合成孔径压缩感知超声成像中的高效能稀疏字典设计

针对目前常见的稀疏字典缺乏针对性,在合成孔径医学超声成像中的应用效果不佳,难以在低压缩率下保证重构图像质量的问题,本文设计了一种高效能的稀疏字典。根据超声回波信号是由发射脉冲信号经过不同延时衰减后叠加的特点,利用发射脉冲作为基函数构造稀疏字典,回波信号在该稀疏字典确定的变换域中具备很好的稀疏性,理论上能使其稀疏表示系数的稀疏度等于超声阵元接收到的反射回波数。通过FieldⅡ对简单点目标和复杂目标的仿真结果表明:在相同的重构算法和压缩率下该稀疏字典重构的平均绝对误差明显小于常见的稀疏字典,其值仅为DWT的几分之一,DFT和DCT的几十分之一,能让回波信号以更低的压缩率实现相同的恢复效果。本文最后使用体模的实际采集数据对算法的实际效果进行检测,实验结果也与仿真结果基本一致。基于该稀疏字典的压缩感知算法可以进一步减少合成孔径成像所需存储的数据量、降低系统的复杂度。      

互相关自适应加权的医学超声成像算法研究

根据超声成像系统的超声回波信号互相关性,提出互相关自适应加权超声成像算法.该算法根据散射点回波信号之间的空间相关性设置加权系数,对不同位置的散射点进行自适应加权成像,从而降低了成像系统的旁瓣,抑制了相关性较差的噪声.通过Field II仿真的点目标和吸声斑目标处理结果表明该方法成像的横向和纵向分辨率高,成像速度快.相对于延时叠加(DAS)算法,该算法对散射点成像可提高对比度16 dB,对于吸声斑成像可提高对比度0.85 dB.最后采用完备数据集进行实验,结果表明该算法成像分辨率优于DAS算法,对比度提高了17 dB.     

圆合成孔径声呐多点定位运动补偿

圆合成孔径声呐(CSAS)的成像性能受平台运动误差影响而下降,利用单侧回波可估计CSAS基阵的斜距误差,但单侧回波在小测绘带时无法估计升沉误差,针对此问题,提出了一种利用单侧回波信号的声呐平台三维运动估计和补偿方法。首先,对CSAS在不同观测角度的目标回波取极大值获得目标回波的到达时间;其次,基于多个点目标的到达时间建立CSAS三维定位模型;然后利用列文伯格-马夸尔特方法对声呐三维坐标进行估计;最后将位置估计结果与时域反投影成像方法结合实现对目标的成像.仿真结果表明:该方法能精确估计声呐平台运动误差,其空间坐标的估计误差小于仿真信号波长的1/8,从而精确补偿了CSAS在不同空间采样点上的阵元回波时间差,显著提高了目标成像质量。湖上试验结果表明,该算法能够实现对CSAS的运动误差补偿。仿真和试验结果均验证了方法的可行性和有效性。      

高斯回波模型在超声回波模拟中的应用及其迭代算法的讨论

本文将超声回波模拟成被高斯白噪声污染的高斯回波，并建立关于一系列估计参数的非线性模型，如回波的频带宽度，到达时间，中心频率，相位及回波幅度。这些参数对于超声无损检测中材料检测、目标探测、目标分类、及速率测量等密切相关。运用高斯-牛顿算法进行迭代计算，该算法能以较少的迭代次数计算出较为精确的参数向量，但该算法对到达时间T的初值设置较为敏感。     

环形阵列超声断层扫描高分辨率成像方法

介绍了针对环形阵列超声断层扫描的高分辨率全波形反演成像方法,不同于传统的延时叠加和渡越时间断层扫描,该方法充分考虑声传播的透射、反射、散射等特点,通过数据匹配的方法重构目标体声学参数图像。开发了环形超声换能器阵列构建超声断层扫描成像系统,并设计了适合检验成像算法的数值和物理模型,分别使用基于射线理论的回波延时叠加和初至波渡越时间断层扫描、以及基于波动理论的全波形反演方法对目标体进行图像重建,并对比其原理、分辨率、计算量等方面的特点,系统性分析了波动方法相对传统射线方法的优点和不足。结果表明,全波形反演这类充分考虑接收信号多种物理特征的方法可用于较高分辨率的声参数成像,适合与传统方法结合设计有效的超声断层扫描成像系统。     

湍流大气中随机粗糙表面激光回波空间相干性仿真

根据广义Huygens-Fresnel原理,推导了von Karman湍流谱条件下激光回波复相干度的理论解析式;基于湍流相位屏分步传输算法和随机粗糙目标表面模型,实现了激光回波光场的仿真计算.首先通过镜面反射回波光场的仿真分析,验证了算法的正确性;然后基于1.1 km的均匀传输路径,综合分析了随机粗糙目标表面特性和路径湍流强度对回波光场复相干度的影响.结果表明:回波光场的空间相干性随目标表面高度均方根的增大而降低,随目标表面相关长度的减小而降低;当表面相关长度远小于大气相干长度时,回波相干性会被严重破坏.该研究可为目标表面特性或利用已知表面获取路径湍流状态的相干探测提供有益的参考.     

光谱编码成像系统的数据压缩技术

利用时间拉伸显微成像系统观察并记录非重复动态随机现象,在其超高成像速度和高空间分辨率下必定会产生大量的数据。一种基于差分检测和游程编码的数据压缩方法,可以有效地解决时间拉伸成像系统的数据存储问题。差分检测可以消除连续相同的信号,只检测出相邻信号的差异,从而提高游程编码算法的有效性。实验中,采用扫描频率为77.76 MHz的时间拉伸显微成像对分辨率板、人红细胞和人乳腺癌细胞线性扫描成像。实验结果表明,数据压缩比可以达到8.47,对比分析发现经过差分检测方法可以获得更高的压缩比。另外,通过计算重建后的图像与原图的结构相似性(SSIM)值发现,经过数据压缩后高质量的图像可以被重建。     

脉冲压缩与互相关联合的合成孔径声呐回波时延补偿

针对合成孔径声呐中阵元相位不一致导致互相关时延补偿算法对成像质量提升效果有限的问题,提出了一种脉冲压缩与互相关联合的回波时延补偿算法.该算法利用脉冲压缩回波的互相关对原始回波相位畸变进行校正,实现粗补偿;联合脉冲压缩与偏移相位中心算法实现精细时延补偿,对不同合成孔径位置各阵元回波时延差实现了较为准确的估计,增强了成像效果。试验数据经该算法处理后,回波时延得到较为精确的补偿,地貌成像结果的亮度、对比度等统计特性得到不同程度的提高,且纹理细节增多;典型线缆目标的成像聚焦加深,成像长度误差约由5%减小为0.8%。试验结果显示,该算法对互相关时延补偿方法改进效果明显,验证了算法的可行性、有效性。      

基于改进八叉树的三维点云压缩算法

针对大数据环境下,三维模型的传输和存储需求,提出了一种基于八叉树的三维点云有损压缩算法。该算法改进了八叉树分割的停止条件,可以在适当的深度停止分割并确保体素大小合适。同时在分割的基础上通过建立K邻域,利用简单有效的统计方法去除原始点云的离群点。在数据结构上,对每个节点分配位掩码,通过操纵位掩码,在遍历时对数据查询和操作,并优化随后的点位置编码。该算法可以有效地移除离群点和表面杂点,并在区间编码上提高了点云压缩效率。实验结果表明,该算法较完整地保留了三维点云数据的关键信息,取得了良好的压缩率并缩短了压缩时间。     

大规模数值模拟数据的无损压缩

针对数值模拟数据的特点，对数据进行预处理后，进一步采用小波变换和洗牌（Shuffle）算法，再使用通用无损数据压缩算法Gzip对科学数据进行无损压缩。数据预处理：为了使用第二代小波变换，数据的预处理就是使用内存拷贝的方法把浮点数转换为整数，对数据的符号位、指数部分和尾数部分分别进行了预处理。由于科学数据数值变换平缓的原因，高频部分通常比低频数据小若干数量级，该数据预处理方法在小波变换及逆变换过程中不会出现溢出现象。     

Reducing spatial data using an optimized regressive discrete Fourier series

With the development of optical measurement techniques it is possible to obtain vast amounts of data. In vibrometry applications in particular operational deflection shapes are often obtained with very high spatial resolution. Fortunately, many techniques exist to reduce (approximate) the measurement data. One of the most common techniques for evaluating optical measurement data is by means of a Fourier analysis. However, this technique suffers from what is known as leakage when a non-integer number of periods is considered. This gives rise to non-negligible errors, which will obviously hamper the accuracy of the synthesized shape. Another technique such as a discrete cosine transform, used in the widely spread -jpeg standard does not suffer these anomalies but can still prove erroneous at times. One of the more recent approaches is via a so-called regressive discrete Fourier series (introduced by Arruda) which suffers one disadvantage. The problem statement is nonlinear in the parameters and needs a priori information about the deflection shape. This can be resolved by using the optimized regressive discrete fourier series (ORDFS), introduced in this article, which uses a nonlinear least squares approach. In this article the method will be applied in particular to the reduction of data for laser vibrometer measurements performed on an inorganic phosphate cement (IPC) beam (1D), as well as on a car door (2D). The proposed technique will also be validated on simulations to illustrate the properties concerning compression ratio and synthesized mode shape error. The introduced method will be bench marked for compression ratio and synthesized deflection shape error with all prior mentioned techniques.     

一种基于高分辨力遥感器的星上数据压缩方法

随着航天技术的发展,遥感器星上数据压缩变得越来越重要。提出了一种基于小波变换的压缩编码方法。首先对遥感图像进行双正交小波变换;然后在分析SPIHT、矢量量化、分形三种具有代表性的编码方法的基础上,提出了“带固定零区间的标量量化+改进游程编码”的编码方法;最后进行了试验。此方法具有以下特点:图像重构后具有较高的PSNR(40dB);算法简单,易于硬件的实现,实时性好;压缩比约为5∶1,但仍有提升的空间。此方法可满足高分辨力遥感图像对测量精度和实时性的要求,压缩比虽然未达到8∶1,但高于法国SPOT5的3∶1和美国Ikonos_2的11∶2.6。     

基于复合抖动技术的三维轮廓数据压缩方法

三维扫描技术获取的三维轮廓数据体积较大,不便于三维数字博物馆数据的实时传输和显示。针对这个问题,提出了基于复合抖动技术的三维轮廓数据压缩方法。首先,利用虚拟结构光投影系统实现三维轮廓数据的图像存储。然后,对图像的不同通道使用不同的抖动技术进行处理,并将结果保存到8位灰度图中主要的3位或者3幅1位逻辑图中,实现三维轮廓数据的进一步压缩存储。实验表明,与提供的三维模型格式相比提出方法的数据压缩比高达75.9∶1,非常适合三维轮廓数据的实时传输、重建与显示。     

成像光谱技术中机上实时数据压缩方法研究

根据成像光谱仪的特点用一系列串行的一维压缩实现二维图像压缩.提出了在机上对原始数据从光谱方向进行实时压缩的“二真值线性预测方法”.该法可做到压缩/复原后光谱特征信息不丢失.实验表明,对大多数光谱数据在1%的重建精度下,至少可获得2:1的压缩比.从总体而言,压缩后原始数据率可降低3～4倍.文中还对压缩数据的编码和误差进行了讨论.     

Wavelet-based ultra-high compression of multidimensional NMR data sets

The application of a lossy data compression algorithm based on wavelet transform to 2D NMR spectra is presented. We show that this algorithm affords rapid and extreme compression ratios (e.g., 800:1), providing high quality reconstructed 2D spectra. The algorithm was evaluated to ensure that qualitative and quantitative information are retained in the compressed NMR spectra. Whilst the maximum compression ratio that can be achieved depends on the number of signals and on the difference between the most and the least intense peaks (dynamic range), a compression ratio of 80:1 is affordable even for the challenging case of homonuclear 2D experiments of large biomolecules.     

Unsupervised classification-based hyperspectral data processing: lossy compression

The existing lossy compression algorithms play an important role in reducing the cost of storage equipment and bandwidth for hyperspectral (HS) application. However, none of the lossy compression algorithms considers the real-time classification of HS data. In this paper, we present a new lossy compression method for HS data that aims to optimally compress in both spatial and spectral domains and simultaneously maximize classification performance. For this target, Harsanyi–Farrand–Chang (HFC) and k-means++?algorithms are applied to achieve a spectral library and an index matrix for HS image. Spectral angle mapping and Euclidean distance are used to update the spectral library and the index matrix. The experiment results indicate that the proposed method has a good classification performance. The results also reveal that the proposed method works well in real-time classification and compression of HS data with a large volume and achieves a high compression ratio. It is noteworthy to mention that the superiority of our method in compression becomes more apparent as the volume of HS data grows. Consequently, the proposed method has a strong advantage in HS applications that require both compression and classification.     

Multiple-view fluorescence optical tomography reconstruction using compression of experimental data

We report on the experimental demonstration of a fast reconstruction method for multiview fluorescence diffuse optical tomography by using a wavelet-based data compression. We experimentally demonstrate that the use of data compression combined with the multiview approach makes it possible to perform a fast reconstruction of high quality. A structured illumination approach, guided by the compression scheme, has been adopted to further reduce the acquisition time. The reconstruction algorithm is based on the finite element method, and hence is suitable for samples of any arbitrary shape.     

端元提取技术在高光谱图像压缩中的应用

高光谱图像海量数据如何实现大比例有效压缩是限制其应用的主要问题之一,而现有有损压缩方法存在大压缩比与光谱特性信息准确保留的矛盾,即使现有最优有损压缩方法也不能够得到令人满意的结果。文章基于混合像元分解的思想提出基于端元提取技术的数据有损压缩方法来解决该矛盾,首先用顶点成分分析(VCA)方法提取场景中地物的端元光谱,根据各端元与观测像元之间的光谱间余弦角相似性度量方法估计各端元的丰度,接着对端元光谱及丰度数据进行无损压缩,最后利用JPEG2000有损压缩方法对高光谱图像的所有单波段图像进行空间维大比例有损压缩。AVIRIS高光谱图像的仿真结果表明,压缩比得到大幅度提高,光谱信息得到有效恢复。在实现压缩比为50∶1时,大部分像元的光谱角误差在2%左右。     

LiDAR data reduction assisted by optical image for 3D building reconstruction

The focus of LiDAR data compression and reduction has been raised in recent years due to the dramatically huge amount of points cloud. To improve the disposal efficiency of LiDAR data, a data reduction method assisted by optical image for 3D reconstruction of building facade is proposed in this article. The method involves a series of procedures of “2D feature line extraction – 3D feature line converting – buffer area of LiDAR points”. The main issue here is finding out a LiDAR point dataset around the 3D feature segment converted from 2D feature lines, which benefits to improving the efficiency of post-processing based on LiDAR data. The reduced LiDAR data can be obtained with reliable structures and accurate geometric position. Furthermore, the experiment of the LiDAR data reduction was conducted over the Xingyuan Building in Nanjing Normal University. The reduction method in this paper proved to be suit for the regular objects modeling, on the basis of the reliable and rigorous mathematical models and computer algorithms, and was a fundamental and useful approach to improve the efficiency of 3D reconstruction of building with the higher modeling accuracy.     

FPGA-based architecture for real-time data reduction of ultrasound signals

This paper describes a novel method for on-line real-time data reduction of radiofrequency (RF) ultrasound signals. The approach is based on a field programmable gate array (FPGA) system intended mainly for steel thickness measurements. Ultrasound data reduction is desirable when: (1) direct measurements performed by an operator are not accessible; (2) it is required to store a considerable amount of data; (3) the application requires measuring at very high speeds; and (4) the physical space for the embedded hardware is limited. All the aforementioned scenarios can be present in applications such as pipeline inspection where data reduction is traditionally performed on-line using pipeline inspection gauges (PIG). The method proposed in this work consists of identifying and storing in real-time only the time of occurrence (TOO) and the maximum amplitude of each echo present in a given RF ultrasound signal. The method is tested with a dedicated immersion system where a significant data reduction with an average of 96.5% is achieved.