基于投影寻踪的高光谱图像异常检测并行算法

投影寻踪方法能有效提取数据中的非高斯结构凸显异常信息,但在求解最优投影方向时存在计算量大、运行时间长的问题,为提高处理效率,提出一种机群环境下的并行算法.选用偏度和峰度组合作为投影指标,将所有像素光谱作为特定投影方向集依次搜索,求解最优投影方向.在并行计算各候选方向投影指标时,分割图像数据分布存储于各机群结点,数据子块朝候选方向并行投影后,将指标计算式变形分解,使各结点在指标计算过程中所需数据均为本地数据,解决数据局部性问题,并采用一种"轮流作主"的机制提高算法负载均衡程度.利用实用型模块化高光谱仪数据在机群系统上进行测试,达到了较好的加速效果,表明该并行算法具有良好的并行性能.     

易于硬件实现的遥感图像两倍压缩算法

结合干涉多光谱遥感图像航天应用的要求,参考CCSDS标准无损压缩算法,提出了一种易于硬件实现的干涉多光谱遥感图像无损／近无损压缩算法。该算法在对12bit干涉多光谱图像进行2倍压缩时压缩性能与JPEG 2000标准算法相当,但算法复杂度和所需存储空间大大降低,在单片FPGA内部即可实现全部压缩功能,可以很好地满足航天应用的要求。     

Sumbandilasat—An operational technology demonstrator

Technology advances in sensor, digital technology and a standardised modular satellite bus are enabling a new generation of 80 kg micro-satellites with a better than 6.5 m GSD multi-spectral performance, to be specified, built and deployed with a dedicated launch within 12 months. The result of the standardised modular bus is lower cost, higher reliability and fast deployment. Operational remote sensing with a micro-satellite is thus within reach of individual organisations for dedicated missions. Sumbandilasat (pioneer in the Venda language) is a second generation satellite technology building on the expertise obtained in the Sunsat small satellite programme. The components used to build Sumbandilasat are the result of a technology development program of more than 3 years. Sumbandilasat is an operational technology demonstrator with more than 90% newly developed or improved subsystems and a compact refractive imager as a precursor to the MSMISat satellite with the same multi-spectral band set. The scalable, standardised modular satellite bus architecture enables satellites with a mass of 80–450 kg to be adapted to the specific mission requirements with minimum new engineering effort.     

FADS系统在飞翼布局飞行器上的应用研究

针对飞翼布局飞行平台对高精度迎角、侧滑角的依赖性,设计了一种采用安装于机头表面的测压孔和机翼前缘的测压孔相结合的FADS系统应用方案,给出了其空气动力学模型,并将该FADS系统与惯性系统相结合来提高机动情况下大气数据的测量精度。由于该系统以FADS测量结果为基础,所以可以保证最后输出结果的精度要求,而且还克服了FADS系统延时的影响,其计算复杂性基本没有增加,满足实时性要求,易于工程实现。     

Aerosols and clouds in the upper troposphere–lower stratosphere region detected by GOMOS and ACE: Intercomparison and analysis of the years 2004 and 2005

Satellite-based limb occultation measurements are well suited for the detection and mapping of polar stratospheric clouds (PSCs) and cirrus clouds. PSCs are of fundamental importance for the formation of the Antarctic ozone hole that occurs every year since the early 1980s in Southern Hemisphere spring. Despite progress in the observation, modeling and understanding of PSCs in recent years, there are still important questions which remain to be resolved, e.g. PSC microphysics, composition, formation mechanisms and long-term changes in occurrence. In addition, it has recently become clear that cirrus clouds significantly affect the global energy balance and climate, due to their influence on atmospheric thermal structure.     

Lower-mesospheric inversion layers over brazilian equatorial region using TIMED/SABER temperature profiles

Lower-mesospheric inversion layers (MILs) were studied using the temperature profiles observed by TIMED/SABER over Cariri (7.5°S, 36.5°W), Brazil, in 2005. A total 175 MILs were identified with the maximum occurrence in April and October and the minimum in January and July. The lower MIL is located in a height region from 70 to 90 km, with the peak at around 83 ± 4 km with the temperature of 205 ± 5 K, and the thickness of 4–10 km. The results show large amplitudes of MILs during equinoxes and minimum in solstices, with a clear semiannual variation. A general feature of lower MIL in monthly mean profile was observed twice a year, one from February to May, and the other from August to October with a downward shift of the top level. These results suggest that formation and long persistence of MIL is an important factor to investigate propagation of atmospheric gravity waves in the mesosphere-lower thermosphere (MLT) region.     

Advances in the Researches in Middle and Upper Atmosphere in 2006—2008

This paper summarizes the research results obtained by Chinese scientists and/or through international collaborations in 2006—2008.Specifically,this paper focus on the research fields in the middle and upper atmosphere,including developments in facilities and instruments,and the advancements in scientific issues.     

超谱遥感图像降维方法研究现状与分析

随着成像光谱仪的发展 ,超谱遥感图像的研究已进入到一个新的阶段———对获取的超谱数据进行有效处理和利用的阶段。目前的处理方法主要集中在对超谱图像的数值分析处理上 ,比如大气校正、降低数据维数、信息提取、分类与压缩等方面。而超谱图像降维方法的研究是做好后继处理的一个关键步骤 ,降维方式的正确选取与使用 ,对于发展和完善那些针对超谱海量数据和丰富信息特点的算法和软件有极大的好处。文章从波段选择、划分数据源、特征提取和融合等 4个角度对目前超谱图像的各种降维方法进行了综合归纳和分析。力图为超谱图像处理寻找突破点 ,加强此领域的研究力度     

基于在轨组装维护的多载荷协同观测卫星平台构型设计

针对高轨遥感卫星多载荷协同观测和一致测量应用需求,在前期总体技术方案研究的基础上,通过开展现有在轨组装航天器和高轨卫星公用平台的构型调研,提出了一种适用于多载荷协同观测需求的平台构型设计方案。平台利用CZ-5运载火箭发射并在轨展开,采用“网格式中心承力筒+板+桁架”式的总体构型,开展了主传力路径、贮箱承载结构、网格式中心承力筒等的细化设计工作,并进行了结构产品的初步设计,建立了平台结构整星的有限元模型。仿真分析表明:平台结构整星的各向一阶频率均能满足运载火箭的频率要求。基于标准化、通用化的平台接口设计,使其具备可扩展及可维修的能力,大大提升了平台鲁棒性。基于全网格筒的主承力设计,增强平台结构承载能力,并改善平台内部热环境。该设计方案可为基于在轨组装维护的高轨遥感大平台结构的设计提供有价值的参考。     

面向月球科研站任务的地月准实时遥操作模拟验证系统设计与关键技术研究

针对月球科研站任务地月遥操作需求和特点,设计地月准实时遥操作模拟验证系统,对其中的关键技术海量多尺度遥感数据环境感知、空地协同高精度定位、基于混合现实的预测仿真等进行了研究。提出了基于海量多尺度遥感数据的联合处理方法,环境感知数据融合可处理米级至厘米级环境感知结果;设计基于空地多运动平台协同定位的原型软件,空地协同地标约束下,定位精度不低于轨道器影像1个像素;开发了带有力觉反馈的预测仿真验证平台,在2~3 s变时延约束下可在仿真环境下协同演示遥操作过程,真实还原从端的遥操作作业场景,提高遥操作的执行效率和安全性。为地面人员的方案设计、操作手训练、故障处置等提供实时支持,为我国未来实施月球科研站任务筑牢基础。