Wyniki wyszukiwania - Biblioteka Nauki

1

An Image Authentication Based on Discrete Fourier Transform

100%

Kieu T. D. , Chang C.-C.

|

tom Vol. 97, nr 4

369-379

EN

The advances of network technologies and digital devices facilitate users to exchange multimedia data over the public networks. However, this also raises significant concerns about how to protect sensitive multimedia data from being illegally copied and unauthorized modifications. Thus, this paper proposes a fragile watermarking method to detect illegitimate alterations of the watermarked data. The proposed method embeds a grayscale watermark image into a grayscale cover image in a block-by-block manner by using discrete Fourier transform. Experimental results show that the proposedmethod can successfully and exactly detect and localize any tampered regions of the watermarked image.

2

Full-Searching-Equivalent Vector Quantization Method Using Two-Bounds Triangle Inequality

100%

Chang C.-C. , Chen C.-C.

|

tom Vol. 76, nr 1-2

25-37

EN

The encoding process of vector quantization (VQ) is indeed computational complex and time consuming. Compared with actual Euclidean distance computation, some inequalities can generate estimations with less computation to filter out the impossible codevectors as well as to reduce the computation time. In this paper, we introduce a new estimation for the Euclidean distance using two-bounds triangle inequality. The experimental results show that our proposed scheme can reduce Euclidean distance computation by 71% to 94% for full search. Having been proved, our proposed scheme can reduce the computing time by 42% to 51%.

3

An Image Retrieval Method Based on Color-Complexity and Spatial-Histogram Features

100%

Wu H.-C. , Chang C.-C.

|

tom Vol. 76, nr 4

481-493

EN

This paper proposes two kinds of image features. One feature is the spatial-histogram feature. It combines the color histogram feature and the information about the dimensional position of pixels in an image to record the distribution of the pixels' colors that are present in different spatial positions within an image. The other image feature proposed in this paper is the color-complexity feature, which can be used to describe the change of pixel colors in the image. From the experimental results, ANMRR value is provided and we observe that the image retrieval system based on these two kinds of image features can provide a fairly good accuracy rate in image retrieval. Moreover, it has the capacity to tolerate errors; that is, for images that are damaged by rotation, shift, or color variant attacks, their similar image pairs can still be retrieved from the image database. Thus, the accuracy and flexibility of the image retrieval system are drastically improved.

4

Tolerant Key Assignment for Enforcing Complicated Access Control Policies in a Hierarchy

100%

Chang Y.-F. , Chang C.-C.

|

tom Vol. 76, nr 1-2

13-23

EN

In 1998, Yeh et al. proposed a flexible key assignment scheme for enforcing complicated access control policies in a user matrix model. Later, Hwang indicated that Yeh et al.' scheme is susceptible to some security flaws. Thus, Lin et al. proposed a key assignment scheme for enforcing complicated access control policies in a hierarchy. However, there exist drawbacks in Lin et al.'s scheme: lack of efficiency and a large variation of the keys. Hence, we propose an efficient key assignment scheme in a hierarchy for enforcing the complicated access control policies. What is more, the secret key of one class is allowed to be changed several times without influencing the derivation key in our proposed scheme.

5

Reversible Data Embedding Based on Prediction Approach for VQ and SMVQ Compressed Images

100%

Chang C.-C. , Lin C.-Y.

|

tom Vol. 74, nr 2,3

189-207

EN

Reversible steganography allows an original image that has gone through the embedding process to be completely restored after the extraction of the embedded data. In this paper, we propose a reversible scheme with a high embedding capacity for VQ compressed images. Our reversible method is based on a prediction strategy and takes advantage of the local characteristics of the image. Since the location map is usually a necessary part of a reversible scheme, two methods, shifting and relocating, are also proposed to reduce the size of the location map. As the experimental results show later, our method outperforms previous schemes in terms of embedding capacity and image quality. To be more specific, with low distortion, the embedding capacity of the proposed methods can be higher than one bit per index value.

6

VQ Codebook Searching Algorithm Based on Correlation Property

80%

Chang C.-C. , Lu T.-C. , Ya J.-B.

|

tom Vol. 76, nr 1-2

39-57

EN

In this paper, we shall present two efficient codebook-matching schemes with a locally adaptive vector quantizer (LAVQ) and a search-ordering coding vector quantizer (SOCVQ) to be applied in the Vector Quantization (VQ) encoding system. The proposed codebook-matching schemes exploit the correlation property between adjacent image blocks so that the process of codebook matching can be speed up. Simulation results show that the time complexity of the proposed schemes is lower than those of PCA, Multipath TSVQ, and DPTSVQ. Moreover, the image quality of the proposed methods is close to those processed by the other methods.

7

Secret Image Hiding and Sharing Based on the (t,n)-Threshold

80%

Chang C.-C. , Lin C.-Y. , Tseng C.-S.

|

tom Vol. 76, nr 4

399-411

EN

This study proposes a new secret image sharing method based on the ( t, n)-threshold. The sharing method is completed through three main steps: (1) quantizing the secret image, (2) generating n shadows from the quantized image, and (3) sharing the n shadows among n cover images. Our goal is both to attain good stego-image and secret image quality preservation effects and to limit the size of each shadow image down to 1/ t of that of the secret image for convenient transmission and storage. In the first step, the secret image is divided into non-overlapping 4 × 4 blocks that are quantized according to the results of block testing. Each block, after the block testing procedure, is judged to be either a smooth block or a non-smooth one. Smooth blocks can be neatly restored after being hidden into the cover image, while non-smooth blocks can be somewhat lossy, but such modification is perceptually invisible. The second step is to apply Shamir's ( t,n)-threshold, which uses a ( t-1)-degree polynomial to generate n shadows so that the secret image can be restored from any t out of n shadows. The last step involves a modulo operation to hide the n shadows in n cover images. According to our experimental results, the PSNR values of the reconstructed secret images by our method center around 39 dB, outperforming other recent methods.

8

High Capacity Data Hiding for Grayscale Images

80%

Chang C.-C. , Chou Y.-C. , Kieu T. D.

|

tom Vol. 86, nr 4

435-446

EN

High embedding capacity and low image distortion factors are main concerns of information hiding researchers. No scheme can achieve these two factors simultaneously. The purpose of achieving either high embedding capacity or low image distortion varies from application to application. This paper presents a high embedding capacity information hiding method for grayscale images. The proposed method converts binary secret data into secret digits in the base-5 numeral system first. Next, one secret digit in the base-5 numeral system is embedded into one cover pixel. The experimental results show that the proposed method can be used to embed two secret bits into one cover pixel with the imperceptible image distortion. Comparing with some other works, we found that our method indeed outperforms all others in terms of embedding capacity.