SGG重力场球谐系数正则解的误差估计方法

重力梯度为重力位的二阶导数,可以通过星载梯度仪进行观测。重力场球谐函数系数可以通过正则化方法由重力梯度算出。本文在对正则化方法分析的基础上提出了估计球谐函数系数正则解误差的方法,为我国今后发射重力梯度卫星提供技术准备。     

Regularization of Motion Equations with L-Transformation and Numerical Integration of the Regular Equations

The sets of L-matrices of the second, fourth and eighth orders are constructed axiomatically. The defining relations are taken from the regularization of motion equations for Keplerian problem. In particular, the Levi-Civita matrix and KS-matrix are L-matrices of second and fourth order, respectively. A theorem on the ranks of L-transformations of different orders is proved. The notion of L-similarity transformation is introduced, certain sets of L-matrices are constructed, and their classification is given. An application of fourth order L-matrices for N-body problem regularization is given. A method of correction for regular coordinates in the Runge–Kutta–Fehlberg integration method for regular motion equations of a perturbed two-body problem is suggested. Comparison is given for the results of numerical integration in the problem of defining the orbit of a satellite, with and without the above correction method. The comparison is carried out with respect to the number of calls to the subroutine evaluating the perturbational accelerations vector. The results of integration using the correction turn out to be in a favorable position.     

SPECTRAL DECOMPOSITRION FORMULA FOR REGULARIZATION SOLUTION OF ILL-POSED EQUATION

1　IntroductionAwell posedequationhasobviousmathematicalandphysicalmeaningandfitsallofthefollowingconditions:a)solutionexists,b)solutionisunique,c)so lutionisstable .Ifanequationviolatesanyoftheabovethreeconditions,itiscalledincorrectlyposedequation .Ifthesolutionisn’tunique,thecorrespondingequa tioniscalledrankdeficientequation .Ifthesolutionisn’tstable,thecorrespondingequationiscalledill posedequation .Somekindofconstraint,forexam ple ,minimumnormconditionmustbeaddedinor derthattherankdef…     

高光谱数据截断加权核范数稀疏解混

受仪器和观测条件限制,高光谱数据易受噪声污染,给数据解译带来挑战。针对传统稀疏解混模型抗噪性能差的问题,本文提出一种截断加权核范数稀疏解混方法,利用高光谱图像像元之间的相关性减轻噪声对丰度估计的干扰。该方法借助低秩表示在挖掘数据内在低维结构方面的优势,在稀疏解混中加入基于截断加权核范数的低秩约束,并结合加权稀疏技术,在稀疏正则项中引入空间邻域权重。截断加权核范数对丰度矩阵的奇异值向量分段处理,可以更好地实现丰度矩阵的低秩逼近,使丰度图像保持空间一致性并保留更多细节信息,空间加权策略则增强了丰度图像的空间连续性。模拟高光谱数据、Cuprite矿区真实数据和红树林高光谱数据实验表明,与其他先进的稀疏解混方法相比,所提方法具有更好的抗噪性,能够提高解混精度。     

低信噪比核磁共振T2谱反演算法研究（英文）

正则化因子的选择方法决定了正则化方法的稳定性和准确度。在分析改进的奇异值分解法与正则化方法的关系的基础上,给出一个正则因子计算公式。数值模拟试验表明改进的奇异值分解法和正则化方法适应低信噪比。当信噪比低于30时正则化方法比奇异值分解法结果更好,当信噪比高于30时奇异值分解法更优。采用本文提出的正则因子的正则化方法可以适用于实际核磁测井的低信噪比(5相似文献   

选权拟合法解的特性探讨

选权拟合法是解决大地测量中的不适定问题的一种方法,是对吉洪诺夫正则化方法的改造。在推导一般正则化解的偏差计算公式并回顾了选权拟合法的基本原理和公式的基础上,推导了选权拟合法解的一个重要性质:只要被约束的部分参数估值无偏,其余的也无偏;该性质说明利用选权拟合进行参数估计的结果是有条件无偏的。这个不同于一般正则化解的重要特性可以用于设计加快GPS短基线快速定位双差模糊度解算策略。恰当利用选权拟合法,用实测数据算例分析了GPS基线分量的先验信息的偏差大小对模糊度解算的影响。     

波场延拓反Q 滤波的正则化方法

基于波场向下延拓的反Q 滤波是补偿大地衰减效应的一种常用方法,但其振幅补偿的不稳定性限制了该方法在中、深层地震资料处理中的应用。针对于此,提出波场延拓反Q 滤波的正则化方法,该方法不仅较好地解决了地震波能量补偿的不稳定问题,同时对经过品质因子较低地层的深层地震波有着较好的能量补偿效果。模型试验和实例应用均表明: 与常规稳定的反Q 滤波方法相比,基于正则化法的反Q 滤波不仅具有较强的抗干扰能力和较好的振幅补偿能力,而且有效地提高了中、深层地震资料的分辨能力。     

基于改进高斯-牛顿法的位场向下延拓

本文针对位场向下延拓的不适定问题,在分析最优化算法中高斯-牛顿法基本原理及滤波函数滤波特性的基础之上,提出基于正则参数指数递增计算方法和残差最小步长准则的改进高斯-牛顿法。基于理论重力模型和航磁实测数据的对比实验表明,改进后得到的自适应迭代法具有相对较高的位场向下延拓精度和很好的收敛性。     

基于Extrapolation Tikhonov正则化算法的重力数据三维约束反演

通过研究重力数据三维反演解的病态性,利用基于拉格朗日插值方法的Extrapolation Tikhonov正则化方法来解决反演中解的不唯一性和不稳定性问题,该方法最大限度的减少了因正则化参数的引入而在反演结果中介入的误差,同时详细讨论了基于三种选择原则的正则化双参数的具体选择方法,模型试算结果表明,与原Tikhonov方法相比,该方法提高了反演的拟合精度.其次,为了消除核函数随深度增加而快速衰减对反演结果的影响,本文改进了前人的重力数据三维反演深度加权函数,改进后的加权函数与原函数相比能更好的识别异常体底部密度分布特征,对于埋深较深的异常体具有较好的识别效果,更好的解决了由近地面趋肤效应作用引起的密度分布不均的问题.同时,利用上下限约束函数限制每一个立方体的密度差范围,并应用于多组人工合成模型.结果表明:该反演方法能准确地获得正演模型的预设参数范围和位置.     

Building segmentation and outline extraction from UAV image-derived point clouds by a line growing algorithm

ABSTRACT

This paper presents an approach to process raw unmanned aircraft vehicle (UAV) image-derived point clouds for automatically detecting, segmenting and regularizing buildings of complex urban landscapes. For regularizing, we mean the extraction of the building footprints with precise position and details. In the first step, vegetation points were extracted using a support vector machine (SVM) classifier based on vegetation indexes calculated from color information, then the traditional hierarchical stripping classification method was applied to classify and segment individual buildings. In the second step, we first determined the building boundary points with a modified convex hull algorithm. Then, we further segmented these points such that each point was assigned to a fitting line using a line growing algorithm. Then, two mutually perpendicular directions of each individual building were determined through a W-k-means clustering algorithm which used the slop information and principal direction constraints. Eventually, the building edges were regularized to form the final building footprints. Qualitative and quantitative measures were used to evaluate the performance of the proposed approach by comparing the digitized results from ortho images.