2012 Vol. 37, No. 8
Reinterpreting the engineering and life reliability in geography information sciences,for the first time,we put for ward the concept of reliable spatial analysis,with an analysis of their characteristics.The reliable spatial analysis is proposed as a new research direction for GISci,and correspondingly the existing primary scientific problems,research issues,the foundational theories,and the potential application areas are identified.This research emphases on a whole process of spatial analysis with emphasis on reliability.In a broad sense,we describe a paradigm advance from uncertainties into reliability in spatial analysis.
Taking Poyang Lake wetland as the research object,using the theories and technologies of GIS,RS,GPS,spatial database,three-dimensional virtual simulation and hydrodynamic theories,we constructed the fundamental GIS for Poyang Lake wetland with the integration of expert knowledge,model and decision making functions.To a certain extent,the system solved the contradictions in wetland economic development—different sectors(e.g.pollution and hydrological monitoring,flood forecasting,land planning) are closely related,but isolated from each other.
According to the characteristics of zonal fragmentation and the existing methods of transformation and optimization based on cross-border topological join,we propose a hybrid parallel scheduling strategy with the principle of static and dynamic combination.Firstly,the excutive sites of the fragment joins are determined using the Move Small rule.Secondely,the pre-allocated task queue on each site is reranked on the principle of larger area of FMBR priority.Finally,the fragment join tasks in each queue are redistributed in order to improve load balance by comparing the costs of three typical task sheduling cases.The experimental result shows that this method solves the load-balancing problem in effectively and further improves the performance of distributed spatial query.
A hierarchical approach is proposed for the semantic classification of islands,which is based on the theory of formal concept analysis.First of all,the new islands concepts are obtained through the combination of the extension and intension of the original islands concepts.Then,the semantic hierarchy is established on the basis of the super-subclass relations among the new islands concepts.Finally,the concept lattice for hierarchical semantic classification of islands can be obtained by means of the formal concept analysis.
We propose a method for multi-resolution representation of 3D complex building model with features preservation based on improved volumetric techniques.Firstly,the polygonal model of 3D complex building is converted into volumetric representation,and the voxels are stored in an octree.Then,various resolution cell models can be generated by clustering the octree cells with a basis on quadratic error metric.Finally,the multi-resolution isosurfaces are extracted and the features are reproduced using normals as expected LOD models.The experimental results show that the proposed method has the characteristics of less calculation and better ability of preserving main structure features of original building.
According to the geometric imaging principle of new type area mapping camera constructed by multi CCD in single field-of-view,the basic method of image geometric processing is proposed for surveying and mapping application.Based on the nature of combined area camera,the digital image error is proposed,which includes the errors of focal length,principle point coordinate and optical distort etc,and the error of the size of area CCD and the camera assembly are considered,and the camera self-calibration is done by bundle adjustment,the effect of the quality on surveying and mapping can be deceased.The method is tested by aerial photogrammtry in Hanzhong field,and the effectiveness of the proposed method is validated by the test result.
Buildings being not orthogonal to the light of sight of SAR are often marked as volume scattering in the three-component decomposition.A new unsupervised classification algorithm is proposed by introducing normalized circular-pol correlation coefficient,which can improve the discernment of non-reflection symmetric structure on the basis of three-component decomposition.The experiment using E-SAR L band PolSAR image around DLR proves the effectiveness of the proposed method.
Hyper-spectral remote sensing imagery provides a large amount of spectral and structure information.However,these availabilities challenge the traditional spectral segmentation methods which may cause salt and pepper effect and low information extraction accuracy.In order to overcome this disadvantage,texture information is proposed into feature space.A 3D-Gabor filter is used to represent the spectral/spatial properties of hyper-spectral data.Thus multi-scale,multi-oriented texture features are extracted.And feature energy from 3D feature points is projected into subspace with PCA which can represent input data with lower dimensional feature vectors.Then the image segmentation is constructed by k-means clustering.Following these steps,the initial residential areas can be obtained,but with many deficiencies including the existence of holes and useless patches.To resolve these problems,a morphological space based method is used to dissolve these residential patches.The experiment on PHI-3 data demonstrates the utility of the algorithm for residential areas recognition.
SEaTH algorithm is an object-oriented feature selection algorithm based on inter-class distance.We present an improved SEaTH—ISEaTH.It overcomes some limitations of existing SEaTH algorithms.The ISEaTH algorithm evaluates the features according to the relation between features,inter-class distance and intra-class distance respectively,and integrates multiple estimation results to obtain the best feature subset.Compared SEaTH and ISEaTH with QuickBird data in Kashi area of Xinjiang province,the experimental results show that the ISEaTH algorithm not only reduces feature dimensions,but also improves the classification accuracy,thus is a much more efficient object-oriented feature selection algorithm.
A new infrared target detection approach based on principal background suppression is presented.Through analyzing the spatial-temporal correlation of thermal imagery,the principal component analysis technique is utilized to suppress the background clutters.A spatially related fuzzy ART neural network is applied to build the spatial-temporal background models and detect targets by the suppressed frame information.The experiments have been carried out and the results show that the F1 measurement of our proposed approach is up to 94.2%.It is able to suppress background clutters and highlight targets effectively,and it is capable of detecting targets in complex thermal scene accurately.
A method for image edge detection is presented based on Deng interrelatedness and the classical Prewitt edge detection operator.The method takes the Prewitt operator as referential sequence,calculates the Deng interrelatedness of the image pixels between the compared sequence and referential sequence in the eight neighborhoods,and detectes the edge according to the Deng interrelatedness.The experimental results show that the proposed method can detect useful information of image edge more accurately and has some anti-noise ability,and also proves that edge extracted by the combined method has more completion and consecutiveness and fewer non-edge points.The method is also applied to edge detection of remote sensing image.
The radiance reflected from neighboring slopes is calculated with an empirical Lambertian model.An assessment model is developed to analyze relative error of target reflectance when neglecting incoming irradiance is neglected from surrounding terrain which will cause the reflectance of target pixel to be overestimated.The overestimation of reflectance will be stronger in shaded areas,which could bring more saturation value in topographically corrected image.The results show that radiance reflected from neighboring vegetation in near-infrared spectrum and neighboring snow in the visible and near-infrared spectrum should be accounted when we compute the reflectance of targets.Especially in the case of deep valleys,the overestimation of target reflectance could be more than 10% when the radiance is neglected.
A new JFD scheme is presented,where a new encryption strategy based on selective content encryption is put forward to enhance security,and a new measure for choosing the fingerprinted area caused by structure distortion is proposed to reduce fingerprint's influence on image quality,while a new fingerprints generating and controlling way is proposed to make the fingerprinted image resist collusion attack.The experimental results show the scheme's effectiveness.
We discuss a real-time quality control(QC) procedure for precise point positioning(PPP) with emphasis given to the detection and repair of clock jump,cycle slip detection and robust estimation.Four schemes with different QC procedures are designed to test the effects on PPP.The results show that clock jump should be first compensated during the preprocessing phase since it may cause the cycle slip detection invalid and seriously affect the precision and efficiency of zero-difference PPP.The union of MW and GF combinations can detect the vast majority of cycle slips which is more accurate and reliable than any single one.With an improved robust estimation,gross errors and their impact on positioning are effectively suppressed,which led to significant improvements in the accuracy and reliability of PPP.
Based on the special relationship of dual-frequency carrier phase observations,GNSS dual-frequency integer relationship constrained ambiguity resolution(FirCAR) is proposed,which enlarges the equivalent wavelength of carrier phase in local integer rang,to achieve a fast integer ambiguity resolution.With the code pseudorange and dual-frequency carrier phase observations,the proposed method resolves integer ambiguities satellite by satellite independently.The correctness and effectiveness of this method have been tested by two baselines of different lengths;and the influence of cut-off angle is also analyzed.
Dam deformation monitoring system(DDMS) which benefits from GNSS continuously operating reference system(CORS) is established.It consists of two reference stations and five monitoring stations.We design a star topology network architecture and TCP/IP protocol for real-time data stream transmission.A data processing software with the function of quality control,coordinates estimation and transformation,then accuracy evaluation and session time setting are realized.One-year operating results demonstrate that DDMS has higher stability.The probability of providing effective solutions is up to 98% without data loss,and 2-hour-session solution can achieve 1 mm,0.7 mm,1.7 mm at north-south,east-west and vertical component of station respectively.The accuracy can be improved to 0.8 mm,0.5 mm,1.4 mm respectively while the session time extends to four hour.It is concluded that the system can provide a technically advanced and cost-effective infrastructure for dam deformation monitoring analysis.
The Mw 9.0 Japan Earthquake occurred in east coast of Japan in March 11,2011.Surface deformation process of Northeast area of China is inversed with the data of 6 GPS stations from CMONOC.Far-field cosesimic displacement filed of continental area of China is determined with the data of 160 GPS stations of CMONOC.The results indicate that cosesimic displacement of NorthEast area of China is the largest,on the average close to 20 mm,the average cosesimic displacement of Beijing and surrounding area is close to 10 mm,far-field area average is close to 5 mm,the major direction is east.This earthquake has no effect on South China(including the South China Sea) and the Qinghai-Tibet area of China.The results will be helpful for the research of dynamic characteristic of earthquakes and determination of the tendency of future earthquakes.
DIODE is a real-time on-board orbit determination software,embedded in the DORIS receiver since 1998.Frequency biases between beacons and satellites USO are the max errors in DORIS measurement data.The purpose of this paper is to find a method to estimate the frequency biases.To start with,we describe the state and measurement equations about DORIS real-time orbit determination.Then,we present a method to estimate the frequency biases.Using our method and considering the 40×40 Earth gravitational fields,solid tide,moon and sun attractions,we analyze ten days' measurements data of SPOT-5.The results show that the radial component RMS is better than 30 cm,which correspond to the SPOT-5 real time orbit accuracy,and the 3D RMS is better than 80 cm,which achieves the specifications of SPOT-5.
A relativistic differential VLBI(ΔVLBI) mathematical model for CE-1 in Earth-Moon transfer orbit is derived.Based on the model,ΔVLBI time delay observations are formed by differencing CE-1 time delay observations with a simulated radio source's time delay observations.Then unknown parameters including orbit of CE-1 and Earth Orientation Parameters(EOPs) are estimated under the least squares adjustment,in optimal parameter apriori accuracy,and the best arc length is determined according to the inner and externally coincident accuracy of each parameter.Finally,the accuracy of each parameter is given.
Based on the physical process of atmospheric signals,an algorithm has been developed to estimate the altitude-related MERIS water vapor and to correct the atmospheric effects on InSAR.The algorithm can significantly reduce the influence of atmospheric phase distortions on ASAR interferogram,and further explore the ground deformation and other geophysical signals.According to the experimental result,this altitude-related MERIS water vapor interpolation can reach 4.25% improvement averagely,relative to traditional MERIS integrated water vapor interpolation.Furthermore,this algorithm may provide an avenue for mitigate atmospheric effects in ASAR interferogram in mountainous areas,which appeals to a wide community of geophysicists.
Based on the idea that getting cloud boundary height by using mutation of relative humidity profile when it comes into or out of clouds,we obtain the statistical characteristics of cloud-layer thickness in East Asia by using COSMIC occultation wet-air data from Jan.2007 to Jan.2009.The results show that minimum detecting altitude of COSMIC is lower than 3 km height,which mainly affects the retrieval of cloud-layer thickness of part middle-and low-level clouds.Occurrence frequencies of clouds along the direction of latitude and longitude show different distribution characteristics,and as altitude increases,frequencies of various types of clouds also have different variation regularity.Average thickness of all clouds in East Asia is 1.78 km,and single-layer cloud has the same distribution with highest-layer cloud,and two-layer and other-layer clouds have similar geographic distribution.Single layer cloud is relatively uniform in spring and autumn but the differences in summer and winter are relatively large.Two-dimensional frequency distributions of cloud-top height and layer thickness of highest-layer cloud in summer and winter are different,which mainly indicates that there are more thicker high clouds existing in upper troposphere in summer than in winter.
The electron density values of one day in Hong Kong obtained by ionospheric tomographic are divided into 6 groups.Then,the 6 groups data are fitted modeling and forecasting by neural network method.The results show that the accuracy of electron density value forecasted by the tomographic and neural network model is significantly higher than the electron density value provided by IRI2007 model.The precision of the electron density forecast value within 30 minutes and 60 minutes can respectively arrive at 0.45 TECU and 1.34 TECU.
Traditional solution to 3D coordinate transformation problem is the optimal estimation at the cost function of the least squares(LS)for residual vector without consideration of point covariance.Whereas ordinary least squares and total least squares methods can not work well in the heteroscedastic cases.Then unweighted and weighted TLS methods are introduced and compared.The data experiments indicate that unweighted TLS method and LS method have the consistent results and the block row-wised WTLS method has the same results as the maximum likelihood estimator(MLE),and the element-wised method is not so accurate as MLE due to neglecting the correlations of the matrix elements.
Vibration isolation is one of key problems for absolute gravimeter development.We find out that the measurement error caused by self vibration is up to about mGal,which cannot be neglected.We analyze the experimental data,and try to find the reason and influence model of self vibration.Thus a new technique to improve the measurement accuracy of absolute gravimeter is provided.
We propose a new DiffS algorithm which is suitable for detecting difference between different version SVG/XML format maps.The algorithm reduces the sum of nodes and levels of parsing tree by defining macro element.By defining dynamic array as data structure and then using top-down level-order traversal approach to match two different version parsing trees,the result of DiffS is an edit script which records the update operations from the first version of SVG map to the second version.The algorithm time complexity of DiffS is low.The application verification shows that the algorithm is effective.
We study P-median location model and hierarchical location problem,present a regional multi-airports location model based on hierarchical P-median location problem.The proposed model is experimented with multi-airports location suitability in Guangdong Province.A GIS-based computational approach is applied to compute the model.The experimental results are interpreted to provide a valuable decision tool for decision-makers in the area of airport planning.
The spatial relationships and geometry features of line group based on spatial statistical character are described.The conceptual neighborhood network of topological relationships between lines is used to define topology similarity between line groups.The directional mean is applied to define direction similarity between line groups,their circular variance is utilized to define distance similarity between line groups.Combined with the length,average length,density and tortuosity of line group,a computation model for similarity assessment between line groups is established.Experiment results show that the similarity assessment is generally consistent with surface features and intuitive cognition of human.
There are some fast and effective methods related to image processing technique.But they are unsatisfactory in resisting the noise and in reducing the redundant information.For this situation,a new method of extracting terrain feature lines is proposed by the mathematical morphology.Considering the characteristics of the ridges and valleys,two kinds of operators are designed by basic morphological operation.Then the new method is built on the operators with iterative forms.The experimental results show that the ridge-lines and valley-lines extracted by the method fit the real terrain.And it can resist some small-scale noise.
A method for D-TIN or constrained D-TIN generation by integration of distance transformation(DT) and region adjacency graphics(RAG) is proposed.Its basic idea is that Delaunay triangulation is the dual structure of the Voronoi diagram.Concretely,DT is used to obtain Voronoi diagram while RAG is used to find spatial relationships between discrete points on building Voronoi diagram,which is helpful to from the D-TIN.Compared to the typical method with which mathematical morphology is used,our proposed method is capable of increasing the efficiencies approximately 57.70% and 56.64% on average for D-TIN and constrained D-TIN generation respectively.
We improve traditional difference expansion algorithm by modifying the formulas in the embedding and data extracting procedure,it need not calculate the mean coordinate values of the vertex pairs,only need embed the coordinate value of the first point along with the payload.The experimental results show that the algorithm can recover the original maps from the watermarked ones and attract the embed watermark,and it is capable of hiding more secret at low map distortion and the data error is more controllable,and the data has better confidentiality.