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。     

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。     

基于迭代的傅里叶望远镜图像重构方法

为了提高最终图像的质量,结合传统的直接重构算法和相位重构算法,提出了一种基于迭代的傅里叶望远镜图像重构方法。对反演得到的傅里叶频谱进行逆傅里叶变换得到直接重构图像,取直接重构图像的阈值图像作为相位重构算法的初始输入,通过迭代能够获得质量更高的图像。成像仿真实验结果表明,成像信噪比为50倍时,与直接重构图像相比,迭代重构图像斯特列尔比从0.82提高到0.88,峰值信噪比从17dB提高到19dB;在成像信噪比为100倍时,迭代重构图像斯特列尔比从0.89提高到0.93,峰值信噪比从20dB提高到22dB。     

一种基于编码孔径的压缩传感图像重构算法的改进

针对压缩传感理论应用于实际系统成像时重构图像质量随图像采样率变化的问题,通过对正交匹配追踪算法进行改进,提出了一种利用空间光调制器实现编码孔径成像的压缩传感图像重构方法。该方法对传统的正交匹配追踪法迭代计算中已选入支撑集的列向量进行标记,并在下一次迭代计算中予以排除,从而减少了重构时间。在此基础上提出了将测量矩阵分别按行和按列排列进行重构后平均的图像增强算法。增强算法在达到同样重构质量时,减小了图像采样率,有利于图像数据的传输和存储。仿真实验验证了方法的有效性和稳定性,可为压缩传感技术的应用提供技术参考。     

基于压缩感知的三维单分子定位显微成像方法研究

本文建立了一种三维压缩感知模型以实现对高密度荧光分子图像的快速三维定位。首先,根据荧光显微的三维点扩展函数成像理论,设计测量矩阵,并建立压缩感知模型。接着,对荧光显微成像过程进行了模拟,并采用凸优化方法(CVX)、正交匹配追踪(Orthogonal Matching Pursuit,OMP)算法和同伦算法对建立的压缩感知模型中模拟生成的图像进行了定位分析,分别从恢复率、定位精度、重构时间几方面进行了对比。最后,采用同伦算法对模拟的生物样品和实验室采集的细胞进行了三维定位,并获得了三维超分辨图像。对比结果表明:在重构密度和定位精度接近的情况下,同伦算法比CVX方法的重构速度快2个数量级。同伦算法较OMP算法的定位精度要高一倍。采用同伦算法来实现三维的超分辨荧光显微成像在节约计算时间、实现实时成像方面具有一定的意义。     

基于邻域相似度的压缩感知鬼成像

为了提高压缩感知鬼成像的成像质量以及解决低采样率条件下成像失真度高的问题,提出一种基于邻域相似度的鬼成像(NSGI)方案。邻域相似度体现在图像像素间的关联性,携带关于物体结构的重要信息,在分析压缩鬼成像原理的基础上,利用邻域相似度来评价待探测目标。根据贪婪算法的原理,采用邻域相似度优化图像重构过程,并设置相关度阈值降低计算的复杂度。仿真和实验结果均表明,与传统方法相比,该方案可以在低采样率条件下获得高质量低失真度的图像,有利于推动鬼成像技术的实用化。     

基于压缩感知成像系统的动态背景去噪

为了消除背景光强动态变化引起的非线性误差,提出了基于压缩感知成像系统的动态背景去噪算法.算法将测量中不同背景光强下所得测量值的平均值之差作为补偿系数,通过补偿系数消除动态背景噪声对压缩感知成像系统的非线性影响.仿真结果表明,在总数为900次的采样过程中,背影噪声动态变化300次时,算法能够将重建图像的峰值信噪比由29.5dB提高到62dB;在动态背景噪声的影响下,本文算法能够大大提高目标图像的可读性,提高成像质量,增加压缩感知成像系统的鲁棒性.     

双臂对称性对关联成像算法的影响

为了分析在实际成像过程中双臂误差对不同关联成像算法的影响。理论上推导了双臂横向误差和轴向误差对传统关联成像算法(Ghost Imaging,GI),差分关联成像(Differential Ghost Imaging,DGI)和基于压缩感知的关联成像算法(Compressive Sensing Ghost Imaging,CSGI)的影响;在不同的双臂横向误差和轴向误差下利用三种算法分别对灰度图像Lena和二值图像Stripes进行了重构仿真实验,并将重构结果的峰值信噪比(peak signal to noise ratio,PSNR)进行对比。当双臂横向行列均错位4个像素点时,Lena的GI重构结果PSNR最多下降0.766dB,DGI下降2.928dB,CSGI下降5.858dB;Stripes的GI重构结果PSNR最多降低0.644 2dB,DGI降低0.943dB,CSGI降低5.975dB;随着双臂轴向误差的增大,Lena的GI重构最多降低了0.895dB,DGI降低4.751dB,CSGI降低7.005dB;Stripes的GI重构PSNR最多降低1.126dB,DGI降低2.913dB,CSGI降低6.43dB。实验结果表明,三种算法在抗双臂误差能力方面,GI表现最好,DGI其次,CSGI最弱。     

一种数字微镜阵列分区控制和超分辨重建的压缩感知成像法

提出一种压缩感知成像框架结构.该结构采样端用新建的采样矩阵实现数字微镜阵列分区控制,可增强信息获取的准确性,测量得到与新数字微镜阵列对应的压缩采样值;重构端由采样值优化重构出低分辨率图像后,根据分区控制过程建立压缩感知理论框架下的超分辨重建模型,利用梯度稀疏约束优化算法进行求解,恢复出原高分辨率图像.实验结果表明:数字微镜阵列分区控制与超分辨重建相结合的方法可以明显降低压缩感知成像系统的计算量,缩短成像时间,并且具有较高的图像重构质量.     

基于卡尔曼滤波的双约束CUP-VISAR压缩图像重构算法

针对从基于压缩超快成像（Compressed Ultrafast Photography,CUP）的任意反射面速度干涉仪（Velocity Interferometer System for Any Reflector,VISAR）中获得的压缩图像中重构出冲击波二维条纹图像的问题,提出一种基于卡尔曼滤波的双约束图像重构算法。该算法首先基于条纹图像具有的稀疏性和平滑性,将问题转化为基于小波与全变分双先验约束的优化问题,然后,考虑到实际成像的噪声问题,采用加权卡尔曼滤波对图像已有信息进行预测和调整,最后将卡尔曼滤波引入二步迭代阈值算法的迭代过程中,进而求解该双约束优化问题,实现压缩图像的精确重构。在大噪声仿真实验中,该算法重构图像的峰值信噪比和结构相似度分别提高了4.8 dB和14.81%,显著提高了图像重构质量。在实际实验中,该算法重构出了清晰的冲击波条纹图像,且将冲击波速度最大相对误差降低了9.57%和平均相对误差降低了2.2%,验证了该算法的可行性。     

Computational integral imaging reconstruction scheme of far 3D objects by additional use of an imaging lens

In this paper, we propose a novel performance-enhanced computational integral imaging reconstruction (CIIR) scheme by additional use of an imaging lens. In the proposed scheme, elemental images can be obtained by using a simultaneous pickup scheme of far three-dimensional (3D) objects from the lenslet array in both real and virtual image fields. And additional imaging lens produces an image shift effect of 3D objects located far away from the lenslet array and improve the visual quality of reconstructed images in CIIR by overcoming limitation of pickup range in integral imaging. To show the usefulness of the proposed system, some experiments are carried out for real 3D objects and its results are presented.     

A 3D image encryption technique using computer-generated integral imaging and cellular automata transform

A novel three-dimensional (3D) image encryption approach by using the computer-generated integral imaging and cellular automata transform (CAT) is proposed, in which, the two-dimensional (2D) elemental image array (EIA) digitally recorded by light rays coming from the 3D image is mapped inversely through the virtual pinhole array according to the ray-tracing theory. Next, the encrypted image is generated by using the 2D CAT scrambling transform for the 2D EIA. The reconstructed process is carried out by using the modified computational integral-imaging reconstruction (CIIR) technique; the depth-dependent plane images are reconstructed on the output plane. The reconstructed 3D image quality of the proposed scheme can be greatly improved, because the proposed encryption scheme carries out in a computer which can avoid the light diffraction caused by optical device CIIR, and solves blur problem caused by CIIR by using the pixel-averaging algorithm. Furthermore, the CAT-based encryption algorithm is an error-free encryption method; CAT as an orthogonal transformation offers considerable simplicity in the calculation of the transform coefficient, that is, it can improve the quality of the reconstructed image by reducing energy loss compared with the traditional complicated transform process. To show the effectiveness of the proposed scheme, we perform computational experiments. Experimental results show that the proposed scheme outperforms conventional encryption methods.     

Resolution-enhanced 3D image correlator using computationally reconstructed integral images

In this paper, a novel resolution-enhanced three-dimensional (3D) image correlator using the computationally reconstructed integral images is proposed in order to extract target object’s 3D location data in a scene. Elemental images of the reference and target objects are picked up by lenslet arrays and using these elemental images, reference and target plane images are reconstructed on the output plane by means of a modified computational integral imaging reconstruction technique. Then, through cross-correlations between the reconstructed reference and the target plane images, 3D location data of the target object can be extracted from the correlation outputs. With the purpose of showing the feasibility of the proposed method, some computational and optical experiments on the target objects in space are carried out and the results are presented.     

A novel three-dimensional digital watermarking scheme basing on integral imaging

A novel integral imaging-based three-dimensional (3D) digital watermarking scheme is presented. In the proposed method, an elemental image array (EIA) obtained by recording the rays coming from a 3D object through a pinhole array in the integral imaging system is employed as a new 3D watermark. The EIA is composed of a number of small elemental images having their own perspectives of a 3D object, and from this recorded EIA various depth-dependent 3D object images can be reconstructed by using the computational integral imaging reconstruction (CIIR) technique. This 3D property of the EIA watermark can make a robust reconstruction of the watermark image available even though there are some data losses in the embedded watermark by attacks. To show the robustness of the proposed scheme against attacks, some experiments are carried out and the results are discussed.     

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     

Three-dimensional visualization of objects in scattering medium using integral imaging and spectral analysis

This paper presents a three-dimensional visualization method of 3D objects in a scattering medium. The proposed method employs integral imaging and spectral analysis to improve the visual quality of 3D images. The images observed from 3D objects in the scattering medium such as turbid water suffer from image degradation due to scattering. The main reason is that the observed image signal is very weak compared with the scattering signal. Common image enhancement techniques including histogram equalization and contrast enhancement works improperly to overcome the problem. Thus, integral imaging that enables to integrate the weak signals from multiple images was discussed to improve image quality. In this paper, we apply spectral analysis to an integral imaging system such as the computational integral imaging reconstruction. Also, we introduce a signal model with a visibility parameter to analyze the scattering signal. The proposed method based on spectral analysis efficiently estimates the original signal and it is applied to elemental images. The visibility-enhanced elemental images are then used to reconstruct 3D images using a computational integral imaging reconstruction algorithm. To evaluate the proposed method, we perform the optical experiments for 3D objects in turbid water. The experimental results indicate that the proposed method outperforms the existing methods.     

针对多散斑图的差分压缩鬼成像方案研究

鬼成像方案实现所需设备、成像的质量以及成像所花的时间是决定鬼成像技术可实用化的重要因素. 本文提出一种针对多散斑图的差分压缩鬼成像方案. 该方案通过连续探测多个独立的散斑图, 降低了热光鬼成像方案对探测器高时间分辨力的要求; 通过采用差分方法, 抑制了背景噪声和其他噪声源的干扰; 通过使用压缩感知重建算法, 有效地降低了鬼成像所需时间并同时提升成像的质量. 数值仿真结果表明, 对于二灰度“N” 图, 本方案在8000次的采样情形下与多散斑图鬼成像方案35000次采样的结果相比, 均方误差降低了96.9%、峰值信噪比提升15.1 dB. 对于八灰度“Pepper”图, 本方案与多散斑图鬼成像方案相比, PSNR提升11.4 dB. 本方案可降低探测设备的要求、提高成像质量、降低重建时间, 具有广阔的应用前景.     

Compressed sensing MR image reconstruction using a motion-compensated reference

Compressed sensing (CS)-based methods have been proposed for image reconstruction from undersampled magnetic resonance data. Recently, CS-based schemes using reference images have also been proposed to further reduce the sampling requirement. In this study, we propose a new reference-constrained CS reconstruction method that accounts for the misalignment between the reference and the target image to be reconstructed. The proposed method uses a new image model that represents the target image as a linear combination of a motion-dependent reference image and a sparse difference image. We then use an efficient iterative algorithm to jointly estimate the motion parameters and the difference image from sparsely sampled data. Simulation results from a numerical phantom data set and an in vivo data set show that the proposed method can accurately compensate the motion effects between the reference and the target images and improve reconstruction quality. The proposed method should prove useful for several applications such as interventional imaging, longitudinal imaging studies and dynamic contrast-enhanced imaging.     

Enhanced compression of integral images by combined use of residual images and MPEG-4 algorithm in three-dimensional integral imaging

In this paper, we proposed a novel approach to enhance the compression rate of integral images by combined use of the residual images generated from the sub-images and the MPEG-4 algorithm. In the proposed method, elemental images picked up from a three-dimensional object are transformed into sub-images, and these sub-images are sequentially rearranged with a spiral scanning topology and the first sub-image is assigned as the reference image. Then, by sequentially computing the differences between the reference image and other consecutive sub-images, a sequence of residual images is generated. Here, the residual images together with the reference image are modeled as the consecutive video frames just like a conventional moving picture. Finally, these residual images are compressed with the MPEG-4 algorithm. Experimental results show that compression efficiency of the proposed method has been improved up to 61.56% as compared to those of the JPEG-based compression scheme and up to 151.54% as compared to those of the conventional method averagely.