ANALYSIS OF REMOTE SENSING IMAGERY IN ENGINEERING GEOMORPHOLOGY

An image interpretation technique known as “indicational analysis” is applied to engineering geomorphology–i.e., image characteristics visible on remote sensing imagery are used to infer (indicate) the presence of features that are not directly visible. In this case the identification of particular relief forms provides a basis for inferring the level of risk posed by geomorphological hazards to major construction projects such as railroads, pipelines, and highways. A key aspect of the procedure is a multiscalar approach, in which different components of the overall store of information brought to bear on a problem are obtained at different levels of interpretation. Translated from: Geomorfologiya, 1987, No. 2, pp. 35–42.     

USE OF REMOTE SENSING IMAGERY IN MULTIPURPOSE MAPPING

This paper focuses on remote sensing's role in support of diverse mapping tasks, including those of interest to soil scientists (land use, soil, and botanical mapping). It analyzes new opportunities and challenges provided by remote sensing imagery with respect to map compilation and revision: the interdisciplinary yet goal-oriented character of map themes, the large stock of data to be processed systematically for coherent information yield, the numerous linkages of map elements to be represented on a series of closely interrelated maps, and the need for standardization and a unified approach to map design and symbolization. Translated by Edward Torrey, Alexandria, VA 22308 from: G. V. Dobrovol'skiy and V. L. Andronikov, eds., Aerokosmicheskiye metody v pochvo-vedenii i ikh ispol'zovaniye v sel'skom khozyaystve: sbornik nauchnykh trudov [Remote Sensing Methods in Soil Science and Their Utilization in Agriculture: A Collection of Scientific Works]. Moscow: Nauka, 1990, pp. 16-21.     

DIGITAL PROCESSING OF REMOTE SENSING IMAGERY IN SOIL MAPPING

The authors describe strategies for the utilization of digital image processing in the compilation of soil maps at scales of 1:1,000,000 and 1:25,000 for the trans-Volga region of Russia. Particular attention is devoted to an exploration of the effects of variations in soil humus content, texture, and degree of erosion upon spectral brightness coefficients. Results of the mapping demonstrate the promise of digital image processing in the differentiation of a number of chernozem soil subtypes. Translated by Edward Torrey, Alexandria, VA 22308 from: Aerokosmicheskiye metody v pochvovedenii i ikh ispol'zovaniye v sel'skom khozyaystve: sbornik nauchnykh trudov [Remote Sensing Methods in Soil Science and Their Utilization in Agriculture: A Collection of Scientific Works]. Moscow: Nauka, 1990, pp. 214–225.     

AUTOMATED MAPPING FROM THEMATIC INTERPRETATION OF SCANNER IMAGERY

The process of the automated compilation and plotting of thematic maps based on selected methods of cartographic representation is examined. The authors argue that the essence of logical and mathematical processing of digitized space imagery is formalized, objective cartographic representation of the results of image interpretation. General procedures of map compilation are operationalized based on a geomorphological mapping project using Landsat MSS data for the Moscow Region.     

ERRORS IN THE INDICATION OF GEOSYSTEMS FROM REMOTE SENSING IMAGERY

After a general discussion of principles of remote sensing indication and image recognition theory, the particular errors in this process are analyzed individually: errors in data recording; errors in automated image analysis; errors in interpretation of the content of geosystems; errors in the selection of interpretation keys; and finally errors in generalization and the extrapolation of results. The total error in geosystem indication from remote sensing imagery can be determined by an analysis of these particular errors. Translated by Edward Torrey, Alexandria, VA 22308 from: Geografiya i prirodnyye resursy, 1988, No. 4, pp. 98-107.     

THE USE OF REFLECTANCE OF OIL-POLLUTED SOILS IN MONITORING FROM REMOTE SENSING IMAGERY

A method is presented whereby soil reflectance on remote sensing imagery is used to estimate the level of soil pollution by oil production operations. This method can be used in lieu of the dangerous and toxic solvents now widely employed, and has the additional advantage of applicability to much larger areas. It is based on the construction of a graph and derivative equations depicting the relationship between the spectral reflectances of soils and the soil hydrocarbon content of sample soil plots. Translated by Edward Torrey, Alexandria, VA 22308 from: G. V. Dobrovol'skiyand V. L. Andronikov, eds., Aerokosmicheskiye metody v pochvovedenii i ikh ispol'zovaniye v sel'skom kho-zyaystve: sbornik nauchnykh trudov [Remote Sensing Methods in Soil Science and Their Utilization in Agriculture: A Collection of Scientific Works]. Moscow: Nauka, 1990, pp. 161–165.     

高分辨率遥感影像在重庆市土地利用动态监测中的应用

本文阐述了高分辨率遥感影像应用于重庆市土地利用动态监测的技术过程及方法。详细介绍了监测过程中运用到的遥感影像几何纠正、融合、变化信息提取等处理技术,同时针对重庆市复杂的地形情况给出了一定的建议。为遥感技术在重庆市国土资源规划、管理中的深入应用提供一定参考。     

REMOTE SENSING IMAGERY AND ECOLOGICAL PROBLEMS OF COASTAL ZONES

Materials derived from aerial and space surveys provided the basis for compilation of a series of geomorphological, geological, morphostructural, and lithodynamic maps as well as a map showing sources of manmade pollution. A derivative map of geoecological regions based on these maps makes it possible to estimate qualitatively the ecological state of any coastal region. The Buor-Khaya Gulf of the Laptev Sea is used as a test case. Translated by Edward Torrey, Alexandria, VA 22308 from: Geomorfologiya, 1995, No. 4, pp. 23-34.     

成矿信息遥感影像模式的判读及统计分析在金矿床预测中的应用

本文以河南省西部熊耳山-外方山区的金矿找矿靶区预测工作为例,阐述了综合运用遥感、物化探信息,建立与控矿地质条件相一致的“线环同现模式”、“多元同现模式”的原理和应用效果;介绍了利用熵函数计算遥感成矿信息的矿化有利度和矿床统计预测过程。     

PRINCIPLES OF MAPPING THE SOIL-VEGETATION STRUCTURE IN WEST SIBERIA BASED ON REMOTE SENSING IMAGERY

Procedures for mapping the soil-vegetation structure of West Siberia on the basis of remote sensing images are described. Four principal stages are involved: interpretation of individual structures on the images; determination of the functions of the interpreted structures; classification of soil-vegetation structures; and mapping of the territory by types of soil-vegetation structures (micro-, meso- and microstructures). Particular attention is given to detection of soil-vegetation hydromorphism using remote sensing imagry, important in the swampy Western Siberian environment, and the special problems related to image scale. Translated by Edward Torrey, Alexandria, VA 22308 from: Geografiya i prirodnyye resursy, 1990, No. 1, pp. 138–145.     

REMOTE SENSING APPLICATIONS IN AGRICULTURE AND SOIL SCIENCE PART II: REMOTE SOIL MONITORING: ORGANIZATION OF SOIL MONITORING BASED ON REMOTE SENSING IMAGERY

A paper addressing general issues of concern in organizing a system of soil monitoring based on remote sensing imagery covers such topics as general principles which must be observed, the selection of indices optimizing the acquisition of useful information, and the adoption of suitable methodologies. A subsequent section of the paper focuses on the types of information to be derived from soil monitoring in arid regions and areas susceptible to heightened soil erosion. Translated by Edward Torrey, Alexandria, VA 22308 from: G. V. Dobrovol'skiy and V L. Andronikov, eds., Aerokosmicheskiye metody v pochvovedenii i ikh ispol'zovaniye v sel'skom khozyaystve: sbornik nauchnykh trudov [Remote Sensing Methods in Soil Science and Their Utilization in Agriculture: A Collection of Scientific Works]. Moscow: Nauka, 1990, pp. 154–161.     

INTERPRETATION OF REMOTE SENSING DATA: THE FOREIGN EXPERIENCE

A report by Soviet delegates to the 15th Congress of the International Society for Photogrammetry and Remote Sensing provides a forum for a Soviet assessment of foreign achievements in the practical application of remote sensing in various fields of economic activity, particularly oil and gas exploration, prospecting for metallic minerals, forestry, and oceanography. Special emphasis is given to digital image processing for land-use classification, a field in which Soviet researchers are becoming increasingly involved. The Congress is described as contributing greatly to the development of common data processing methodologies and the exchange of information and opinions on common problems. Translated from: Geodeziya i kartografiya, 1985, No. 4, pp. 53-58.     

约束型贝叶斯网络在遥感图像解译中的应用方法研究

采用一种带有约束条件的贝叶斯网络来构造分类器,即特征节点被约束为类节点的子节点,子节点间允许有不同的连接关系,并将约束型贝叶斯网络的几种典型模型--NB、TAN、BAN用于遥感图像的解译中.通过评价结构的似然函数得到网络结构,综合特征节点和类别节点的拓扑和概率统计信息学习得到分类器.将这些模型用于多光谱和高光谱影像的分类,并就其性能进行探讨.     

THE POTENTIAL FOR VISUAL AND AUTOMATED INTERPRETATION OF FOREST VEGETATION WITH SCANNER IMAGERY

Feature classification maps derived from visual and automated methods of interpreting band-specific and composite imagery from the “Fragment” multispectral scanning system are compared in the study of vegetation and related features along the Gulf of Riga. The automated method, featuring a two-stage unsupervised/supervised classification algorithm developed at Moscow University (see MSRS, 1984, No. 3, pp. 255-261) provided for enhanced discrimination of wetland areas, farm land, and settlements, as well as for the elimination of extraneous components (especially the above) visually classified as deciduous forest. Translated from: Vestnilk Moskovskogo Universiteta, geografiya, 1988, No. 3, pp. 49-57 by Jay K. Mitchell, PlanEcon, Inc., Washington, DC 20005.     

覆盖区与铀矿化有关的遥感信息解译

覆盖区遥感隐伏地质信息的解译一直是遥感地质工作中难以解决的问题。本文主要讨论如何运用遥感数据处理技术结合各种资料综合分析的方法,提取覆盖区隐伏地质层体信息。通过遥感影像信息、透视信息①、地貌特征信息及水系格式信息等的综合解译,对覆盖层本身的影像信息及反映被覆盖的地质层体的遥感信息进行了研究区分,成功地解译了覆盖区与铀矿化有关的各种隐伏控矿因素的遥感信息。现以伊犁盆地为例,通过解译及综合分析,详细阐述如何解译覆盖区隐伏的控制铀矿分布的产铀建造;圈定了影响产铀建造产出条件及矿化富集条件的隆起和凹陷;同时对控制盆地形成和改造的断裂构造及影响铀成矿作用的相对稳定和相对活动区进行了分析。实践表明,遥感方法不仅对揭示该覆盖区的铀矿分布规律有意义,而且对指导该区进一步铀矿普查与找矿也是有意义的。     

基于SVM的多源遥感影像面向对象建筑物提取方法

在分析支持向量机(Support Vector Machine,SVM)分类技术和机载激光雷达(LIDAR)数据、航空影像特征的基础上,提出了基于SVM的LIDAR数据和航空影像的面向对象建筑物提取方法。结果表明,该方法充分利用了多源影像的互补信息,能够得到更高的信息提取精度,准确而快速地更新地理空间数据库,是一种有效的面向对象建筑物提取方法。     

自组织神经网络在遥感影像分类中的应用研究

竞争学习网络与Kohonen神经网络相比,由于不考虑邻域神经元,其网络结构相对简单。采用这种简化的网络结构,并对其学习算法进行改进,用最大、最小距离法设置的初始聚类中心来代替随机初始中心。实验结果表明,用改进的竞争学习网络对遥感影像进行非监督分类,在分类精度和效率上都有较大的提高。     

基于神经网络模型的遥感影像几何校正研究

在遥感影像几何校正方法中,通常认为精度最高的是共线方程模型.针对共线方程模型定向参数解算过程中误差方程的病态问题,提出了利用基于控制点的神经网络方法进行高分辨率遥感影像几何校正方法,并从理论上进行了可行性分析.实验证明,在具有一定数量控制点作为训练样本的条件下,应用BP和RBF神经网络进行遥感影像几何校正,可以达到比共线方程模型更高的精度;神经网络模型能够自动抑制含较大误差控制点对模型纠正精度的影响,在实际应用中可以提高几何校正效率.