基于Poisson-Markov分布最大后验概率的多通道超分辨率盲复原算法

在Poisson和Markov随机场分布假设基础上,改进了存在噪声时多通道互限制原则,提出了一种基于最大后验概率判据的多通道图像盲复原算法,可在无需已知通道内离散点扩散函数性质、类型和具体分布等先验知识的情况下,在图像幅度限制约束空间和离散点扩散函数能量恒定约束空间内,通过对图像和离散点扩散函数的交替迭代,使两者收敛到全局最优解,最终复原出超分辨率图像.对计算机模拟和实际存在微平移、微离焦情况的图像处理表明：算法在不同信噪比水平和非命中PSF尺寸下具有良好的图像复原效果和稳定性. 关键词： 泊松-马尔科夫分布 最大后验概率 多通道 盲复原     

Dynamic deformation image de-blurring and image processing for digital imaging correlation measurement

This paper proposes a method for de-blurring of images captured in the dynamic deformation of materials. De-blurring is achieved based on the dynamic-based approach, which is used to estimate the Point Spread Function (PSF) during the camera exposure window. The deconvolution process involving iterative matrix calculations of pixels, is then performed on the GPU to decrease the time cost. Compared to the Gauss method and the Lucy–Richardson method, it has the best result of the image restoration. The proposed method has been evaluated by using the Hopkinson bar loading system. In comparison to the blurry image, the proposed method has successfully restored the image. It is also demonstrated from image processing applications that the de-blurring method can improve the accuracy and the stability of the digital imaging correlation measurement.     

基于哈特曼-夏克波前传感器的模糊图像复原方法

离焦模糊图像的清晰度较低,因此必须对其进行复原。传统方法通常采用圆盘或高斯函数来近似离焦造成的点扩散函数,复原效果不够理想。为此,提出利用哈特曼-夏克波前传感器探测离焦波前,根据所得波前计算光学系统的点扩散函数,并采用Richardson-Lucy算法对模糊图像进行复原。搭建了实验用的光学系统,采集了离焦模糊图像以及相应的波前信息,获得了清晰的复原图像,并利用客观图像评价方法对退化图像和复原图像进行了评价,同时与传统方法得到的复原图像进行了比较。实验结果表明,该方法能精确重建点扩散函数,有效提高图像的质量。     

点扩散函数高斯拟合估计与遥感图像恢复

为了减轻或消除航天遥感相机成像过程中图像退化造成的模糊,突出图像的特征目标,对获取的图像进行了恢复处理。首先,采用陷波滤波器在频率域对遥感图像进行了去噪预处理。然后,通过图像中具有刀刃边缘的地物估计成像系统的退化函数,即点扩散函数;同时,利用高斯拟合对估计的点扩散函数进行校正。最后,利用拟合后的点扩散函数,采用自适应维纳滤波对图像进行恢复。实验结果表明：陷波滤波器基本消除了图像中叠加的条带噪声。与原图相比,细节图像恢复后其方差增大4.395,灰度平均梯度增大1.799,Laplacian梯度增大10.014,图像目视效果更清晰。高斯拟合的点扩散函数用于遥感图像恢复,减轻了图像模糊,使图像细节突出,纹理清晰,利于图像的判读和分析。     

基于光学系统像面分割的图像恢复方法研究

实际光学系统的点扩散函数随视场的不同而变化,因此基于线性不变成像系统的图像恢复方法在处理由大视场系统得到的图像时,恢复效果不理想。若把成像系统作为线性空变系统来进行处理,根据等晕成像原理,如将降质图像分割为不同区域进行复原的话,则可以有效地提高图像的恢复效果。针对区域间边界不连续性的问题,提出了区域延拓的解决方法,通过数字仿真,验证了图像分割处理与区域延拓方法的可行性。     

X光能点、放大倍率及针孔尺寸对空间分辨的影响

X光针孔成像是惯性约束聚变(ICF)研究中重要的诊断方法,对其点扩散函数的计算可用于图像重建和系统空间分辨的判断。对菲涅耳衍射公式进行了化简,分析了X光能点、针孔尺寸及放大倍率对针孔点扩散函数的影响。实验在保证成像能获得足够高信噪比的条件下,通过模拟获得在最佳空间分辨时所要的针孔大小、放大倍率和X光能点等参数。在流体力学不稳定性的静态样品定标实验中,通过模拟获得了针孔的调制传递函数(MTF),结合实验测量的结果反推获得分幅相机本身的MTF值。同时采用测刀边函数的方法获得了分幅相机本身的刀边函数,进而得到相机在各空间频率下的MTF值。两种方法得到的分幅相机MTF值一致,验证了通过菲涅耳衍射模拟X光针孔成像的可行性。     

顾及光谱混叠的高光谱相机任意方向侧扫几何成像

以移动平台的线推扫式高光谱相机横向推扫成像不同于以卫星平台或飞机平台的竖直摄影成像方式,其属于水平方向推扫竖直方向成像,存在投影面选择、空间方位确定以及光顾混叠等问题。针对上述情况,详细推导了适合于线阵高光谱相机地面推扫的影像像素地理参考模型,可以进行任意方向的地面横向推扫成像;结合影像的地面采样间隔大小及推扫成像的区域范围,给出了地理参考后影像格网划分的一般过程;同时,考虑到在影像格网划分过程中存在的舍入误差,以及在影像校正过程中采用传统直接光谱采样可能造成的光谱掺杂问题,提出了基于地理参考后相邻像素重叠面积作为权重系数进行加权融合的改进光谱采样方法。最后进行了大量地面横向推扫成像实验,并依据上述算法进行了影像畸变校正,验证了线阵影像几何畸变校正算法的有效性及鲁棒性,同时,对校正好后的影像选用多个样本点进行了光谱数据验证,实验表明改进后的光谱采样方法明显优于直接光谱采样算法,为同类产品的地面应用提供参考。     

Application of multi-frame approach in single-frame blind deconvolution

A new single-frame blind deconvolution algorithm for the linear shift-invariant imaging system is presented. The algorithm processes the partial images segmented from one single degraded image by multi-frame approach to recover the point spread function (PSF). Then a deconvolution method is employed to restore the whole image with the recovered PSF. In addition, in order to improve the fidelity and resolution of the recovered PSF, the coprimeness of the partial images is utilized. Results of simulated and real atmospheric turbulence degraded images using the algorithm are reported.     

空变环形编码孔成像模拟与图像复原

环形编码孔成像技术具有高的探测效率和信噪比,是一种解决低强度脉冲辐射源成像较好的技术。基于该技术,利用Geant4建立环形编码孔中子成像的模拟过程,获取6个不同位置的点扩散函数(PSF)和编码图像。根据空间移变图像分块原理,将图像分成矩形和圆形分块,每一块图像用RL迭代法复原,去除边界明显畸变的像素,这些像素强度由相邻的图像块像素到边界距离的加权系数叠加而成。模拟结果表明,该方法提高了图像复原效果,能够更好地诊断射线区域的空间分布情况。     

TDI线扫相机光学通道建模与仿真

卫星相机建模与仿真研究不论在民用还是军事领域都具有重要的理论研究和实际应用价值。根据TDI线扫相机的工作原理建立了一个TDI线扫相机光学通道的数学模型,该模型包括从景物到TDI CCD离散图像的几何投影关系和能量传递关系两部分,对于确定的成像系统,给定景物参数和光学系统参数,就可以确定对应图像传感器上的离散图像。为了验证模型,对一幅卫星图像进行了计算机仿真。建立的模型可以用来研究当成像条件改变时对图像的影响,进一步还可以用来发展虚拟卫星相机系统。     

Multi-frame partially saturated images blind deconvolution

When blurred images have saturated or over-exposed pixels, conventional blind deconvolution approaches often fail to estimate accurate point spread function (PSF) and will introduce local ringing artifacts. In this paper, we propose a method to deal with the problem under the modified multi-frame blind deconvolution framework. First, in the kernel estimation step, a light streak detection scheme using multi-frame blurred images is incorporated into the regularization constraint. Second, we deal with image regions affected by the saturated pixels separately by modeling a weighted matrix during each multi-frame deconvolution iteration process. Both synthetic and real-world examples show that more accurate PSFs can be estimated and restored images have richer details and less negative effects compared to state of art methods.     

Point Spread Function Estimation for a Turbulence-Degenerated Image Based on Àtrous Wavelet Transform

The observed object images are seriously blurred because of the influence of atmospheric turbulence. The restoration is required for the reconstruction of turbulence degraded images. Point spread function (PSF) estimation, an essential part of image restoration, has no accurate estimation algorithm at present. Based on the àtrous wavelet, we deduce a novel PSF estimation algorithm. First, the àtrous wavelet at varying scales is transformed. Then, on the basis of the relation among the local maxima of the modulus of the wavelet coefficients at different scales, the Lipschitz exponent of the wavelet coefficients is computed, thus the variance of a PSF is estimated. From this estimated variance, one is able to obtain the PSF. Consequently, the object image can be restored. Experimental results show that the proposed method is highly effective with good performance.     

高速TDI CCD亚像元动态成像系统分析

亚像元动态成像技术是目前实现相机小型化及提高相机空间分辨率的有效方法。将亚像元动态成像技术引入高速TDICCD相机中,通过特殊设计的光学拼接方法,使相机在不影响扫描速度及不改变TDICCD器件尺寸和相机焦距的情况下,可大大提高空间分辨率。为评价该TDICCD亚像元动态成像系统的成像质量,对成像系统的调制传递函数(MTF)进行了分析,分析结果表明该TDICCD亚像元动态成像技术可以将系统的空间分辨率提高1.6倍以上。     

Design of a very high frame rate camera based on an asynchronous CCD driving method

A very high frame rate camera is designed based on an innovative CCD driving method. The CCD driving method is mainly implemented on frame transfer CCDs. Asynchronous drive timing sequences are applied in the image and storage section of the CCDs. Several rows of the charge in the image section are binned onto the same row in the storage section, and there are the same number of images to be stored in the storage section before they are read out. Based on the new driving method, the frame transfer CCDs can work at a very high frame rate in acquiring burst images though the reading speed remains at a lower level. A very high frame rate camera is designed in this paper. The innovative CCD driving method is mainly of concern. An e2v's CCD60 is adopted in the camera system, whose full size resolution is 128 × 128, and the up most frame rate is 1000 Hz in the conventional CCD driving method. By using the presented method, the CCD60 based imager is capable of operating at up to 40000 frames per second (fps) at a recognizable resolution of 128 × 32. Comparing cameras using traditional binning and region of interest technologies, the frame rate is normally less than 5000 fps while the resolution is only 32 × 32 left.     

利用维纳滤波改善声透镜光声成像系统的分辨率

为了克服衍射效应对光声成像系统分辨率的限制,需要采用逆卷积方法进行图像反演.从理论上分析了声透镜成像原理,模拟仿真了声透镜的点扩展函数对声透镜成像系统分辨率的影响和维纳滤波解卷积方法复原光声成像的过程,并利用自搭建的声透镜光声成像系统进行了深入的实验研究,得到了物平面上相距4 mm和3 mm的两个黑胶带点的直接成像光声...     

High-resolution two-dimensional image upconversion of incoherent light

We consider a technique for high-resolution image upconversion of thermal light. Experimentally, we demonstrate cw upconversion with a resolution of more than 200 × 1000 pixels of thermally illuminated objects. This is the first demonstration (to our knowledge) of high-resolution cw image upconversion. The upconversion method promises an alternative route to high-quantum-efficiency all-optical imaging in the mid-IR wavelength region and beyond using standard CCD cameras. A particular advantage of CCD cameras compared to state-of-the-art thermal cameras is the possibility to tailor and tune the spectral response leading to functional spectral imaging.     

基于TV正则化和局部约束的遥感图像恢复

阐述了基于总变分理论和基于像元亮度局部约束的退化图像恢复算法,为利用二者的优点获得更好的恢复效果.把总变分方法和局部约束方法结合在一起,提出了一种新的混合恢复算法.对最小二乘问题进行总变分正则化约束,形成迭代公式,在迭代过程中对所获得的结果利用局部均值和局部方差进行局部约束.实验中对退化的遥感图像分别用总变分约束方法恢复和本文提出的方法进行恢复,结果表明,该方法具有良好的图像恢复能力,图像恢复效果有了明显的提高.     

基于MTF的时间延迟积分CCD成像系统同步误差分析

通过像移调制传递函数MTF分析了TDI CCD行扫速率与运动像不同步时对相机成像质量的影响。为验证速度失配对TDI CCD相机成像质量的影响，针对应用环境，利用高对比度靶标对TDI CCD动态成像进行了仿真实验和动态测试，结果表明，在奈奎斯特频率范围内，只要将相机动态成像系统的同步误差控制在±2%的范围内，并使图像时钟处于连续多相的工作方式，就可以基本上消除同步误差对成像质量的影响。     

基于B样条插值算法的亚像元技术的研究

亚像元动态成像技术是目前实现卫星相机小型化及提高相机空间分辨率的一种有效方法。在不改变TDICCD相机的焦距、成像距离,以及TDI CCD器件的像元尺寸的前提下,亚像元动态成像技术可以提高相机的空间分辨率。利用B样条插值算法,研究了TDI CCD相机的图像的亚像元动态插值问题,由两幅相机输出的原始图像经过算法得到比原图像分辨率高的图像,并对比了该算法与其它几种算法的效果,分析结果表明,该算法较其它算法得到的高分辨率图像的效果更佳。     

基于Walsh-Hadamard变换的单像素遥感成像

本文提出了基于Walsh-Hadamard变换的单像素成像方案, 并从理论分析、模拟仿真和实验验证三方面分别验证了该方案的可行性. 实验上实现了350-900 nm波段对 距离500 m和5000 m自然目标的128×128 像素成像, 成像速度0.5帧/秒. 研究并讨论了单像素相机方案与计算量子成像方案的差异与共性, 在此基础上分析了基于Walsh-Hadamard变换的单像素成像方案的优势与局限性. 研究表明本方案同时适用于单像素相机和计算量子成像. 由于单像素成像适用于应用在如红外热成像、微波成像等波段, 因此在阵列探测器灵敏度或工艺达不到要求时存在优势. 本文所提出的方案使得单像素成像技术向实际应用迈进了一步.