多源遥感影像像素级融合分类与决策级分类融合法的研究

首先探讨了基于像素的多源遥感影像高频调制融合法，根据成像系统特性和Heisenberg测不准原理，设计的高斯滤波器对高分辨率影像滤波的方法是合理有效的。在研究BP神经网络的基础上，采用动量法和学习率自适应调整的策略，提高了BP神经网络学习算法收敛速度，并增强了算法的可靠性。提出并实现了多源遥感影像像素级融合分类与决策级分类融合两种分类方法，并进行了比较。采用Landsat TM3，4，5和航空SAR影像进行试验，结果表明两种分类方法是行之有效的，均适用于多源遥感影像分类。     

Enhancing group resolution of TM6 based on multi-variate regression model and semi-variogram function

［1］Castleman K R.Digital image processing.Englewood Cliffs,NJ:Prentice Hall,1996 ［2］Carr J R,Miranda F P.Spectral and texture classification of single and multiple band images.Computers & Geosciences,1996,22(8):849～865 ［3］Chen S P,Zhen W.Conciseness of remote sensing mineral resources exploration.Science and Technology Publishing House,1994(in Chinese) ［4］Costanitini M,Farina A,Zirilli F.The fusion of different resolution of SAR images.Proceeding of IEEE,1997,85(1):139～146 ［5］Dong Q,Fang H L.The use of variogram in remotely sensed images.Journal of Remote Sensing and Application,1997,12(1):7～13(in Chinese) ［6］De Jong S M,Burrongh P A.A fractal approach to the classification of Mediterranean vegetation types in remotely sensed images.PE & RS,1995(61):1 041～1 053 ［7］Fang H L,Qian G H.Fusion of ADEOS-AVNIR panchromatic and multispectral image data using principle component analysis.Journal of Remote Sensing and Application,1998,13(3):48～53(in Chinese) ［8］Franklin S E,Wulder M A,Lavigne M B.Automated derivation of geographic window size for use in remote sensing digital image texture analysis.Computers & Geosciences,1996,22(6):665～673 ［9］He J G,Zhu C G.Methods for data fusion between satellite-boarded SAR and multi-satellite remote sensing.Journal of Earth-science Information,1997 (16):29～33(in Chinese) ［10］Jia Y H.A data fusion method for spatial resolution enhancement of remotely sensed multi-spectral images.Journal of Remote Sensing and Application,1997,12(1):19～33(in Chinese) ［11］Jin G L,Qiu Z C.A research on information amount of multi-spectral images.Acta Geodaetica et Cartographica Sinica,1992,21(2):101～107(in Chinese) ［12］Kang Y H.Theories of data fusion.Xi‘an:Xi‘an Electronic University Press,1997(in Chinese) ［13］Li H,et al. Multi-sensor image fusion using the wavelet transform.Graphical Models and Image Processing,1995,27(3):235～244 ［14］Liu J G.Digital image processing of remotely sensed imagery data.Imperial College of Science,Technology and Medicine,1997 ［15］Liu J G,McM J.Moore:Pixel block intensity modulation: adding spatial detail to TM band 6 thermal imagery.Int.J.Remote sensing,1998,19(13):2 477～2 491 ［16］Lou Z,Zhu C G.Multi-variate statistics fusion of TM images.Journal of Aero-computational Technology,1998,28(3):40～42(in Chinese) ［17］Peng W N.Statistical methods for geo-data processing.Wuhan:Wuhan College of Geology,1983(in Chinese) ［18］Richard J R.Remote sensing digital image processing.an introduction,Berlin:Springer-Verlag,1999 ［19］Wang R S.Image understanding.Changsha:National Defense University Press,1995(in Chinese) ［20］Winkler G.Image analysis.Random Fields and Dynamic Monte Carlo Methods (A Mathematical Introduction),Berlin:Springer-Verlag,1995     

多波束与单波束测深数据的融合处理技术

本文在全面分析和总结多波束海洋测深主要误差源的基础上，提出通过相邻条带测深数据融合处理进行多波束测深系统偏差补偿方法；并提出以单波束测深数据作为控制，进一步提高多波束测深整体测量精度的数据处理方案；详细讨论了数据融合处理中的数值解算可行性和稳定性问题，相应提出了两步平差方法。本文最后使用我国自行研制的条带测深系统实测数据验证了上述方法的有效性和可靠性。     

GEOSCIENCE DATA IN THE INFORMATION AGE: ACQUISITION,DISSEMINATION AND THEIR APPLICATIONS

Recent dramatic advances in technologies such as computer,data management,and communication have revolutionized the way geologists collect,disseminate,and utilize geoscience data.After summarized the characteristics and fusion of geo-data,several major field data capture systems adopted in North America have been introduced.Geoscience information dissemination and distribution via the web technology and GIS package have also been discussed.In addition,multisources geo-data integration and their applications to mineral exploration,geological hazard risk assessment,and national land resources protection and utilization have been addressed by concrete examples.     

低稀释比制样技术与多类型地质样品X射线荧光分析方法研究

研究制定了在X射线荧光光谱分析中适用于多类型地质样品的低稀释比制样技术。确定了四硼酸锂和偏硼酸锂混合试剂作为熔剂，使用m样品:m熔剂=1:3的稀释比制样，对4种不同种类的地质材料进行了实验，并采用基本参数法和理论α系数法校正基体效应，得到了准确度较高的定量分析结果。     

靶场光电经纬仪多台交会测量数据的一种处理方法

给出了用多台光电经纬仪进行交会测量获取空中目标弹道轨迹的一种方法。该方法以每2台光电经纬仪交会测量方差的倒数为权将所有的交会测量结果进行了融合处理，并通过仿真试验进行相应的验证，进一步说明该方法的有效性。     

Image Processing on Geological Data in Vector Format and Multi-Source Spatial Data Fusion

The geological data are constructed in vector format in geographical information system (GIS) while other data such as remote sensing images, geographical data and geochemical data are savedf in raster ones. This paper converts the vector data into 8 bit images according to their importance to mineralization each by programming. We can communicate the geological meaning with the raster images by this method. The paper also fuses geographical data and geochemical data with the programmed strata data. The result shows that image fusion can express different intensities effectively and visualized the structure characters in 2 dimensions. Furthermore, it also can produce optimized information from multi-source data and express them more directly.     

基于IHS变换和小波变换的遥感影像融合

在遥感影像融合中，IHS变换法与小波变换法具有互补性，文中把这两种方法结合起来，提出了一种基于IHS变换与小波变换的影像融合方法。通过对具体影像的实验证明，该方法是有效的，达到了预期的目的。     

遥感与地球物理数据的融合处理及其地质应用——以上饶地区为例

在上饶地区１∶５万区调构造地质研究中，运用了航磁数据专业化处理及遥感—航磁数据图像融合处理方法，通过对融合图像深部构造和隐伏地质体的特征信息提取，增强了对区域基底构造的认识能力，揭示了该区中生代盆地下垫面的深部构造的产出状态特征，建立了区域深部地质构造的结构框架。     

基于Bayes融合法的多源遥感影像分类

提出了顾及各数据源成像模型、上下文关系模型和可靠性的基于Ｂａｙｅｓ融合分类的方法，并采用该方法对ＬａｎｄｓａｔＴＭ和航空ＳＡＲ影像进行土地利用分类试验。结果表明：同单独ＳＡＲ影像分类结果相比，融合分类法将分类精度提高了２０％。