基于形态特征的目标匹配是地图空间认知、地表位置自动感知等领域的重要方法.然而,由于不同数据源的遥感影像提取的同一个空间目标在空间尺度和局部形态上存在一定的差异,现有形态匹配算法难以获得较高的匹配精度.本文以等角边界距离编码及相似性算法为核心,提出了面向遥感影像的面状图斑匹配方法.①从面目标的质心以10.为间隔,顺时针方向从质心向外引36条射线,获取面目标的边缘点,形成等角边界距离编码;②提出了与该编码相适应的几个综合形态指标,包括自相似度、圆形度、形态复杂度和质心偏移度,并建立了面状图斑相似度算法;在此基础上,结合近邻目标的形态相似度,构建了遥感影像面状图斑匹配算法;③从不同数据源的在线遥感影像上提取中国西部湖泊与全球湖泊,开展面状图斑匹配实验.通过实验,发现本文的面状图斑相似度算法相对于按距离采样的相似度算法,召回率提升了 13.8％;在添加近邻目标相似度约束的基础上,面状图斑匹配精度达到90％以上.通过算法适应性分析,发现该算法在一定的投影变形下仍能保持一定的匹配精度;如果空间目标及其邻域目标的总体形态和分布接近,不同空间尺度的匹配精度保持在80％以上.本文的研究为地图空间认知、地表位置机器自动感知提供了新的思路和方法.

由于二维水质模拟属于计算密集型运算,如何在保证精度的条件下,提升水质模拟的性能,是流域水质模拟必须解决的关键问题.为此,从性能和精度两方面开展了流域水质模拟研究.在水质模拟性能方面,设计了通用计算框架下的流域模拟网格划分方法.将河道及边缘网格作为模拟网格,降低了网格规模,使流域级别的精细化河道水质模拟成为可能;以此为基础,将水质模拟分解为水速增量计算、水流迁移计算和污染物扩散模拟三步,并分别设计并行化算法;在水质模拟精度方面,引入了稳定流场.将模拟起始时刻的上游流量作为稳定水源,通过模型迭代建立的稳定流场近似模拟流域前期水情.最后,通过牡丹江中游河段2年的观测数据验证,发现通用计算使得水质模拟性能显著提升,传统水质模拟几个小时才能完成的任务几分钟内即可完成;同时,方法达到传统水质模拟相近的精度,这使流域实时水质模拟成为了现实.

近年来,人类活动导致气候异常,暴雨洪涝灾害频发.建立有效的径流汇流模型模拟流域径流过程具有重要意义.然而,流域前期水情误差成为进一步提高模拟精度的瓶颈.为此,以纳维-斯托克斯方程(N-S方程)为指导,提出一种基于稳定流场的流域径流汇流模型模拟方法.首先,通过对N-S方程分解,建立了栅格模式下的径流汇流模型;然后,针对模型存在"汇流水流方向异常"问题,加入水流混合模拟环节,包括栅格单元水速增量修正和邻域栅格单元水流混合模拟.实验表明,当上游供水稳定时,模型能产生稳定的水流场;最后,以岷江上游为实验流域,将模拟时段24 h前的小时上游流量代入模型创建稳定流场,模拟流域前期水情;利用600 h的实测降水数据和上游流量数据进行径流汇流模拟,获取流域出口水文站的模拟水位序列,并与实测水位对比,发现算法能提高流域径流模拟精度,在暴雨洪涝灾害的预警预报中具有一定的理论意义和实践价值.

为了提高传统径流汇流模拟的时效性,提出了一种基于通用计算的径流汇流模型.模型采用纳维-斯托克斯作为基础方程.首先,文章探讨了利用通用计算进行径流汇流模拟的实现方法并设计了模拟计算流程.然后,以涪江流域中段为研究区域,将流域内25个常规站和区域站的实况降水数据为数据源,分别利用本文径流汇流模型和FloodArea模型对流域进行径流模拟,并将两种模型模拟结果与水文站实测数据进行对比分析.结果发现,基于通用计算的径流汇流模型不仅在模拟效率上相对于FloodArea模型有很大程度的提高,而且模拟结果具有更小的水位变化误差,与水文站实测水文数据具有更好拟合效果.模拟时效性和结果准确性的同时提升表明本文的径流汇流模型对暴雨洪涝预警预报具有重要的意义.

基于分布式水文模型的径流模拟以像元为单元进行并行计算并逐小时迭代,计算量大,故将OpenCL引入流域径流模拟中,以提高径流模拟效率.由于普通计算机显存有限,面对流域范围大、分辨率高的情况,一般以分块计算方式解决;虽然分块计算能解决显存不足问题,但频繁的显存和内存的数据交换降低了通用计算性能.该文尝试将分布式计算与通用计算相结合,将多台计算机显存资源整合,以避免频繁的显存与内存数据交换.首先将流域数据分为单台计算机能够存储和处理的子块,然后通过网络送至服务器不同的计算机上调用OpenCL计算;每执行完一次径流模拟,服务器相关节点通过网络交换计算后的边缘数据.实验证明,分布式计算和通用计算相结合能快速完成大流域的径流汇流模拟,具有一定的应用价值.

Pearson-Ⅲdistribution is widely applied in various fields such as meteorology, hydrology, agriculture and so on. Pear-son-Ⅲprobability density equation has three parameters to be estimated, including the shape parameter of alpha, the scale parame-ter of beta and the initial parameter of alpha-0. The precision of estimating the three parameters is one of the key focuses when us-ing Pearson-Ⅲdistribution to forecast rainstorm intensity in actual applications. Without considering the effective range, the tradi-tional method used in estimating Pearson-Ⅲdistribution??s three parameters may cause alpha-0 to be less than zero, thus making the predicted rainstorm intensity to be negative. This is impossible to happen when predicting the real physical quantities such as the rainstorm intensity, flow velocity, water depth and so on. Strict theoretical derivation has proved in this paper that the results of rain-storm intensity estimation would be nonnegative as long as alpha-0 maintained nonnegative. Based on the above conclusion, an im-proved algorithm aiming at fixing the drawbacks of the traditional estimation algorithm was proposed in this paper. Firstly, a series of parameters are estimated by iterative calculations based on the Least Square Method. Secondly, the sum of the squared deviations was calculated for each parameter group and each group was sorted by its value for the sum of the squared deviations. On that basis, different alpha-0 values are calculated for each group respectively to exclude groups with negative alpha-0 values. Finally, one group of parameters which has the minimum sum of the squared deviations was selected as the optimal combination for the three pa-rameters of Pearson-Ⅲ. In the end, this algorithm was verified by estimating the rainstorm intensity in Sichuan Province. The result shows that the proposed algorithm can assure the alpha-0 to always be greater than or equal to zero and the value of rainstorm inten-sity to always be in a reasonable range. Furthermore, it makes more weather stations pass the distribution fit test with a 95%confi-dence level. In short, the improved algorithm is more reasonable and accurate than the traditional algorithm, and it provides a refer-ence to the preparation of city rainstorm intensity formula.

In order to analyze the performance of different algorithms for data interpolation,an evaluation system is setup.Using the visual development tools-Surfer software that integrated in VB 6.0,12 kinds of interpolation methods can be used to implement data interpolation.Geosciences data,such as contour,three-dimensional,etc,can be displayed rapidly for different users;moreover,quantitative assessment of the results of interpolation can be achieved.Taken the average temperature of HuNan providence as an example,experiment results shows that the performance of interpolation algorithms differ with specific conditions.Meanwhile,spatial distribution and tendency of the data can be predicted and summarized from the obtained overlay chart.

Geometric rectification of topographic map depends on the intersection coordinates of kilometers grid.It′s possible to make the automatic rectification if the intersections can be extracted and their coordinates be defined accurately.This paper proposed a method of real-time and high-accuracy geometric rectification.Firstly,a rapid algorithm is applied to detect the grid lines from topographic map based on an array to decide whether each pixel is on grid line or not.When detecting the first pixel of each row of the binary topographic map,the array is initialized which includes the former pixels of detected angle within half detected length.To detect the latter pixels of each row the algorithm just needs pop-up the bottom pixel of the array and adds a new pixel of detected angle at half detected length.Secondly,the intersections are extracted accurately on the detected image by pattern recognition and Hough Transform formula.Unfortunately,the result includes many invalid points due to the interference of other line elements and noise on the map.Considering the parallel character of kilometers grid and each of them has an equal interval,a key algorithm based on Hough Transform is proposed to filter the right points from extracted points.As a result,invalid points are removed and the coordinates of intersections are computed automatically when the user gives one projective coordinate among the final points.The study shows that the method can be applied to calibrate the topographic map effectively and accurately.

This paper presents a new remote wireless monitoring system,which the current 3G and 3S technology integration,the use of 3G high-speed wireless access technology provides,3S provides high-definition remote sensing images,as well as global positioning and the corresponding service platform to realize remote control and monitoring.The system has control convenience,no duty,able to adapt to a variety of complex environments,real-time observation and so on.

A simple framework for embedded wap browser,with modular,customizable,easy to expand the properties.And focuses on the analysis WML,script processing,display interface module to achieve specific,the browser can be achieved basic wap browser functionality,with high efficiency,easy-to-achieve characteristics.

The measure and analysis of surface cover spectrum are very important for theoretical as well as application study of remote sensing.The spectrum data of the different surfaces including wheat with different cover,soils with different condition,water body,asbestos tile,cement ground,and stone ground are gotten by using SE590 hyperspectral radiometer in Tengchong county of Yunnan province.The spectrum curves of the surfaces are made and the spectrum feature analyzed based on the spectrum data.It is shown that:the reflectivity of wheat increases as its cover decreases in the wavelength ranging from 400 nm to 700 nm;the reflectivity of wheat increases as its cover increases in the wavelength ranging from 800 nm to 1 000 nm;the reflectivity of dry soil is lower than that of the moisture soil in the wavelength ranging from 368 nm to 696 nm;the reflectivity of dry soil is higher than that of moisture soil in the wavelength ranging from 726 nm to 1 100 nm;as for road surfaces,in the wavelength ranging from 400 nm to 1 000 nm,cement ground and asbestos tile have highest reflectivity,stone ground has middle reflectivity,contracting to the lowest for water body;and,when the wavelength is from 373 nm to 632 nm,it is found that the slope of spectrum curve for cement ground is higher than those of asbestos tile and stone ground.

A quick thinning algorithm based on labeling regions is proposed.This algorithm belongs to gradually erosion method,and although the shortcoming of gradually erosion method cannot be avoided,some improved measures have been proposed to make up for it.Experimental result shows that the algorithm is effective,and it lays a solid foundation for quick and automatic line vectorization.

It is of great importance to the study of the dynamic extension of a city and its peripheral towns for simulating,predicting,adjusting and controlling urban extension,and for rationally protecting the farmland and using land resources.The extension of Chengdu and its peripheral towns is studied here by using LANDSAT TM satellite images respectively acquired in 1987 and 2000.Firstly,The images are registered to the GIS data.Secondly,the Different Building Index(DBI) is formulated by LANDSAT TM7-TM4 on the basis of analyzing the remote sensing image mechanism of the cities and towns and their background,which reflects the situation of construction.The model of extracting Chengdu city and its peripheral towns from DBI image is formulated by the threshold method.The spatial data of Chengdu city and its peripheral towns for the two periods is extracted from the remote sensing images of the two periods by calculating DBI and applying the model.Thirdly,the sprawl data of Chengdu city and its peripheral towns are obtained by overlaying the spatial data of the two periods.Finally,the sprawl data of each analytical unit obtained by masking or overlaying the sprawl data with the administrative boundary,the third circular road and the highway around the city are analyzed.It is shown that the spatial data of the two periods can be effectively extracted by using DBI and the extraction model.The GIS technology is especially suitable for analyzing the sprawl of the city in each unit.The sprawl area in the highway around the city is about 11132 hm2,which is 1.3 times of its original area.The sprawl area in the third circular road is about 6546 hm2,which is 0.9 hm2 of its original area,and mainly occurrs in the western,northwestern and southwestern parts,and most of which are external extensions.Between the third circular road and the highway around the city,the sprawl mainly occurs in the northwestern and southwestern parts,expanding mainly along the road.The sprawl pattern is mainly caused by the construction of high technology area and the spatial difference of investment and residential environment.In the five districts,Wuhou district is of the highest sprawl,which is 2.5 times of its original area.The sprawl area is over two times of its original area for each peripheral town,which is higher than the sprawl multiples of the city.

It is difficult to automatically and accurately extract vector data from thematic map because of annotations.Therefore,a modal of removing annotations from thematic map named Background Painting is discussed in this paper.Firstly,the color difference equation LCH is applied to the thematic map to identify annotations.When the maximum thresholds of ΔL(lightness difference) and ΔC(chroma difference) and ΔH(hue difference) are defined,the pixels of thematic map fall into two groups:background pixels and annotation pixels.As a result,the thematic map can be considered as a binary image.Then,a method similar to the algorithm of dilation is used to remove annotations.Because mainly used for binary image,the algorithm of dilation is developed in this paper.The annotation pixel value is replaced by its nearest background value but not a single value.Finally,the model for removing annotations from thematic map is developed based on ArcObjects and integrated with ArcGIS.The study shows that the model can be applied to remove annotations from thematic map effectively,so that the vector data can be obtained automatically and accurately from thematic maps by using ArcScan of ArcGIS.

A method for classifying contour line on the topographical maps based on the angle detecting is presented.It can detect the subtle changes of the line width and keep the continuity of the contour line as far as possible.This method can produce a better base map for the vectorization.