Cheng-Hsing Yang

Adaptive Data Hiding in Edge Areas of Images With Spatial LSB Domain Systems

This paper proposes a new adaptive least-significant- bit (LSB) steganographic method using pixel-value differencing (PVD) that provides a larger embedding capacity and imperceptible stegoimages. The method exploits the difference value of two consecutive pixels to estimate how many secret bits will be embedded into the two pixels. Pixels located in the edge areas are embedded by a k-bit LSB substitution method with a larger value of k than that of the pixels located in smooth areas. The range of difference values is adaptively divided into lower level, middle level, and higher level. For any pair of consecutive pixels, both pixels are embedded by the k-bit LSB substitution method. However, the value k is adaptive and is decided by the level which the difference value belongs to. In order to remain at the same level where the difference value of two consecutive pixels belongs, before and after embedding, a delicate readjusting phase is used. When compared to the past study of Wu et al.s PVD and LSB replacement method, our experimental results show that our proposed approach provides both larger embedding capacity and higher image quality.

Reversible data hiding of a VQ index table based on referred counts

This paper presents a new reversible VQ-based hiding scheme that can recover the original VQ compressed codes after data extraction. Our scheme sorts a VQ codebook using the referred counts. The VQ codebook is then divided into 2^B clusters and half of these clusters are used to embed secret data, in which B denotes the size of the secret data embedded into each VQ index. Compared to Chang et al.s scheme, which divides a sorted VQ codebook into 2^B^-^1x3 clusters and uses the front one-third clusters to embed secret data, our method can embed more data. Moreover, indicator, index exchanging, and side-match prediction schemes are proposed to further improve our scheme. Under the same sorted VQ codebook, the experimental results demonstrate that our data hiding algorithm has higher capacities and better compression rates.

Fractal curves to improve the reversible data embedding for VQ-indexes based on locally adaptive coding

The VQ reversible data embedding technology allows an original VQ coding to be completely restored after the extraction of embedded data. In this paper, we propose a new reversible scheme based on locally adaptive coding for VQ-compressed images. The fractal Hilbert curve is applied to replace the traditional trace of processing the VQ index table. The VQ index table is pre-processed to create a fractal Hilbert curve. Following the curve to process the VQ index table can get better compression rates in the data embedding procedure. Besides, compared to Chang et al.s scheme, which compressed the inputted VQ index value only when the to-be-embedded bit b is 0, our method performs compressing operations in both cases that the to-be-embedded bits b are 0 and 1. The experimental results show that the proposed method has the best compression rate and the highest embedding capacity compared with other reversible VQ embedding methods.

Varied PVD+LSB evading detection programs to spatial domain in data embedding systems

Image steganographic schemes based on the least-significant-bit (LSB) replacement method own the character of high capacity and good quality, but they are detected easily by some programs. Pixel-value differencing (PVD) approaches are one kind of methods to pass some program detections, but PVD approaches usually provide lower capacities and larger distortion. Accordingly, the combined method of PVD and LSB replacement was proposed to raise the capacity and the quality of PVD approaches over the past literatures. In this paper, not only we contribute a new exploration in spatial domain to benefiting both the capacity and the fidelity on the basis of varied LSB+PVD approaches, but also the risk of the RS-steganalysis detection program is evaded. Furthermore, proof works are conducted to proclaim the correctness of the general LSB+PVD method. Following up, the varied LSB+PVD approach is therefore applied to our scheme.

Anti-Forensics with Steganographic Data Embedding in Digital Images

E-forensics investigates and extracts confidential information from electronic products; in other words, the anti-forensics indicates that evidences in those products are imperceptible and undetected. This work presents an anti-forensic steganography method that can embed and extract messages from images. Highlight of exploiting modification direction (HoEMD) and adaptive EMD (AdEMD) methods use the module operation and take into account of the sensitive nature of a human visual system. The HoEMD approach exploits the pixel directions. A pixel with a larger change implies more pixel directions and, ultimately, a larger embedding capacity. The pixel differencing in the proposed AdEMD method is used to evaluate whether the pixel located in the edge area can tolerate a larger change than that of the pixel location in a smooth area. To successfully extract a message, the differencing value is maintained on the same level before and after data is concealed; a delicate adjusting phase is used as well. In contrast with the PVD and LSB replacement method of Wu et al., LSB substitution method, and Lee and Chens data hiding scheme based on modulus function, the proposed steganography system has a larger embedding capacity and a higher image quality. Effectiveness of the proposed steganography schemes over that of a previous blind steganalyzer is demonstrated using the statistical attack of Chi-square analysis.

Analyses of Pixel-Value-Differencing Schemes with LSB Replacement in Stegonagraphy

This paper shows that a steganographic scheme based on Wu et al. s pixel-value differencing (PVD) and least-significant-bit (LSB) replacement method is too conformable to the LSB approach. Their PVD+LSB method can embed secret data larger than the PVD method, however, it is detectable by Fridrich et al. s method but PVD method is not. Some of the shortcomings of the PVD+LSB method are shown in this paper. At the same time, some techniques are proposed for improving the image quality of the PVD+LSB method without sacrificing the capacity. All proposed techniques are demonstrated by experimental results.

Data embedding in image-media using weight-function on modulo operations

Multimedia hiding system is to embed message behind the specified media, but it is still kept normal in media representations via human sensitive organizations without causing imperceptibility. In this article, we propose a data hiding system by means of flexible exploiting modification directions to achieve safer message concealments in image-media. In our scheme, n cover-pixels are flexibly chosen on modulo operations to embed a secret s, where n = ⌈ log3(s) ⌉. The varied pixel values associated with the chosen n pixels are only changed among [-1, 1]. Because the numbers of adjustable pixels are much greater than the pixels in the past scheme, our scheme is able to obtain a higher embedded ratio in response to the capacity requirements of information hiding systems. In addition, we also applied the statistics-steganalyzers to demonstrate that our scheme has accomplishment not only higher capacity but also kept the robustness against the blind steganalyzers.

Capacity-Raising Steganography Using Multi-Pixel Differencing and Pixel-Value Shifting Operations

In order to provide large embedding capacity and to minimize distortion for the stego-image, a steganographic method using multi-pixel differencing is presented in this paper. It takes into consideration four pixels of a block, and the differences between the lowest gray-value pixel and its surrounding pixels are used to determine the degree of smoothness and sharpness for embedding the secret data. If the difference values are large in a block, and the block is located in the sharp area then more data can be embedded. On the other hand, if the block is located in the smooth area less data can be embedded. The multi-pixel differencing method is employed in our scheme. We also propose the pixel-value shifting method to increase the image quality. The experimental results show that our scheme has a large embedding capacity without creating a noticeable distortion.

Grouping strategies for promoting image quality of watermarking on the basis of vector quantization

In this paper, a novel vector quantization (VQ)-based digital image watermarking scheme is proposed which retains high image quality even after embedding. The general scheme of watermarking is to embed a representative digital watermark in the protected image to claim the ownership. The robustness of watermark is important; however, the imperceptibility of embedded images also cannot be ignored. In previous papers, existing greedy technique to group codewords often results in low quality of stego-images. As an improvement to the existing techniques, two new grouping strategies, codebook-sort grouping and bipartite-match grouping are proposed in this paper to improve the quality of stego-images. Moreover, a new VQ encoder, called as shortest-group encoding method is proposed to replace the traditional VQ encoder in the purpose of embedding. Finally, in order to improve the grouping flexibility, two kinds of weights are considered in our approach. Experimental results show that all of the above approaches can improve the quality of stego-images on VQ-based watermarking.

Histogram-Based Reversible Information Hiding Improved by Prediction with the Variance to Enhance Image Quality

Reversible data hiding based on prediction-based is a good technique that can hide message bits into digital images with low distortion. In this paper, we propose a reversible data hiding method based on interleaving prediction and local complexity for enhancing stego-image quality. The thresholds and local complexity are used to determine which predicted error should join the group of pixel shifting or message concealing. If the local complexity is smaller than thresholds, the predicted error will be taken for message hiding or pixel shifting, otherwise, if the local complexity is larger than thresholds, the predicted error will quit joining the process of data concealing and pixel shifting. Therefore, more pixels will avoid executing the process of pixel shifting, resulting to images with lower distortion. The experimental results show that our image quality is superior to other approaches at the same payload.