基于lp范数的压缩感知图像重建算法研究

图像重建是光学成像、光声成像、声纳成像、核磁共振成像、 天体成像等物理成像领域中的关键技术之一. 近年来提出的压缩感知理论指出: 对稀疏或者可压缩信号进行少量非自适应线性投影,投影信号含有足够的信息, 从而能对信号进行高概率重建. 压缩感知已被应用于多种物理成像系统. 将罚函数法和修正Hesse阵序列二次规划方法相结合, 并采用了分块压缩感知思想, 提出一种基于l_p范数的压缩感知图像重建算法. 以cameraman, barbara和mandrill图像为例, 采用该算法进行图像重建. 首先, 在不同采样率下对图像重建. 即便采样率低至0.3时, 也能获得高达32.23dB的信噪比, 重建图像清晰可辨. 验证了该算法的正确性. 其次, 将该算法与正交匹配追踪算法进行对比, 在采样率达到0.5以上时, 能够获得高信噪比的重建图像, 成像时间也大为减少, 特别是采样率为0.7时, 成像时间减少88%. 最后, 与现有基于l_p 范数的压缩感知图像重建算法进行对比, 计算结果表明在成像质量有所提高的基础上, 成像时间大为缩短. 关键词： 图像重建 压缩感知 罚函数 修正Hesse阵序列二次规划     

基于压缩感知与扩展小波树的自适应压缩成像

为了在现有的采样条件下,通过新的压缩采样方式获得计算量小且质量更好的图像,提出了基于压缩感知与扩展小波树的自适应压缩成像方法。首先将图像投影到分区控制的DMD上,获得图像在低分辨率下的测量值,并通过压缩感知重构算法重构出低分辨图像,接着利用扩展小波树预测重要小波位置,通过DMD在小波域采样获取图像的细节信息,最后由小波逆变换恢复高分辨率图像。将该方法与最小化全变分算法（TVAL3）和近来提出的基于扩展小波树的自适应成像算法（EWT-ACS）效果进行对比,实验结果表明,以boat图像为例,在压缩感知采样率为0.75,整体采样率为10%的无噪声条件下,该方法相较于TVAL3、EWT-ACS算法信噪比提高了4.63 dB和2.87 dB,在附加噪声条件下成像效果也较好。该方法能极大地降低压缩感知重建算法的运行时间,同时减少采样次数,具有较好的抗噪性。     

一种基于重叠块的遥感图像压缩算法

以大图像块或整个图像为处理单元的图像编码算法需要大量的内存来缓存图像,且编码过程中也会消耗大量内存,这种直接分块算法往往带来方块效应,影响图像的恢复质量。提出了以重叠块为单位的提升小波变换的方法,重叠分块可减小编码器对大块内存的需求,同时还可去除分块引入的方块效应。在变换中提出了多级并行分解方法,提高了分解效率。在对重叠块提升小波变换后的子带进行了统计分析,采用了DPCM与SPIHT相结合的方法。对直接分块、重叠分块、不分块算法进行了对比实验。结果表明,经重叠分块算法压缩的遥感图像具有较高的恢复质量。     

一种新的基于纹理和空间分布特征的图像检索

提出一种新的基于纹理和空间分布特征的图像检索方法.将检索图像分块,采用平移和尺度不变小波对各图像子块进行分解,在改进的快速小波直方图算法基础上提取图像子块的小波直方图,并提取每个图像子块的小波信息熵和三阶中心距作为纹理特征.对小波信息熵和中心矩特征进行高斯归一化,并利用特征向量的欧氏距离计算图像的纹理和空间特征的相似度.基于纹理图像库和自然图像库的检索试验表明,该方法比基于快速小波直方图算法和对数极坐标变换检索算法具有较高的检索准确度.     

双树小波变换与小波树稀疏联合的低场CS-MRI算法

压缩感知理论常用在磁共振快速成像上,仅采样少量的K空间数据即可重建出高质量的磁共振图像.压缩感知磁共振成像技术的原理是将磁共振图像重建问题建模成一个包含数据保真项、稀疏先验项和全变分项的线性组合最小化问题,显著减少磁共振扫描时间.稀疏表示是压缩感知理论的一个关键假设,重建结果很大程度上依赖于稀疏变换.本文将双树复小波变换和小波树稀疏联合作为压缩感知磁共振成像中的稀疏变换,提出了基于双树小波变换和小波树稀疏的压缩感知低场磁共振图像重建算法.实验表明,本文所提算法可以在某些磁共振图像客观评价指标中表现出一定的优势.     

基于自适应正则化的超分辨率重建方法

传统正则化超分辨重建得到的图像往往存在过度平滑或伪信息残留的问题,结合超分辨重建模型对重建图像伪信息的产生进行了分析,针对传统方法的不足提出了基于图像区域信息自适应的正则化方法,通过图像的区域信息将图像划分为平滑区与非平滑区域,对不同区域选用不同的先验模型进行约束。同时考虑人眼的视觉感知特性,结合区域信息实现正则化参数的自适应选取。实验结果表明该方法在抑制重建图像伪信息的同时能有效保护细节,效果要优于传统方法与单一的先验模型约束,对于红外与可见光图像重建效果的提升提供了一定的理论参考。     

基于信号相关性的自适应时域视频压缩感知重建方法

提出了一种基于信号相关性的自适应时域视频压缩感知重建方法,在超时间分辨率视频成像过程中自适应地判断物体的运动量并有针对性地重建信号。该方法将所观察到的图像分成不同运动量大小的区域,然后利用由视频样本训练得到的对应的字典重建这些区域;在视频重建阶段,将编码曝光图像快速分块重建,再计算各帧图像块之间的相关系数,通过相关系数估计局部图像运动量,根据估计的运动量选择训练字典并重建图像。仿真实验结果表明,该方法能准确地获得视频图像的运动分布信息,在降低重建时间的同时提高了重建质量。     

光谱去相关技术在高光谱图像小波压缩中的应用

随着数据量的不断增长,如何有效压缩高光谱图像成为影响其普及应用的一个关键问题。近年来,小波压缩技术已经被证明是高光谱图像压缩方法中很有发展前景的一个,但由于其对高光谱图像特性的利用较为有限而使其性能的进一步提升受到了限制。文章根据高光谱图像的光谱特征,提出了一种基于光谱去相关的高光谱图像小波压缩方法,设计了分块预测方法来同时去除光谱间相关性和空间相关性,并将其应用于小波压缩方法之中。首先,将高光谱图像分为几个具有高谱间相关性的图像块。然后推导出各块中波段的近似成比例的特性,并在各块分别进行基于这一特性和超光谱图像其他特性设计波段预测编码。最后,将预测用的参考波段和预测后获得的偏差数据,通过小波编码技术进行压缩。实验结果表明,所设计的方法与目前先进的超光谱压缩技术相比其性能有显著的提升。与AT-3DSPIHT算法比较,最高PSNR或SNR提升幅度均能达到4.2 dB左右。此外,此方法在低比特率下的优势也十分突出。     

分块稀疏信号1-bit压缩感知重建方法

1-bit压缩感知理论指出:对稀疏信号进行少量线性投影并对投影信号进行1-bit量化,该1-bit信号包含足够的信息,从而能对原始信号进行高精度重建.然而,当信号难以进行稀疏表达时,传统1-bit压缩感知算法无法精确重建原始信号.前期研究表明,分块稀疏模型作为一种特殊的结构型稀疏模型,对于难以用传统稀疏模型进行表达的信号具有较好的表达作用.本文提出了一种针对分块稀疏信号的1-bit压缩感知重建方法,该方法利用分块稀疏的统计特性对信号进行数学建模,通过变分贝叶斯推断方法进行信号重建并在光电容积脉搏波(photoplethysmography)信号上进行了实验验证.实验结果表明,与现有1-bit压缩感知重建方法相比,本文方法重建精度更高,且收敛速度更快.     

结合基于梯度的振铃评价算法的总变分最小化图像分块复原法

为了消除退化函数随空间变化发生变化模糊图像分块复原法子块之间的不平滑拼接缝,提出了一种结合了基于梯度的振铃评价算法梯度振铃评价(GRM)的总变分(TV)最小化分块复原法.根据图像分布及退化类型将模糊图像划分为矩形、环形或其他形状的子块,图像子块之间要留有一定的重叠区;然后对每一个图像子块进行复原,GRM方法是基于图像梯度结构相似度的图像质量评价算法,以GRM作为TV复原算法迭代过程中的收敛条件,可以更好地控制复原图像的振铃;最后去除复原图像子块含振铃波纹的重叠区,拼接得到完整图像.并以矩形分块及环形分块为例,证明该方法可以很好地抑制图像边界振铃效应,克服分块复原法本身的缺陷,得到拼接平滑的完整图像.     

Learning-based compressed sensing for infrared image super resolution

This paper presents an infrared image super-resolution method based on compressed sensing (CS). First, the reconstruction model under the CS framework is established and a Toeplitz matrix is selected as the sensing matrix. Compared with traditional learning-based methods, the proposed method uses a set of sub-dictionaries instead of two coupled dictionaries to recover high resolution (HR) images. And Toeplitz sensing matrix allows the proposed method time-efficient. Second, all training samples are divided into several feature spaces by using the proposed adaptive k-means classification method, which is more accurate than the standard k-means method. On the basis of this approach, a complex nonlinear mapping from the HR space to low resolution (LR) space can be converted into several compact linear mappings. Finally, the relationships between HR and LR image patches can be obtained by multi-sub-dictionaries and HR infrared images are reconstructed by the input LR images and multi-sub-dictionaries. The experimental results show that the proposed method is quantitatively and qualitatively more effective than other state-of-the-art methods.     

动态测量的高光谱图像压缩感知

针对高光谱图像相邻波段之间具有强光谱相关性的特点,为了提高高光谱图像压缩感知的重构效果,本文提出一种利用边缘信息设计动态测量率的压缩感知算法。首先,通过随机投影的分块压缩感知方法对每个图像块以固定测量率采样,重构出单波段图像作为其他波段的先验信息,并对其提取出图像边缘区域;然后,根据每个图像块边缘信息的丰富程度来自适应分配测量值。在固定总测量数的前提下,对不同图像块分配不同的测量次数。最后,利用分配好的测量次数对其余波段进行采集和重构。仿真结果表明,在相同总测量数情况下,本文提出的动态测量算法重构出的高光谱图像质量（PSNR）与传统固定测量压缩感知策略相比提高了1~4 dB,相比较下的重构时间也减少,在成功重构高光谱图像的基础上更增强了细节处的图像质量。     

Infrared super-resolution imaging based on compressed sensing

The theoretical basis of traditional infrared super-resolution imaging method is Nyquist sampling theorem. The reconstruction premise is that the relative positions of the infrared objects in the low-resolution image sequences should keep fixed and the image restoration means is the inverse operation of ill-posed issues without fixed rules. The super-resolution reconstruction ability of the infrared image, algorithm’s application area and stability of reconstruction algorithm are limited. To this end, we proposed super-resolution reconstruction method based on compressed sensing in this paper. In the method, we selected Toeplitz matrix as the measurement matrix and realized it by phase mask method. We researched complementary matching pursuit algorithm and selected it as the recovery algorithm. In order to adapt to the moving target and decrease imaging time, we take use of area infrared focal plane array to acquire multiple measurements at one time. Theoretically, the method breaks though Nyquist sampling theorem and can greatly improve the spatial resolution of the infrared image. The last image contrast and experiment data indicate that our method is effective in improving resolution of infrared images and is superior than some traditional super-resolution imaging method. The compressed sensing super-resolution method is expected to have a wide application prospect.     

Undersampled MRI reconstruction with patch-based directional wavelets

Compressed sensing has shown great potential in reducing data acquisition time in magnetic resonance imaging (MRI). In traditional compressed sensing MRI methods, an image is reconstructed by enforcing its sparse representation with respect to a preconstructed basis or dictionary. In this paper, patch-based directional wavelets are proposed to reconstruct images from undersampled k-space data. A parameter of patch-based directional wavelets, indicating the geometric direction of each patch, is trained from the reconstructed image using conventional compressed sensing MRI methods and incorporated into the sparsifying transform to provide the sparse representation for the image to be reconstructed. A reconstruction formulation is proposed and solved via an efficient alternating direction algorithm. Simulation results on phantom and in vivo data indicate that the proposed method outperforms conventional compressed sensing MRI methods in preserving the edges and suppressing the noise. Besides, the proposed method is not sensitive to the initial image when training directions.     

光子计数集成成像的压缩与重构

集成成像技术作为一种重要的裸眼三维显示技术,在完整记录三维场景信息的同时,庞大的数据量给传输和存储带来了压力。为了实现图像的有效压缩和重构,根据光子计数集成成像的特点,基于分布式压缩感知理论,提出用于图像压缩与重构的方案。该方案将图像分为参考图像和非参考图像两类,对其设置不同的测量率并分别进行重构。为保证非参考图像的重构质量,提出一种联合重构算法。该算法首先对非参考图像进行分块测量,依据与参考图像之间的相关性进行图像块分类,然后结合参考图像测量值信息构建新的测量矢量,利用新的测量矢量完成初次图像重构。为了进一步提升图像重构质量,对初次重构结果进行二次残差补偿重构,获得最终重构结果。最后通过设置不同的测量率进行了大量实验,实验结果表明,所提算法在测量率为0.25时,图像重构质量可以达到30 dB,测量率为0.4时,图像质量可以达到35 dB,算法性能具有一定的优越性。     

Image Encryption Scheme Based on Multiscale Block Compressed Sensing and Markov Model

Many image encryption schemes based on compressed sensing have the problem of poor quality of decrypted images. To deal with this problem, this paper develops an image encryption scheme by multiscale block compressed sensing. The image is decomposed by a three-level wavelet transform, and the sampling rates of coefficient matrices at all levels are calculated according to multiscale block compressed sensing theory and the given compression ratio. The first round of permutation is performed on the internal elements of the coefficient matrices at all levels. Then the coefficient matrix is compressed and combined. The second round of permutation is performed on the combined matrix based on the state transition matrix. Independent diffusion and forward-backward diffusion between pixels are used to obtain the final cipher image. Different sampling rates are set by considering the difference of information between an image’s low- and high-frequency parts. Therefore, the reconstruction quality of the decrypted image is better than that of other schemes, which set one sampling rate on an entire image. The proposed scheme takes full advantage of the randomness of the Markov model and shows an excellent encryption effect to resist various attacks.     

An adaptive fusion approach for infrared and visible images based on NSCT and compressed sensing

A novel nonsubsampled contourlet transform (NSCT) based image fusion approach, implementing an adaptive-Gaussian (AG) fuzzy membership method, compressed sensing (CS) technique, total variation (TV) based gradient descent reconstruction algorithm, is proposed for the fusion computation of infrared and visible images.Compared with wavelet, contourlet, or any other multi-resolution analysis method, NSCT has many evident advantages, such as multi-scale, multi-direction, and translation invariance. As is known, a fuzzy set is characterized by its membership function (MF), while the commonly known Gaussian fuzzy membership degree can be introduced to establish an adaptive control of the fusion processing. The compressed sensing technique can sparsely sample the image information in a certain sampling rate, and the sparse signal can be recovered by solving a convex problem employing gradient descent based iterative algorithm(s).In the proposed fusion process, the pre-enhanced infrared image and the visible image are decomposed into low-frequency subbands and high-frequency subbands, respectively, via the NSCT method as a first step. The low-frequency coefficients are fused using the adaptive regional average energy rule; the highest-frequency coefficients are fused using the maximum absolute selection rule; the other high-frequency coefficients are sparsely sampled, fused using the adaptive-Gaussian regional standard deviation rule, and then recovered by employing the total variation based gradient descent recovery algorithm.Experimental results and human visual perception illustrate the effectiveness and advantages of the proposed fusion approach. The efficiency and robustness are also analyzed and discussed through different evaluation methods, such as the standard deviation, Shannon entropy, root-mean-square error, mutual information and edge-based similarity index.     

基于谱间线性滤波的高光谱图像压缩感知

根据高光谱图像较强的谱间相关性,提出一种基于谱间线性滤波的高光谱图像压缩感知方法.高光谱图像进行压缩重构时,利用相邻波谱的谱间相关性,对重构的当前帧与前一谱段的重构图像进行谱间线性滤波,降低了重构帧的噪音信息,纠正了重构帧的轮廓信息,从而提高重构质量.在进行谱间线性滤波时,保留重构帧的低频系数,高频系数与前一波谱重构图像的高频小波变换系数进行线性加权求和,达到滤波的效果.通过实验表明,该方法能够有效提升图像重构质量,并降低重构时间.     

Image Encryption Scheme with Compressed Sensing Based on a New Six-Dimensional Non-Degenerate Discrete Hyperchaotic System and Plaintext-Related Scrambling

Digital images can be large in size and contain sensitive information that needs protection. Compression using compressed sensing performs well, but the measurement matrix directly affects the signal compression and reconstruction performance. The good cryptographic characteristics of chaotic systems mean that using one to construct the measurement matrix has obvious advantages. However, existing low-dimensional chaotic systems have low complexity and generate sequences with poor randomness. Hence, a new six-dimensional non-degenerate discrete hyperchaotic system with six positive Lyapunov exponents is proposed in this paper. Using this chaotic system to design the measurement matrix can improve the performance of image compression and reconstruction. Because image encryption using compressed sensing cannot resist known- and chosen-plaintext attacks, the chaotic system proposed in this paper is introduced into the compressed sensing encryption framework. A scrambling algorithm and two-way diffusion algorithm for the plaintext are used to encrypt the measured value matrix. The security of the encryption system is further improved by generating the SHA-256 value of the original image to calculate the initial conditions of the chaotic map. A simulation and performance analysis shows that the proposed image compression-encryption scheme has high compression and reconstruction performance and the ability to resist known- and chosen-plaintext attacks.