Automatic generation is a critical procedure in remote sensing image mosaicking and a key technique for producing digital orthophoto map (DOM). The seamline is closely related to the object integrity and color consistency in the image,so the automatic generation of seamline determines directly the quality and efficiency of mosaic processing, which thus becomes a hot research topic both in China and abroad. In this paper,the authors described the relative principle of the seamline generation and made a summary of four methods for automatic generation of seamline,which included methods based on overlap area image difference,identical points,auxiliary data and morphology. The prospect of further optimizing the seamline is forecasted on the basis of comparing and analyzing the advantages and disadvantages of these methods.

The mechanism of MTF(modulation transfer function) and SNR(signal noise ration) degradation in the spaceborne TDI CCD(time-delayed-integration charge coupled device) optical camera imaging process was studied in this paper. Using the mathematic models at different stages as nodes and the signal transmission and conversion process as the framework, the authors studied the descriptive means for the signal and simulation method of MTF and SNR degradation in every node. The MTF and SNR simulation flow of synthetically response was brought forward on the basis of FFT and random noise of given distribution creating arithmetic. The method mentioned above can be applied to the MTF and SNR comprehensive response simulation in one or several links. The simulation test of TDI CCD camera was conducted crossing orbit MTF and along orbit MTF and SNR by using this method.

According to the displaying problem of the high radiation resolution remote sensing image(HRRRSI) in ordinary equipment, this paper presents a suitable new algorithm for the displaying of the HRRRSI. This algorithm based on image high radiation resolution large gradient field gradient attenuation can be used to accurately display the detailed information of HRRRSI on common display device. The technical route of this algorithm is as follows: Firstly, the brightness dynamic range of the true HRRRSI is compressed. Then,the compressed remote sensing image is reconstructed to get low dynamic domain image. Finally,through a comparison with the commonly used linear stretching processing results, the effectiveness of this algorithm is varified. The results show that the algorithm can achieve considerable dynamic range compression of HRRRSI and can also reserve the details of the image information and inhibit general edge effect.

The analysis of multi-source high resolution remote sensing images is important for Earth surface surveying,rapid response to hazard,and military detection. The problem as to how to obtain high accurate coordinates of ground points is essential for the analysis of multi-source remote sensing images. Hence,under the frame of rational polynomial coefficient(RPC)theory,the authors established an adjustment model of multi-source images based on RPC. The model can correct the error of initial RPC model system and obtain more accurate positional results. For proving the accuracy of the model,the authors made use of IKONOS and GeoEye-1 panchromatic stereo images. The adjustment calculation results show that the adjustment model of the multi-sensor remote sensing images based on RPC is correct and independent of ground control point. The positioning accuracy of multi-sensor adjustment is more precise than that of homologous sensor adjustment. The model proposed in this paper has a lot of potential applications.

In order to accurately identify ice or snow and obtain the range, this paper presents a new recognition method based on transition region feature for high spatial resolution panchromatic remote sensing imagery. Firstly, the high reflection region including snow or ice and cloud was extracted by K-means cluster analysis. Secondly, the transition region was segmented by SUSAN edge detection. Then, the average, variance and thickness were chosen as the transition region feature vectors to differentiate ice or snow pixels as the target boundary. Finally, the snow or ice area was obtained by edge growing and region filling. The "Beijing-1" high spatial resolution panchromatic remote sensing image was selected to identify the ice or snow area by transition region feature, and the recognition precision reached 97.39%. A comparison of the experimental results with those of other methods shows that the accuracy of the transition region feature analysis is obviously improved. The application analysis indicates that the method of ice and snow recognition based on transition region feature can obtain higher precision of results and more details of edges, and can also provide the references for separating the cloud and snow and extracting the snow line.

With the case II waters in the Taihu Lake and Yellow River estuary as the research object and seven images of the advanced wide-field sensor(AWiFS)and linear imaging self-scanner(LISS)of Indian satellite as the basic data,the authors theoretically illuminated and experimentally evaluated the scale-depended properties of pseudo dark target pixel for dark target atmospheric correction. The results of the study show that:1 with the scale-downing method,the false dark pixel can be divided into several sub-pixels,each of which at least includes one dark pixel; 2 the problem whether there are dark pixels suitable for atmospheric correction or not is a conclusion vaguely containing scale properties; 3 there are about 8.98% bias between the reflectance of false dark pixel of AWiFS and that of LISS sensors in the Taihu Lake and Yellow River estuary,because of the different scales of the pixels; 4 the linear model (y=0.996 x-0.003 1)can be used to correct the apparent reflectance of false dark pixel of AWiFS to that of LISS,and the regression error is only 1.86%.

In view of the low accuracy of the single feature-based method for target recognition in SAR image, a multi-feature decision-making level fusion method based on SVM and D-S evidence theory was proposed.After a series of image processing, the texture feature, Hu invariant moments feature and peek feature were extracted from the target image. Then the targets were classified according to each type of features utilizing SVM, and the results were used as evidence to construct the basic probability assignment. Conclusively, D-S combination rule of evidence was used to achieve fusion, and final recognition results were given by classification thresholds. The method is used for recognizing three-class targets in MSTAR database, and the recognition rate arrives at 95.5%. Experimental result shows that the method is effective for SAR images target recognition.

The coastal terrain of Zhejiang is complex. It is impossible to use the traditional way to measure the tideland heights of many beaches or offshore islands. With UAV remote sensing technology and tidal observation, the tideland height can be measured, and the difficult problems of artificial measurement can be solved, thus greatly improving the efficiency. Firstly, in combination with existing data, the aerial region and tide information are determined.Secondly,the UAV aerial operations and tidal observations are implemented outside synchronization. Finally, the boundary between tide and coast are interpreted and drawn with high-precision remote sensing image data (DOM), the tidal observation data are interpolated to update tideland height. The results show that this technology can quantitatively describe the tideland height information and is hence a new method of tideland height measurement.

Aimed at solving the problem of registration of 3D laser multi-view scanning point cloud,this paper presents a new synchronization registration approach based on the spatial similarity transformation principle. Firstly,the original transformed parameters are figured out between each survey station and reference coordinate system with the Rodrigues matrix,then adjustment is made according to the relationship of survey stations so as to get the adjusted transformation parameters by means of synchronization registration model put forward in this paper and realize multi-view point cloud synchronization registration. Finally, the registration results verify the algorithm put forward in this paper by the target-based registration tool in Realworks Survey software. The precision of the coordinate transformation model and the error source of the point clouds registration were analyzed through a comparison with Realworks Survey software registration. The results show that the program of 3D laser multi-view scanning point cloud data synchronization registration is feasibility and practical.

Aimed at solving the problem of crack detection on buildings, the author has designed a method of crack detection based on the temperature observation above buildings by using the infrared thermal imaging technology. The method utilizes Matlab and other software to conduct related processing of thermal images collected by external industry so as to enhance the display of cracks, thus achieving the purpose of detection and simplifying the image processing process. An analogue experiment on a chapped building in Hangzhou was carried out to examine the reliability of this method. The experimental results show that the theory and method put forward by this paper have some practical value.

2D scatter plot of remote sensing image,which is the best graphical representation method,expresses the structural information of the combination of two bands in spectral space; nevertheless, previous work only focused on such problems as the background - abnormal subspace model in the dot matrix space of spectral data,the estimation of the elliptical geometry parameters of 2D scatter plot based on the Gaussian distribution model of combination of 2 bands’ probability density,and the relationship between the background,interference and alteration anomaly of remote sensing image to the spatial structure of 2D scatter plot. Further studies show that the type classification of 2D scatter plot of remote sensing image is based on the physical meaning,and it can not only locate the background,interference and alteration information of the remote sensing image but also direct the selection of the method for the extraction of the remote sensing alteration information. At last,the authors discussed different kinds of 2D scatter plot practically used in the extraction of the remote sensing alteration information and,with practical examples, explained the theoretical basis and the application results of the 2D scatter plot used in the extraction of the remote sensing alteration information,which is based on the physical meaning.

A new algorithm of soil moisture inversion was developed by using ANFIS based on ALOS/PALSAR data. Firstly, the surface backscattering characteristics in bare region were simulated and the relationship between the backscattering coefficient and the surface roughness was built in consideration of the actual situation. Secondly, the surface roughness in the study area didn’t change due to the fact that the effect of backscattering coefficient brought by surface roughness was constant. On such a basis, a model of soil moisture inversion was built by using such methods as ANFIS model, BP neutral network, multiple linear regressions and multiple nonlinear regressions, and then the simulating data was used to validate the accuracy of this model. The result shows that the estimated soil moisture error is 0.030, and the relative error is 14.5%. Therefore, the ANFIS method can be used to conduct inversion of soil moisture in bare region with higher precision.

The timely and accurate automatic extraction of coastline from satellite remote sensing imagery is one of the important applications of remote sensing technology and has great significance for management planning of the sea area. Because the spectral characteristics of coastal water are susceptible to regional environment,the traditional method of normalized difference water index(NDWI)threshold segmentation may easily misclassify water as land in the process of separation of land and water, which will seriously affect the accuracy of shoreline extraction. In this paper,on the basis of NDWI model,the authors proposed an automatic coastline extraction method based on classification sample auto-selection and support vector machine(SVM). Firstly,through the NDWI calculation and global threshold segmentation,the initial water distribution information is obtained. And then,the classification samples are selected automatically under the control of NDWI information. Thirdly,the water are separated from the land by using SVM classifier. The last step is to fill small terrestrial water body units and track coastline automatically. The experimental results show that this method can effectively enhance the capability of coastal water identification and improve the accuracy and automation of the coastline extraction from remote sensing imager.

According to the visual characteristics of vegetation, the vegetation areas in the high resolution remote sensing images can be extracted accurately without any transcendental knowledge by using the tonal characteristics of the image itself. Following a brief introduction to the principle of the method, the authors add the NDVI vegetation index to the method which could previously only be applied to the bright tonal, thus overcoming the shortcomings of misjudgments of non-vegetation areas and poor accuracy of vegetation extraction. Morphological closing operation is conducted before the detection of the vegetation region contour. The improved method can resolve the misjudgments of non-vegetation areas significantly, qualifying it for a certain extraction accuracy even in the image whose tonal is somewhat dark, thus expanding the application of the method.

The problem as to how to form binary or multi mask image by ERDAS’s Visualization Modeler according to the condition with mult-interference information has been seldom dealt with in previous papers; nevertheless, this problem is really the key to deleting the interference information and extracting the useful information. In this paper, the authors discussed the method for forming binary or multi mask image by ERDAS’s Visualization Modeler, extracted and analyzed the iron-stained and hydroxyl anomaly alteration information based on the multi mask image and principal component analysis technology. The results show that the interference information or the false information can be deleted and the useful information can be extracted from the remote sensing image by masking based on the multi mask image, and that the high quality basic data can be provided for the extraction of the alteration information from the remote sensing image.

In classification of remote sensing images without any training samples, the choice of training samples from other representative images might be the only direct way; nevertheless, due to the difference of radiometric environments, the classification training samples from one image could not be well representative of other images. It is known that labeled samples from one image may not be effective for classifying others with high accuracy. In view of the above problem, a novel semi-supervised transcductive support vector machine(TSVM)method is proposed. The authors first chose a large quantities of unlabeled samples from the images which need to be classified in an unsupervised way, then extracted the inherent construction information of different classes in the feature space. Next, with the help of semi-supervised learning theory, the authors trained a classifier which was pre-trained by the labeled samples from another image in a recursive way, and at last an optimized classifier was obtained. It should be noted that two images involved in the method must have familiar land covers and acquired times. Classification experiments of SPOT5 and QuickBird remote sensing images were undertaken by the authors, and the classification results prove that the method proposed in this paper can effectively realize the sample extending application in both time and spatial domain.

In order to obtain high accuracy DSM data, the authors put forward the method for construction of DSM by using irregular ground points and contour vector line of buildings. First, the point clouds are separated into ground points and non-ground points through adaptive TIN filter method. Second, the building points are separated from the non-ground points according to the height texture. Then, the edges of buildings are extracted by Canny operator based on the depth images of the building which have been gray-quantized; afterwards, the azimuth clustering Comparative Law is used to regulate the building edge. Finally, the DSM could be built by using irregular ground data and contour vector line of buildings. An experiment was performed with a group of LiDAR data obtained in Liuzhou. The results show that the information of buildings’ edges can be more accurate in the DSM constructed by the proposed method in this paper.

The most important limiting factor is probably the strong atmospheric wet delay in the application of InSAR to monitoring those landslides located in special location complex topography and rainy climatic condition areas, such as Shuping landslide in Three Gorges Area in China. The atmospheric delay phase could even cover its deformation phase in the area of Shuping landslide at times. To overcome this default of D-InSAR, the authors carried out a correction by the correlation between wet delay phase and elevation, which probably is a considerable atmospheric correction method. Therefore, based on space partial correlation, this paper established the best-fit-function model between wet delay phase and corresponding DEM in the area of Shuping landslide. In fact, wet delay is closely related not only to the elevation but also to the distance along the radar in some cases. In Fig.1 the function of 'Ⅰ’ zone depends on the elevation, while that of 'Ⅱ’ zone is a function of both the elevation and the X (range direction). In view of the least-square method, two best-fit-function models would be found under the condition of the minimum mean square deviation without obvious variation. Finally, the simulated atmospheric delay phase formed by optimal correction models with elevation or/and distance values is removed from the unwrapping phase. And it clearly gives the result of the location, the size and the distribution of the Shuping deformation field within the 11 days. In conclusion, this method could effectively reduce the wet delay phase of the interferogram and is of significance for monitoring the slow deformation of landslides in the Three Gorges Area.

The object-oriented method was chosen to extract land use information from UAV image. Through pretreatment of the original UAV images, the appropriate partition parameters were selected to perform multi-scale segmentation in the experimental area; according to the terrain classification characteristic differences the classification hierarchy system of multi-scale segmentation and the surface feature characteristic extraction rules were established to extract land use information in the respective optimal segmentation scale layer. The result shows that the application of object-oriented multi-scale segmentation technology can extract the land use information rapidly and accurately by using the texture, shape, size, and spatial relationship information of high-resolution UAV image.

In this paper, aerial images of the 1970s, SPOT5 satellite images of 2003 and RapidEye images of 2009 were respectively selected to establish interpretation keys based on field survey data. Then three phase remote sensing images with different sensors were interpreted by ArcGIS, and land change maps from 1970s to 2003 and from 2003 to 2009 were obtained by overlapping interpretation results. The change and mutual transformation of lands were analyzed, and their change rules and causes were obtained. The analytical results are summarized as follows: Firstly, from 1970s to 2003, the main change of lands in the Shizuishan mining area was that the vegetation coverage area and the cultivated land were transformed to residential area and wasteland as a result of infrastructure construction and soil erosion. In addition, the areas of gangues, coal piles and water increased because of mining activities. Secondly, the area of vegetation coverage showed the fastest increasing rate from 2003 to 2009. On the contrary, the areas of surface features of the mine, such as gangues and coal piles, decreased sharply, which was closely related to integrated renovation of the mining geo-environment.

Regional dust pollution,resulting from airborne transmission of dust formed by coal mining,endangers the health of the surrounding residents. Therefore, the development of the methods for dust monitoring and controlling is extremely important. In this study, Wansheng, located in the southeast of Chongqing City,was selected as the study area,and ASD Field Pro Ⅲ spectrometer was employed for spectrum collection. A comparative analysis of spectral data was carried out on the basis of the anomalies of the vegetation overall spectral reflectance and small absorption peaks at the wavelength of 1 000 nm and 1 200 nm,and then a relative level of dust pollution was retrieved. The results show that the dust pollution in the southern part of the Nantong coal mine and the southwestern part of the Yanshitai coal mine in Wansheng district is the most serious. The crops impacted by dust mainly include leaf beet,cabbages and Chinese mallow. The results obtained by the authors provide fundamental basis for further researches on dust pollution based on quantitative remote sensing.

Using the TM images of Qiemo oasis, Xinjiang Uygur Autonomous Region, obtained in August of 1998, 2001, 2006 and 2011, the authors built the complex vegetation index VBSI based on FCD model. This index contains vegetation index (NDVI), bare soil index (BI), and shadow index (SI). Furthermore, the vegetation coverage in Qiemo oasis in 1998, 2001, 2006 and 2011 was calculated separately according to the dimidiate pixel principal and then the vegetation coverage graduation maps were compiled. Based on statistical diagrams and transfer matrices, the authors have drawn some conclusions: 1 The overall vegetation coverage area increased 742.62 km² in the 13 years, and Qiemo oasis was expanding; 2 The vegetation coverage showed a fluctuation change during the period. Vegetation was restored well from 1998 to 2001 and from 2006 to 2011. The area of recovered vegetation (vegetation transferred to higher coverage) was 2.98 and 1.79 times the area of degraded vegetation (vegetation transferred to lower coverage) from 1998 to 2001 and from 2001 to 2006; Nevertheless, it degraded heavily from 2001 to 2006, the area of degraded vegetation was 435.3 km², whereas the recovered vegetation area was only 5.41 km²; 3 The vegetation coverage change in Qiemo oasis was influenced by such factors as temperature, precipitation in spring and summer, population and anti-desertification projects.

The purpose of this study is to provide the basis for further study of desertification origin and the protection of ecological environment in this region. The morphological characteristics of Kumtagh sand dunes in Beishan of Xinjiang Uygur Autonomous Region were analyzed, the Landsat ETM images of morphotype of Kumtagh sand dunes were cut out, and the sharp images were also enhanced. On the basis of SRTM data,the digital elevation mode for Kumtagh sand dunes were obtained by means of three-dimensional analysis module of ArcGIS software. Simple wind figures were drawn according to the local terrain and physiognomic characteristics. The characteristics of each type of sand dune were analyzed in detail. This study results that sand dunes in Kumtagh form a long and narrow "S" shaped ridge due to the unique natural geographic features,and the dune morphology is diverse and complex. The main types of sand dunes are crescent dune chains and complex longitudinal sand dunes. The single crescent dunes are scattered, the linear dunes are less distributed,and the dendrite dunes are in the early development stage.

The situation of the desertification in the middle reaches of the Yellow River is critical. In order to understand the dynamic characteristics and influencing factors of the desertification, the authors analyzed the remote sensing images from 4 different periods (i.e., 1975, 1990, 2000, 2007) in the study area and performed the statistical analysis by means of MapGIS 6.7 and SPSS. Principal factor analysis was employed to find the main factors influencing the desertification. The results show that from 1975 to 2007 the area of desertification in some areas decreased by 2.43?10³ km², but the total area of desertification still accounted for over 60% of the total study area. The geomorphic type played an important role in the desertification. The desertification area of wind-borne landforms accounted for 43.1% of the study area. In a short period, desertification caused by human activities and villainous climate in the middle reaches of the Yellow River accounted for 52% and 48%, respectively.

This study combines the use of multi-temporal satellite remote sensing images with the GIS spatial analysis method. The study of the changing characteristics of Fangchenggang coastline has important practical significance. The TM images of Fangchenggang in two phases were obtained in November 2000 and October 2010, First, preprocessing, extraction of feature information of coastline and spatial analysis were conducted for two-phase TM images of Fangchenggang. Then, the CA model was established in accordance with the relevant controlling factors and the corresponding neighbor rules. On such a basis, the Monte Carlo method was used in combination with the controlling factors to make judgment so as to finalize the conversion of cellular state. The actual and prediction image overlay analysis in 2010 shows that the number accuracy of prediction coastline is 83.65%, and the spatial location accuracy is 93.45%, whose values are in the range of allowable error. It is shown that the use of the CA model to predict the shoreline in Monte Carlo method is practical. Finally, the prediction of the Fangchenggang coastline in 2020 was realized based on the combination of CA model algorithm with Matlab technology.

Rapid industrialization and urbanization have resulted in extensive changes in land use and coastlines in Bohai Bay coastal zone. For extraction of coastlines and land classification, Landsat TM\ETM images in 2000, 2005 and 2010 were used. Based on the methods of man- machine interactive interpretation and GIS spatial analysis, the authors made a comprehensive analysis of the changes of the coastlines and spatial variation of land use. The extent, speed and spatial variation of the changes of the coastlines and land use were revealed by the combined method of quantitative and qualitative analysis. It can be concluded that Bohai Bay coastal zone had been changed greatly during the last decade. The land and coastlines were growing toward the Bohai Sea by about 322 km² and 331.6 km², respectively. Land for housing and other construction purposes increased; unused land and wetlands decreased greatly; Tianjin Port showed the most significant sea-beach wetland change, and the decrease reached 89 percent as compared with the situation in 2000; in other wetlands such as salt pans of Hangu in Bohai Bay the decline rate was the largest (55 percent); the coastline and the land of Tianjin Port had the fastest growing rate during 2000-2005; from 2005 to 2010 the reclamation of Caofeidian Port was the fastest. The main driver of these changes is the result of over-exploitation of human economic activities.

As it is difficult to transfer the mapping results of remote sensing from the interpretation of geological disasters in 3D virtual environment without third-party software, this paper introduces a method for automatic mapping of the results from the remote sensing interpretation of geological disasters based on a 3D GIS platform named iTelluro. Based on the mapping specification of the remote sensing interpretation from thematic maps, we can acquire the remote sensing image layers, vector layers, geological and geographic layers and interpretation topic layers from the image database, and then choose 3D GIS layers according to the scale and mapping range, set the rendering styles of the thematic map and finally realize the method in the 3D remote sensing interpretation system of geological disasters. This method can improve the efficiency of the mapping results from the remote sensing interpretation of geological disasters, and can provide the information products for prevention and control of geological disasters as well as for tackling the emergency event.

In order to effectively manage and utilize mass spatial data accumulated in researches on resources and environments remote sensing, the authors designed the methods for storage, retrieval and access of these mass spatial data. In accordance with the mass spatial data of different types and characteristics, approaches like spatial database, file directory and Web Service were applied for storing and delivering data. On the basis of two-step spatial query algorithm, the retrieval method of mass spatial data based on the spatial location was designed to retrieve the data stored in integrated spatial databases as well as data delivered in forms of Mapping and Web Service, so as to resolve the problem of fast data search. Methods of tile and hierarchy and visible area grid index were used to optimize the performance of transferring, loading and displaying data, and relative performance tests were executed. The designed approach solved the problems of data management, data precise location and data fast transportation that appeared in the application of massive spatial data via investigating strategies of data storage, data retrieval and data accessing and appearing. On the basis of this approach, the basic framework for resources and environments remote sensing emergency response and monitoring infrastructure has been constructed.

Based on an analysis of the research progress of multisource spatial-data integrating, this paper puts forward an idea to integrate multisource image intelligence spatial-data just in the pattern form which accesses spatial data directly by using SuperMap SDX+ spatial-data engine. The contents system structure of the spatial database was designed and the accessing pattern and the integrating model were analyzed. Then the integrating technological frame structure was designed based on the study of GIS with its application superiority and the technological system structure of Seamless Integration of Multisource Spatial Data. Its Database Prototype was actualized so as to prove the rationality of the proposed theory. The research achievement has resolved preferably image intelligence spatial data integration characterized by multi-source, multi-format and multi-modality. On the basis of data integration, the application of target intelligence inquiry and assistant image interpretation by contrast and target theme-mapping was actualized.

With the purpose of shortening planning road updating period and prompting data quality so as to realize data sharing, the authors designed and developed the planning road data management system in Beijing, studied the design of centerlines of planning road condition coding table and curve identifier, curve position identifier, and normal curve identifier in point attribute table of planning road, realized the functions of curve element calculation, and derived redline graph editing and graph output on planning road centerlines. Having been approved by the experts, the system is used in Beijing Institute of Surveying and Mapping. It is proved that the system is well designed and can realize the expected aim and produce better social and economical efficiency.