2014 Vol. 39, No. 5
Objective First,we expound the background of rise of UAVRSS in this paper.Then,we discuss thefoundation,problem,research progress and trends for development of UAVRSS from unmannedaerial platform,flight control and navigation,data transmission and storage,data processing,sensortechnology,airspace usage policy,and so on.Third,we set forth the necessity and significance of de-velopment of UAVRSS through UAVRSS applications and practices in many related industry domain.At last,we make recommendations on development of UAVRSS from science and technology policy,industry policy,financial policy and so on,respectively.
Objective On the basis of the analysis of vehicle-borne laser point cloud features,aground filteringmethod based on the adaptive mathematical morphology is proposed.A three-dimensional virtual gridis used to organize data.In the morphological iterative process,the filtering window is optimized byannular structure and the morphological opening operation is controlled by a local adaptive slopethreshold.After having obtained the initial low point,the ground point cloud data is extracted basedon neighborhood slope.Finally,experiments are processed using two different regions in the pointcloud data validiting the algorithm.
Objective In this paper,a marked point process based method is used to extract buildings from air-borne LIDAR data.At first,a Gibbs energy model is build according to the geometric feature of theobject in the point cloud data.This model contains both a data coherence term which fits the objects tothe data and a prior term which incorporates the prior knowledge of the object geometric properties.Then the previously defined model is optimized by the RJMCMC(Reverse Jump Markov Chain MonteCarlo)algorithm and simulated annealing algorithm.Finally,fine processing removes the terrestrialpoints,noise points and tree crown points of the extracted objects which are mistakenly extracted asbuildings,while combining adjacent objects.The method was tested with three different aerial LiDARdata sets from ISPRS.The results show that our method is capable of efficient and robust building ex-traction.
Objective Stripe noise can be found in many bands of Terra MODIS data,especially in the 28th band.Additionally,the problem gets more serious and complex with time changes.In many cases,the datasuffers from detector,detector stripes and mirror stripes noise,this paper proposes an interpolationfitting destriping algorithm and an detrending destriping algorithm both based on a moment matchingmethod.Experimental results and quantitative analyses verify the efficacy of these two new destripingalgorithms.
Objective Homogeneous regions or edges are important structural information for object recognitionand extraction in high resolution remote sensing images.This paper considers the homogeneous re-gions and edges from the perspective of spatial dependence,which is a measure of the spatial associa-tion between the pixel values in the image.Spatial dependence is one of the spatial characteristics ofhigh resolution images.Based on the measure to spatial dependence using local spatial statistics(localMoran’s I,local Geary’s C and Getis),this paper proposes a simple,effective method of extractingspatial salient structures(homogeneous regions or edges)which adopts a new technique of 3Dthresh-olding for spatial dependence intensity.Comparative experiments show the potential and performancedifferences of three statistics in modeling spatial dependence and extracting spatial salient structures.
Objective After the launch of the new generation high-resolution SAR satellites,high quality SAR da-ta of up to meter resolution became available.However,severe layover problems occur especially incomplicated scenarios like dense urban areas,which make high-resolution data interpretation more dif-ficult.SAR tomography(TomoSAR)aims at retrieving distribution of scatterers in the elevation direc-tion and retrieving the corresponding reflectivity inside one resolution cell.In this way,TomoSARcan achieve real and unambiguous 3Dreconstruction.First the basic principles of SAR tomography areillustrated.Then,a Butterworth filter based singular value decomposition method is proposed for To-moSAR processing.Third,experimental results about Berlin Debis Tower are analyzed in detail bycomparing them with real building heights,the estimation precision reaches the meter level.
Objective A physically-based model for calculating radiance reflected from neighboring slopes is devel-oped.The path radiance and atmospheric transmittance of reflected terrain radiance between two adja-cent slopes are considered.The model computes terrain irradiance illumination and its ratio to total ir-radiance as quantitatively supported by DEM data and atmospheric parameters.According to an exper-iment with spot images and the principle of terrain irradiance,terrain irradiance illumination dependson the surface reflectance of the adjacent objects,terrain,the total irradiance reaching the adjacentslopes and other factors.The radiance reflected from neighboring slopes usually is small,but it has tobe accounted in high surface reflectance areas.The ratio of terrain irradiance to total irradiance de-pends on direct irradiance,diffused irradiance,and terrain irradiance,which are received by a targetpixel.The ratio of terrain irradiance to total irradiance of low illumination pixels is larger than that ofits surrounding pixels.For a shadow pixel,the ratio of terrain irradiance to total irradiance of theweak scattering bands is larger than that of the strong scattering bands.
Objective A post-processing method is proposed based on the theory of multiple-point geostatistics.The method extracts prior spatial structures from a training image,and infers the pattern distributionand correlation of classes.A spatial correlation model can be established from training image,which ispreferable to the traditional two-point-based variogram model.An experiment was performed on aLandsat TM image,wetlands with a complicated distribution were extracted.The method was com-pared to the spatial filtering and the contextual Markov random field(MRF)classifier.This approachincreases overall classification accuracy,and has advantages when dealing with classes that have curvi-linear distributions.
Objective Natural hazards and disasters in China have increased in magnitude and frequency in recentyears as consequences of the fast urbanization promotion and global climate change,which has greatlythreatened to the sustainable development of social economy.The widespread devastation,economicdamages and loss of human lives,caused by numerous forms of natural emergency situations,are be-coming more serious.The emergency surveying and mapping service system(ESMSS),which is ableto make practical contributions to planning and operation of civil protection and disaster reduction,isurgently needed.The paper probes into some key techniques involved in the construction of theESMSS at first.Taking applications in“4·20”Lushan earthquake as an example,the article deeplyanalyses the problems in data acquisition,transmission,processing,distribution,sharing of the cur-rent ESMSSs and their working mechanism.Then,the construction method of the proposed task-driven ESMSS is presented in details.Finally,the key techniques including focused service mechanismare expounded.
Objective Road network generalization and analysis have been hot issues in the geographical informa-tion science.The stroke in a road network is defined as a set of one or more road segment in a non-branching,connected chain,based on the good continuity principle.The stroke plays an importantrole in road network generalization,analysis,pattern recognition and schematic map generation.Ex-isting studies focus on the stroke generation algorithms.However,the formalization of stroke genera-tion at the conceptual level is absent.This paper first formalizes the stroke generation as clusteringproblem,and then presents a hierarchical clustering based stroke generation algorithm.Time com-plexity and some properties of the proposed algorithm are analyzed in detail.Finally,the algorithm isverified using the Shenzhen road network at the 1∶50 000scale.
Objective Web service and map generalization can improve syntactic and semantic interoperability how-ever their collaboration remains unsolved.On the one hand,standards for geospatial information serv-ices are generic and do not consider the specific field,such as map generalization;on the other hand,the generalization of knowledge is too complex to systematically extract accurate information to enrichthe semantics of a map generalization service.This paper aims to investigate the classical“operator-al-gorithm-threshold”decision mode,introducing an implicit correspondence between the mode and WPSinteraction process,and further extending this to service chaining.Based on this implicit relationship,four kinds of information suitable for enriching generalization services are summarized and analyzed indetail.Finally,this semantic information mentioned above is used to extend the WPS specification.The semantic information in this study is concise and accurate enough to be embedded in the WPS in-teraction process hierarchically and thus improves the semantic interoperability of a map generalizationservice while taking the OGC standard into account.
Objective In order to prevent the gaps in coordinate datum between different regional continuously op-erating reference systems(CORS),we analyze the causes and introduce a method for coordinate da-tum unification.Using experimental data from the Hubei and Hunan CORS,the results show that thesystematic deviation reaches 17.9±4.5mm between these two datums.The proposed method establi-shes a unified coordinate datum to eliminate the systematic deviation and builds connections betweeneach coordinate datum.The validity of our method was supported by empirical evidence.Finally,some suggestions about coordinate datum unification for a regional CORS network combination are putforward.
Objective To address the problem of fitting a straight line in three-dimensional space,since the equa-tion is a six parameter equation,not a simple linear relationship,the traditional least squares methodcannot be used to solve it.In this paper,a new method of space line based on the total least squares isproposed.Firstly,the number of parameters was decreased from six to four by changing the standardequation of the straight line,then re-expressed the equation in the form of a matrix.Therefore,thefitting problem was transformed to the parameter-solving problem in total least squares.Further,thefitting four parameters were obtained using a TLS iteration,and the six parameters of the space lineswere recovered through a backtracking method.An experiment in the paper verifies the effectivenessand applicability of the new method.
Objective By selecting different number of IGS tracking stations,ERP parameters are estimated fromGPS observations and the combined observations of GPS and GLONASS,respectively.The influencesincurred by increasing the number of stations and adding GLONASS observations to the estimatedERP parameters are analyzed by comparing results to IGS published values.Thirdly,the ERP param-eters are estimated with SLR data from GLONASS satellites.Finally,ERP parameters are estimatedwith joined results from SLR and GNSS.This research shows that not only that the systematic errorsof ERP or High-frequency ERP estimated from GNSS observations are ameliorated,but also the sta-bility of estimated ERP is improved greatly when using the joined results from SLR and GNSS.
Objective Initial alignment is one of the key technologies of the strapdown inertial navigation system.The applications of the strapdown inertial navigation system however,are directly affected by the ac-curacy of initial alignment.Kalman filtering is an effective algorithm for SINS initial alignment,butthe optimal estimates are based on the filtering model and the noise covariance matrices which are al-ready known.This paper focuses on the simplified autocovariance least-squares noise estimation meth-od in the strapdown inertial navigation system’s stationary initial alignment.The proposed method es-tablishes a relationship between unknown measurement noise and the autocovariance.The noise co-variance can be estimated by solving it as a linear least squares problem.The proposed method esti-mates measurement noise and corrects INS attitude by iterative calculation.Simulation results showthat the proposed method performs very well in noise covariance estimation and strapdown inertialnavigation system initial alignment.
Objective In GPS/INS integrated navigation,the low precision of the INS dynamic model reduces nav-igation accuracy.Based on Newton interpolation,a multi-order dynamic model algorithm for GPS/INSintegrated navigation is proposed.First,the detail algorithm of Newton interpolation is introduced.Based on this a Gauss-Markov model of the inertial system is modified to realize the multi-order mod-el.The design formulas of the observation and dynamical models are presented.Finally,an actual cal-culation was performed to test the validity of new algorithm.The results of the experiment indicatethat when compared with the Gauss-Markov model,the inertial dynamic model for GPS/INS integrat-ed navigation based on Newton interpolation can improve the position and attitude precision effective-ly.At the same time,an analysis of the experiment shows that a 4-order Newton interpolation modelnot only enhances model precision,but also reduces model complexity,which provides a good refer-ence for order selection in the Newton interpolation model.
Objective In connection with huge GNSS network data processing,has three strategy approches,awhole huge net data process,a non-difference method and a divide the network to small networksprocess,we propose to divide a network and analyse the limitations of this module based on theGAMIT subnet module and present a move grid density method.The method is researched and ana-lyzed in detail using data from CMONOC.The results indicate that the baseline repeatability and siteprecision is more than methods based on the GAMIT subnet module method’s result,while the algo-rithm weakens the unreasonable situation in subnetting,enhancing the reliability of a subnet struc-ture.
Objective Usually,it takes a long time to get the double-differential ionoshperic delay using dual-fre-quency observations.To solve this problem,based on the three-frequency carrier,a new ionoshpericdelay estimation method for long distance is proposed.First,the optimal combinations for ionoshpericdelay resolution are selected according to error characteristics of the different combinations.Then,in-itial ionoshperic delay is solved using optimal combinations ambiguity.At last,an idea which is simi-lar to carrier-smoothed code through hatch filter is introduced to improve the initial ionoshperic delay.Experimental show results that the estimation accuracy of double-differential ionoshperic delay can bein 2cm with only dozens of epochs or even less.It realizes the real-time and high accuracy resolutionfor long distance double-differential ionoshperic delay.
Objective Using COSMIC occultation data in 2011,the atmospheric temperature profiles in Antarcticaare inversed,and the tropopause parameters(temperature and altitude)are extracted.The spatial andtemporal Antarctic tropopause variations are analyzed quantitatively.In Antarctica,the temperaturelapse rate tropopause is more accurate than the coldest point tropopause,and the occultation method isthe same as ozonesonde and radiosonde.The Antarctic tropopause shows one wave structure with con-trary phase,temperature ranges from 200Kto 230K,the height from 9km to 11km.Antarctic trop-opause disappears in winter and spring,and occurs inversion layer in summer and autumn.The tropo-pause temperature in winter and spring shows significant gradient feature.In latitude direction,thetemperature is low near the pole,high around;in longitude direction,the temperature is lower in theWest Antarctica.
Objective A coastal low water model with high precision has important application value.Aiming toexploit the characteristics of multi-source and precision difference between different tidal data for con-structing coastal low water models,a method based on multi-source tidal data is presented in the pa-per.By utilizing multi-source tidal data,such as tidal gauge data and low water data calculated basedon satellite altimetry data,interpolation method of tidal data is modified by assigning a weight todifferently sourced data,depending on the precision information,thus the coastal low water model isconstructed considering the precision difference between different source data.Experimental resultsdemonstrate that the proposed method can utilize multi-source tidal data and improves the quality ofcoastal low water models,which can meet the demand for special applications such as extracting lowtide lines.
Objective To establish a regional seamless chart datum and achieve seamless vertical datum conversion,thispaper starting from the datum calculation principle,and according to the tidal wave propagation characteristic,we propose a new method of constructing regional seamless chart datum based on the tidal harmonic constantsinterpolation.Firstly,the constants of unknown points in waters can be calculated by interpolation with usingthe known tide gauges.Secondly,we establish seamless chart datum based on these points with using a math-ematical model.When compared with the traditional geometric method,it significantly improves the construc-tion precision and overcomes the precision unstablity defect.Experiments in Yangtze Estuary were performed,and the results show that the method is correct and feasible.Finally,a seamless chart datum in Yangtze Estu-ary was established.
Objective Based on the Kalman filter function and stochastic models with ambiguity parameters,a newalgorithm is proposed for determining the parameters of the Kalman filter in RTK positioning.Theactual measured short baseline data of GPS/BDS is processed by a self-compiled GPS/BDS RTK posi-tioning program using this algorithm.It employs only a few epochs to give centimeter level positioningresults.Positioning results for three modes using BDS,GPS and BDS/GPS are compared and ana-lyzed.In case of short baselines,the RTK positioning accuracy of GPS/BDS is not improved noticea-bly relative to GPS or BDS.However it takes less time to get a fixed solution for GPS/BDS.This a-nalysis provide reference point for further GPS/BDS positioning research.
Objective Modeling and forecasting of the geomagnetic variation field is the important research topic ofgeomagnetic navigation and space environment monitoring.According to the chaotic feature of geo-magnetic variation time series,a combined forecasting model based on modified ensemble empiricalmode decomposition(MEEMD)-sample entropy(SampEn)-least square support vector machine(LSS-VM)is proposed.Firstly,the geomagnetic variation time series is decomposed into a series of geo-magnetic variation subsequences with obvious differences in complex degree using MEEMD-SampEn.Then,the forecasting model of each subsequence is created with LSSVM using the optimal model pa-rameters.Finally,the simulation is performed by using the real data collected from the geomagneticobservatory.The results show that the forecasting value of the MEEMD-SampEn-LSSVM model canclosely keep up with the trend of geomagnetic variation field,and obviously better than the other twomodels.The mean absolute error of the model forecasting three hours is 1.63nT when Kplessthan 3.