基于压缩量化与邻域空间LBP算子的图像哈希算法

目的为了解决哈希算法的感知鲁棒性与伪造检测能力不高的问题,提出基于特征压缩机制与邻域空间局部二值模式的紧凑图像哈希算法。方法首先利用2D线性插值技术,对输入图像进行预处理;嵌入Ring分割技术,将其变为二次图像;再利用Gabor滤波技术对其完成过滤;考虑到图像的颜色特征与其内在的空间关系,基于局部二值模式LBP设计邻域空间LBP算子,提取滤波图像的特征;构建特征压缩量化准则,输出紧凑的哈希二值数组;迭代Logistic映射,输出随机序列,通过量化每个序列值输出密钥流,通过构建动态引擎设计分段异加密模型,实现紧凑哈希序列的加密,获取图像哈希;最后计算原始哈希序列与待检测哈希序列的Hamming距离,实现图像信息的安全认证。结果与已有的哈希生成机制相比,文中算法所输出的哈希序列更紧凑,对旋转、伽马校正等篡改操作具有更好的感知鲁棒。结论所提哈希技术具备较高的安全性,在包装图标检索、信息检测等领域具有较好的价值。     

基于颜色矢量角度直方图与DCT压缩的鲁棒哈希算法

目的为了改善哈希序列对篡改内容的正确识别率和鲁棒性,提出基于颜色矢量角度直方图耦合离散余弦变换压缩的鲁棒哈希算法。方法结合插值运算和Gaussian滤波器,对图像实施预处理,使其对于任意的认证目标均可产生一个固定长度的哈希序列;引入极坐标变换LPT(Log-polar transform),对规范尺寸图像完成坐标变换,输出二次图像;借助SVD(SingularValueDecompostion)方法来处理二次图像,输出全局鲁棒特征;提取规范尺寸图像中所有像素分量的颜色矢量角度,并计算其对应的直方图,形成色彩感知特征;将全局与色彩感知特征组合,将其视为1D哈希序列;基于离散余弦变换DCT(Discrete Cosine Transform),对1D哈希序列实施压缩,输出一组交流系数;引入Logistic映射,根据其输出的混沌序列来设计差异加密方法,对交流系数实施扩散,从而形成目标哈希序列。基于l2范数距离和优化的识别阈值,对待检测图像内容的真伪实施判别。结果试验结果显示,相对于已有的哈希机制而言,所提方案具备更强的鲁棒性,对各类几何攻击均有更高的检测准确度。结论所提哈希方法具有较高的鲁棒性与识别正确率,对商标检索、信息防伪等领域具有较好的参考价值。     

架构于抗混叠轮廓波的红外图像插值放大算法

为了在放大红外图像时保持边缘或轮廓的正则性,提出一种基于抗混叠轮廓波变换的图像插值放大算法.该算法首先结合抗混叠塔式滤波器组和方向滤波器组,构造出抗混叠的轮廓波变换;然后将原始图像的小波域线性插值结果看成是放大图像的初始估计,输入到一个迭代过程;在每次迭代中,将放大图像看成是理想高分辨率图像的含噪逼近,并对其实施抗混叠轮廓波变换,根据变换系数的稀疏性约束实现降噪处理;最后,经过若干次迭代得到理想的红外放大图像.实验表明,对于测试图像,经过迭代处理后峰值信噪比平均提高了0.837 dB;且该算法在视觉质量上明显优于双线性插值算法及基于小波的算法.     

基于方向金字塔分解与稳定几何失真校正的鲁棒图像水印算法

目的为了提高水印算法的抗几何攻击能力,并兼顾较高的鲁棒性与不可感知性,设计一种基于方向金字塔分解与稳定几何失真校正的鲁棒图像水印算法。方法首先,引入方向金子塔,对载体图像完成分解,输出对应的低通与高通子带;将低通子带分割为一系列的非重叠块;根据载体的亮度、纹理与边缘掩码,计算水印嵌入强度,最大程度地平衡水印图像的不可感知性与鲁棒性;设计水印嵌入方法,将经过Arnold映射加密后的水印嵌入到非重叠子块中,通过修改载体的方向金子塔分解系数,获取水印图像;将不同的攻击类型作用于水印图像,建立训练样本;再利用方向金子塔分解训练样本,计算高通子带的高斯-厄米矩能量,将其视为特征矢量;再利用特征矢量对模糊支持向量机完成训练,以预测几何失真参数,准确校正受攻击的水印图像;设计水印检测机制,从水印图像中复原水印。结果实验数据表明,与当前图像水印方案相比,所提算法具有更高的抗几何变换能力,以及较好的不可感知性与鲁棒性,其提取的水印失真度最小,对应峰值信噪比保持在40dB以上。结论所提水印算法具有较高的鲁棒性和视觉隐秘性,在版权保护、信息防伪等领域具有一定的参考价值。     

基于融合鲁棒特征与多维尺度变换的紧凑图像哈希算法

目的为了增强哈希序列对任意旋转角度的鲁棒性与识别能力,提出一种基于融合鲁棒特征与多维尺度变换的紧凑图像哈希算法。方法首先,利用双线性插值来固定图像的哈希序列长度,获取规则尺寸的图像;借助高斯低通滤波对规则图像完成过滤操作,消除噪声污染和插值误差对哈希生成的影响;将滤波图像转换到YCbCr颜色空间,提取亮度Y分量,增强哈希对亮度调整的鲁棒性;利用极坐标变换LPT方法处理亮度Y分量,输出二次图像;引入SVD机制来分解二次图像,获取其抗旋转的鲁棒特征;同时,根据Fourier变换与残差机制,获取Y分量的局部显著特征;随后,将这2种特征组合,形成融合鲁棒特征,将其视为中间哈希序列;引入多维尺度变换,对中间哈希序列完成压缩,获取紧凑哈希;基于Logistic映射,完成紧凑哈希序列的加密,形成目标哈希;通过计算真实图像与待认证图像之间哈希序列对应的Hamming距离,根据预设阈值,完成图像识别。结果测试数据表明,较已有的哈希方案而言,所提方案拥有更高的鲁棒性和更紧凑的哈希长度,呈现出更为理想的ROC曲线,在多种攻击下,其稳定的正确识别率保持在96%以上。结论所提哈希方案拥有良好的鲁棒性与敏感性,在包装图标检索、信息水印等行业具备较好的应用价值。     

基于颜色属性相关性的图像色差公式的研究

CIE 色差公式不适合用于复杂图像对间色差预测,为更准确地评价再现图像与原稿之间的颜色差别,提出了一个基于颜色属性相关性的图像色差计算算法。算法的图像色差的计算,采用对比度敏感度函数(CSF)去除图像内容中视觉难以觉察的信息,降低图像内容的空间复杂性,通过对图像对间颜色三属性的相关性分析,确定各色彩属性差的权重。心理物理实验证实,色差公式计算值与主观评价的评分是一致的,表明该算法可用于评价图像对间的颜色差别,也可用于评价压缩方式、图像增强算法以及色域映射算法等的性能。     

基于自适应步长双参数正则化算法的超声波过程层析成像图像重建

提出了一种新的自适应步长双参数正则化算法,对超声波层析成像系统检测浆体浓度分布进行图像重建。该算法利用转换矩阵将超定解作为先验信息,嵌入到正则化泛函中,避免重建图像被过度平滑,不仅成像速度较快且重建图像具有较高分辨率。仿真实验结果表明,相比于Tikhonov正则化算法以及Landweber算法,自适应步长双参数正则化算法重建图像的相关系数有明显提高并且边界信息更加可靠。     

改进的共轭梯度算法的图像重建算法

针对电容层析成像图像重建问题的病态性,在Tikhonov正则化的基础上,以正则化解决方案的规模和给定数据的质量为理论依据,引入一个数学变换,提出了一种新的正则化方法,该方法克服了常规正则化方法扰乱原系统的缺陷。同时,将ECT物理模型进行规范化,并对共轭梯度算法进行改进。仿真实验表明,改进的共轭梯度算法的成像质量高于LBP算法、 Tikhonov正则化算法和共轭梯度算法,并利用相关系数进行了验证。     

一种基于正则化和Watson视觉模型的图像指纹算法

大量多媒体应用的发展使得数字图像很容易地被非法操作和篡改,提出一种基于图像正则化和视觉特性的图像指纹算法,可以有效地实现图像的认证和识别.首先对图像进行正则化预处理,消除几何形变对图像的影响,然后对图像进行分块DCT变换,利用Watson视觉模型对DCT系数进行处理,增大人眼敏感的频域系数在计算图像特征时的权重,经过量化形成最终的指纹序列.在图像指纹序列生成过程中,加入密钥控制,提高了指纹的安全性.实验结果表明,该方法的冲突概率在10-7数量级,对JPEG压缩、旋转、缩放等操作具有较好的稳健性.     

一种基于压缩感知的全变分图像去噪算法

针对传统图像去噪方法的不足,提出了一种基于压缩感知的全变分正则化图像去噪算法,利用基于压缩感知算法中的TVAL3算法对含噪图像进行图像重构和噪声去除.通过对比该算法与OMP(Orthogonal matching pursuit,OMP)与SP(Subspace pursuit,SP)算法的峰值信噪比和重构时间,发现在采样率为0.4和0.8时,该算法的峰值信噪比提高都在3dB以上,时间方面也有明显提升;随着采样率的提升,算法所需的迭代次数越来越少;采样率为0.4时,所需的迭代次数为78次,但采样率在0.8时,迭代次数减少到57次,所需时间越来越短.实验结果表明,该算法的重构效果明显优于其它压缩感知重建算法,能够很好地进行图像重构和噪声去除.     

Perceptual image hashing using local entropies and DWT

Abstract

Image hashing is an emerging technology in multimedia security. It uses a short string called image hash to represent an input image and finds applications in image authentication, tamper detection, digital watermark, image indexing, content-based image retrieval and image copy detection. This paper presents a hashing algorithm based on the observation that block entropies are approximately linearly changed after content-preserving manipulations. This is done by converting the input image to a fixed size, dividing the normalised image into non-overlapping blocks, extracting entropies of image blocks and applying a single-level 2D discrete wavelet transform to perform feature compression. Correlation coefficient is exploited to evaluate similarity between hashes. Experimental results show that the proposed algorithm is robust against content-preserving operations, such as JPEG compression, watermark embedding, Gamma correction, Gaussian low-pass filtering, adjustments of brightness and contrast, scaling and small angle rotation. Similarity values between hashes of different images are small, indicating good performances in discriminative capability.     

Feature Fusion Multi-View Hashing Based on Random Kernel Canonical Correlation Analysis

Hashing technology has the advantages of reducing data storage and improving the efficiency of the learning system, making it more and more widely used in image retrieval. Multi-view data describes image information more comprehensively than traditional methods using a single-view. How to use hashing to combine multi-view data for image retrieval is still a challenge. In this paper, a multi-view fusion hashing method based on RKCCA (Random Kernel Canonical Correlation Analysis) is proposed. In order to describe image content more accurately, we use deep learning dense convolutional network feature DenseNet to construct multi-view by combining GIST feature or BoW_SIFT (Bag-of-Words model+SIFT feature) feature. This algorithm uses RKCCA method to fuse multi-view features to construct association features and apply them to image retrieval. The algorithm generates binary hash code with minimal distortion error by designing quantization regularization terms. A large number of experiments on benchmark datasets show that this method is superior to other multi-view hashing methods.     

Robust Image Hashing via Random Gabor Filtering and DWT

Image hashing is a useful multimedia technology for many applications, such as image authentication, image retrieval, image copy detection and image forensics. In this paper, we propose a robust image hashing based on random Gabor filtering and discrete wavelet transform (DWT). Specifically, robust and secure image features are first extracted from the normalized image by Gabor filtering and a chaotic map called Skew tent map, and then are compressed via a single-level 2-D DWT. Image hash is finally obtained by concatenating DWT coefficients in the LL sub-band. Many experiments with open image datasets are carried out and the results illustrate that our hashing is robust, discriminative and secure. Receiver operating characteristic (ROC) curve comparisons show that our hashing is better than some popular image hashing algorithms in classification performance between robustness and discrimination.     

基于边缘曲线光顺的图像放大算法

针对常规插值方法在图像放大时图像轮廓模糊的问题,提出了一种基于边缘的图像放大算法。算法包括边缘曲线光顺和插值放大两个步骤。第1步使用能量优化方法,从离散的边缘曲线数据恢复成连续而光顺的边缘曲线,这些曲线将图像分为若干个区域;第2步基于这些光顺曲线,针对不同的区域分别进行图像插值放大计算。使用该方法得到的放大图像边缘清晰光顺,实例证明该算法产生的图像质量高于常规算法。     

基于生成对抗网络的文物图像修复与评价

目的 为了有效解决文物图像的不易保存和物理方法修复困难等问题,提出一种基于生成对抗网络的图像修复算法。方法 文中算法主要分为2个阶段,第1阶段通过Canny边缘检测器提取图像已知部分的边缘信息,利用1个生成器和1个鉴别器修复图像缺失边缘。第2阶段将第1阶段生成的边缘作为先验信息,通过1个生成器和2个鉴别器修复图像缺失部分。2个鉴别器由整体鉴别器和局部鉴别器组成,整体鉴别器用来评估修复后的图像整体连贯性,局部鉴别器用来查看待修复区域为中心的小区域局部一致性。结果 与传统算法对比,文中算法在提高生成图片纹理质量的基础上保证了全局语义结构一致性,在客观指标（峰值信噪比和结构相似性）上,较其他方法有更好的效果。结论 文中算法可以有效修复文物图像的缺损部分,尤其是结构复杂的大范围缺失,取得了良好的视觉效果,表明该算法有良好的修复性能。     

SVD-based digital image watermarking using adaptive generated watermark

In this paper, an evolutionary-based algorithm for digital image watermarking is proposed. Robustness and imperceptibility are two important properties of digital image watermarking which conflict with each other. A singular value decomposition (SVD)-based intelligent algorithm is applied to make a trade-off between robustness and imperceptibility. In the proposed algorithm, first, the original image is divided into 8×8 blocks. Then some special blocks are transformed into discrete cosine transform (DCT) domain. SVD decomposition of DCT coefficients of special blocks is the next stage of the method. Finally, a binary watermark image is embedded into the singular values by a quantisation method. The major property of the proposed method is the generation of a binary watermark image using genetic algorithm to solve the optimisation problem between robustness and imperceptibility. So the watermark could be variable and image adaptive. The simulation results show improvement in robustness of the proposed method against several attacks such as noise contamination, filtering, joint photographic experts group (JPEG) compression and some geometrical attacks like cropping, rotation and rescaling in comparison with a recently similar existing work.     

可见光热反射成像测温中图像配准技术研究

为了消除被测件与测温仪器相对位移对光热反射成像测温结果的影响,研究了亚像素量级的图像配准算法,配合以基于压电陶瓷的三轴纳米位移台,实现了对位移的实时修正.采用傅里叶变换、频域卷积及sine函数近似等方法完成了能够实现亚像素级图像配准的算法,实时计算被测件与测温仪器的相对位移;搭建了一套光热反射成像测温实验装置,采用了三...     

基于改进几何哈希法的矢量图形检索

针对图形文件以矢量方式记录几何实体的特点,提出了一种基于改进几何哈希法的图形检索算法。与图像的对应处理方式相比,该算法充分利用图形的矢量信息,以实体基准点和实体的准确几何形式描述几何实体;在检索过程中,每个图形实体仅有基准点参与排序和查询,哈希表数据组织形式简单,查询速度快;通过验证图形实体属性,保证了查询结果的准确性。最后以矢量图形文件为图例,验证了方法的有效性。     

An efficient and secure partial image encryption for wireless multimedia sensor networks using discrete wavelet transform,chaotic maps and substitution box

Wireless Sensor Networks (WSN) is widely deployed in monitoring of some physical activity and/or environmental conditions. Data gathered from WSN is transmitted via network to a central location for further processing. Numerous applications of WSN can be found in smart homes, intelligent buildings, health care, energy efficient smart grids and industrial control systems. In recent years, computer scientists has focused towards findings more applications of WSN in multimedia technologies, i.e. audio, video and digital images. Due to bulky nature of multimedia data, WSN process a large volume of multimedia data which significantly increases computational complexity and hence reduces battery time. With respect to battery life constraints, image compression in addition with secure transmission over a wide ranged sensor network is an emerging and challenging task in Wireless Multimedia Sensor Networks. Due to the open nature of the Internet, transmission of data must be secure through a process known as encryption. As a result, there is an intensive demand for such schemes that is energy efficient as well as highly secure since decades. In this paper, discrete wavelet-based partial image encryption scheme using hashing algorithm, chaotic maps and Hussain’s S-Box is reported. The plaintext image is compressed via discrete wavelet transform and then the image is shuffled column-wise and row wise-wise via Piece-wise Linear Chaotic Map (PWLCM) and Nonlinear Chaotic Algorithm, respectively. To get higher security, initial conditions for PWLCM are made dependent on hash function. The permuted image is bitwise XORed with random matrix generated from Intertwining Logistic map. To enhance the security further, final ciphertext is obtained after substituting all elements with Hussain’s substitution box. Experimental and statistical results confirm the strength of the anticipated scheme.     

Addressing for letter‐oriented keys

Abstract

This paper presents a new algorithm for hashing. In this approach each key (x, y) is denoted as a pair of letters, and it is extracted from each letter‐oriented key in the key space according to a certain pre‐determined extraction algorithm. For the keys (x, y), our hashing function hashes as follows:

These functions allow single probe retrieval from minimally sized space of letter‐oriented keys. Various areas can be applied, such as searching reserved words in compilers and filtering frequently used words in natural language processing.