2012 Vol. 37, No. 7
As one of the most important variables of topography,the quality of extracting flow length has a serious effect on soil and water conservation,soil erosion and environmental assessment studies.The error of extracting flow length from grid-based Digital Elevation Model is mainly from original DEM and extracting algorithm.Based on the principle of data independent DEM interpretation analysis method,this paper quantitatively studies the error spatial distribution of flow length which is extracted from grid-based Digital Elevation Model.The result shows that the direct calculation algorithm,which is widely used,has a better effect on the calculation of the catchment area than the valley area,and at the water outlet the error achieves its maximum;flow length of the valley area contains some system error.
In this paper,Yuzhong County in Lanzhou City was selected as study area.An approach based on Goal Programming and Simulated Annealing was applied for land use zoning in this area,with land suitability,planning coordination and spatial compactness as zoning objectives.Results show that the zoning plans under different scenarios reflect the goals of decision-makers.The new generated plan over-performs the original one in the aspects of zoning objectives and meets the requirements of agricultural production,economic development and ecological protection.
In this paper,firstly,the expression of spatiotemporal is implemented;secondly,spatiotemporal product-sum model is analyzed and deduced into another one which can be calculated;thirdly,the spatiotemporal weight coefficient matrix based on ordinary kriging is put forward and the spatiotemporal interpolation is done based on this matrix;lastly,the performance of spatiotemporal interpolation is tested by cross validation.The experiment shows that spatiotemporal interpolation researched in here can provide more information on spatiotemporal random domain with some certain accuracy so as to achieve the spatiotemporal interpolation results.
This paper presents a method to partition the scattered points in planar domain based on grid-statistics and adaptive binary-partition.The basic idea of the method is as follows.First,build a grid field which can cover over all planar points to create the spatial index of the points.Then,figure out the position of every point in the grid field,such as the row and column place in the field,meanwhile,reserve the number of points in each grid.Finally,set the number of iterations as the parameter of terminating to partition the scattered points with proposed rules.This method can divide the data point-set into several groups flexibly and the result is relatively balanced.Besides,the computational efficient of the algorithm is relatively high and not too much extra storage space is needed.The method proposed in this paper can be used in the existing spatial analysis algorithms using partitioning-strategy and can be further applied to the related parallel algorithms.It also provides reference for adaptive partition of other types of spatial elements.
The similarity measurement of spatial relations is the foundation of some many application problems in geographic information science,for example,it is an important restriction to keep the consistency of spatial relations before and after the cartographic generalization.In this paper with the adjacency relation similarity of building group as the research object,and on the basis of Delaunay triangulation network and Voronoi diagram with the tools,based on the similarity theory,establishes two kinds of similarity models to use for quantifying calculation and estimatinge of adjacency relation,including the similarity models of adjacency relation based on the same characteristic value and the similarity model of adjacency relation based on the iso-distance-relationship contour.
Because of the curse of dimensionality,the traditional similarity measurement has been unfit for the high dimensional space.And the equally spaced partition can't represent the data distribution,so the similarity measurement based on that can't compute the similarity between high dimensional data reasonably.Aimed at these problems,the existing improved similarity measurements for high dimensional space are introduced firstly,and the problems are analyzed.Then,improves the similarity measurement PIDist(X,Y,kd) based on unequally spaced partition of each dimension.Finally,the experimental result of clustering heart-statlog and vehicle data sets provided by UCI proves the validity of the proposed similarity measurement.
In this paper,an efficient ray tracing algorithm based on image theory is proposed,which is based on image theory,with high computational efficiency by using database search technology.Unlike the conventional image method,our methods have two steps: direct search and reverse search.Virtual source tree is created by directing search,after creating the virtual source tree,by inversing search;it can find all valid ray-tracing paths from transmitter to receiver.The proposed method can be applied to urban 3D environments with arbitrary building layouts and heights.Finally,we chose a community to test the accuracy of our method and the predictions are in good agreement with the measurements.
Aiming at improving the utilization of polarimetric information,a new edge detector for polarimetric synthetic aperture radar(PoLSAR) images based on adaptive optimal polarimetric contrast enhancement and ratio of averaging was proposed in this paper.The optimal polarization is adaptive to window location and edge templates.Therefore the new method is robust.Experimental results of simulated and real PolSAR data both revealed that the performance of the new detector was improved.
This paper presents an approach to integrating remote sensing inversion and data assimilation to better estimate soil moisture over a large area.In support of the Microsoft.Net and ESRI ArcEngine,the C# and RSI/IDL are used to develop software modules to implement the models for soil moisture estimation from MODIS and AMSR-E data.A software package for soil moisture retrieval and data assimilation is developed by integrating the inversion modules,data assimilation modules and GIS modules.This application offers a tool for soil moisture estimation over a large area using remotely sensed data and proposed models,enabling related management departments to operationally map large-area soil moisture for assessing ecological and environmental quality.
As is known to all,high resolution remote sensing imagery contains plenty of spectral and spatial information which represent objects' inner physical attributes and outer distribution and composite structure respectively.According to its characteristics mentioned above,we proposed an building extraction method based on spatial-spectral information on high resolution remote sensing imagery.On the basis of feature unit extraction through multi-scale segmentation and vectorization,we selected building samples automatically from feature units according to characteristics like shape and spectral information.Then,we preceded the building area recognition on the whole image using the template constructed by buildings' shape,and done edge detection and thinning inside the building regions to extract the outline of building.Last,we implemented the algorithm mentioned above,and proved its validity through experiments.
Building characteristic information,especially building height extraction,is of great significance in urban remote sensing application.A method for building height determination via radar backscattering characteristics is studied,using only one spaceborne high-resolution SAR image.The building height can be evaluated by the strength of the double-bounce backscattering of a building,looking angle,orbit,and material parameters based on the geometrical optics(GO) backscattering model.The approach provides good results for typical buildings,indicating the validity of the proposed method.With the continuous development of high resolution SAR system,there are various possible applications in urban building characteristic extraction.
A new algorithm considering edge information in dual tree complex wavelet transform(DT-CWT) domain was proposed in this paper.In the multiscale geometry transforming domain of DT-CWT,the edge coefficients were determined by analyzing both the transitivity of interscale coefficients and the correlation of intrascale coefficients.The edge and non-edge coefficients in DT-CWT domain were shrunk with different thresholds by using the Bayes bivariate shrinkage function.The experimental results show that the algorithm proposed in this paper obviously restrains the interferogram noise and also greatly maintain the edges and details information comparing with the traditional way of soft-threshold denoising in wavelet domain.
Considering strip adjustment as key technology to eliminate systematic errors,we propose a method of 3D surface matching without control points.Using the least Z-difference(LZD) and least normal distance(LND)matching algorithm to realize.The experimental results show that the accuracy of LND and LZD algorithm can both meet the requirements of project accuracy;LZD's total efficiency is lower than LND's,but its accuracy is higher.Compared to commercial software TMatch,LZD and TMatch accuracy are at the seme level,and our algorithm can successfully complete the task of strip adjustment when TMatch software failure.
With wide application of GNSS in the military and the economy area,a large number of observations recorded every day.These data need to store in the disks and transfer on the internet.In order to reduce data storage space and improve network transmission efficiency,it is necessary to study the compression algorithm for GNSS observations.In this paper,sixty-hex is put forward for digital storage,Recursive differential operation algorithm is used for GNSS observations compression.We have developed the software GnssRAR(V1.0),which can compress GNSS observation data.Experiments show that the compression ratio of GnssRAR can achieved 80% with very good compression.
This paper combines the model's dynamic analysis method with multi-scale transformation method of statistical characteristics.At first data-block transformation technique is used to change the original state equation into the new state-block equation,and the original measurement equation is expressed in the form of data-block.Then,the multi-layer wavelet transformation technique is used to the original measurement vector to achieve the new measurement vector which goes into the Kalman filter with the state-block equation.At last,this paper builds up the multi-scale distributed fusion estimation algorithm according to the above results of Kalman filter.When the above algorithm is used to GPS/SST/SIN integrated navigation system,the simulated results show that this algorithm can increase the filtering precision.
Based on the fact that the half-sum linear combination of code and phase observations may remove the first-order ionospheric error,a mathematic model of the combined GPS and GLONASS precise point positioning(PPP) using UofC ionosphere-free combination is developed.The UofC PPP model estimates individually ambiguity parameters on two frequencies,which facilitate further obtaining integer solutions of ambiguities.The positioning results obtained from the UofC model and the traditional model in which code observations on two frequencies are used to form ionosphere-free combinations are compared using GPS/GLONASS observation data from IGS tracking stations.The statistical results indicate that the UofC model performs slightly better in terms of the horizontal position accuracy.But the difference about the three-dimensional positioning accuracy is insignificant.
Based on attitude error equations of typical dual-axis-index Inertial Navigation System(INS),influence of angular motion of vehicle is theoretically analyzed.The angular motion,coupled with scale factor and installation error,can deteriorate precision of INS,and produce a new type of error.Scale factor error leads to DC component,which is equivalent to un-compensable constant drift of gyroscope,and has great effect on INS.Installation error leads to AC component,which is equivalent to periodic oscillating drift of gyroscope,and has limited influence on INS.Results of experiment verifies the mentioned conclusions.
The estimate of marine kinematic GPS slant-path water vapor(SWV) is researched using Precise Point Positioning(PPP) technique in this paper.Based on the current study,this paper proposes a new method which takes into account the single difference residuals between satellites.The validity of this new method is verified by the comparison between the result of MM5 and that from the ship-borne kinematic GPS data.The result shows that the new method gives 1.2 mm RMS in SWV relative to MM5 result.That improves 1mm RMS in SWV compared to the current method.
To address the problems of multiple solutions and indefinite ring length in the classical independent closed loop algorithm,using the closed ring-shaped information matrix and taking into account the side element,this paper introduces a new algorithm based on matrix operations.The resulting closed loop meets all the requirements of the least independent closed loops.Through comparing the results of two algorithms in GPS network closed loop researching,the uniqueness of the result and the shortest length of ring are verified.
The system error contained in the temporary tidal information would have a great impact on the water level correction and computation of the theoretical bathymetrical datum,and not easy to find,especially the zero point of pressure-type tide gauges gradually sinking.This paper presents a self-segmented tide sequence matching method which is based on the least squares.Through analysis system error type,determination of length of sub-section and specific time of sinking,algorithm research of detection and repair,ultimately implement the accuracy detection and precision repair of different type of tidal system error in the temporary tide gauge,and tested by the experiment.
The characters of 7.5 a timing series of dual-frequency timing for millisecond pulsar PSR J1939+2134 are studied for the first time.DE421 is suggested by analyzing the effects on the precision of pulsar timing with different planet ephemeris.The long-term stability of pulsar time and that of atomic time is evaluated using σz(τ) method,and it is found that the long-term stability of pulsar time is always improved with time elapsed and up to 10-15 in 6 a.Hence,pulsar time has obvious advantage as the time standard for keeping time and time service.
In the present,the relationship between the resolution of theoretical minimum angle and the accuracy of angle measurement were not gived,the scanning interval constraint conditions which is used to obtain the angular resolution that equal to the scanning interval and spot diameter size is also presented.To resolve these problems,the relationship between the conditions of the angular resolution equal to the scanning interval and spot diameter size and quantification accuracy and the relationship between the resolution of theoretical minimum angle and quantification were deduced,at the same time,these simplified relationship formula was proposed.The constraint conditions and theoretical minimum angle resolution of 29 kinds of scaners is caculated.The conclusion have directly significance on the angular resolution's calculation and the scanning interval setting.
Using 146relative soil moisture observe stations data from 2003to 2008in the Beijing、Tianjing and HeBei area,we have established a spherical cap harmonic model of the annual trend.Spherical cap pole located in longitude 115°,latitude 38°.The radius of the spherical cap is 5degree while the maximum degree expands to 8degree.According to the spheical cap harmonic model,we calculate and analyze the characteristics of the relative soil moisture annual linear spatial distribution in 10cm 、 20cm and 50cm depth.
Differential Synthetic Aperture Radar Interferometry(DInSAR) can only measure one-dimensional deformation of surface along Radar line-of-sight(LOS).However,surface changes always exhibit three-dimensional(3D) characteristics.This paper presents a method to estimate 3D deformation of L'Aquila earthquake by combining multi-platform DInSAR LOS data based on the imaging geometry of Radar.The characters of the surface reflected by the 3D deformation are very consistent with the geological exploration.The 3D deformation was also compared with real GPS survey,and the results demonstrate that the 3D surface deformation fields are reliable and with high accuracy.
The moving average methods,including circumferential method and grid method,are extrapolated to gravity anomaly separation from traditional Cartesian coordinate system to spherical coordinate system.The effect of gravity anomaly separation depends upon the selected window size of moving average.In practice,a reasonable window size is estimated by the priori knowledge and trial.In gravity anomaly separation,there is the false anomaly,which can be removed by iteration.Model testing and actual application show that this method for gravity anomaly separation is better than truncation method between different degree of gravity field model to separate gravity anomaly,and the regional and residual gravity anomaly of the Earth have good coherence with actual geological tectonics.What's more,this method can be applied to global or local region.
This paper proposed a technological flow based on data mining technique for coping with difficulties in processing mass and various landslide-monitoring data effectively.Baijiabao Landslide was taken as a case study to analyze the feature information of deformation and its inducing factors by utilizing Apriori algorithm right after processing monitoring data qualitatively.It is demonstrated that the proposed technological flow can process landslide-monitoring data effectively and discover useful knowledge for understanding the circumstance and inducing-factors of landslide's deformation.
A new wavelet thresholoding model is proposed for denoising the deformation data,on the basis of the regular models.In order to verify the correctness and feasibilities of the method,theory analysis and experiments were performed.Results show that the proposed method can achieve higher SNR and smaller RMS.
Parallel algorithm of Harris corner detection based on the core concept of Multi-GPU is proposed,so that time-consuming Gaussian image convolution filtering part during the whole image corner detection process can be implemented by many parallel threads.Finally,implementation of this SIMT parallel algorithm using GPU mechanism of shared memory and constant memory and pinned host memory in CUDA is detailed.The Experiments show that the parallel algorithm of Harris corner detection based on Multi-GPU demonstrates substantial improvement up to 60 times speedup than the serial algorithm running in the CPU,is with high efficiency compared with CPU counterpart algorithm,and show great potential for large-scale data processing in real-time processing.