BS-ACEBOC: a generalized low-complexity dual-frequency constant-envelope multiplexing modulation for GNSS

GPS Solutions - In order to satisfy the need for combining signals in a constant-envelope modulation, we propose a new dual-frequency constant-envelope modulation (DCEM), called asymmetric...     

Analysis of geometry-free residuals in case of traveling ionosphere disturbances and their impact cycle slip detection

GPS Solutions - We discuss the impact of a traveling ionospheric disturbance (TID) on cycle slip detection for the geometry-free (GF) combination. Based on the GF combination algorithm, we deduce...     

空间灰度相关在影像纹理分类中的应用及分析

介绍了空间灰度相关法的11个纹理特征。试验表明,仅3个是描述性强的纹理特征。通过用t检验法来分类纹理,发现该方法具有良好的\     

内核地球模型的形变场和地球参考系

讨论了内核地球模型的形变场（ 位移、速度） 和地球参考系的确立, 为内核地球动力学的研究提供了依据。     

Contribution of new AAM data source to δLOD excitation

Data sets of the changes of the length of day (ΔLOD) measured by space geodetic techniques and of the atmospheric angular momentum (AAM) derived from global meteorological data by the National Meteorological Center (NMC) and the National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) are used to reanalyze and study the excitations of ΔLOD, and to examine and compare the contribution of each AAM series to ΔLOD. The main results are as follows. 1. The AAM reanalyzed with the assimilated global meteorological data by NCEP/NCAR are more accurate and have lower noise than the original AAM derived by NMC. The NCEP/NCAR-based AAM is more consistent with the changes of the length-of-day series. 2. The NCEP reanalysed AAM data may better explain the non-tidal LOD variations on intraseasonal to interannual time scales, especially on the quasi-biennial time scale. The atmosphere cannot, however, explain all LOD variations; other excitation sources are possible. 3. The effects of atmosphere on the estimated values of tides for ΔLOD components up to a month are weak. The preliminary results of the annual and semiannual tides can be estimated after removing the effects of atmosphere from ΔLOD. Received: 27 May 1998 / Accepted: 22 March 1999     

遥感图像分类与后处理综合技术研究—基于约束满足神经网络方法

遥感图像计算机分类的精度问题是阻碍计算机遥感信息处理系统实用化的一个关键问题。将分类后处理中的分类结果平滑过程模型化为约束优化问题,采用神经网络方法把分类结果平滑过程与遥感图像分类过程结合起来,提出了基于约束满足神经网络的遥感信息分类与后处理综合技术。实验表明该方法可明显提高森林类型划分、土地利用调查等遥感应用专题的分类精度。     

用模糊ARTMAP算法对CBERS-2数据进行分类

用模糊ARTMAP（fuzzy adaptive resonance theorymap）神经网络算法对CBERS-2数据进行了分类实践。首先介绍了模糊ARTMAP神经网络的算法原理和具体训练分类过程；然后用2004年9月新疆石河子地区的影像数据进行土地利用分类试验，并将分类结果与基于统计的最大似然法（MLC）、反向传播神经网络（BP）的分类结果作比较，总分类精度比MLC和BP算法分别提高9．9％和4．6％。结果表明，模糊ARTMAP对试验区CBERS-2影像上的裸地识别能力很强，对高分辨率的CBERS-2影像可获得很好的分类结果。     

高分辨率机载SAR影像判读实验

近年来机载SAR系统在国内得到快速发展。由于SAR系统具有不同于光学传感器的成像机理,因而可以提供全新的数据获取手段和不同的地物信息。利用高分辨率机载SAR影像进行解译时必须熟悉其成像机理和图像信息特点,充分了解SAR图像解译标志的特点和各类地物的解译规律。本文以高分辨率机载SAR数据为基础,对比光学影像,从SAR成像特点和地物散射特性出发,对SAR影像进行了目视解译,分析了SAR影像在地形地物要素识别中的应用潜力。     

摄影测量数据库的设计与建立

摄影测量数据库主要存储和管理数字地图成果数据，包括数字高程模型数据（DEM）、数字矢量地图数据（DLG）、数字影像地图数据（DOM）和数字栅格地图数据（DRG）。本文主要介绍了摄影测量数据库的设计、建立以及应用分析等内容。     

Efficient interpolations to GPS orbits for precise wide area applications

For precise real time or near real time differential GPS positioning in a wide or global area, precise GPS orbits or, alternatively, precise orbital corrections with respect to a reference orbit, such as GPS broadcast ephemerides, must be used. This work tests orbit interpolation methods, in order to represent the GPS orbits and orbital corrections accurately and efficiently for these and other GPS applications. For precise GPS orbits given in the SP3 format at the 15 min interval, numerical tests were conducted using Lagrange and Chebyshev as well as trigonometric polynomial functions. The results have demonstrated that the 19- or 20-term trigonometric function is apparently the most efficient interpolator for a 12 h GPS orbital arc, achieving 1 cm level 3D interpolation accuracy that can meet the requirements of most precise applications. The test results also demonstrated that the 9-term trigonometric function always yields optimal interpolation for a 2 h GPS orbit arc, in terms of interpolation errors, compared to the results when using a different number of terms for the same function or one of the other tested polynomial functions. This is evident from the minimal performance degradation when using the 9-term trigonometric function to interpolate near or at the end of a data interval. By limiting interpolation to the center 15 min to 1.5 h of a 2 h orbit arc, thereby eliminating the need to interpolate near the ends of that interval, users can opt for more terms (11 and 13) or different interpolators to further improve interpolation accuracy. When interpolating the orbital corrections with respect to the GPS broadcast ephemeris, all the tested interpolation functions of 3- to 9-term yield the same suitably accurate results. Therefore, a 3- to 5-term trigonometric function is arguably sufficiently accurate and more efficient for GPS orbital correction messaging in wide area and real time positioning.