2007 Vol. 32, No. 12
A method for computing the value of eave correction by using the single close range image of a building is introduced.In the method the digital lean image and the conditions of geometry and restriction of the framework appearance of building are used.According the cross ratio invariability and perspective reversion,the relation among the vanishing points,the spatial beelines and their projections in image is determined,and the width of the eave can be calculated.Experimental results show that it meets the requirements.
Adopt a two-steps strategy to achieve the automatic 2D-3D pose estimation of sheet metal part.The generalized point adjustment with extreme point constraint,which based on lines extracted from multiple baseline images is also introduced to accurately reconstruct and measure the industrial sheet metal part.
An approach for estimating urban imperviousness surface percent(ISP) through the synergistic use of Landsat ETM+ and high-resolution imagery is presented.In the proposed approach,the ensemble leaning of CART analysis based on Boosting is utilized to construct the ISP predict model,furthermore,to obtain the subpixel ISP results at 30 m resolution.The experiment shows this approach yields ISP estimation performance compared to that of the traditional estimation method based on single CART,and its correlation coefficient of predicted versus actual ISP reaches 0.91 with an average error 11.16%.
The method is proposed to extract the residential areas on panchromatic remote sensing images based on Na ve Bayesian network.In this procedure,the conditional probabilities and prior probabilities are obtained from learning the positive and negative samples.On the basic,unknown regions that have not the category information will be classified into residential areas and non-residential areas by learning probability network.Experimental results of extracted residential areas on the images of SPOT5 are presented to illustrate the merit and feasibility.
To meet the requirements of efficient management and web publishing for marine remote sensing data,a spatial database engine,named as MRSSDE,is designed independently.The logical model,physical model,and optimize method of MRSSDE are discussed in detail.Comparing to the ArcSDE,which is the leading product of spatial database engine,the MRSSDE proved to be more effective.
Evaluation strategies of terrain generalization approaches for regular-grid DEM data were introduced firstly.Four kinds of generalization methods were used to generalize the same region of DEM data separately,which the scale of DEM was changed from 1∶50 000 to 1∶250 000.And then the generalized results were evaluated in three areas: elevation distribution,derivative slopes and contours' overlay analysis.The factors about how these methods affected the precision of the generalization results were also analyzed.Lastly the structured generalization method was considered to be the best method with the highest precision of these four kinds of regular-grid DEM generalization methods.
Spatial database updating is a bottleneck in GIS,which needed to be resolved urgently.Some data updating modes and technologies are discussed,and a multi-scale data updating scheme is presented based on CHT_EUR spatial database model.It can automatically and effectively identifies those changed data in updating by some methods such as spatial feature matching,comparing attribute and implement linkage updating of multi-scale spatial data based on linkage-relation which constructed by geographical object,settled the disaccord of multi-scale data.The scheme is significant to maintain the current spatial data and perfect data product system.
With the characteristic of 3D laser scanning on road and the distribution analysis of points cloud's normal vectors,the filtering algorithm based on fuzzy clustering of normal vectors is proposed.Three different types of real road data are put into experiment.It proves that the algorithm gives good filtering effect to non-road data produced in the road scanning.The algorithm provides a new method for filtering points cloud data.
A technical framework of DCT-based GIS vector data compression is proposed.DCT is applied to transform vector data into combination of cosine curves,while in this combination coefficients are quantified to integers before lossless coding to compressed bit-stream.Experimental results show that this compression framework works well for naturally formed linear vector data with very high compression ratio while its distortion is well controlled by predefined thresholds.
A novel approach is presented to solve feature subset selection based on ACO(ant colony optimization algorithm).The approach has the ability to accommodate multiple criteria such as accuracy and cost of classification into the process of feature selection and find the effective feature subset for texture classification.A classifier based on minimum distance is described to classify two types of texture images with feature subset selected by ACO and extracted by PCA(principal component analysis) respectively.Experimental result illustrates that the algorithm can reduce feature dimension,speed the classification of image and improve the recognition rate compared to PCA.
The index structures of HR+-tree,SEST spatio-temporal index and snapshot-increment index are blended,and a new spatio-temporal access method is proposed,which distinguishes the shared pointer and non-shared pointer.This method is named as EHR-tree,and is possible to not only process time slice and interval query,but also queries about events.The EHR-tree implementation uses a HR-tree structure for storing snapshots and a log data structure for storing increments and events that occur between consecutive snapshots.
Using the IGS precise ephemeris,the acceleration of GPS satellites can be computed,which can be regarded as the result of all forces.Based on mechanical models,all kinds of perturbation accelerations can be computed.The accelerations of two methods are compared and analyzed.It's concluded that the difference between mechanical models and true forces of satellite is evident.
The on-orbit measurement performance of GLAS is summarized,and the precision of the data products is analyzed.With the high-accuracy ICESat measurements and the GPS ground surveys in 21st CHINARE inner ice sheets research,the validity of Dome A area defined in 21st CHINARE and the measured high-point of Antarctic inner ice sheets is analyzed.
This paper first introduces the methods of calibrating antenna phase center offset and variation and the principle of correcting antenna phase center,then describes the algorithm in GAMIT.A sample is computed to consider the impact of phase center variation on the results of GPS positioning.Finally some feasible conclusions and suggestions with respect to antenna phase center correction in high precision GPS applications are listed.
A new algorithm,named as several-epoch algorithm (SEA) is studied in the short baseline GPS rapid positioning using several epoch single frequency phase data.Based on TIKHONOV regularization theorem,a new regularizer,which has explicitly physical meaning,is chosen to mitigate the ill-condition of the normal matrix.The accurate float ambiguity solutions and their MSEM(Mean Squared Error Matrix) are obtained using several epoch single frequency phase data.Combined with LAMBDA method,SEA can fix the integer ambiguities correctly and quickly using MSEM instead of the covariance matrix of the ambiguities.At last,the baseline vector is resolved.The effect of the new algorithm and LS is illustrated by one example.
The calibration of onboard accelerometer using dynamic method and the simulated orbit is studied in detail.Based on the calibration results from different reference gravity model,integrating the satellite orbit again and compared to the simulated orbit respectively,the maximum difference is on the order of meter,which shows that the impacts of the reference gravity model on the calibration of accelerometer cannot be negligible.The further investigation demonstrates that a better strategy is to calibrate the accelerometer and estimate gravity model parameters simultaneously to remove the aliasing error of the reference gravity model.
On the basis of the combined bias interpolation(CBI) method of GNSS network RTK,a modified combined bias interpolation(MCBI) method is proposed.MCBI computes the troposphere delay using the troposphere model and estimates the first order ionosphere delay.Then MCBI separates the combined bias into three parts: the first order ionosphere delay,the troposphere delay computed by troposphere model and residual error.The example shows that the precision of MCBI is higher than that of the CBI and very convenient to use.
The theory and method for validating satellite clock offset using SLR and pseudorange data are presented.The precision of this method is evaluated,and the feasibility is validated.The clock precise of forecasting broadcast ephemeris is validated using the data of 2005.Some useful conclusions are presented.
The deficiency of the current technique of general land use planning layout in our country is discussed,and the standard cellular automata's cellular meaning,rule definition etc are expanded.The land use planning layout model is presented under neighbor match promoting,constrained with suitable,constrained with inheritance,constrained with target of land use planning.And by taking Dongxing city of Guangxi autonomous region as an example,the simulatation of the model and the result analysis are made.
Using the output benefit coefficient,consumption and return coefficient,pollution and substitution coefficient and intensive cultivation coefficient,the benefits of cultivated land utilization is analyzed in quantitative way.Policies are proposed from the perspectives of financial investment and property-right system,which are different from the traditional modes.
A new method for extracting and identifying pointshaped symbols of scanned topographic maps based on map algebra is presened.The extracting results of element maps are optimized and the lines which do not meet the features of point symbols are filtered according to curve density and line density.Weighed distance function,and identify symbols using the method of template matching are put forward according to the function results.The algorithm can work out the problems of extracting and identifying point symbols in the condition of character & line conglutination or complex background,and it achieves a high level of recognition accuracy.
Analyzing the basic principle of automatic hill-shading and methods for enhancing the 3D expression of relief in traditional hill-shading,exposure and slope are considered to be important factors to the 3D expression of hill-shading.A proper change to gradient or grayness according to slope and exposure can be helpful to improve visual effect of hill-shading map.Two methods are introduced to modify gradient and grayness in computer-aided hill-shading.Practices prove that these methods are effective.