基于高阶马尔可夫随机场及非线性压缩感知的相位恢复算法

在编码衍射成像系统中,为精确重构复图像的幅值和相位,需获取大量的编码衍射图样,导致数据采集时间长.为减少编码衍射图样的数量,本文基于非线性压缩感知理论框架,利用高阶马尔可夫随机场统计先验模型,提出了一种鲁棒相位恢复算法.该方法将复图像的幅值和相位分别进行正则化,并将数据保真项与幅值和相位正则项结合作为代价函数,采用Heavy-Ball算法求解所对应的非凸优化问题.实验结果表明,本文算法在编码衍射图样较少的情况下仍能获得较高的图像重构质量,且对噪声鲁棒.     

采用拉普拉斯尺度混合先验的结构化近似消息传递算法

为了准确有效地实现自然图像的压缩感知重构,提出一种使用拉普拉斯尺度混合（Laplacian Scale Mixture,LSM）先验的结构化近似消息传递（Approximate Message Passing,AMP）算法.利用LSM模型构建AMP算法的高阶统计约束,将压缩感知重构问题转化为先验信息估计问题和奇异值最小化问题.首先,用LSM分布刻画相似块矩阵奇异值的稀疏性,其中该稀疏性指示了图像块的相似性,因此LSM模型被用来描述图像的非局部相似结构;然后,通过期望最大化算法估计LSM模型的尺度参数,得到可靠的先验信息;最后,由AMP算法求解奇异值最小化问题,实现图像的精确重构.实验结果表明,提出的结构化AMP算法的图像重构质量优于多种主流的压缩感知图像重构算法.     

基于无监督栈式降噪自编码网络的显著性检测算法

针对现有的显著性检测算法检测目标类型单一、通用性差的问题,提出一种基于无监督栈式降噪自编码网络的显著性检测算法.该算法利用无监督栈式降噪自编码网络（Stacked Denoising Auto Encoder,SDAE）在多个尺度对原始图像进行稀疏重构,将原始图像与SDAE网络重构图像之间的差作为显著图,二值化后的显著图作为显著性目标检测结果.在SDAE网络训练过程中,将原始图像作为原始数据,网络重构的图像作为观察数据.为了提升网络训练效率,首先利用无监督逐层贪婪方法训练同结构的深度信念网络（Deep Belief Network,DBN）,将训练得到的DBN网络参数设为SDAE网络的初始参数,再计算原始数据与观察数据之间的互信息作为网络收敛代价,利用反向传播进行网络参数微调.实验表明,该网络模型可以完成多类型目标的显著性检测,具有通用性好,准确度高等优点.     

线性相位余弦调制滤波器组研究

图像子带编码要求滤波器组具有线性相位特性，非线性相位特性可对图像编译码带来影响。现研究了一种具有线性相位的最大抽取FIR余弦调制滤波器组，该滤波器组中每一个滤波器都具有线性相位，且由原型滤波器经余弦序列调制得到。经推导这种滤波器组可以设计成具有近似准确重构（NPR）特性或准确重构（PR）特性，文中分别给出了两种情况下滤波器组应满足的条件。通过对一幅图像的二维可分离滤波实验说明了当原型滤波器满足一定条件时滤波器组是准确重构的。     

基于卡通纹理模型的相位恢复算法

相位恢复是指仅利用图像的傅里叶幅值对原始图像进行恢复。由于傅里叶幅值中包含的信息量较少,当图像的过采样率相对较低时,传统的相位恢复算法无法实现图像的有效重构。因此如何利用合适的先验知识来提高图像重构质量是相位恢复的一个关键问题。该文将卡通-纹理模型用于相位恢复,利用全变差(TV)和双树复数小波(DTCWT)两种稀疏表示方法将图像分解为卡通成分和纹理成分,并提出了基于交替方向乘子法(ADMM)的有效求解算法。实验结果表明,该算法能有效提高图像重构质量。     

基于全变差正则化的相位恢复算法

相位恢复问题是指仅通过信号傅立叶变换（或其它线性变换）的幅值恢复原始信号.由于相位信息的缺失,该问题是一个不适定问题,因此需利用先验知识确保精确重建.本文基于非线性压缩感知框架,提出利用自然图像在梯度算子下的稀疏性进行相位恢复的算法.该算法将全变差正则项融合到基于支撑约束和幅值约束的相位恢复问题中,并利用交替方向乘子法（ADMM）对所对应的非凸优化问题进行求解.实验结果表明,该算法明显优于HIO,RAAR等经典的相位恢复算法,并对噪声具有鲁棒性.     

物理可实现的相位编码压缩成像

 压缩成像是压缩感知理论最重要的研究领域之一.在分析压缩成像中实际测量矩阵与测量值约束的基础上,提出一种基于4-f光学架构的物理可实现的频域相位编码压缩成像方法.该方法利用两路光学架构之间的补偿实现相位编码压缩成像中测量值的非负记录,然后从该测量值精确恢复原图像,解决压缩成像应用中理论与实际物理约束之间不一致的问题.该方法可以单次曝光获得充分的测量值精确重建原图像,不需要其它附加信息,是压缩成像物理实现的一种非常可行的方案.模拟实验证明该方法可以有效地实现图像的压缩测量与超分辨率重建.     

一类冗余比为2的二维DFT调制滤波器组的设计

本文构建了一类冗余比为2的二维线性相位的双原型离散傅立叶变换（DFT）调制滤波器组。利用原型滤波器的多相位分解,推导出了该滤波器组的完全重构(PR)条件。基于该PR条件,我们将滤波器组的设计归结为一个关于原型滤波器的多相位分量的无约束优化问题。由于原型滤波器是线性相位的,多相位分量之间具有一定的关系,因此我们可以简化该优化问题。仿真结果验证了滤波器组PR条件的正确性。同时,仿真表明了优化算法的有效性,设计所得的滤波器组重构误差很小、频率特性较好,基本满足实际应用的需要。      

用于空间转换的巴克ISAR信号形成与图像重构

讨论了基于巴克相位编码调制发射信号和三维逆合成孔径雷达（ISAR）结构的ISAR概念。ISAR信号形成可理解为3D图像功能向2D信号功能的非对称空间转换,而图像重构被认为是2D信号功能向2D图像功能（重构的图像）的空间逆转换。经证实,这种图像重构算法由距离压缩互相关、方位压缩傅里叶变换（被认为是第一级运动补偿）和更高级相位校正组成（被认为是更高级运动补偿,通称自聚焦程序）。引入熵作为图像成本功能,以评估相位校正功能的多项式系数。通过数值实验以验证ISAR的几何结构,信号形成模式和图像重构算法。     

基于分块压缩感知的图像全局重构模型

已有的基于分块压缩感知（Block Compressed Sensing,Block CS）的图像重构模型采用相同的测量矩阵以块×块的方式获取数据,解决了传统CS方法中测量矩阵所需存储量较大的问题,但由于采用分块重构,没有考虑到图像的全局稀疏度,出现了大量的块效应。本文分析了图像分块重构产生块效应的三个主要原因:块稀疏度不均匀、频谱泄漏和块尺寸受限,提出了一种基于Block CS的图像全局重构模型。该模型在编码端采用高斯随机矩阵逐块作非相关测量;在解码端,引入排序算子,重新构造测量矩阵,该测量矩阵既适合于进行全局重构,又适合于分块测量的CS观测值,并仍与图像的稀疏矩阵高度不相关,所以其可充分利用图像的全局稀疏度进行CS重构。仿真实验表明,所提出的全局重构模型有效地消除了块效应现象,并且对块尺寸的变化有较强的鲁棒性。      

New insight at level set & Gaussian mixture model for natural image segmentation

Level set method and Gaussian Mixture model (GMM) are two very valuable tools for natural image segmentation. The former aims to acquire good geometrical continuity of segmentation boundaries, while the latter focuses on analyzing statistical properties of image feature data. Some studies on the integration between them have been reported due to their complementarity in the last 10 years. However, these studies generally supposed that the image-featured data density distribution of every segmented domain is independent with each other and can be separately approximated by Gaussian model or GMM, which conflicts with the fundamental idea of GMM clustering-based image segmentation. To remedy this problem, we give a new insight at image segmentation objective under the combined framework between Bayesian theory and GMM density approximation. Thereby, a novel level set image segmentation method integrated with GMM (GMMLS) is proposed. Then, the theoretical analysis on GMMLS is given, in which some valuable results are demonstrated. At last, several types of natural image segmentation experiments are reported and the corresponding results indicate that GMMLS can obtain better or at least equivalent performance compared with existing relevant methods in almost all cases.     

Feature fusion via dual-resolution compressive measurement matrix analysis for spectral image classification

In the compressive spectral imaging (CSI) framework, different architectures have been proposed to recover high-resolution spectral images from compressive measurements. Since CSI architectures compactly capture the relevant information of the spectral image, various methods that extract classification features from compressive samples have been recently proposed. However, these techniques require a feature extraction procedure that reorders measurements using the information embedded in the coded aperture patterns. In this paper, a method that fuses features directly from dual-resolution compressive measurements is proposed for spectral image classification. More precisely, the fusion method is formulated as an inverse problem that estimates high-spatial-resolution and low-dimensional feature bands from compressive measurements. To this end, the decimation matrices that describe the compressive measurements as degraded versions of the fused features are mathematically modeled using the information embedded in the coded aperture patterns. Furthermore, we include both a sparsity-promoting and a total-variation (TV) regularization terms to the fusion problem in order to consider the correlations between neighbor pixels, and therefore, improve the accuracy of pixel-based classifiers. To solve the fusion problem, we describe an algorithm based on the accelerated variant of the alternating direction method of multipliers (accelerated-ADMM). Additionally, a classification approach that includes the developed fusion method and a multilayer neural network is introduced. Finally, the proposed approach is evaluated on three remote sensing spectral images and a set of compressive measurements captured in the laboratory. Extensive simulations show that the proposed classification approach outperforms other approaches under various performance metrics.     

Image denoising via patch-based adaptive Gaussian mixture prior method

Although the expected patch log likelihood (EPLL) achieves good performance for denoising, an inherent nonadaptive problem exists. To solve this problem, an adaptive learning is introduced into the EPLL in this paper. Inspired from the structured sparse dictionary, an adaptive Gaussian mixture model (GMM) is proposed based on patch priors. The maximum a posteriori estimation is employed to cluster and update the image patches. Also, the new image patches are used to update the GMM. We perform these two steps alternately until the desired denoised results are achieved. Experimental results show that the proposed denoising method outperforms the existing image denoising algorithms.     

Automatic image annotation based on Gaussian mixture model considering cross-modal correlations

Automatic image annotation has been an active topic of research in the field of computer vision and pattern recognition for decades. In this paper, we present a new method for automatic image annotation based on Gaussian mixture model (GMM) considering cross-modal correlations. To be specific, we first employ GMM fitted by the rival penalized expectation-maximization (RPEM) algorithm to estimate the posterior probabilities of each annotation keyword. Next, a label similarity graph is constructed by a weighted linear combination of label similarity and visual similarity by seamlessly integrating the information from both image low level visual features and high level semantic concepts together, which can effectively avoid the phenomenon that different images with the same candidate annotations would obtain the same refinement results. Followed by the rank-two relaxation heuristics over the built label similarity graph is applied to further mine the correlation of the candidate annotations so as to capture the refining annotation results, which plays a crucial role in the semantic based image retrieval. The main contributions of this work can be summarized as follows: (1) Exploiting GMM that is trained by the RPEM algorithm to capture the initial semantic annotations of images. (2) The label similarity graph is constructed by a weighted linear combination of label similarity and visual similarity of images associated with the corresponding labels. (3) Refining the candidate set of annotations generated by the GMM through solving the max-bisection based on the rank-two relaxation algorithm over the weighted label graph. Compared to the current competitive model SGMM-RW, we can achieve significant improvements of 4% and 5% in precision, 6% and 9% in recall on the Corel5k and Mirflickr25k, respectively.     

基于自适应小生境混合遗传算法的说话人识别

为了解决传统高斯混合模型(Gaussian Mixture Model,GMM)对初值敏感,在实际训练中极易得到局部最优参数的问题,本文提出了一种GMM参数优化的新方法.将小生境技术与最大似然估计融入到遗传训练过程,形成了一种新的混合算法,缓解了遗传算法产生的"早熟"现象,提高了算法的局部搜索能力.采用自适应策略来控制交叉和变异算子,同时在适应度评价中融入了其他用户的区分性信息,提高了模型的分类精度,增强了GMM的泛化能力.实验表明,与传统和改进的两种方法相比,本文的方法都可以得到更优的模型参数,使得系统的识别率进一步提高.     

Multiscale fragile watermarking based on the Gaussian mixture model.

In this paper, a new multiscale fragile watermarking scheme based on the Gaussian mixture model (GMM) is presented. First, a GMM is developed to describe the statistical characteristics of images in the wavelet domain and an expectation-maximization algorithm is employed to identify GMM model parameters. With wavelet multiscale subspaces being divided into watermarking blocks, the GMM model parameters of different watermarking blocks are adjusted to form certain relationships, which are employed for the presented new fragile watermarking scheme for authentication. An optimal watermark embedding method is developed to achieve minimum watermarking distortion. A secret embedding key is designed to securely embed the fragile watermarks so that the new method is robust to counterfeiting, even when the malicious attackers are fully aware of the watermark embedding algorithm. It is shown that the presented new method can securely embed a message bit stream, such as personal signatures or copyright logos, into a host image as fragile watermarks. Compared with conventional fragile watermark techniques, this new statistical model based method modifies only a small amount of image data such that the distortion on the host image is imperceptible. Meanwhile, with the embedded message bits spreading over the entire image area through the statistical model, the new method can detect and localize image tampering. Besides, the new multiscale implementation of fragile watermarks based on the presented method can help distinguish some normal image operations such as JPEG compression from malicious image attacks and, thus, can be used for semi-fragile watermarking.     

反射式空间光调制器的相位调制特性测试

为了测试德国HOLOEYE公司LC-R2500型102像素×768像素反射式空间光调制器的纯相位调制特性,采用了基于双光束干涉原理的测量方法,得到了输入图像灰度与相移量对应关系.将相位恢复算法编码后的纯相位图像作为空间光调制器的输入信息,进行了理论分析和实验验证.结果表明,光学再现像与数值模拟结果相吻合,验证了纯相位调制特性测量的准确性.     

一种基于特征值和级联聚类的协作频谱感知方法

为了提高低信噪比(Signal-to-Noise Ratio, SNR)下频谱感知的性能，使用模糊C均值(Fuzzy C-means, FCM)和高斯混合模型(Gaussian Mixture Model, GMM),提出了一种基于特征值和级联聚类的协作频谱感知方法。从接收信号的协方差矩阵中提取特征值构造特征向量，通过在三维空间中执行聚类得到信道是否可用的分类模型，此过程无需获得主用户(Primary User, PU)信号以及噪声功率的先验信息，避免了复杂的门限计算。FCM聚类用于优化GMM聚类的初始参数，有效解决了在低SNR下GMM容易陷入局部最小值的问题。仿真结果表明，该方法降低了GMM的收敛时间并提高了模型分类的准确性，与其他主流方法相比能够有效提升频谱感知的性能。     

基于统计模型和KL距离的纹理图像检索

为了进一步提高纹理图像的检索性能,提出了一种基于统计模型离的纹理特征提取算法。根据小波分解的特点,从小波系数角度出发,以每个子带的小波系数系数直方图分布特性作为纹理特征,采用混合高斯模型和一般高斯模型分别对低频和高频信息进行描述,利用最大似然估计规则将特征提取和相似计算结合起来,采用KL距离进行度量。与一般高斯模型方法比较,该算法具有检索精度高等特点。理论分析和在纹理图像检索的对比实验数据说明该算法在纹理特征提取方面的性能较一般高斯模型方法提高了5%。