Extracting Residential Area from TM Imageon the Basis of Knowledge DiscoveredYANG Cun-jian(Institute of Remote Sensing Applications,The Chinese Academy of Sciences Beijing100101China)ZHOU Cheng-hu(Institute of Geography and Resources,LREIS,The Chinese Academy of Sciences Beijing100101China)Abstract:Rapidly and accurately acquiring residential area is very important for evaluating the damage in-volved in such aspects as building and property. It is cost-labor and cost-time to investigate residentialland in field because of the large extent distribution of the residential area. Remote sensing technology pro-vides important information for investigating residential land,but how to rapidly get residential land fromimage is a problem urgently to need be solved. Therefore,The Landsat TM image covering Chengdu plainis selected for case study because of the fine background of the residential land. The spectrum knowledgeof two type residential area is discovered, which is used to develop two formulas with combination of rela-tionship knowledge between residential, water and roads. It is shown that not only the residential landconsisting of building with cement and tile roof but also the residential land consisting of building withgrass roof can be extracted with the accuracy of 85 percent.Key words:Knowledge discovering, Residential area extraction, Landsat TM image

Experimental Studies about Methods on Optimal Bands Selection ofHyperspectral Remote Sensing DatasetsLIU Jian-ping, ZAHO Ying-shi, SUN Shu-ling(The Graduate School,University of Science&Technology of China.Beijing100039China)Abstract:The intrinsical relationships of the“entropy”methods on optimal bands selection in multispectralremote sensing data interpretation are analysed,and limitations of using these methods in hyperspectral re-mote sensing data interpretation explained.The optimal bands selection principles and methods based onthe classes distinguishability are put forward.The validity,accuracy,limitations and computational com-plexity of the processing methods concerned are illustrated throuh experiments.Key words:Hyperspectral remote sensing data, Optimal bands selection, Integrated entropy, Distin-guishability

The Tentative Study on the Status Quo of the Dawenliu Nature Reserveat the Yellow River Bayou Using the Remotely Sensed ImageLIU Qing-sheng, LIU Gao-huan, YE Qing-hua(State Key Laboratory of Resources and Environmental Information System,The ChineseAcademy of Science,Beijing100101,China)CHANG Jun(Population,Resource and Environment Academy,Shandong Normal University,Jinan, 250014China)Abstract:The Dawenliu nature reserve is one of national nature reserve, located at the bank of the YellowRiver, which is the newest land and changes most quickly. It is the habitat of many kinds of the rare birdsand the natural experiment field to study the evolution of the salt vegetation. So it is important to investi-gate its land use status quickly and exactly. On the base of the Logistic spectral unmixing program whichwas developed by us, the land use status quo of the Dawenliu nature reserve is surveyed using the LandsatTM image and field GPS data. Firstly in order to reduce the interferential factors, we cut the registeredTM image by the registered vector data of the border of the Dawenliu nature reserve in 1997. Because TMband six has low spatial resolution, it' s not used. Secondly the atmospheric calibration of the subset imageis done by the Internal Average Relative Reflection in order to reduce the path radiation and partial terraineffects. Thirdly seventy-two training endmembers of eight surface cover types are selected. Fourthly theimage is spectrally unmixed by the logistic model. As the result, the abundance images of the endmembersconstituting the area of a pixel are gotten, and the classification image is given, and the areas of the eighttypes of the surfaces are obtained. Compared with the spatial data in 1997, the land use status quo changesremarkably. Finally make the conclusions and discussions.Key words:Dawenliu nature reserve, land use status, spectral unmixing

Study on Variation of Land Surface Temperature in ChinaBy Using NOAA/AVHRR DataYAN Hao, WANG Chang-yao, NIU Zheng, YANG ZhenLARSIS,Institute of Remote Sensing Applications,The Chinese Academy of Sciences,Beijing, 100101ChinAbstract:By using a suitable split window algorithm on large scale, the authors retrieved the yearly andmonthly land surface temperature (from 1982 to 1993) from the thermal bands of NOAA～AVHRR dataset over China. The result indicates that the variation of LST has some features and is different from thatof air temperature.As to LST, the hottest part is in Northwest drought region, the higher LST part is in Southeast re-gion, and the low LST parts are in Tibet plateau and Northeast region. Conversely, as to air temperature,the hottest part is located at Southeast region, the higher part is located at Northwest drought region, andthe low air temperature parts are suited at Tibet plateau and Northeast region. Moreover, there is no 0OCisotherm of annual LST over China, but there exists obvious 0OC isotherm of annual air temperature. Fur-ther, the 0OC isotherm of LST in January lies higher in latitude than that of air temperature. And EasternTibet and Western Tibet have different seasonal features of LST.Our work indicates that using thermal band of NOAA-AVHRR is a feasible way to study land surfacetemperature. The variation of land surface temperature in China has great denoting significance in thestudy of the change of land cover in China and the impact of global change on China.Key words:Land surface temperature(LST), NOAA/AVHRR, Variation of temperature

JIANG Cheng, JIN Ji-yun, ZHANG Wei-li(Key Laboratory of Plant Nutrition Research of MOA,Soil and FertilizerInstitute of CAAS,Beijing100081,China)Abstract:The intention of site-specific management is to optimize grower inputs on areas much smallerthan the entire field. These areas may be as small as a few square meters in size. To manage a field onsuch a scale, data would have to be collected on a similar or smaller scale. To collect the data by handwould be very time consuming, labor intensive and destructive. In recent years, as an important compo-nent of site-specific soil management in precision agriculture, remote sensing developed greatly. Satelliteremote sensing is a promising technique, which could provide the essential, real-time and spatially continu-ous crop information for site-specific crop management. Specially, recent advances in the spatial, spectraland temporal resolution of remote sensing as well as potential positive changes in availability of remotelysensed data may make it a profitable tool for more farmers. However, higher cost still hinders itswidespread application in agriculture production. In this paper, TM remote sensing image with relativelylow cost was studied for the feasibility as a important information resource for evaluating crop yield vari-ability and designing the management unit of soil nutrients. The results showed that Vegetation Index(VI) could reflect the wheat characteristics of each growth stage. Two VI (NDVI and RVI) showedgreater spatial variability in accordance with the wheat yield in the field and the largest spatial variabilityoccurred in the late growth period of wheat. NDVI and RVI of three important growth stages of wheat(heading, stooling and jointing stages) were significantly correlated and the significant correlation betweenwheat yield and NDVI on Nov.18 (stooling stage) was found. Moreover, NDVI and RVI could supply im-portant information on some yield parameters, such as 1000 kernel weight and spike number. This re-search was preliminary and some other detailed studies would be needed in future.Key words:Precision agriculture, Yield variability, TM remote sensing, Vegetation index

Landscape Pattern Analysis Based on NOAA/AVHRR DataTaking the Mainstream of Tarim River Basin as An ExamplePENG Ru-yan1, WANG Rang-hui2, SUN Bao-sheng1(1.Resource and Environment Dept.,Xinjiang Engineering College,Xinjiang Urumqi, 830008China;2.Xinjiang Institute of Ecology and Geography,The Chinese Academy of Sciences,Urumqi, 830011China)Abstract:Based on NOAA/AVHRR data processing with gray level scaling, histogram equalization, multi-channels composition and so on, the macro-landscape patterns in the mainstream of Tarim River basin areanalyzed using the spatial function of IDRISI software. The results show that landscape types is relativelysimple in the mainstream of Tarim River basin. Meanwhile, every type has low fragmentation and goodconnectivity, but their evenness is bad. This is a typical vulnerable ecological landscape pattern in conti-nent river basin in arid zone.

ZHANG Xiao-yu1, CHEN Yu-ying2, SU Zhan-sheng1,ZHOU Hui-qin1, MA Yu-ping3(1.Ningxia Meteorological Institute, 2.Ningxia Meteorological Observatory,3Ningxia Meteorological office,Yinchuan750002,China)Abstract:The frost occurred is lies on many factors, such as condition of weather, landform, soil texture,bacterial population and vegetation itself etc. To monitor the frost by using remote sensing, we have tosolve many problems, for example, the precision of temperature retrieval and distribution of vegetation byusing NOAA data. In addition, we have to consider to both the aerosol and humidity effect on the temper-ature retrieval. Due to above mentioned reasons, monitoring the frost by remote sensing is rather compli-cated and difficult.Based on the analysis of the characteristic and the law of the frost, using the data of NOAA andweather station during 1991 to 2000,as well as using the data of the frost damage, there are three methodswere introduced to monitor the frost in Ningxia in this paper. The first is difference vegetation index(DVI) which monitoring the frost by means of the difference of vegetation index between fore-frost and af-ter-frost. The second is temperature reference (TR) which monitoring the frost by compare to the differ-ence between the temperature retrieval and frost index of crops. The third is the area of cold valley(ACV) which obtained ACV by integral the temperature below the frost index. We have examined themonitoring precision of those methods, the results show that each is available to monitor the different typeof frost in Ningxia.Key words:Vegetation index, AVHRR, Minimum temperature, Frost, Remote sensing

The Application of GPS in the Airborne LaserSearch for Underwater TargetZENG Yi, LIU Wen-yu, MA Yong(Dept of Electronics and Information,HuaZhong University of Science&Technology,Wuhan430074,ChinaAbstract:That how to search the target quickly and effectively and get its exact position information is animportant part in the detection of underwater target with airborne laser. So we developed a method, whichapplied GPS in the airborne laser search for underwater target. In this method, we regarded GPS signal asa parameter of detection signals and executed real-time transportation and storage with them. Based on it,we could accomplish highly effective search by cooperation of multi-airborne laser and position every detec-tive signal exactly. This made detection information fully used even in off-line analysis. Such method hasbeen simulated in computer and its feasibility has been tested.Key words:GPS, Search, Airborne laser, Simulation, Efficiency of search

A Review of Regional Soil Salinization Monitoring by Remote SensingGUAN Yuan-xiu, LIU Gao-huan(The State Key Laboratory of Resources and Environment Information System,Institute of GeographicSciences and Natural Resources Research,The Chinese Academy of Sciences,Beijing100101,China)Abstract:Globally, an estimated 9.5×108hm2of land are affected by salinity and sodicity. Soil salinizationis the major land degradation problem of the world. The information on saline soil is prerequisite for salinesoil amelioration and sustainable development of agriculture. By virtue of synoptic view of large area in dis-crete bands of the electromagnetic spectrum, multi-spectral data acquired from spaceborne platforms, re-mote sensing has been found to be useful in detection, mapping and monitoring of salt-affected soils. Theinternational and national development of soil salinization monitoring based on remote sensing is intro-duced. Overseas, satellite remote sensing has been used to monitor salt-affected soil since 1970' s. Multi-spectral and multi-temporal data was been widely used in the 1980' s. During the time, visual interpreta-tion approach was generally used. Since the 1990' s, data has been enriched. Research category becomeswider than before. But visual interpretation is still the major approach used to extract information on salinesoil. In China, satellite remote sensing has been used to monitor saline soil since 1980' s. It is 10 years lat-ter than overseas. Fundamental technical, monetary and data handling limits act collectively to constrainthe information extraction on saline soils. The methods of regional soil salinization monitoring are visualinterpretation and digital image processing. The approaches of saline soil dynamic detecting are pixel-by-pixel comparison and post-classification comparison. The digital image processing will be the major meansin the future. The progress achieved in remote sensing technique and digital image processing will greatlyenhance the capability to monitor saline soil.Key words:Regional soil salinization, Dynamics, Monitoring by remote sensing

Remote Sensing Application Research in Ecological EnvironmentConstruction of the West ChinaZHOU Ri-ping(Remote Sensing Application Institute,Aerophotogrammetry&Remote SensingBureau of China Coal,Xi'an710054,China)Abstract:Ecological environment and water resource are two key problems in the west development. Tra-ditional methods are difficult to develop environment investigation and monitoring quickly due to broad ter-ritory, mountainous topography, bad environment condition and inconvenient traffic. Remote sensingtechnology, which has been made rapid progress during recent decade, can develop ecological environmentsurvey and search underground water. In these fields, traditional methods can not substitute for remotesensing technology. In this paper, the author demonstrate the technological methods and routes of ecologi-cal environment investigation and monitoring and underground water search in the diversified areas bymeans of remote sensing technology, it can be taken as inference for government officials, experts andtechnicians.Key words:Ecological environment, Water Resource, Dynamic monitoring, Geographical information sys-tem

The Production of Objective Region with the Shape FactorGUAN Jia-ying, LI De-ren, GUAN Ze-qun(Information Engineering Institute,Wuhan Technical University ofSurveying and Mapping.Wuhan430079,China)Abstract:In the processing of the remote sensing imagery interpretation, the needs of selecting the objectsthat we are interested become more and more important. But much of the work is still accomplished byhandwork presently. In order to change this situation, to improve the automatic analysis of the computer,this paper makes some discussion of the objects' automatic detection that we are interested.In color edgedetection, we used to applying the same method as monochrome images processing, thus we can' t takegood advantage of much color information. The idea of using vector order Prewitt operators to edge detec-tors can help us make full use of the color information. On the other hand, many Remote Sensing imagesare multi-wave-band. Therefore, using vector order Prewitt operator is more useful than conventional Pre-witt operators in processing the imagery of Remote Sensing.The“Snake”algorithm is a famous active contour model, which is often used when we need a precisecontour of an object in image processing field. This method will be more reliable if we consider the shapefactor-the relationship among all the Snake points in the procedure of finding the minimum energy posi-tion.According to these approaches, this paper introduces the concept of“vector order Prewitt operators”into the snake algorithm with the shape factor. During the processing, it makes some improvement in thegradient direction algorithm. That is, it can judge the orientation of a pixel' s gradient directly, withoutcomputing the exact degree of the gradient. Consequently this helps to reduce the amount of computation.In addition, we make some argument about the threshold. We found that make threshold in different se-quences will have different effects. The effects are shown in this paper.The effectiveness of the proposed method and its value of practicality can be confirmed by the experi-ments.Key words:Vector order Prewitt operators, Snake algorithm, Shape factor, Edge detection, Objective se-lection

An Application of the Hough Transform for Removingthe Random-phase Jam in the Remote Sensing ImageGUO Qiang, CHEN Gui-lin, Tong Wei-qi(Shanghai Insititute of Technical Physics,The Chinese Academy of Science,Shanghai200083China)Abstract:Remote Sensing Images have been applied extensively in the aspects of meteorology, militray andmanufacture. However, Because of the influences of some subsystems, a little correlation jam slip into thechannels inevitably during the imaging of the MCSR, which will damage the validities of the mensurableanalysis to the remote-sensing Images and the precise identification to the remote-sensing ground-objects inlager degrees. So the correlation jams must be removed or suppressed. Based on the analysis of the meritsof the correlation jam as well as comparision of several removel ways, the article presents the Hough trans-form algorithm upon the two-demension frenquency space analysis, its application in remote-sensing imageprocessing and its performances detailly. Experiental results show, compared with the methods of neigh-bour average, median filter and classical frenquency space filter, this algorithm is better in the removingrandom-phase jam of the images and reserves richer fine textures and edge information. On the otherhand, this algorithm makes it possible that the PSNRs of the processed images by this algorithm are higherthan the other removal ways above and the subjective evaluations to the processed images are better too.At the same time, this algorithm has also been applied to the real-time acquisition and processing of aus-tronautic remote sensing system and got the successful results.Key words:Random-phase, Correlation jam, Hough transform, Two-demension frenquency space analysis

Remote Sensing Image Classification Based On Fuzzy GaussFunction Neural NetworkMAO Jian-xu, WANG Yao-nan(College of Electrical and Information Engineering,Hunan University,Changsha410082,China)Abstract:Information on land-cover/land-use gives important clues for both more effective planning andthe wiser exploitation of natural resources. Get useful information from remote sensing image of large ar-eas is a time-consuming process. Computer-aided classification has provided an alternate, effectivemethod. Improving classification accuracy is a key issue in remote sensing image classification. The maindrawback of traditional remote sensing information classification methods is its low precision. In recentyears the artificial neural network technology and fuzzy technology have been developed and applied to re-mote sensing data classification problem. Compared to the conventional statistical classifier, the neuralnetwork classifier is a non-parametric method and has the capacity of self-learning and high robust. Thefuzzy technology has the superiority in logical inference and can be used in high level information process-ing. In this paper, considering the features of remote sensing images, we synthesize the advantages ofthese two methods and proposed a remote sensing image classifier using fuzzy gauss function neural net-work. In the system, fuzzy technique and neural network technique are combined, and the fuzzy inferenceis realized by neural network. It using the learning ability of neural network to adjust the membershipfunction and fuzzy rule, which endue the system with the distinguished capability of self-adapt and self-learning. We applied the proposed method to construct a classifier for remote sensing image classificationexperiment research. Experimental results show that the fuzzy gauss function neural network classifier canbe used in remote sensing image classification, and its classification precision is superior to that of the con-ventional maximum-likelihood.Key words:Fuzzy neural network, Gauss function, Remote sensing image classification

SAR Image Segmentation Based On Markov Random Field ModelZHANG Cui, LI Sudan, WANG Zheng-zhi(Remote Sensing Group,Department of Automatic Control,NUDT ChangSha410073,China)Abstract:In this paper, we propose a SAR image segmentation algorithm based on Markov Random FieldModel. The model captures the image contextual information by making the assumption that the propertiesof a pixel are dependent upon nearby pixels and are independent of distant pixels. A two-order neighbor-hood system is used in this paper. The segmentation result is obtained by applying ICM(Iterative Condi-tional Mode)algorithm. The ICM is an iterative algorithm. During the iteration sequence, each pixel is la-beled according to the MAP(maximum a posteriori)criterion. After each pixel labeling step, the parame-ters of each segmentation class' probability density function are re-estimated from current set of pixel la-bels. The influence of speckle noise is further reduced by an outlier rejection method. Using this method,outlier pixels are excluded from current pixels' neighborhood. The algorithm is applied to SAR imageryfrom the MSTAR(Moving and Stationary Target Acquisition and Recognition)datasets. The result showsthat this algorithm can efficiently reduce the influence of speckle noise, and segment the image into threeparts of target, shadow and background. The experimental results are very satisfactory.Key words:SAR image segmentation, MRF model, ICM algorithm, outlier data

Analysis of the Soil Erosion under the Different Environment Condition——for Instance in Chongqing CityYANG Cun-jian1,2, ZHANG Zeng-xiang1, WANG Siyuan1(1.Institute of Remote Sensing Applications,The Chinese Academy of Sciences,Beijing100101China)(2.Institute of Ecology and Geobotany,Yunnan University,Kunming650091China)Abstract:In order to discover the relationship between the soil erosion status and its environment factorssuch as temperature, rainfall, vegetation index, topographic index, environment index, the analysis of thesoil erosion under the different environment condition in Chongqing City is discussed here. Firstly, a soilerosion index is defined here to describe the heavy of soil erosion in per unit. Secondly, all of theenvironment data was rasterized into 100M * 100M including topographic index, annual averagetemperature, accumulating temperatures above zero and ten degree, annual rainfall, aridity, humidity, hotindex, water index, hot and water index, environment quality comprehensive index. Thirdly all of theenvironment data were used to overlay with the soil erosion data with grid size of 100 m*100 m. Finally,the statistics analysis function of the GIS is used to discover the relationship between the soil erosion andits environment factors.Key words:Geographic information system, Soil erosion, Environment factors

Land Use Dynamic Monitoring for SuXiChangRegion by Remote SensingSUN Dan-feng(Land Resources dept.,China Agricultural University,Beijing100094China)ZHOU Guang-yuan,YANG Ji-hong(The Academy of Land Survey&planning of China,Beijing100081China)Abstract:According to the Land and resources Ministry project of 1999 land use dynamic monitoring byremote sensing, the SuXiChang region has been finished the land use change between 1998 and 1999 byusing the TM and SPOT PAN images .Adopting the multi-sensor data overlay and fusion method toenhance the change information, which mainly includes the image spatial correction, the IHStransformation and overlay method to fusion the TM and SPOT image, which enhances the changeinformation combining the spectral and textural characters, then creating the 1998 land use map and thechange image via the automatic classification method ,in which the no change pixel value is zero and changepixel value is one, and then using the visual interpretation aided by computer to determine the 1999 landuse type of change region. Combining the change region and 1998,1999 land use type to create the land usedynamic map, calculating the transfer velocity of arable land importantly and reached the conclusion-city,industry and roads were the main parts occupying the arable land ,the average velocity of arable landtransfer to industry in this region was 2.72‰,and the arable land loss was serious ,the rate of urbanexpansion was still high in this region. So establishing the SuXiChang region land use dynamic monitoringand warning system can acquire the land resource quantity and using information, necessary for landresource management. The method to enhance the change information and using the multi-sensor datafusion and automatic classification and visual interpretation only for change region improved the dynamicmonitoring speed and guaranteed the accuracy of the results.Key words:SuXiChang region, Multi-sensor fusion, Land use change

The Investigation of Geology Using Remote Sensing onPailong-layue Chachang of Chuanzang RoadLU An-xin1, SHI Zhen-tao1, BAO Han-zhang2, WANG Jian1, WANG Li-hong3(1.Cold and Arid Regions Environmental and Engineering Research Institute,The Chinese Academy of Sciences, 2.Water Resource and HydroelectricityInvestigation and Design Research Institute of Gansu Province,3.Gansu Center for Remote Sensing Lanzhou730000,China)Abstract:Geological hazards such as landslides, rockfalls and debris flows happen frequently in ChuanzangRoad and bring about great damage to peoples′life and property. How to make clear the geological statusis important to the design and maintenance and disaster' s father of Chuanzang Road .The real status ofChuanzang Road could be analyzed macroscopically by using remote sensing such as aerial photographs andLandsat TM data. As a powerful tool for spatial analysis and spatial database management, geographicalinformation systems (GIS) are drawing more and more attention in the field of the geological investigation.In this paper, the aerial photographs and Landsat TM data combining the technique of geographyinformation system were used to investigate the geology of Chuanzang Road by image geometry correctionand image enhancement and spatial analysis. It is predicted that the integrate application of remote sensing(RS) and geographical information systems (GIS) will promote the development of geologicalinvestigation.Key words:Chuanzang road, RS, GIS, Geological investgation

Using the TM Image for Monitoring Land Desertification WU Wei1,2 (1.School of Resources and Environment,Lanzhou University,Lanzhou730000,China) (2.Cold and Arid Regions Environmental and Engineering Research Institute, The Chinese Academy of Sciences,Lanzhou730000,China) Abstract: Desertification is land degradation characterized by wind erosion mainly resulted from the excessive human activities in arid, semiarid and part of sub humid regions in the North China. It is effective that using the TM image for monitoring land desertification. In this paper, the author discusses: ①the methodology of monitor desertified land using remote sensing data,②some influencing factors on the monitoring results,③the TM image characteristics the index system of the desertified land. Based on above work, an indicator system of desertification had been established. This system uses the changes of surface features as the main indicators and also considers the changes of soil, vegetation and ecosystem etc. These indicators have widespread representation and are easy to identify. A practice of monitoring and assessing the land desertification had been carried out in the Mu Us Region, Northern China, which had brought out the development and state of land desertification in different periods in the region. The desertified land area in the Mu Us Region had decreased 1 936 km2(from 32 586 km2in 1987 to 30 650 km2 in 1993). That means during this period the process of desertification was in the situation of reversion, and there were 277 km2desertified land had been controlled annually. Based on such methodology, much more work could be conducted on the development, distribution and disastrous processes of desertification in North China using remote sensing and GIS techniques, making desertification research enter quantitative analytical and economic assessment stage. Key words:Land desertification, Monitor and assessment, TM image

The Application of Satellite Remote Sensing Technique onThe Survey of Soil Erosion In Northwest HebeiQIAO Yan-xiao(Remote Sensing Centre of Hebei Province,Shijiazhuang050021,China)Abstract:On the basis of analyzing all the factors effectting soil erosion,first,each factor relative to soilerosion is interpreted using landsat-5TM image.Second,all basic erosion pattern spots are interpretedaccording to comprehensive features of soil erosion on TM image too.Third,for each basic soil erosionpattern spot,through a multi-factor judgment model used for judgemet soil erosion degrees,a numericalvalue representting soil erosion degree of the spot is worked out by all the factors on the spot with differentweight on soil erosion.Final,all the spots obtain their numerical value,and the degrees of all the spots aredetermined by the value.The study result shows that the soil erosion in Northwest Hebei is very seriously.With the exception of slight degree,the total areas of the others different kinds of erosion degrees accountfor 60.5 per cent of the whole area.The soil erosion of the basin areas is more serious than the mountainareas. Multi - temporal comparison analysis has discovered that the general trend of soil erosion isweakening because of continued soil erosion control for over ten years.Key words:Soil erosion, Satellite remote sensing, Basic erosicn pattern spot, Comprehensive judgment

Advances in Study on Microwave Remote Sensing of Soil MoistureGAO Feng1,2, WANG Jie-min1, SUN Cheng-quan2, WEN Jun1(1.Cold&Arid Region Environment&Engineering Institute,The Chinese Academy of Sciences,Lanzhou730000,China)(2.The Scientific Information Center for Resources and Environment,The Chinese Academy of Sciences,Lanzhou730000,China)Abstract:It is important to hydrology, meteorology and agriculture to develop practical satellite remotesensing algorithms for regional soil moisture. Active microwave remote sensing of soil moisture is analyzedbriefly. Then the principle, algorithms and research advances and trends of passive microwave remotesensing of soil moisture are reviewed in detail in this paper. With launch of a series of satellites with activemicrowave sensors, active microwave remote sensing of soil moisture will be emphasized. Passivemicrowave remote sensing of soil moisture has a long research history, and its retrieval algorithms weredeveloped well, so it is an important tool to retrieve large scale soil moisture information from satellitedata in the future. Combination of active and passive microwave remote sensing of soil moisture will be afocus in the future.Key words:Microwave remote sensing, Soil moisture, Brightness temperature, Backscatter coefficient

Physiological Cognition Model and MethodologicalSystem for Remote Sensing ImageLUI Jian-cheng, ZHOU Cheng-hu(LREIS,Institute of Geographical Science and Natural Resources Research,The Chinese Academy of Sciences,Beijing, 100101China)Abstract:To build geographical cognition model has been regarded as one of the most important point inthe research of Geo-understanding and analysis for remote sensing image. In this article, the theories andmethods of integrating Geo-scientific knowledge, Geographical information and Remote SensingInformation in the processing and analyzing of spatial data are synthetically discussed, and thephysiological vision cognition model is proposed based on two basic computation theory, including neuralcomputation and evolutionary computation theory. Within the proposed model, three main components,including the physiological Geo-vision emulation procedure with Artificial Neural Network (ANN) andGenetic Algorithms, Geo-optimization model for remote sensing image, and knowledge-integratedintelligent cognition system, are mainly included. Furthermore, several detailed physiological visioncognition algorithms for remote sensing image and their applying way are introduced based on the proposedframework.Key words: Remote sensing image, Neural computation, Evolutionary computation, Physiologicalcognition, Geoscientific knowledge

Study on Fusion Algorithms of TM and SPOT ImagesCHEN De-chao, ZHOU Hai-bo, CHEN Zhong-yuan, WU Jian-ping(East China Normal University,Urban&Environmental Remote Sensing OpenLaboratory,Shanghai200062,China)Abstract:With the development of remote sensing technology, there appear more skills for acquiringremote sensing database. In the same region, remote-sensing data obtained from different sensors are ableto provide multiple time and multiple resolution images to have formed image pyramid. One of effectivemethods discussed in text is the data fusion used for processing different sources of image data. Thismethod is tested in the present Landsat and SPOT dataset. To make best use of TM′s spectrum andSPOT' s resolution, four data fusion algorithms are compared in the present study, including weighingfusion, K-L transform, K-T transform, and feature-based fusion. In terms of information content andclass accurateness, the feature-based fusion is selected as the optimum one.Key words:TM, SPOT, Image Fusion, Quantificational Evaluation

A HT-based Algorithm of Multi-Resolution Image Fusion LU Lim-ing, WANG Run-sheng (ATR State Key Lab.,National University of Defense Technology,Changsha410073,China) Abstract:The notion of image fusion was conceived about 1970s and fusion algorithms pay much attention to information optimization. It has been widely used in image processing of remote sensing because of its superiority over the traditional methods. After having analyzed algorithms of remote sensing image fusion using multi-resolution analysis(MRA), this paper expatiates on relations between Gaussian-Laplacian pyramids and Hadamard transform. A new algorithm based on predicting HT coefficients is developed and the experiments are performed multi-resolution images. The studies show that the algorithm hight MRA/ HT has good performance in preservation of the spectral fidelity as well as in resolution enhancement, especially acting on the bands with different spectrum. Most of the fusion algorithms of remote sensing image are based on MRA method for the purpose of improving resolution of multispectral images. The fashionable algorithms are divided into two typical classes: Gaussian-Laplacian pyramids and wavelet-based algorithms. In course of implementation, we found their common limitations that the algorithms could not work well for two images without spectral superposition. MRA/HT put forward in the paper solved the problem partially. We use panchromatic band image of SPOT to sharpen three multispectral bands in the experiments. As the result, MRA/HT promises to accurately maintain the spectral characteristics of the original low-resolution image XS3 which has little spectral overlap with panchromatic band image. Key words: Multi-resolution analysis (MRA), Image fusion, Gaussian-laplacian pyramid, Hadamard transform(HT).

Auto-generalization of Maps in the GIS EnvironmentMA Rong-hua1, PAN Tao2, XIE Shun-ping1(1.Dept of Urban&Resources Sciences,Nanjing,Jiangsu, 210093;2.Xinchang Railway Cor.LTD,Nanjing,Jiangsu, 210093China)Abstract:The new era of study on auto-generalization has started based on the GIS visualization and spatialanalysis tools. On the one hand, the connotation of automated map generalization has been broadened. Onthe other hand, automated map generalization advanced the GIS development, and enlarged the applicationfield of GIS. In fact, it is the important part of GIS system or its functions to the automated mapgeneralization . Approach and means of automated map generalization are more and more plentiful andgreat progress has been made. The paper points out geo-features-oriented auto-generalization supported bydifferent layeres in GIS. The experiment suggests it is practical, but most problems are still left to besolved, and its total solution must be based on the knowledge generalization.Key words:Auto-generalization, GIS, Layered, Geo-featurs

Geoinformatics in Environmental ManagementDU Jing-long, ZHANG Zhi-min, ZHANG Tian-qiao(ECNU Urban&Environment Opening Lab of Ministry of Edncation,ShangHai200062China)Abstract:Over the past number of years there has been a growing usage of technology in environmentalmanagement. and indeed the greater capacity of the improving technologies to provide effective tools forenvironmental management.Geographic information technologies, or geoinformatics, which differentiatethemselves through their ability to handle the spatial component of data, clearly have a key role inhandling environmental data. Environmental processes have a definite spatial element which needs to bespecifically considered if those processes are to be understood and managed.Geographical Information Systems (GIS) are powerful information management and analysis tools.The integrated digital approach identifies different stages: data collection and field support technologies;digital ' workbench' technologies which include the integration of GIS with other techniques for dataanalysis; and the transfer and ultimate utilisation of processed information. GIS is just one of a suite oftools required to meet the tasks involved in each of these stages For efficient environmental managementwith an improved likelihood of achieving specified objectives, an integrated suite of processes from datacollection, to data handling and analysis, to information presentation, is required.Key words:GIS, GPS, RS, DGPS, Windthrow, Precision Agriculture

Construction of Urban Outdoor AdvertisementInquiring System On WebGISLI Wei, XIE Wen-jun, WEI Yu-chun, XU Xiu-ying(College of Geographic Science,Nanjing Normal University,Nanjing210097China)Abstract:Urban outdoor advertisement is an important constitution of urban spatial information. It is themain stream of current development and study on urban geographic information system as to how toeffectively manage urban spatial information using geographic information system. It is of greatsignificance to set up an inquiring information system of urban outdoor advertisement based on Internet bymeans of combining the urban geographic information system with Internet technology. By the comparisonand analysis of all kinds of WEBGIS, the system framework of construction of urban outdooradvertisement based on Internet technology is advanced in this paper firstly. Then the system is furtherconstructed where Java language was adopted to develop Java Applet to realize the system functions.Through sending a Java Applet request command to the server from the client, the Java applet will bedownloaded to the client browser, the server accept the request, and then do the correspondingprocessing, at last provide the results in JPEG or GIF files to the client and display the outputs by JavaApplet. The main functions of the system include the graphic display of urban outdoor advertisement,query from graphics to attributes and the attributes to graphics, spatial information searching and statisticanalysis.Key words:WebGIS, Urban Outdoor advertisement, Urban GIS134　　　　　　　　　　　　　　　遥　感　技　术　与　应　用　　　　　　　　　　　　　第16卷

A Study on Distribution of Vally Category in HuangshuiRiver Based Upon GISWU Qaunyuan, KONG Min, HOU Wei(Shandong Normal University,Jinan250014China)MENG Fanhai(Longkou Bureau of Water Conservancy,Longkou264200China)Abstract:The Huangshui River Drainage Area is taken as an example in this essay,quantitative analysis tothe landscape characteristic of the drainage area by using GIS software ARC/IMFO, the classifiedfoundation,method and procedure of the type of the drainage area' s underlyingsurface are explained,andthe classified results are rationally evaluated.Key words:GIS, Drainage area, DTM, DEM

Polarimetric SAR InterferometryYANG Zhen, YANG Ru-liang(Institute of Electronics,Chinese Academy of Scienses,Beijing100080,China)Abstract:The development of radar polarimetry and radar interferometry is advancing rapidly. With radarpolarimetry, the textural fine-structure, target orientation, systemmetries and material constituents canbe recovered with considerable improvement above that of standard amplitude-only radar; with radarinterferomety, the elevation structure of a target can be explored. In Polarimetric Synthetic ApertureRadar Interferometry (Pol-InSAR) imaging, it is possible to recover such co-registered textural andspatial information from Pol-InSAR data sets simultaneously. Pol-InSAR technique significantly improvesthe performance of conventional scalar radar interferometry by introducing radar polarimetry technique intoSAR interferometry. Polarimetric SAR interferometry generates different interferograms of the samescene. Each interferogram is associated with a particular scattering mechanism. The interferometricanalysis of a full coherent polarimetric data set enables the combination of the final structure properties andthe spatial information of the target. Polarimetric Synthetic Aperture Radar Interferometry is a potentialtechnique, which not only allows a more sophisticated physical interpretation of scattering mechanism butalso improves the accuracy of the interferometric measurements. In this paper we introduce thedevelopment of this technique and its basic principles, and then discuss some key problems that must besolved. Finally, we present the applications and developing tendency of Pol-InSAR technique.Key words:Synthetic aperture radar, Polarization, Interferometry, Polarimetric SAR interferometry, Theoptimum polarimetric coherence

Remote Sensing Systems on Board Autonomous Miniature AircraftCHEN Hong-bin1, MA Shu-qing2, WANG Gai3, XUAN Yue-jian1,PAN Yi3, LI Qian3, FAN Ke-ping1, WANG Pu-cai1, LU Da-ren1(1.LAGEO,Institute of Atmospheric Physics,The Chinese Academy of Sciences,Beijing100029,China;2.Atmospheric And Meteorological Observatory,CMA,Beijing100081,China;3.Institute of Meteorological Sciences,Jiangxi330046,China)Abstract:Aircraft is an important platform for remote sensing. This paper presents a remote sensingsystem onboard autonomous miniature aircraft which has been developed principally for atmosphericsounding. The advantages of this system are summarized as follows: (1) high flexibility in fieldexperiments with no requirements of specific runaway and weather conditions for taking off or landing;(2) relatively economical cost without any dangers for operators; (3) adequate clarity of the aerial photo-pictures and video images obtained with quite high resolution; (4) superimposition of GPS information onvideo imagery. So, this very economical and flexible system of remote sensing can be used in many fields,for example, in monitoring landuse, grassland phytomass, and weather or geological disasters.Key words:Autonomous miniature UAV; Aerial photography; Airborne video system; Remote sensing ofgrassland

Compression of Hyperspectral Image Based on Spectral ClassificationWANG Zeng-zhu, LIU Tong-huai, HUANG Lu(Department of Electronic Science and Technology,The University ofScience and Technology,Hefei,230026,China)Abstract: A new method based on spectral classification is proposed for hyperspectral image datacompression. Original hyperspectral image is reduced to a cluster table and pixel values that are nowindices to the cluster table.To be able to perform classifying,it is very important to define aprecise measureby which minimum error is acquired.This paper describes three techniques for spectral classification at thecost of some loss in fidelity of the reconstructed data.They are Wavelet Transform, Euclidean Distance,and Maximal Distanc. Wavelet Transform gets the best reconstructed image with the most time expense.The performance of Euclidean Distance is inferior to Wavelet Transform,but using much less time,Maximal Distance gets the worst result ,while taking the least time.The compression/decompressionquality can be evalulated using Signal-to-noise Ratio(PNSR).It is experienced that these three techniquesdescribed can achieve compression ratios of greater than 45:1 with average PNSR of 39.65db,41.33db,43.45db.Key words:Hyperspectral image, Spectral classification, Data compression, Wavelet transform

The Implementation of Moving Target Imaging ProcessorZHANG De-feng, WANG Yan-fei, YANG Ru-liang(The Institute of Electronics,The Chinese Academy of Sciences,Beijing100080,China)Abstract:Aimimg at the implementation of moving target imagination(MTI), a new algorithm which ishighly compatible with the stationary target imaging methods is presented, and a real-time imagingprocessor of moving targets according to the algorithm is developed based on DSP technology. Thealgorithm, the implementation of moving targets imaging processor and the resulted moving targets imageare introduced in the article.Key words:Synthetic aperture radar, MTI/MTD, DSP

Automatically Remove Impulse Noise from Telemetric Terrain DataYIN Wei-ming, LIU Jian, TIAN Jin-wen(Department of Electronic and Informational Engineering,HUST,Wuhan430074,China)Abstract:As for impulse noise removal from terrain data, traditional algorithms either usually requirethresholds manaully selected , or can' t effectively preserve useful signal when filtering noise. To resolvethe problem, This paper presents a new adaptive algorithm . The algorithm, which based on static theoryand fuzzy mathematics, evaluates the average and variance of each local data group (or named filterwindow), and then uses the results to automatically select thresholds for each window to detect andsmooth impulse noise. The experimental results show clearly that the growth of GSNR using thisalgorithm is much higher than those of other commom algorithms and similar algorithms whenever densityof impulse noise is below 5%, and the algorithm is more suitable for processing steadily changed siginal.Key words:Terrain telemetry, Impulse noise detection, Adaptive filter, Fuzzy membership function,Statics

Spatial Detect Technique Applied on Earthquake's Impending ForecastGUO Zi-qi1, HU Gui-wen1, QIAN Shu-qing2(1.Laboratory of Remote Sensing Information Sciences,Institute of Remote Sensing Applications,The Chinese Academy of Sciences,Beijing100101,China;2.Institute of Geophysics,Chinese Seismological Bureau,Beijing100081,China)Abstract:The impend forecasting of strong earthquake is still an unsolvable scientific question in theworld. Based on our analysis of the disadvantages of present earthquake forecasting, this article discussedthe probability of spatial detect technique that is used in earthquake' s forecasting, and then summarizedthe status, extent and some unanswered questions of those new techniques and methods in domestic andinternational applications. In earth' s crust, where there is the broken proceeding, there is the changing themedia of the energy between inside and outside. The phenomena can be caused by crust stress changingunder ground, temperature up on earth surface and low frequency electromagnetic emission in atmosphereby interface match of various solid earth and atmosphere circle. Furthermore, it is possible to acquirestrong earthquake information from space remote sensing data. In so far, three sorts of strongearthquake' s information can be detected using various sensors of satellites. There is crust deforming ingeologic body, thermal abnormal on surface and electromagnetic over atmosphere. We can surely believethat successive, dynamic, systematic and multi-parameter stereoscopy observation will be a developmenttendency of earthquake' s impending forecasting in future. Through spatial detect technique, we can realtime capture the correlative dynamic information before earthquakes. Low price, covering more area,faster, more information, more strong directing perception and continual observing are characteristics ofapplication of remote sensing technology, which compensate the inadequate of seismic stations on theground and improve present system of earthquake' s monitoring and forecasting. There is fairly reason totrust that the short-term predicting precise degree of earthquake information must be improved.Key words: Impending forecast, Earthquake precursor, Spatial detect technique, Remote sensingtechnology

On Forest Volume Estimate Research Based onRS and GIS through Ridge EstimateLI Chong-gui1, SHI Qiang1, ZHAO Xian-wen2, SI Lin2(1.Shenzhen Polytechnic,Shenzhen518055,China;2.The Research Institute of Forestry,CAF,Beijing100091,China)Abstract:In this paper, based on the analysis of possible limitation of using LS estimation predicting forestvolume, the ridge estimate is put forward. Meanwhile, by means of the method that makes the deviationof the total predicting volume minimum, the ridge parameter is obtained through computer simulation.The real example indicates that when there are some multi-collinearities between the RS and GISinformation that effecting volume estimation, ridge estimate is prior to LS estimate in forest volumepredicting.Key words:Ridge estimate, Variance inflation factor, Multi-collinearity, Forest volume

An Application of Brovey Image Fusion in Land CoverInvestigation in Yinda Irrigated RegionWANG Jian1, LU An-xin1, GUO Ting-tian2, MIAO Chun2(1.Cold and Arid Regions Environmental and Engineer Research Institute,The Chinese Academy of Sciences,Lanzhou730000,China;2.Gansu Yinda Irrigate Project Management Center,Lanzhou730000,China)Abstract:Image fusion is one of widely used techniques to enhance image information of remote sensing,which could be used in different geometric and quality images to generate a new image with same geometricfeature and quality by transform of mathematics and physics. The investigation of land cover in Yindairrigated region was selected as cast site. According to the characteristics of land types and the requirementof spatial resolution of the investigation, panchromatic Spot image and multi-spectral TM image wereused to generate new false color composite image with 10m spatial resolution by using Brovey imagefusion. The precision of image was analyzed and tested. The results show that the fusion image could meetthe demand of application.Key words:Image fusion, Land cover investigation, False color composite image

Development,State-of-the-art andApplications of Ultra-Wideband RadarLI Hai-ying, YANG Ru-liang(Institute of Electronics,The Chinese Academy of Sciences,Beijing100080,China)Abstract:Ultra-Wideband(UWB) radar, well known for its high range resolution,powerful penetration,low probability of intercept and robust jamming immunity, is paid more attention by militaryrequirements,commercial usage and environment protection. The foliage penetration (FOFEN) radar andground penetration radar (GPR) are prominent military applications of UWB. The FOFEN radar is used todetect large tactical vehicles hidden in forested areas, and GPR mainly to seek landmines and unexplodedordnance (UXO). Synthetic Aperture Radar(SAR), noted for its high azimuth resolution, in conjunctionwith UWB technology,will improve its range resolution with sub-surface imaging capability, so UWB SARcan product very high resolution image for both national defense and remote sensing.First, the definition of UWB is put forward, then the history of UWB is summarized, and the state-of-the-art of UWB SAR is illustrated by some examples of practical UWB SAR systems. Furthermore,some characteristics and applications of UWB radar are discussed. In past decade years, some difficultissues, which are not resolved yet, are listed. In conclusion, the trends of technology and its applicationsare made clear.Key words:Ultra-Wideband (UWB), Synthetic aperture radar(SAR), High range resolution(HRR)

Imaging Simulation Analysis of 3-D Imaging AltimeterZHANG Hong-yuan, ZHANG Xiang-kun, XU Ke, LIU He-guang(Center for Space Science and Applied Research,The Chinese Academy of Sciences,Beijing100080,China)Abstract: This paper is part of the Chinese 863-2 project“Key Technology of Spaceborne ThreeDimentional Imaging Altimeter (CIALT: China Imaging Altimeter)”. CIALT system is to be used forboth ocean and land imaging with different modes and it is different from traditional altimeters because itgets height information by the interference of two complex-value images in two channels and it can gethigh-resolution 2-D images by synthetic aperture technology. CIALT can be used in many fields. In thispaper, both 2-D and 3-D imaging algorithms of CIALT system are described at first. Then according tothe algorithms, this paper introduces a simulation system that can be used to get the raw data of targetsand to process it into radar images. Using this simulation system, the outputs of point target and facetarget are illustrated. Further more, the factors such as system parameters, windowing functions, focusdepth, range curvature and instability of Doppler parameters that influence the resolution of output imagesare discussed in detail in this paper. In the end, some simple results of 3-D imaging are introduced.Key words:Imaging altimeter, Matched filter, I/Q demodulation, Doppler frequency shift

Feature Selection in Synthetic Aperture Radar ImagesZHAO Xiao-jie, CHONG Jin-song, WANG Hong-qi(Institute of Electronics,The Chinese Academy of Sciences,Beijing100080,China)Abstract:This paper discusses the wide used feature in image classification. The principle of featureselection is given after all the feature are analyzed and many kinds of feature combination are tested. Atlast the features which are fit for SAR urban images classification are given.Key words:Synthetic aperture radar, Feature selection, Textural feature

Optimal Segmentation of SAR ImagesLI Su-dan, ZHANG Cui, WANG Zheng-zhi(Department of Automatic Control,National University.of Defense Technology,Changsha410073,China)Abstract:In this paper, we first introduce the probability distribution function of SAR image, and get thelikelihood function of this image. Then we proposed proper object function combining the likelihoodfunction and boundary constrains function. The boundary constrains function is used to avoid irregularregion boundary by introducing neighbor structure information. The best segmentation is the segmentationwhich makes the object function minimize. The algorithm runs step by step as follows: At first, the rawimage is segmented into block regions with specified size as the initial segmentation, every region presentsfor one class; Then we randomly adjust the pixels between two adjacent regions, the object functiondifference caused by this adjustment is used to determine the acceptance of the adjustment. This is aoptimal problem, in order to achieve global optimal, simulated annealing algorithm is used. when thetemperature gradually tends to zero,we can get the optimal segmentation which makes the object functionminimize. At last, a merge method based on region similarity is needed to make small similar regionsmerged into large region, which makes segmentation looks more rational. We applied this algorithm tosome SAR images and got satisfied segmentation results.Key words:SAR, Segmentation, Simulated annealing