Pei-Yu Lin

Adaptive watermark mechanism for rightful ownership protection

Watermarking is used to protect the integrity and copyright of images. Conventional copyright protection mechanisms; however, are not robust enough or require complex computations to embed the watermark into the host image. In this article, we propose an adaptive copyright protection scheme without the use of discrete cosine transformation (DCT) and discrete wavelet transformation (DWT). This novel approach allows image owners to adjust the strength of watermarks through a threshold, so that the robustness of the watermark can be enhanced. Moreover, our scheme can resist various signal processing operations (such as blurring, JPEG compression, and noising) and geometric transformations (such as cropping, rotation, and scaling). The experimental results show that our scheme outperforms related works in most cases. Specifically, our scheme preserves the data lossless requirement, so it is suitable for medical and artistic images.

Distortion-free secret image sharing mechanism using modulus operator

The (t, n)-threshold has been extended to secret image sharing due to its practicability. In this article, we provide a novel version that employs the modulus operator to embed the secret share into a host image. The simulator shows that the modulus operator is useful for decreasing shadow image distortion. Using Rabins signature cryptosystem, participants can detect if a cheater exists in the cooperation. In particular, the new mechanism permits involved members to restore a lossless secret image and to reconstruct a distortion-free host image.

Sharing a Secret Two-Tone Image in Two Gray-Level Images

In this paper, we shall propose two spatial-domain image hiding schemes with the concept of secret sharing. The two new schemes use the two-out-of-two visual secret sharing technique to generate two shares for hiding a secret two-tone image. These two secret shares are embedded into two gray-level cover images by the proposed embedding scheme. To decode the hidden messages, we can superimpose the extracted shares from the secret-share-carrier images (namely the embedding images). The advantages of our scheme are simple computation and good security, and thus it is very suitable for applications involving low power verification systems. Besides, our scheme can support two participants to share one secret two-tone image. According to our experimental results, the two proposed schemes are capable of offering satisfactory embedding image quality

Protecting the content integrity of digital imagery with fidelity preservation

Fragile watermarking is applied to protect the content integrity of digital images. The main concerns related to watermarking include retaining the quality of the watermarked image and retaining the ability to detect whether any manipulation has occurred. Because recent watermarking techniques seriously distort the quality of the protected image after embedding the authentication code into the image content, attention has been drawn to how to satisfy both the need for image fidelity and detection ability. To account for the influence from both essentials, a novel algorithm is proposed in this article. The new scheme utilizes a weighted-sum function to embed (n + 1) authentication bits into a block with 2n pixels by modifying only one original pixel with (±1). With fewer authentication codes, the new process can protect the content of the image. The experimental results demonstrate that the approach can guarantee the fidelity of the watermarked image while retaining tamper-proof functionality.

Dual Digital Watermarking for Internet Media Based on Hybrid Strategies

This paper presents a novel dual watermarking mechanism for digital media that embeds a recognizable pattern into the spatial domain and an invisible logo into the frequency domain. Undoubtedly, visible watermarking is important for protecting online resources from unauthorized reproduction. Due to the visibility of embedded patterns, however, watermarked digital media are vulnerable to the inpainting attack and common signal processing operations. Utilizing hybrid strategies, simulation results show that the novel method can resist these attacks. In particular, the new mechanism allows legal subscribers to restore an unmarked image, whereas other dual watermarking schemes do not. This feature makes it suitable for protecting artistic and valuable media.

Preserving robustness and removability for digital watermarks using subsampling and difference correlation

Watermarking techniques are applied to digital media to protect their integrity and copyright. The embedding of a watermark, however, often distorts the quality of the protected image. This may be intolerable since the protected media is for preserving artistic and valuable images. Hence, engineers have proposed removable solutions permitting authorized users to restore watermarked images to unmarked images with satisfactory quality. Unfortunately, these mechanisms cannot resist signal processing attacks to protect the ownership. In this article, we propose a novel watermarking mechanism by utilizing pair-difference correlations upon subsampling and the technique of JND. This new approach can guarantee the robust essentials of watermarking schemes. Experimental results reveal that the new method outperforms others in terms of restored image quality. More specifically, this novel approach can resist various attacks to which related works are vulnerable.

A Sudoku-based Secret Image Sharing Scheme with Reversibility (Invited Paper)

A common drawback of the image sharing with steganography approaches is that the revealed secret image is distorted due to the truncation of the grayscale secret image. To lossless reveal the secret image in the ( t , n )- threshold, we provide a novel sharing scheme in this article. Moreover, the original host image can be recovered by the embedded shadow images. To accomplish the above purposes, the proposed scheme derives the secret shadows and generates the meaningful shadow images by adopting the sudoku. In the new scheme, the sudoku grid is setting to 16  16 and divided into sixteen 4  4 blocks. Thus, we can embed 4  ( t -1) secret bits into each pixel pair of the host image. Besides, the embeddable secret capacity can be improved according to the threshold t in the ( t , n )-threshold sharing system. The experiments show that the shadows can be successfully camouflaged in the host image with satisfactory quality. The distortion of the embedded host pixels is limited within range [0, 3]. Moreover, the proposed scheme provides a large capacity for embedded secret data.

Prevention of cheating in visual cryptography by using coherent patterns

A new VC scheme with cheating-prevention ability is proposed.It reveals solid pattern with moderate-scale on partial-superimposed image.The proposed method verifies shares without any additional authentication shares.The proposed method endows conventional VC with cheating-prevention ability. This paper proposes a simple method for establishing a visual cryptographic (VC) scheme with the ability to prevent cheating. Given the n base-shares generated in a conventional (t,n), 2≤t≤n, VC scheme, an authentication pattern stamping process was designed to encode the n base-shares to get n verifiable shares that exhibit the following properties: (1) knowledge of less than t verifiable-shares gets no secret information, (2) the secret can be revealed by completely superimposing t or more verifiable-shares, and (3) the authentication pattern can be revealed by partially superimposing any pair of verifiable-shares. Theoretical proof and computer simulation for the proposed method are provided. The proposed method has smaller pixel expansion than previous cheating prevention VC schemes, and the verification process is fulfilled without resorting to any additional dedicated verification share. It can be attached easily to any reported VC scheme to endow legitimate users with the capability of detecting faked shares provided by malicious participants.

Contrast-Adaptive Removable Visible Watermarking (CARVW) mechanism

Abstract Engineers have proposed many watermark mechanisms for protecting the content of digital media from unauthorized use. The visible watermark scheme indicates the copyright of digital media posted over the Internet by embedding an inconspicuous but recognizable pattern into media. However, the embedding process often results in serious distortion of the protected image. Since the strength of the watermark in conventional methods mainly depends on the feature of protected media, this may lead to unsatisfactory transparency of watermarked images. This paper proposes a removable solution for visible watermark mechanism. By adopting the subsampling technique, the method proposes a contrast-adaptive strategy to solve this problem. This method can also guarantee the essentials of general visible watermark schemes. Experimental results show that the proposed method outperforms related works in terms of preserving the quality of the restored image.

A Color Image Authentication Method Using Partitioned Palette and Morphological Operations

Image authentication is applied to protect the integrity of the digital image. Conventional image authentication mechanisms, however, are unfit for the palette-based color images. Palette-based color images such as GIF images are commonly used for media communications. This article proposes a palette-based color image authentication mechanism. This novel scheme can guarantee the essentials of general authentication schemes to protect palette-based color images. Morphological operations are adopted to draw out the tampered area precisely. According to the experimental results, the images embedded with the authentication data still can preserve high image quality; specifically, the new scheme is highly sensitive to altered areas.