改进的模糊聚类遥感影像分割

聚类是数据挖掘的重要分支之一,引入模糊理论的模糊聚类分析为显示数据提供了模糊处理能力,在许多领域被广泛应用。本文应用考虑邻域关系的约束模糊C均值(Fuzzy C-Means with Constrains,FCM_S)算法,将邻域像素引入到目标函数中,进而有效地利用邻域像素信息,提高分割精度。本文应用FCM_S算法对模拟彩色纹理图像进行分割,计算其混淆矩阵,定性定量地与FCM算法进行对比分析,证明了该算法的鲁棒性。     

一种结合四元数与直方图的多光谱图像自适应边缘检测方法

针对图像自适应边缘检测的难点和传统多光谱遥感图像边缘检测方法的局限性,提出了一种结合四元数与直方图的自适应边缘检测方法。该方法在四元数空间,利用矢量旋转完成了多光谱图像的边缘检测,并通过直方图统计的方法获取自适应阈值,实现了边缘图像的检测。利用陆地多光谱遥感图像和水域多光谱遥感图像分别进行实验,验证了方法的有效性。     

信息分配系数的自适应滤波在GPS/DR组合导航中的应用

在组合导航中常采用联合滤波器进行多源传感器的信息融合,其中信息分配系数的取值直接影响着联合滤波器的性能。在详细分析子滤波器状态变量估计协方差阵特点的基础上,文中提出了采用两组信息分配系数的自适应联合滤波模型,并将设计的自适应联合滤波模型应用于车载GPS/DR组合导航中。仿真实验表明,导航的位置精度提高约为40%;速度精度提高约为10%。同时也提高了组合系统的容错性和可靠性。     

滑动窗口的普通克立格方法在降水量插值中的应用

利用2009年全国2200个观测站降水量资料,使用滑动窗口的普通克立格方法对降水量资料进行格点化估计。针对滑动窗口的普通克立格方法在降水量格点化应用中存在的问题,设计了3种技术处理的试验方案。比较了全局搜索与方位邻近方法的误差,讨论了最大影响半径及屏蔽效应对插值效果的作用。提出了方位邻近法的样本点选择策略,结果表明,相对传统滑动窗口的普通克立格方法较常使用的全局搜索法而言,方位邻近法显著降低了计算资源的耗用,同时又具有较高的插值精度,特别是在站点密集地区有突出的优势;试验结果也表明:变程为4°～5°的经(纬)线弧长时,在方位邻近法下,我国大部分区域有较好的插值效果;屏蔽效应弱,ε取值为0.1时,降水量插值准确率较高,随着ε增大,插值误差也逐渐增大。     

Improvement in the performance of solar adaptive optics

Adaptive optics （AO）, which provides diffraction limited imaging over a field-of-view （FOV）, is a powerful technique for solar observation. In the tomographic approach, each wavefront sensor （WFS） is looking at a single reference that acts as a guide star. This allows a 3D reconstruction of the distorted wavefront to be made. The correction is applied by one or more deformable mirrors （DMs）. This technique benefits from information about atmospheric turbulence at different layers, which can be used to reconstruct the wavefront extremely well. With the assistance of the MAOS software package, we consider the tomography errors and WFS aliasing errors, and focus on how the performance of a solar telescope （pointing toward zenith） is related to atmospheric anisoplanatism. We theoretically quantify the performance of the to- mographic solar AO system. The results indicate that the tomographic AO system can improve the average Strehl ratio of a solar telescope in a 10＂ - 80＂ diameter FOV by only employing one DM conjugated to the telescope pupil. Furthermore, we discuss the effects of DM conjugate altitude on the correction achievable by the AO system by selecting two atmospheric models that differ mainly in terms of atmospheric prop- erties at ground level, and present the optimum DM conjugate altitudes for different observation sites.     

The Millennium Galaxy Catalogue: bulge–disc decomposition of 10 095 nearby galaxies

We present near-infrared (NIR) adaptive optics-assisted spectroscopic observations of the  CO (Δμ= 2)  absorption bands towards the centre of the giant elliptical galaxy NGC 1399. The observations were made with NAOS-CONICA (on the European Southern Observatory's Very Large Telescope) and have a full width at half-maximum resolution of 0.15 arcsec (14 pc). Kinematic analysis of the observations reveals a decoupled core and strongly non-Gaussian line-of-sight velocity profiles in the central 0.2 arcsec (19 pc). NIR imaging also indicates an asymmetric elongation of the central isophotes in the same region.
We use spherical orbit-superposition models to interpret the kinematics, using a set of orthogonal 'eigen-velocity profiles' that allow us to fit models directly to spectra. The models require a central black hole of mass  1.2^+0.5_−0.6× 10⁹ M_⊙  , with a strongly tangentially biased orbit distribution in the inner 40 pc.     

A STATISTICAL APPROACH TO THE PROBLEM OF MISSING SPATIAL DATA USING A FIRST-ORDER MARKOV MODEL*

Geographical data sets sometimes contain missing observations that need to be estimated. A statistical approach to the problem is discussed for multivariate normal spatial data sets satisfying the first-order spatial Markov property with constant mean, where the information at neighboring or contiguous observed sites is used to estimate the missing values. The completed data are used to estimate the parameters of the distribution. The procedure is iterative. The approach is a special case of the Orchard and Woodbury missing information principle. The paper concludes with an illustrative empirical example using rainfall data from an area of Kansas and Nebraska. The quality of the estimates for different sites are compared.     

Synthetically adaptive robust filtering for satellite orbit determination

The precise orbit information is the premise of the application of satellite, and the precise orbit deter- mination includes dynamic and geometric methods. The dynamic method estimates the position and veloc-ity of a satellite at one epoch by extending the obser-vational arc in order to include more observations of different epochs in it. It is realized by integrating the dynamic equation of the satellite so that the observa-tions at different epochs are related to the particular state. The dy…