高清晰图象水印算法

提出了一个针对图象水印的DCT域的水印算法,该算法基于利用人眼视觉掩蔽特性(HVS)的块分类方法,对原有的此类方法进行了改进,更大挖掘了载体图象的信息隐藏能力。在图象水印的嵌入方法上,采取了将水印图象进行DCT变换从而可以利用各块比例系数的差异嵌入不同频率的DCT系数。同时,对各块的DCT系数进行调整,使得各频率的能量基本相当,间接地提高了信噪比。经过以上几个方面的改进,得到的水印具有很好的清晰度。     

基于PCA和数据融合的遥感图像数字水印算法

提出了一种向遥感图像中嵌入水印以保护其版权的算法。算法将数据融合技术和数字水印技术相结合,首先将全色图像进行小波分解,提取图像分解后的第三级低频边缘特征,利用PCA变换得到边缘特征的第一主分量作为水印信息,将水印与第三级中频进行融合;然后进行小波逆变换得到重构图像;最后采用小波变换和PCA融合法将含有水印的全色图像和多光谱图像相融合。提取水印时使用独立分量分析(ICA)方法。实验表明,该算法可以保护遥感图像的版权和进行真伪认证,且不破坏原始遥感图像的信息和特征,是有效可行的。     

Robust RDWT-ICA based information hiding

We present a new method for watermarking which is based on redundant wavelet transform (RDWT) and independent component analysis (ICA) for authenticity protection of multimedia products. RDWT gives an over complete presentation of the input sequences and functions as a better approximation to the wavelet transform, and therefore, is well-suitable for watermarking. An advantage of using a large logo is the ability to carry redundant information about copyright, increasing the robustness. By exploiting RDWT, we can embed a large binary logo watermark, which has the same size as the host image. The watermark is detected by a new intelligent ICA-based detector, with which we can directly extract the watermark in spatial domain without transformation steps. Our findings are that RDWT-based watermarking out-performs against image compression, and produces less image distortion than conventional DWT. The robustness of the proposed method is demonstrated and tested under various attacks     

基于ICA的小波域图像数字水印方法

利用独立分量分析(ICA)和小波变换,提出一种新的图像数字水印方法。该方法充分利用置乱技术、纠错码技术以及HVS的特性,使得水印的鲁棒性和透明性都得到了保证。使用ICA技术可以更加有效的提取水印。经大量实验证明,本文所提出的方法对常用的图像处理有较强的抗攻击性。     

一种基于混沌和独立成分分析的自适应多重数字水印算法

采用混沌和独立成分分析技术,提出了一种在小波域自适应嵌入多重数字水印的算法。首先根据一个混沌标签序列从原图像中抽取一子图像,并对其进行小波分解,然后将多个互相独立的混沌水印信号依次嵌入到子图像的中频小波系数中,并自适应确定嵌入强度。根据Newman-Pearson准则检测水印是否存在,若存在,采用独立成分分析的方法提取出水印并进一步判断是否为原始水印。水印的检测与提取过程都不需要原始图像。仿真结果表明该算法对各种常见的攻击有着良好的鲁棒性,有效地解决了数字产品的多著作权问题。     

一种基于ICA的脊波域数字水印算法

鲁棒性是数字水印的基本要求之一,水印嵌入的位置和强度直接影响到水印的鲁棒性。脊波变换是新型的多尺度分析方法。本文提出的脊波变换是图像在有限Radon变换(FRAT)的基础上进行二维小波变换,将Arnold置乱后的水印嵌入到图像的脊波系数中,利用独立成分分析(ICA)进行水印提取。实验结果表明,在强烈攻击的情况下,此方法具有良好的水印不可见性,对Gaussian、JPEG、Speckle攻击具有良好的鲁棒性。     

Information recommendation using ICA

We describe an information filtering system using independent component analysis (ICA). A document–word matrix is generally sparse and has an ambiguity of synonyms. To solve this problem, we propose a method to use document vectors represented by independent components. An independent component generated by ICA is considered as a topic. In practice, we map the document vectors into a topics space. Since some independent components are useless for recommendation, we select the necessary components from all independent components by a maximum distance algorithm (MDA). Although Euclidean distance is usually used by MDA, we propose topic selection by cosine-distance-based MDA to solve the mismatch of similarities in information filtering. We create a user profile from the transformed data with a genetic algorithm (GA). Finally, we recommend documents with the user profile and evaluate the accuracy by imputation precision. We have carried out an evaluation experiment to confirm the practicality of the proposed method.This work was presented, in part, at the 9th International Symposium on Artificial Life and Robotics, Oita, Japan, January 28–30, 2004     

On Security of Compressed Gray Scale Image Using Joint Encryption and Data Hiding

ABSTRACT

This paper proposes a novel scheme that integrates quality access control and tracking of illicit distribution of digital image(s) in a single platform. The goal is achieved by (1) modulating some of the valuable Discrete Cosine Transform (DCT) coefficients of the compressed data followed by (2) embedding a binary watermark (fingerprint) as tracking information using Quantization Index Modulation (QIM). The data modulation process serves the purpose of access control so that an unauthorized user is unable to enjoy proper visual quality. On the other hand, embedded watermark tracks illicit distribution. The coefficients to be modulated are selected pseudo randomly using a secret key (K). Before embedding, the watermark is encoded by applying convolution coding that reliably identifies colluder(s) involved in time varying (intelligent) collusion operation. Simulation results have shown the validity of the above claims without affecting compatibility with standard JPEG coding scheme.     

Mixing Two Bibliography Style Languages

This article reports a twofold experience. On the one hand, we show that some concepts related to xml can be used to define a language for formatting a bibliography section put at the end of a printed work. On the second hand, we show that this new language can be used jointly with the bibliography style language of BIBT_EX in the new version of MlBIBT_EX. This should ease the transition between the “old” bibliography style language and the “new” one.     

一种基于DWT的自适应大容量音频水印算法

提出了一个自适应的音频水印嵌入算法。对音频载体和水印图像分别进行了离散小波变换和离散余弦变换,根据人类听觉系统的时域掩蔽效应选择嵌入点,在小波系数中的对应位置嵌入水印信号。嵌入水印后的音频信号觉察不出明显的变化,并且嵌入后的水印具有较高的鲁棒性。     

ICA based identification of dynamical systems generating synthetic and real world time series

Independent Component Analysis (ICA) is a recent and well known technique used to separate mixtures of signals. While in general the researchers put their attention on the type of signals and of mixing, we focus our attention on a quite general class of models which act as sources of the time series, the dynamical systems. In this paper we focus our attention on the general problem to understand the behaviour of ICA methods with respect to the time series deriving from a specific dynamical system, selecting large classes of them, and using ICA to make separation. This study gives some interesting results that are very useful both to highlight some properties related to dynamical systems and to clarify some general aspects of ICA, by using both synthetic and real data. From one hand we study the features of the linear (simple and coupled) and non-linear (single and coupled) dynamical systems, stochastic resonances, chaotic and real dynamical systems. We have to stress that we obtain information about the separation of these systems and substantially how from the entropy of the complete system we can obtain the entropies of the single dynamical systems (so that we also could obtain a more realistic analogic circuit). On the other hand these results show the high capability of the ICA method to recognize the dynamical systems independently from their complexity and in the case of stochastic series ICA perfectly recognizes the different dynamical systems also where the Fourier Transform is irresolute. We also note that in the case of real dynamical systems we showed that ICA permits to recognize the information connected to the sources and to associate to it a phenomenological dynamical system that reproduce it (i.e. Organ Pipe, Stromboli Volcano, Aerosol Index).     

Dynamic aspects of visual modelling languages

A large class of diagrammatic languages falls under the broad definition of “executable graphics”, meaning that some transformational semantics can be devised for them. On the other hand, the definition of static aspects of visual languages often relies on some form of parsing or constructive process. We propose here an approach to the definition of visual languages syntax and semantics based on a notion of transition as production/consumption of resources. Transitions can be represented in forms which are intrinsic to the diagrams or external to them. A collection of abstract metamodels is presented to discuss the approach.     

Independent component analysis of Gabor features for face recognition

We present an independent Gabor features (IGFs) method and its application to face recognition. The novelty of the IGF method comes from 1) the derivation of independent Gabor features in the feature extraction stage and 2) the development of an IGF features-based probabilistic reasoning model (PRM) classification method in the pattern recognition stage. In particular, the IGF method first derives a Gabor feature vector from a set of downsampled Gabor wavelet representations of face images, then reduces the dimensionality of the vector by means of principal component analysis, and finally defines the independent Gabor features based on the independent component analysis (ICA). The independence property of these Gabor features facilitates the application of the PRM method for classification. The rationale behind integrating the Gabor wavelets and the ICA is twofold. On the one hand, the Gabor transformed face images exhibit strong characteristics of spatial locality, scale, and orientation selectivity. These images can, thus, produce salient local features that are most suitable for face recognition. On the other hand, ICA would further reduce redundancy and represent independent features explicitly. These independent features are most useful for subsequent pattern discrimination and associative recall. Experiments on face recognition using the FacE REcognition Technology (FERET) and the ORL datasets, where the images vary in illumination, expression, pose, and scale, show the feasibility of the IGF method. In particular, the IGF method achieves 98.5% correct face recognition accuracy when using 180 features for the FERET dataset, and 100% accuracy for the ORL dataset using 88 features.     

A blind video watermarking scheme based on ICA and shot segmentation

With the development of network and multimedia techniques, the digital video copyright protection issues have become more and more important as digital video can be easily pro- duced, replicated, accessed and distributed. Digital video watermark is the ma…     

基于小波包和DCT变换的双水印技术

目前较流行的水印算法是在图像的时空域或变换域中感知重要的系数嵌入水印。然而感知重要的部分通常是图像的低频部分,当图像经历某些变换操作时,从低频提取的水印会产生严重的质量退化。针对这一问题,该文在结合小波包变换和离散余弦变换的同时,在图像小波包分解后的低频和高频部分选取适当系数分别嵌入同一个水印。水印采用二值图像,在水印嵌入之前对水印图像进行Amold变换以增强算法的隐蔽性和鲁棒性。实验结果表明,这一方案比单重水印方案能够更有效地抵抗JPEG有损压缩、剪切、滤波和噪声等常见的水印攻击,具有更广泛的适应性和更强的鲁棒性。     

A simple ICA-based digital image watermarking scheme

In a digital watermarking scheme, it is not convenient to carry the original image all the time in order to detect the owner's signature from the watermarked image. Moreover, for those applications that require different watermark for different copies, it is preferred to utilize some kind of watermark-independent algorithm in extraction (does not need a priori knowledge of the watermark). In this paper we introduce a novel approach called WMicaT that employs an independent component analysis technique in watermark embedding and extraction. Using a single ‘public image’ that can be made publicly available, the new algorithm is able to extract the watermark without requiring the original image and any information about the watermark. In addition, the watermark is not limited to some specific binary sequences but can be any meaningful image. The WMicaT method, undergoing different experiments, has shown its robustness against many attacks.     

Strategy of information management (?)

Information management seems to be the “in” subject of the decade. However, it remains an elusive target. It is suggested that the problem lies in the assumption that we are dealing with information management, as of it were solely concerned with the supply of information. We argue that it would be more fruitful to think in terms of managing information technology, in view of its perceived increasing strategic interest to organizations. The subject is treated from both an historical and organizational perspective.For the time being, the ideal and reality still seem to be poles apart. On the one hand, we find that, despite a serious lack of sound insight, information technology has already been absorbed by most organizations; on the other hand, it is clear that for today's strategy, simple extrapolation does not suffice. We are learning how to bridge the gap in understanding, so that information technology may be actively exploited to strengthen whatever other activities an organization seems to find important for its competitive advantage.     

基于局部相关匹配的小波变换图像水印方法*

现有局部感兴趣区域的水印嵌入方法在选取感兴趣区域时未考虑水印信号能与载体图像局部相似的特性,本文提出一种基于局部相关匹配的小波变换图像水印方法。该方法利用Arnold变换对水印图像加密,采用相位相关法选取合适的水印嵌入区域,并根据图像纹理的差异在小波的高频和低频系数上分别嵌入水印信息。水印检测时,在嵌入水印的区域提取出多个水印信息,将多个水印信息的加权平均作为检测水印。实验结果表明,该方法嵌入水印的不可见性及抗噪声和压缩的性能较好。     

高稳健性图像水印算法

提出了一个以8位灰度图像作为水印的DCT域水印算法。首先对以往的基于人眼视觉掩蔽特性(HVS)的块分类方法进行了改进，使其能更充分地挖掘载体图像的信息隐藏能力。在图像水印的嵌入方法上，将水印图像进行DCT变换，然后将DCT系数的低频分量嵌入到拉伸因子较大的块中。同时，对水印图像的DCT系数进行调整，使得各频率的能量基本相当，间接地提高了提取的水印图像的信噪比。实验结果表明，所提出的水印算法对多种攻击具有较好的稳健性，尤其对JPEG压缩、旋转和剪切攻击。其稳健性更明显。     

基于DCT域的两种盲水印算法的应用与比较

利用Arnold置乱理论以及混沌序列加密的方法对水印图像进行加密预处理,水印信息嵌入时,提出两种基于DCT变换域的数字水印算法。分别通过修改特定系数关系和中频系数替换的方法实现水印信息的嵌入。对两种水印算法的仿真实验结果进行比较与分析,得出两种水印算法在图像的不可见性方面都表现出良好的性能,在水印检测时,实现了完全盲提取;但在抵抗JPEG压缩、噪声、滤波等攻击方面基于中频系数替换的水印算法表现出更强的鲁棒性。