基于遗传算法的抗几何攻击图像水印方案

现有的大部分鲁棒图像水印算法很难抵抗几何攻击,其原因是同步信息被破坏。因此只要在提取水印前使同步信息得到恢复,水印即可成功提取。本文采用仿射不变特征点匹配方法估计图像所经历的几何变换参数,其中特征点匹配通过遗传算法实现。水印嵌入过程是自适应的,且实现了盲提取。实验表明,本文提出的方法对多数攻击,尤其对几何攻击有很好的鲁棒性。     

图像水印检测分布研究及算法改进

本文首先分析了加法和乘法2种嵌入模型下检测器的理论输出分布,提出影响漏检率和虚警率的因素。进而提出一种改进的嵌入算法,根据水印与图像特征的相关系数提出不同的嵌入策略,使检测结果分布在正负轴两边。该算法在噪声环境下体现出很好的稳健性。     

基于混沌加密的二值图像数字水印算法

本文提出了一种基于混沌加密的DCT域二值图像数字水印算法。该算法首先利用混沌序列将含有版权信息的有意义二值图像水印进行加密,然后利用人眼视觉系统(HVS)的特性将原始载体图像进行块分类,最后将不同强度的水印分量嵌入到不同类的图像块DCT中频系数中去。利用混沌序列对水印进行加密,使水印本身具有安全性,对数字作品进行双重保障。通过高斯噪声、椒盐噪声以及JPEG压缩和剪切等攻击试验,表明了该算法具有较好的鲁棒性。     

一种基于Arnold置乱和小波变换的二值图像水印算法

为了更有效地保护网络上静态图像版权,提出了一种基于Arnold置乱和小波变换的二值图像水印算法。将Arnold置乱过程用于提高其安全性和抗剪切的鲁棒性;然后将置乱后的水印进行一次小波分解,再根据人眼的视觉特性以相应的尺度因子嵌入原始图像的小波变换不同频域中。仿真实验表明该算法安全、有效,兼顾了水印的透明性和鲁棒性。由于该算法是在小波变换域中实现的,故可JPEG2000、MPEG4相兼容,是一种可以推向实用的水印算法。     

A novel fast fractal image coding algorithm based on texture feature

A novel, fast fractal image coding algorithm based on the texture feature is proposed in this paper. Fractal image coding is a very promising technique for image compression. However, it has not been widely used because of the long encoding time and high computational complexity. The most fractal image encoding time is spent in determining the approximate D‐block from a large D‐blocks library by using the global searching method. Clustering the D‐blocks library is an effective method to reduce the encoding time. First, all the D‐blocks are clustered into several parts based on the new texture feature α derived from variation function; second, for each R‐block, the approximate D‐blocks are searched for in the same part. In the search process, we import control parameter δ; this step avoids losing the most approximate D‐block for each R‐block. Finally, the R‐blocks whose least errors are larger than the threshold given in advance are coded by the quad tree method. We have performed a simulation with MATLABR2010a to verify the effectiveness of the proposed algorithm. The experimental results show that the proposed algorithm can be over 6 times faster than the moment‐feature‐based fractal image algorithm; in addition, the proposed algorithm also improves the quality of the decoded image and increases the PSNR's average value by 2 dB. The comparisons demonstrate that this method is better than the fractal image coding algorithm based on statistical features. © 2012 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

Image adaptive RST invariant watermark using pseudo-Zernike moments

Rotation, scaling and translation (RST) attacks can desynchronize watermark detection, which causes failure in many watermarking systems. In this paper, an image adaptive RST invariant watermark (AWPZM) is proposed by using the rotation invariant property of pseudo-Zernike moments (PZM) and oddeven quantization. PZM of the original image is computed first, and then those suitable for watermark generation are selected. Then, magnitudes of them are odd-even quantized to generate the watermark. In detection, a normalized hamming function is employed to determine the similarity of the watermark. Experimental results show its robustness to rotation and scaling. For traditional attacks, such as JPEG compression, added noise and filtering, the similarities are all above 0.95. __________ Translated from Journal of Xidian University, 2007, 34(1): 38–42 [译自: 西安电子科技大学学报]     

Optimum watermark detection of ultrasonic echo medical images

To prevent unauthorized access to a person's medical images, it is widely acknowledged that the security features of confidentiality, availability and integrity should be in place. This paper proposes the watermarking of ultrasonic echo images together with optimal watermark detection, in which pseudorandom noise is added to the images for integrity. The optimum watermark detection is the integration of the generalized Gaussian distribution (ρ‐GGD) and the Cauchy distribution. The results show that the proposed method gives good detection performance. The proposed method not merely achieves optimum detection using the Rao test but also leads to the highest detection probability with JPEG2000 compression. Compared with other detection methods, our proposed method exhibits better watermark detection performance even when the watermark‐to‐document ratio (WDR) is −50 dB. © 2014 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

一种DCT域子采样的鲁棒可去性盲水印

随着多媒体技术的发展,图像信息安全已引起学者的广泛关注。数字水印技术是图像信息隐藏和版权保护的一种有效工具。在子采样和图像分块的基础上,采用新的跳跃取点策略。该策略对隐秘图像连续区域的剪切攻击具有鲁棒性。同时,由于水印的嵌入会对图像造成一定程度的失真,因此尝试对水印进行移除操作。实验结果表明,该方法不仅对噪声和常规的图像处理攻击具有鲁棒性,尤其对剪切攻击具有很强鲁棒性,而且还具有很好的不可视性。     

小波数字水印在电力系统信息安全中的应用

随着电力系统信息化的快速发展,信息安全问题已是电力系统中一个急待解决的难题之一。为保证电力系统中待传递数据的安全性,研究了利用一种基于小波的数字水印技术把电力系统中的传输数据隐藏起来。具体以电力市场环境下,研究采用数字水印技术将电量交易价格隐藏,保证了电量交易价格不被竞争对手截获,从而有效地保证了电力系统信息的安全性。数字水印的实现方案是：数字水印的生成、嵌入和检测。通过Matlab编程和仿真,验证了采用该方法的有效性。     

基于奇偶量化的图像水印嵌入与检测方法

针对数字图像水印在嵌入和提取过程中要求载体图像和水印信息都具有完整性的基本特征,提出了基于奇偶量化的自适应水印嵌入与检测方法.该方法将水印信息的二值图像进行Arnold置乱,嵌入到载体图像DCT的中频系数中,利用Watson视觉模型和JPEG量化表共同自适应地调节量化步长,由于采用奇偶量化,水印提取方法简单.实验结果表...     

一种基于形状的图像检索系统

基于内容的图像检索是当前国内外的一个研究热点.形状是物体的一个重要特征,是刻画物体结构、形态、轮廓的特征量,也是比较难描述的图像特征之一.由于物体形状的自动获取涉及图像中感兴趣目标的分割,所以基于形状检索的图像检索是一项有难度的研究工作.基于形状的检索一般适用于较容易识别的物体.在本文中采用了面积、形状矩等局部或全局特...     

A novel MOSFET with vertical signal‐transfer capability for 3D‐structured CMOS image sensors

We have developed a novel MOSFET that can transfer signals vertically without through‐silicon vias but by using a fully depleted silicon‐on‐insulator (FDSOI) structure with its source region connected to the back electrodes as well as the front ones. A prototype MOSFET fabricated using the backside anisotropic wet etching technique has confirmed that the electrical characteristics measured from the front and the back electrodes are identical. The subthreshold factor S of the prototype was found to be 64.5 mV/decade, suggesting a good switching performance. Since the double‐sided MOSFET has vertical signal‐transfer capability and excellent operating characteristics, it is expected to contribute to developing a More‐than‐Moore type device of three‐dimensional integration such as pixel‐parallel image sensors. © 2014 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

A novel framework for evaluating the image accuracy of dynamic MRI and the application on accelerated breast DCE MRI

Objective

To develop a novel framework for evaluating the accuracy of quantitative analysis on dynamic contrast-enhanced (DCE) MRI with a specific combination of imaging technique, scanning parameters, and scanner and software performance and to test this framework with breast DCE MRI with Time-resolved angiography WIth Stochastic Trajectories (TWIST).

Materials and methods

Realistic breast tumor phantoms were 3D printed as cavities and filled with solutions of MR contrast agent. Full k-space raw data of individual tumor phantoms and a uniform background phantom were acquired. DCE raw data were simulated by sorting the raw data according to TWIST view order and scaling the raw data according to the enhancement based on pharmaco-kinetic (PK) models. The measured spatial and temporal characteristics from the images reconstructed using the scanner software were compared with the original PK model (ground truth).

Results

Images could be reconstructed using the manufacturer’s platform with the modified ‘raw data.’ Compared with the ‘ground truth,’ the RMS error in all images was <10% in most cases. With increasing view-sharing acceleration, the error of the initial uptake slope decreased while the error of peak enhancement increased. Deviations of PK parameters varied with the type of enhancement.

Conclusion

A new framework has been developed and tested to more realistically evaluate the quantitative measurement errors caused by a combination of the imaging technique, parameters and scanner and software performance in DCE-MRI.

     

基于异构描述子的新型高斯混合模型图像自动标注方法

过去几十年以来,各种图像标注算法相继提出。这些方法要么需要很大的计算量,要么标注效果不理想。论文提出了一种基于异构描述子的新型高斯混合模型图像语义自动标注方法。本文的高斯混合模型是采用异构空间来构建的,不同于其他的高斯混合模型。对于每个标注词,在多个特征空间下分别用高斯模型来描述,形成对应子空间的“标注词分描述子”。由于各个分描述子描述不同标注词的能力有很大差别,因此通过机器学习的方法来融合这些分描述子,形成更加有效的“标注词描述子”,从而提高标注的准确率。论文提出的“标注词描述子”可以有效地建立图像高层语义概念与底层视觉特征之间的对应关系,准确地描述标注词的语义内容,从而提高图像的标注性能。通过在COREL数据集上的测试表明了方法的有效性。     

时域抗JPEG压缩数字水印算法的设计

本文首先介绍数字水印的含义、用途和基本特征;然后详细说明了为了隐藏秘密信息,防止针对图像有意或无意的攻击,设计了基于时域的抗JPEG压缩数字水印算法,并讨论了水印嵌入过程中各个参数在调节水印不可感知性、鲁棒性和容量3者关系时的作用;最后,本文列出了不同攻击下水印嵌入的效果和水印提取的结果.其结果表明,该方法可以针对所经受的JPEG压缩程度以及其他攻击的强度,自由调节各个参数,以达到水印不可感知性、鲁棒性和隐藏信息容量之间的平衡,并保证隐藏信息的完整性.     

单图像超分辨率方法综述

图像超分辨率是指从低分辨率图像生成对应的高分辨率图像。它在人脸识别、数字高清电视、视频通信等领域应用广泛。为了对图像超分辨率技术进行深入探索和总结,本文首先回顾了单图像超分辨率的历史,综述了基于非学习的超分辨率方法,其中展开介绍了基于插值和基于重建的方法,接着重点介绍了基于学习的方法,详细分析了基于深度学习的图像超分辨率,具体总结了SRCNN、ESPCN、SRGAN这三种图像超分辨率方法,并将其与递归结构、密集结构、注意力机制网络结构进行对比,之后分析了损失函数和上采样方式在图像超分辨率中的作用,介绍了常用数据集和图像评价指标,展示了图像超分辨率的可视化结果。最后,总结了现有单图像超分辨率方法的进展和不足。     

基于完全反传播神经网络(FCNN)的数字水印新算法

数字水印对于保护和鉴定多媒体信息是一项重要技术,它可以在非法复制和篡改中保护数字媒体.数字水印的嵌入与检测涉及到复杂的程序,其中包括对水印的随机化,选择水印的嵌入位置和检测位置,将随机水印嵌入到特殊位置并在此位置上检测水印等.本文设计了一个完全反传播神经网络方法,水印通过特殊的完全反传播神经网络嵌入于检测水印.区别于传统的方法,水印嵌入到FCNN的神经元上代替了嵌入在宿主图像上,因此数字水印图像几乎与宿主图像一样,大部分攻击不可能降低嵌入水印质量.实验结果表明,此方案能够达到鲁棒性、不可见性与真实性.     

Construction of ubiquitous acoustic spaces using audio watermark technology and mobile terminals

We propose ‘Ubiquitous Acoustic Spaces,’ where each sound source can emit some address information and automatically enable us to access its related cyber space, using mobile terminals. This will be used for a new type of wireless advertisement for mobile phones using sound media. In order to realize this, we have proposed a nearly lossless audio watermarking technique, utilizing two‐channel stereo audio characteristics. This enables contactless asynchronous detection of embedded watermarks through loudspeaker and microphone devices even by using cell phones. We extend the embedding frequency band width and improve the extraction precision. Thus we enable watermark extraction from both stereo channels. Possible embedding data rate can be extended to 80 bps, and almost 100% data can be extracted in case of using PC or PDA devices as terminals. Moreover, we can carry out the watermark extraction even from cell phone loudspeakers or through public telephone networks. From our experiments, almost 100% extraction of 20 bps embedded data could be achieved with these phone devices. In this paper, we describe briefly our newly proposed dual monaural watermark embedding algorithms, and present a prototype system of ubiquitous acoustic spaces using cell phones, based on our experimental results of extraction precision on several capture signal conditions including experiments using cell phones or public telephone networks. Copyright © 2007 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

Increasing the audio watermark data rate in the construction of ubiquitous acoustic spaces

We propose “Ubiquitous Acoustic Spaces,” in which each sound source can emit some address information with audio signals and allow automatic access to its related cyber space, using handheld devices such as cell phones. In order to realize this concept, we have considered three types of extraction methods: acoustic modulation, audio fingerprints, and audio watermarking. We have proposed a novel audio watermarking technique, which allows contactless asynchronous detection of embedded audio watermarks through speaker and microphone devices. However, its embedding data rate was around 10 bps, which was not sufficient for embedding generally used URL address texts. Therefore, we have extended the embedding frequency range and proposed a duplicated embedding algorithm, which uses both the previously proposed frequency division method and the temporal division method together. By these improvements, the possible embedding data rate could be extended to 61.5 bps, and we could extract watermarks through public telephone networks, even from a cell phone sound source. In this paper, we describe our improved watermark embedding and extracting algorithms, and present experimental results on watermark extraction precision under several audio signal capturing conditions. © 2008 Wiley Periodicals, Inc. Electr Eng Jpn, 165(1): 42–51, 2008; Published online in Wiley InterScience ( www.interscience. wiley.com ). DOI 10.1002/eej.20758