its applications in the real word will continue to move ahead.

Reversible Image Data Hiding with Contrast Enhancement

In this letter, a novel reversible data hiding (RDH) algorithm is proposed for digital images. Instead of trying to keep the PSNR value high, the proposed algorithm enhances the contrast of a host image to improve its visual quality. The highest two bins in the histogram are selected for data embedding so that histogram equalization can be performed by repeating the process. The side information is embedded along with the message bits into the host image so that the original image is completely recoverable. The proposed algorithm was implemented on two sets of images to demonstrate its efficiency. To our best knowledge, it is the first algorithm that achieves image contrast enhancement by RDH. Furthermore, the evaluation results show that the visual quality can be preserved after a considerable amount of message bits have been embedded into the contrast-enhanced images, even better than three specific MATLAB functions used for image contrast enhancement.

Reversible Watermarking by Modulation and Security Enhancement

This paper presents a watermarking algorithm that is suitable for the data represented in floating- or fixed-point numbers. By keeping the modulation information, every value in the original object can be restored with the smallest error. For the sake of exact recovery, the redundancy in the watermarked object should be exploited to save the recovery information. We analyze the security of the algorithm and enhance it by adding a pseudorandom (PR) sequence to the watermarked object. The algorithm can be applied in the measurement to embed some useful information in a distortion-free way. We implement it on the Virtual Reality Modeling Language (VRML) models, and the experimental results show its efficacy.

A reversible data hiding method with contrast enhancement for medical images

Reversible data hiding with contrast enhancement is performed on medical images.Pre-processing is improved so that better visual quality can be obtained.Principal values in background are differentiated to improve image quality.Better contrast enhancement effects than the previous work are achieved. In this paper, a reversible data hiding method with contrast enhancement is presented for medical images. Firstly, image background segmentation is performed and the principal gray-scale values in the segmented background are identified. By excluding the corresponding histogram bins from being expanded for data hiding, the contrast of region of interest (ROI) in medical images can be selectively enhanced. Considering the characteristics of pixel distribution, we develop a new pre-processing strategy to reduce the visual distortions that may be caused. With the proposed method, an original image can be exactly recovered from the corresponding enhanced image by hiding the side information within it. The experimental results on a set of medical images show that the visibility of ROI can be improved. Compared with the previous method, the proposed method can achieve more contrast enhancement effects and better visual quality for medical images.

Reversible data hiding in Paillier cryptosystem

Reversible image data hiding in the Paillier cryptosystem has been introduced.Two new reversible data hiding algorithms are proposed for the encrypted images.The potential applications of the proposed algorithms have been discussed.High-capacity reversible data hiding can be achieved in the Paillier cryptosystem. In this paper, reversible image data hiding in the Paillier cryptosystem is investigated. To transmit additional data in homomorphic encryption domain, two algorithms are proposed for different application scenarios. By exploiting the additive homomorphism, high-capacity data hiding can be accomplished with the first algorithm by conducting value expansion on the encrypted pixel values. But the hidden data can only be extracted after image decryption (i.e., in plain-text domain). With the second algorithm, both of data embedding and extraction can be performed in the encryption domain by exploiting the self-blinding property, while the corresponding plain-text values are unchanged. Compared with the reversible data hiding algorithms designed for encryption with a cipher stream, the proposed ones are more applicable in cloud computing without degrading the security level. Thus the additional data can be losslessly transmitted in the different applications of Paillier cryptosystem.

A novel reversible data hiding method with image contrast enhancement

Abstract Recently, several image contrast enhancement methods have been proposed such that the original image can be recovered from its contrast-enhanced version. Hence a flexibility in changing image contrast can be provided when needed. However, the artificial distortions may be introduced into the image content after adopting these methods. Meanwhile, there is lack of using the adequate image quality metrics for performance evaluation. In this paper, a novel reversible data hiding method is proposed for image contrast enhancement. To better preserve image quality, it is restricted that only the adjacent bins in the original image histogram may be merged in the pre-processing. The proposed method has been applied to two image sets and compared with the previous methods. For image quality assessment, the PSNR, SSIM and three no-reference metrics have been adopted in performance evaluation. The experimental results have clearly shown that better visual quality can be achieved with the proposed method. Besides recovering the original images, extra data can be hidden into the contrast-enhanced images and correctly extracted.

Improved steganalysis algorithm against motion vector based video steganography

This paper proposes an improved steganalysis algorithm to detect the secret message hidden in the compressed video. As majority of video steganographic algorithms modify motion vectors (MV) in inter-frame encoding to hide data, aliasing effect may be caused in the distribution of the difference between MVs in two adjacent macroblocks. This phenomenon has been observed in detecting the secret data that were added to the MVs in cover video. To exploit the correlations between the neighboring MVs so as to detect the hidden data more efficiently, we consider the joint distribution of the MV differences between one macroblock and the other two macroblocks neighboring to it. The calculated joint probability mass functions are used to distinguish the stego videos from the non-stego ones. The experimental results show that significant improvement in detection accuracy can be made by using the joint distribution of MV differences instead of the statistics calculated from two neighboring MVs as features.

Reversible image watermarking by rhombus prediction and histogram modification

A histogram based reversible data hiding algorithm is proposed for digital images with the rhombus prediction scheme. With the prediction errors generated in each of the two phases, data embedding is performed by modifying their histogram while the overflow and underflow of pixel values are prevented in pre-process. Blind data extraction and exact recovery of the original image are both enabled by saving the overhead information into the watermarked image. Furthermore, high-capacity embedding can be performed by using multiple pairs of histogram bins to host the embedded data. The numerical results show that good performance can be achieved with the proposed algorithm, especially when the embedding rate is high.

Secure watermarking on 3d geometry via ICA and orthogonal transformation

The technique of independent component analysis (ICA) has been utilized in digital image and audio watermarking. In this paper, it is applied to improve the security of 3D geometry watermarking by using the ICA de-mixing matrix for transformation. An orthogonal transformation (OT) matrix is further employed so that another watermarking scheme is generated. With the same embedding method, the two watermarking schemes are implemented on the test models. The experimental results show that they have similar performance in imperceptibility and capacity. As both of them have the properties of blind extraction and security, the OT-based scheme can be regarded as a simplified version of the ICA-based one with less computational cost.