一种从粗到精的红外和可见光卫星图像配准方法

气象卫星所携带的多种传感器可以获得可见光、红外、多光谱等多模态的卫星图像,目前处理这些多模态图像的一个重要手段是数据融合分析方法,而获取不同模态图像空间对应关系的图像配准是数据融合分析的前提和基础。针对多模态气象卫星图像的配准问题,重点研究红外图像和可见光图像的配准问题,并根据红外图像和可见光图像的特点,提出了一种由粗到精的两阶段配准方法。在粗配准阶段,将Fourier-Mellin变换应用于红外和可见光图像的边缘图像上,并通过变换图像在频域的关系实现了图像配准仿射变换参数的快速计算;在精配准阶段,基于图像的Harris算子检测红外图像和可见光图像的特征点,并通过特征点局部区域的互相关函数实现特征点的匹配,最终通过匹配特征点求得精确配准的变换参数。文章提出的由粗到精的图像配准方法,有效结合了Fourier-Mellin变换对边缘图像配准的高效性和Harris算子图像配准的准确性,是红外和可见光图像配准的一种新方法。利用FY-2D气象卫星获取的红外和可见光图像进行了配准实验,实验结果表明所提出的方法具有良好的鲁棒性和较高的配准精度。     

Double image encryption based on phase-amplitude hybrid encoding and iterative phase encoding in fractional Fourier transform domains

A new method for double image encryption is proposed that is based on amplitude-phase hybrid encoding and iterative random phase encoding in fractional Fourier transform (FrFT) domains. In the iterative random phase encoding operation, a binary random matrix is defined to encode two original images to a single complex-valued image, which is then converted into a stationary white noise image by the iterative phase encoding with FrFTs. Compared with the previous schemes that uses fully phase encoding, the proposed method reduces the difference between two original images in key space and sensitivity to the FrFT orders. The primitive images can be retrieved exactly by applying correct keys with initial conditions of chaotic system, the pixel scrambling operation and the FrFT orders. Computer simulations demonstrate that the encryption method has impressively high security level and certain robustness against data loss and noise interference.     

Noise-Robust Image Reconstruction Based on Minimizing Extended Class of Power-Divergence Measures

The problem of tomographic image reconstruction can be reduced to an optimization problem of finding unknown pixel values subject to minimizing the difference between the measured and forward projections. Iterative image reconstruction algorithms provide significant improvements over transform methods in computed tomography. In this paper, we present an extended class of power-divergence measures (PDMs), which includes a large set of distance and relative entropy measures, and propose an iterative reconstruction algorithm based on the extended PDM (EPDM) as an objective function for the optimization strategy. For this purpose, we introduce a system of nonlinear differential equations whose Lyapunov function is equivalent to the EPDM. Then, we derive an iterative formula by multiplicative discretization of the continuous-time system. Since the parameterized EPDM family includes the Kullback–Leibler divergence, the resulting iterative algorithm is a natural extension of the maximum-likelihood expectation-maximization (MLEM) method. We conducted image reconstruction experiments using noisy projection data and found that the proposed algorithm outperformed MLEM and could reconstruct high-quality images that were robust to measured noise by properly selecting parameters.     

A statistical method for characterizing the noise in nonlinearly reconstructed images from undersampled MR data: The POCS example

The projection-onto-convex-sets (POCS) algorithm is a powerful tool for reconstructing high-resolution images from undersampled k-space data. It is a nonlinear iterative method that attempts to estimate values for missing data. The convergence of the algorithm and its other deterministic properties are well established, but relatively little is known about how noise in the source data influences noise in the final reconstructed image. In this paper, we present an experimental treatment of the statistical properties in POCS and investigate 12 stochastic models for its noise distribution beside its nonlinear point spread functions. Statistical results show that as the ratio of the missing k-space data increases, the noise distribution in POCS images is no longer Rayleigh as with conventional linear Fourier reconstruction. Instead, the probability density function for the noise is well approximated by a lognormal distribution. For small missing data ratios, however, the noise remains Rayleigh distributed. Preliminary results show that in the presence of noise, POCS images are often dominated by POCS-enhanced noise rather than POCS-induced artifacts. Implicit in this work is the presentation of a general statistical method that can be used to assess the noise properties in other nonlinear reconstruction algorithms.     

Dual-mode registration of dynamic contrast-enhanced ultrasound combining tissue and contrast sequences

This study proposes a new method for automatic, iterative image registration in the context of dynamic contrast-enhanced ultrasound (DCE-US) imaging. By constructing a cost function of image registration using a combination of the tissue and contrast-microbubble responses, this new method, referred to as dual-mode registration, performs alignment based on both tissue and vascular structures. Data from five focal liver lesions (FLLs) were used for the evaluation. Automatic registration based on the dual-mode registration technique and tissue-mode registration obtained using the linear response image sequence alone were compared to manual alignment of the sequence by an expert. Comparison of the maximum distance between the transformations applied by the automatic registration techniques and those from expert manual registration reference showed that the dual-mode registration provided better precision than the tissue-mode registration for all cases. The reduction of maximum distance ranged from 0.25 to 9.3 mm. Dual-mode registration is also significantly better than tissue-mode registration for the five sequences with p  -values lower than 0.03$0.03$. The improved sequence alignment is also demonstrated visually by comparison of images from the sequences and the video playbacks of the motion-corrected sequences. This new registration technique better maintains a selected region of interest (ROI) within a fixed position of the image plane throughout the DCE-US sequence. This should reduce motion-related variability of the echo-power estimations and, thus, contribute to more robust perfusion quantification with DCE-US.     

Two-step image registration by artificial immune system and chamfer matching

Image registration is the precondition and foundation in the fusion of multi-source image data. A two-step approach based on artificial immune system and chamfer matching to register images from different types of sensors is presented. In the first step, it extracts the large edges and takes chamfer distance between the input image and the reference image as similarity measure and uses artificial immune network algorithm to speed up the searching of the initial transformation parameters. In the second step, an area-based method is utilized to refine the initial transformation and enhance the registration accuracy. Experimental results show that the proposed approach is a promising method for registration of multi-sensor images.     

Accuracy of image registration between MRI and light microscopy in the ex vivo brain

A multistep procedure was developed to register magnetic resonance imaging (MRI) and histological data from the same sample in the light microscopy image space, with the ultimate goal of allowing quantitative comparisons of the two datasets. The fixed brain of an owl monkey was used to develop and test the procedure. In addition to the MRI and histological data, photographic images of the brain tissue block acquired during sectioning were assembled into a blockface volume to provide an intermediate step for the overall registration process. The MR volume was first registered to the blockface volume using a combination of linear and nonlinear registration, and two dimensional (2D) blockface sections were registered to corresponding myelin-stained sections using a combination of linear and nonlinear registration. Before this 2D registration, two major types of tissue distortions were corrected: tissue tearing and independent movement of different parts of the brain, both introduced during histological processing of the sections. The correction procedure utilized a 2D method to close tissue tears and a multiple iterative closest point (ICP) algorithm to reposition separate pieces of tissue in the image. The accuracy of the overall MR to micrograph registration procedure was assessed by measuring the distance between registered landmarks chosen in the MR image space and the corresponding landmarks chosen in the micrograph space. The average error distance of the MR data registered to micrograph data was 0.324±0.277 mm, only 8% larger than the width of the MRI voxel (0.3 mm).     

Registration method for infrared images under conditions of fixed-pattern noise

This paper proposes a new registration method for infrared images under conditions of fixed-pattern noise (FPN). Conventional registration techniques are susceptible to FPN and it is therefore very desirable to have a registration algorithm that is tolerant to FPN. For this purpose, we utilize the difference of the cross-power spectrum of two discrete shifted images to suppress the noise power spectrum while the shifts information is well preserved. In particular, we show that the phase of the cross-power spectrum difference is a periodic two-dimensional binary stripe signal with the exact shifts determined to subpixel accuracy by the number of periods of the phase difference along each frequency axis. Robust estimates of shifts can be obtained by transforming its discontinuities to Hough domain. Experimental results show that the proposed method exhibits robust and accurate registration performance even for the noisy images that could not be handled by conventional registration algorithms. We have also incorporated this technique to a registration-based nonuniformity correction (NUC) framework, indicating that our registration technique is able to estimate motion parameters reliably, leading to satisfactory NUC result.     

基于特征匹配的快速鲁棒数字稳像

针对手持移动摄像装置拍摄视频序列相邻帧间存在平移、小角度旋转运动,而且易受噪声、光照变化的影响等问题,提出一种基于优化Oriented FAST and rotated BRIEF(ORB)特征匹配的实时鲁棒电子稳像算法。对相邻帧预处理后用Oriented FAST算子检测特征点,再用Rotated BRIEF描述提取的特征点并采用近邻汉明距离匹配特征点对,然后采用级联滤波去除误匹配点对,最后使用迭代最小二乘法(ILSM)拟合模型参量进行运动补偿实现稳像。图像匹配测试和稳像实验结果表明:基于改进的ORB算法的电子稳像方法补偿每一帧的时间均小于0.1 s,定位精度可达亚像素级,能有效补偿帧间平移旋转运动,而且对噪声和光照变化有较强鲁棒性。经稳像处理后,实拍视频质量明显提高,峰值信噪比（PSNR）平均提高了10 db。     

基于小波变换和数据融合技术的弱小目标检测

 鉴于弱小目标检测所固有的难点及常用的单一分辨率下的检测方法还不能准确稳定地检测出目标，提出了一种弱小目标检测新方法。考虑到实际应用中的复杂背景和大量干扰噪声，运用数据融合技术，先对图像进行小波多分辨率分解，然后将不同分辨率下的子图进行最优加权平均融合来检测弱小目标。用实地拍摄的空中弱小目标红外和可见光图像分别进行实验验证，实验图像取256×256像素点阵大小，其中目标占10×10像素左右。结果表明该方法能够准确稳定地检测弱小目标，为后续的跟踪作了很好的铺垫。     

Performance of iterative gradient-based algorithms with different intensity change models in digital image correlation

Digital image correlation (DIC) is a non-contact and powerful tool for whole-field displacement and strain measurement in modern optical metrology. In the DIC technique, the gradient-based algorithm is one of the most commonly used sub-pixel registration algorithms due to its effectiveness and accuracy. Thus, this algorithm has been further investigated and improved by many researchers to increase its accuracy. In the existing gradient-based algorithms, there are three relation models for the gray level intensity of a point in the undeformed and deformed images: namely the constant intensity model, the linear intensity change model and the non-linear intensity change model. However, little quantitative investigation has been conducted to compare their performance. In this paper, three iterative gradient-based algorithms are given using the corresponding intensity change model and the iterative least squares (ILS) solving method. The accuracy, robustness and computational efficiency of the three algorithms for intensity variation are investigated through numerical simulation experiments. The experimental results reveal that the algorithm generated using the non-linear intensity change model is the most accurate, robust and efficient of the three algorithms and the algorithms generated using the linear and non-linear intensity change models have approximately the same accuracy when the intensity variation is small between undeformed and deformed images.     

Wavelet-based edge correlation incorporated iterative reconstruction for undersampled MRI

Undersampling k-space is an effective way to decrease acquisition time for MRI. However, aliasing artifacts introduced by undersampling may blur the edges of magnetic resonance images, which often contain important information for clinical diagnosis. Moreover, k-space data is often contaminated by the noise signals of unknown intensity. To better preserve the edge features while suppressing the aliasing artifacts and noises, we present a new wavelet-based algorithm for undersampled MRI reconstruction. The algorithm solves the image reconstruction as a standard optimization problem including a ?₂ data fidelity term and ?₁ sparsity regularization term. Rather than manually setting the regularization parameter for the ?₁ term, which is directly related to the threshold, an automatic estimated threshold adaptive to noise intensity is introduced in our proposed algorithm. In addition, a prior matrix based on edge correlation in wavelet domain is incorporated into the regularization term. Compared with nonlinear conjugate gradient descent algorithm, iterative shrinkage/thresholding algorithm, fast iterative soft-thresholding algorithm and the iterative thresholding algorithm using exponentially decreasing threshold, the proposed algorithm yields reconstructions with better edge recovery and noise suppression.     

Causal dynamic MRI reconstruction via nuclear norm minimization

This work addresses the problem of online reconstruction of dynamic magnetic resonance images (MRI). The proposed method reconstructs the difference between the images of previous and current time frames. This difference image is modeled as a rank deficient matrix and is solved from the partially sampled k-space data via nuclear norm minimization. Our proposed method has been compared against state-of-the-art offline and online reconstruction methods. Our method has similar reconstruction accuracy as the offline method and significantly higher accuracy compared to the online technique. It is about an order of magnitude faster than the online technique compared against. Our experimental data consisted of dynamic MRI data that were collected at 6 to 7 frames per second and having resolutions of 128×128 and 256×256 pixels per frame. Experimental evaluation indicates that our proposed method is capable of reconstructing 128×128 images at the rate of 4 frames per second and 256×256 images at the rate of 2 frames per second. The previous online method requires about 3.75 s for reconstructing each image. The improvement in reconstruction speed is clearly discernible. Moreover, our method has a reconstruction error that is about half that of the previous online method.     

扫频光学相干层析视网膜图像配准去噪算法

多帧叠加平均处理是去除扫频光学相干层析系统散斑噪声、获得较为清晰结构信息的有效方法,但眼睛的震颤、漂移、微眼跳等生理特性和系统光路特性会使图像之间存在错位,导致叠加效果不佳、结构稳定性差,为此本文提出一种基于灰度分布信息和目标几何信息相结合的配准算法。该方法根据图像平均灰度分布提取包含目标信息的感兴趣区域,通过相位相关算法和基于分段拟合的灰度投影算法的双重作用校正图像的平移变换;通过拟合视网膜上边界作为特征点迭代确定最佳旋转参数,并再次重新估计平移参数,实现图像的刚性配准;最后通过轴向扫描一对一映射法以能量函数为约束条件实现图像的非刚性配准。对活体兔眼进行实验,结果表明,本文算法配准后的叠加图像边界清晰,结构信息增强,信噪比和对比度平均有效提高一倍多。本算法适用于强噪声视网膜B-Scans图像的配准,能满足多种类型OCT系统的叠加成像需要,具有较高的鲁棒性和图像配准精度。     

改进SIFT变换与客观评价结合的图像配准算法

针对SIFT(Scale Invariant Feature Transform)算子在大幅复杂图像中提取的过多不稳定特征点及在只有少量重合区域下图像配准过程中出现的过多误匹配,导致图像配准精度下降。提出一种改进的SIFT算法,在对目标图像提取SIFT特征后,利用双向BBF(Best-Bin-First)匹配算法对提取的特征点进行匹配,采用SIFT描述子的尺度以及梯度方向信息建立最小邻域匹配剔除误匹配点,通过随机抽取一致性算法(RANSAC)进一步筛选匹配点,并利用最小二乘法结合多项式近似拟合出变换模型,利用局部均方根有效值(RMS)评价映射矩阵与实际图像的误差,找出并删除引起误差的误匹配点,迭代至配准图像符合评价标准后,计算出精确变换模型。实验结果表明,该算法提高了大幅复杂图像在少量重合区域时的配准精度。     

基于变分推断的磁共振图像群组配准

为解决基于深度学习的成对配准方法精度低和传统配准算法耗时长的问题,本文提出一种基于变分推断的无监督端到端的群组配准以及基于局部归一化互相关（NCC）和先验的配准框架,该框架能够将多个图像配准到公共空间并有效地控制变形场的正则化,且不需要真实的变形场和参考图像．该方法得到的预估变形场可建模为概率生成模型,使用变分推断的方法求解;然后借助空间转换网络和损失函数来实现无监督方式训练．对于公开数据集LPBA40的3D脑磁共振图像配准任务,测试结果表明：本文所提出的方法与基线方法相比,具有较好的Dice得分、运行时间少且产生更好的微分同胚域,同时对噪声具有鲁棒性．     

Super-resolution reconstruction face recognition based on multi-level FFD registration

Super-resolution (SR) reconstruction is an effective method to solve the problem, that the face image resolution is too low to be recognized in video, but the non-rigid change of deformed face and expression changes greatly affect the accuracy of registration and reconstruction. To solve these problems, a method of multi-level model free form deformation (FFD) elastic registration algorithm based on B spline is proposed. It first use low-resolution FFD grid for global registration, to emphasize the contribution of edge information for registration, we introduce edge registration measure into the sum of squared difference (SSD) criterion. Then divide the global registration image and reference image into a series of corresponding sub-image pairs and calculate the correlation coefficient of each pair; at the same time, we do local registration with high-resolution FFD grid to the small value correlation coefficient sub-image pairs. In the registration process of optimization, the paper uses adaptive step length gradient descent method algorithm based on chaotic variables to improve optimization efficiency. After registration, the algorithm of project onto convex sets (POCS) is used to reconstruct SR face image through several low resolution image sequences, and then recognized these SR face images by support vector machines (SVM) classifier. Experimental results from standard video database and self-built video database show that this method can register and reconstruct face image accurately in the condition of great face deformation and expression change, while the face recognition accuracy is also improved.     

A subpixel motion estimation algorithm based on digital correlation for illumination variant and noise image sequences

One of the challenges in practical subpixel motion estimation is how to obtain high accuracy with sufficient robustness to both illumination variations and additive noise. Motivated by the fact that the normalized spatial cross-correlation is invariant to illumination, we introduce a gradient-based subpixel registration method by maximizing the digital correlation (DC) function between the reference and target frames. Such DC function is remodeled with the presence of image noise, yielding that the correlation coefficient is only sensitive to noise standard variance. To fairly suppress the noise corruption, not only the target frame but also the reference one is reformulated into Taylor gradient expression with half but opposite motion vector. The final solution to motion estimates can be approximated into a closed form by reserving first-order coefficient terms of unregistered motion variables. The error trend of approximated solution is discussed. Computer simulations and actual experiments’ results demonstrate the superiority of the proposed method to the LMSE-based method and ordinary DC method when illumination variations and noise exist. Among the experiments, the influences of real subpixel translation value and noise variance degree on accuracy are studied; correspondingly, an optimized iterative idea for big translations and the recommended noise level adaptive to our method are introduced.     

基于边缘与SURF算子的SAR与可见光图像配准方法

鉴于SAR(synthetic aperture radar)与可见光图像的成像机理存在很大差别,使得其同名特征的提取和配准十分困难,但在某些情况下,这两类图像的边缘存在一定的相关性。提出一种基于边缘与SURF(speed-up robust feature)算子的图像配准方法。通过适当预处理增强图像间的共性,采用综合性能比较好的Canny算子提取两幅图像共有的边缘特征,在边缘图像的基础上提取SURF特征;通过比值提纯法进行特征点粗匹配,RANSAC(random sample consensus)算法剔除误匹配点,计算仿射变换模型从而实现SAR与可见光图像的自动配准。实验结果表明:该算法的正确匹配率为100%,均方根误差为0.852个像素,配准精度达到亚像素水平。