2006 Vol. 31, No. 7
An efficient and fully automatic approach to detect terrain changes and update old DEM simultaneously with outdated DEM and up-to-date aerial stereoscopic pair is presented.The related tests are done with geo-spatial data from two regions in China,and experimental results of both change detection and updating are satisfactory.Compared with existing procedure in production,the approach excels at efficiency markedly,requires less for source data,produces more products,and it is more robust and automatic.
The relation between the concept of remote sensing/GIS uncertainty and the uncertainty of information theory is discussed,to enable quantitative evaluation of the measurement of information and uncertainty of remote sensing and GIS data.The formulation for calculating information content of GIS graphics data is inferred based on the definition of uncertain degree in information theory.With this formulation,the information of geometric position and attribute can be measured in the unit of bit.Furthermore,the information of any geographic target in GIS is measurable by considering the doubtfulness of geometric position and attribute.Besides GIS vector data,the information of remote sensing raster imagery is also discussed.
An improved embedded hidden Markov model structure for face recognition is presented.The model not only improves the recognition rate,but also reduces the train and recognition time costs dramatically.Experimental results show that the highest recognition rate of 100% is achieved.The most exciting thing is that the structure can shrink 30% time costs than traditional structure both on train and recognition stage.
An automatic meteorological data mining approach is presented based on two kinds of heterogeneous satellite datasets,temperature black body(TBB) remote sensing imagery and high resolution limited area analysis and forecasting system(HLAFS) data,aiming to analyze the meteorological evolvement trends and patterns of MCS and discover the hidden correlations and causalities between the environmental physical fields and the eastward evolvement trends of MCS.Research results are provided in the form of decision rules and environmental physical model graphs to express the hidden knowledge effectively and understandably.
The method of airborne laser scanning altimetry footprint point cloud patches segmentation is discussed,a new method used to segment the footprint point cloud of group buildings is given,which is called "scanning strip method".The algorithm and flow chart of the method is demonstrated in detail.The practice shows that the algorithm is efficient,reliable,accurate,and can save great time of segmentation as well.
Based on the fictitious compress recovery approach the fictitious downward continuation is clarified,which has theoretically solved the downward continuation of gravity field.To test the correction of the theory,simulation experiments are executed using EGM96 model.Simulation experiment results support the local fictitious downward continuation approach.Meanwhile,in the case that the observation values contain random errors,the simulation experiment tests show that the local fictitious downward continuation approach is stable,almost not losing the accuracy.
With Europe main land as the region in study,the potential of the new satellite gravity technology,satellite-to-satellite tracking,to improve the precision of regional gravity field model of the SST models is investigated.The deficiencies of these models are discussed.With GPM98C as the reference,the gravity anomaly residuals of several other models,the latest SST global gravity field models(EIGEN and GGM),are computed and compared.The results of the comparison show that in the selected region,some systematic errors with periodical properties exist in the EIGEN and GGM's S serial models in the high degree and order.Some information that is not shown in the classic gravity models is detected in the low and middle degree and order of EIGEN and GGM's S serial models.
The contribution of the IB correction to the annual constituent S_(a) and other principal tides is evaluated.When the IB correction is applied with the MGDR or calculated with an average pressure for each cycle,the discrepancy both are large,and the difference between them is small. But if IB correction is not added to the altimeter ranges,the discrepancy is almost reduced.Through comparing the difference of harmonic constant of S_(a),the average contribution of seasonal air pressure variation in amplitude of S_(a) is(about) 42%,and the seasonal air pressure variation causes the phase lag of S_(a) decreasing 27°.So IB correction should not be added to the altimeter range,and the tide model of S_(a) is essential to the annual variation to be removed.Therefore an S_(a) model covering China sea is established.
On the basis of the analysis of error sources in marine magnetic survey,different error models,which can characterize the change of the systematic errors,and their compensation models are constructed.The solution methods of compensation models are presented.Some questions involved in solving the adjustment problem,such as the choice of error model,the stability of solution of problem and the test of significance of compensation efficiency,etc.,are discussed.A practical survey data set is used as a case study to prove the efficiency and reliability of the compensation methods.
The efficiencies are analyzed and compared to GPS data process using GA with two coding methods.GA with binary coding is prone to carry out GA operators,but its length is longer,EMS memory is larger and stability is weaker than ones of real coding.To integer nature of double-differenced ambiguity,real code were modified to round in coding then made up of individual.The results of the practical data processing show that GA with the modified real coding can be higher in reliability and get success rate on GPS ambiguity resolution than one with binary coding.
A method has been developed to calibrate the positions of the transducer and the transponder in the ultra-short baseline acoustic position system(USBL).The experimental results show that this method is more reliable and practical than that in the commercial USBL system.
Formulas are detailed deduced using Rung-Kutta method for orbit integral of GLONASS satellite motion function,and the integral scope and integral step on the basis of processed result using actual data are analyzed.A comparison is made between fix-step integral and vary-step integral after description of the two methods.
The methods of spatial modeling for fuzzy geographical objects are focused and the following two issues are resolved.Firstly,fuzzy geographical object is defined by the novel object-oriented modeling method.Secondly,fuzzy object class and fuzzy geographical object class are defined by unified modeling language(UML) based on vector data model,and the method of modeling spatial relationship is put forward.Then an integrated UML graph is established to express the conceptual model of fuzzy geographical object as well as the fuzzy topological relationships between objects.Last,the UML graph can be input to the general relation database to finish the database design.
Based on the suitability evaluation readjustment of village land use method of land suitability evaluation,a new evaluation system,hierarchical fuzzy evaluation method,is established for the complicated indexes problem of suitability evaluation of the readjustment of village land use.It makes good effects when it was applied in the case study of some villages and towns,Huangpi District.
The shortcoming of the methods of differential profit and multi-factor comprehensive appraisal is analyzed,and the main thinking and step of gradation by the land price price partition based on spatial clustering is put forward,and an example of Shishi City urban land gradation is given.
A new method of user-defined linetype is put forward to design linear symbol.This method can solve the drawing problems and improve the drawing speed.In addition,with respect to the serious distortion of symbol on the vertices of line,such as symbol breakage,symbol intersection,the methods of symbol self-adaptability and double affine transformation are used to manipulate the distortion,which promotes the visual effect of symbol and meets the need of linear symbol with different styles.
An algorithm for building typification based on data matching between multi-scale databases is proposed.The algorithm is based on the dividing and conquering strategy.The whole map sheet is divided into several parts by road network.Each part is divided into several clusters by means of spatial clustering.The cluster is the basic process unit of building typification.The algorithm supports parallel process and is rapid,effective and adapted to Web mapping.
A new spatio-temporal data model based on graph theory is presented,the organizing strategy of spatio-temporal data based on weighted graph is studied,and a formalized method to describe spatio-temporal semanteme is provided.The application foreground of transformation between different spatio-temporal data models is analyzed with the model.
Considering the characteristic of SVG and the publishing of SVGP,the application project of map cross-media publishing based on SVG is put forward,the implementation of SVG and SVGP in the cross-media publishing of map are discussed,and an experiment is carried out to verify the function of SVG in the cross-media publishing of map.
The spatial data quantity of different formats for a map are compared by experiments.Storaging map by raster format,its spatial data quantity is equal to the other of same size map if not be compressed.If only be compressed,the raster format data quantity has relation to chart speckle area size..bmp file spatial data quantity is the most large in several raster formats,.jpg file and.gif file data quantity are less.But storaging map data quantity by vector format has relation to spatial data attribute..e00 file spatial data amount is the most large in several vector formats.
The time-frequency analysis based on signal decomposition is proposed,after comparing some conventional time-frequency analysis methods such as short-time Fourier transform(STFT) and Wigner-Ville distribution(WVD).The time-frequency analysis based on signal decomposition can avoid that the time and frequency resolution in STFT method cannot be improved at the same time and can solve the cross-term problem in WVD method,and the feasibility of the time-frequency method based on signal decomposition is proved by processing the actual data.