2018 Vol. 43, No. 5
In accordance with the special requirement of determining the external gravity field quickly in ultra-large sea area, a detail analysis is made on the applicability and limitations of the three traditional models, i.e. spherical harmonic expansion model, direct integral model and point mass model, for computing disturbing gravity. And then three modified computational models, i.e. modified direct integral model, modified point mass model and mixed model of direct integral and point mass, are proposed one by one. A numerical test has been made to prove the reasonableness and validity of the suggested models. It is shown that the singularity in three traditional models have been eliminated in the solutions of the modified models. They can satisfy the practical needs of determining the local gravity field quickly in ultra-large sea areas and full height. The accuracy of the modified point mass model is estimated to be about ±1mGal in smooth sea areas, and to be less than ±3mGal in rugged sea areas. The obtained conclusions are valuable to the practical engineering projects.
When the sampling rate of base station is lower than that of the rover station, it's impossible to obtain the coordinates of the rover station at all epochs with a conventional differential GNSS post-processing method. To solve this problem, a method based on PPP model is proposed to construct the virtual observation data of the non-sampling points of base station. The method separates the receiver clock error and tropospheric zenith wet delay from the observation error, and estimates them simultaneously with ionosphere ambiguity, after which the distance between the satellite and the station is obtained by base station's real coordinate. On this basis, the method calculates the residual errors of two adjacent epochs, by which the residual errors of non-sampling epochs are fitted. Eventually, the virtual observation data are generated by the non-sampling epochs' residual errors, the distance between the satellite and the station and the estimated errors. Error characteristic of the virtual observation data is maintained, especially the common error between the base station and the rover station; This method only densifies the data of the base station, which will not affect the rover station. The experiment results show that when base station works with a sampling interval within 30 s, the virtual observation data generated by this method are accordant with the real data. In 30s, 15s and 5s sampling intervals, the standard errors of virtual observations of pseudo-range are about 0.2, 0.1 and 0.05 meters, and virtual observations of carrier phase are about 1.2, 0.7 and 0.2 cycles, respectively; In the case of 30 s sampling interval, the positioning result obtained by this method can still meet the cm level accuracy requirements.
The core-mantle boundary (CMB) is one of the most important physical and chemical interfaces in the earth's interior, through which a variety of interactions occur between Earth's core and mantle. These interactions might have great impacts on Earth's gravity field, rotation and magnetic field. The geoid anomalies is an important observation of the earth's gravity field, which reflects significant information of the earth's interior, such as material density anomalies and interfaces, etc. A formula using the geoid anomalies to invert undulations of core-mantle boundary was derived. The large-scale undulations of the core-mantle boundary were calculated by this formula from degree 2 to 4. The result showed that the amplitude of core-mantle boundary undulations reached ±5 km, which corresponded to that obtained by Morelli using seismic tomography, but showed some differences in the worldwild distribution. We also simulated influences of the density anomalies in the core-mantle boundary to the geoid undulation using a prism model with height of 5 km and base length of 1 000 km. The result showed that it was closed to the observed value of geoid undulations.
It is difficult to model and predict satellite clock offset with conventional approaches. In this paper, an extreme learning machine (ELM) is used to predict satellite clock offset in order to improve prediction accuracy. For the problem that it is arduous to determine the hidden layer structure of ELM neural network, a new algorithm for ELM network structure design is proposed based on the good online classified characteristic of adaptive resonance theory (ART) network. The proposed algorithm employs the clustering characteristic of ART network to design the ELM network structure. The number of hidden layer nodes can be determined adaptively through the similarity comparison of input vector. The experiment results show that the ART-ELM prediction model outperforms the quadratic polynomial model and grey model remarkably.
Using the observation data of GPS sites in southern Qinghai-Tibet plateau in recent 10 years (2004-2014) and combing with the observation data inversion of the GRACE period regional hydrological load changes caused by the vertical displacement, it can be found by GPS and GRACE that there is a vertical displacement time series in a certain correlation. The correlation coefficient of most sites is around 0.7; especially the site of Himalayan orogenic belt, the correlation is higher than others. Most sites, such as CHLM and KKN4, the fitting cycle amplitude and phase are in consistent. Defining RatioWRMS to characterize of GRACE to correct the effectiveness of the GPS vertical hydrological load, the value closing to 1 indicates that GPS and GRACE data have a greater consistency; The study showed that the different areas of GPS sites RWRMS have obvious differences:maximum and minimum value are respectively 0.96 (TPLJ) and 0.24 (XZGE), while the average is 0.64. This is associated with the south Tibetan region of the complexity of the vertical movement, there are many factors (tectonic movement, GIA, the non-tidal atmosphere and ocean, Seasonal Hydrological Load, the tides, etc.) work together.This article research results to further study the south Tibetan region of tectonic movement has important reference value.
As the drought occurred frequently, and remote sensing technology matures, the drought remote sensing monitoring has become an important means of drought monitoring, and the vegetation supply water index(VSWI) products is an important reference for drought monitoring, but its production involves large amount of data, and the processing cycle is long, which influence the timeliness seriously. Based on OSGI service, a distributed production model of VSWI is proposed, which turn the VSWI algorithm into an OSGI module called Bundle. This Bundle can be deployed and installed dynamically in a distributed OSGI platform, which can make full use of the resources in LAN, and be helpful to raise up the operating rate of image processing, as well as lessen the memory footprint, which have been proved in this paper. This model will play an important role in massive remote sensing data processing and emergency monitoring.
The 3" SRTM elevation error in China were methodically evaluated by using sample survey at more than five hundred thousand sites. The characteristics of SRTM errors was identified according to the landscape attributes. The results showed that SRTM error is closely related to different topography and land cover types. The error changes from positive to negative with its absolute value getting bigger with increasing slope; Positive error concentrate in north facing aspect, while negative error in southwest facing aspect; The error increases with increasing vegetation cover fraction; In the regions of glacier, desert and wetland, the mean errors are negative, while in building areas the mean error is positive. Among different factors, slope plays the pivotal role affecting SRTM errors. Additionally, SRTM data has significant abnormity in some areas such as desert and steep mountain, and arise stripes in flat areas.
An approach for damaged rooftops areas detection is proposed based on visual bag-of-words model. First, the building rooftop is segmented into different superpixel areas using simple linear iterative clustering(SLIC) method, then features of color and histograms of oriented gradients are extracted from each superpixel area and the visual bag-of-words (BoW) model is employed to build the semantic feature vectors of damaged and non-damaged area. Finally, damaged and non-damaged parts of rooftop superpixel areas are discriminated using SVM. Experimental results show that the proposed method can be feasible and effective for detection of damaged rooftop areas, which is an important significance for improving the accuracy of overall building damaged detection.
The side-scan sonar towed working mode leads to changing local distortion in types and magnitude along the track, and further results in features' disposition and distortion in mosaic image. In order to guarantee clarity edge of the local features in overlap region of adjacent images, this paper proposes an elastic blocking mosaic method based on speeded-up robust features. Firstly, combining with the information of track line and swath, each stripe image can be sliced to several blocks. Then the SURF image registration is done in each block including features extraction, matching and refining. With the registration point-pairs between blocks from adjacent stripe image, a rigid transformation in global image and a series of elastic transformation in each block image can be done. The method not only eliminates systematic errors of position and heading between two adjacent side-scan sonar images, but also weakens random local distortion and achieves high-accurate registration in local area by corresponding features. At last, the experiment demonstrates the validity of this method, the overall registration accuracy reaches two pixels.
In order to overcome the drawbacks of using either spectral or morphological features for traditional image segmentation methods, a multi-scale image segmentation method using both the spectral and morphological information is proposed. First of all, Differential Morphological Profiles are combined with spectral features to form spectral-morphological characteristics. Then, Hausdorff distance is implemented to calculate the weight of edges based on graph theory and minimum spanning tree algorithm Kruskal is applied to complete the initial segmentation of color images. Finally, the obtained segmentation result is refined by a region merging procedure with the regional heterogeneous criteria proposed in fractal network evolution. Furthermore, object-based Gray Level Co-occurrence Matrix and object-based Pixel Shape Index are proposed on the basis of segmentation results. Experimental results show that the proposed segmentation method is more effective and more efficient than eCognition software 8.0 and Meanshift algorithm. In addition, object-based Gray Level Co-occurrence Matrix and object-based Pixel Shape Index are apparently better than traditional pixel-based methods.
Vegetation structure parameters retrieval and forest ecological monitoring based on satellite large-footprint and full-waveform lidar(light detection and ranging) data is a hotspot in recent years. Since vegetation structure parameters and the optical properties have effects on lidar return waveforms, a waveform simulator specified to vegetations for a satellite lidar is established to investigate the influence in detail. The statisticlaw of the spatial vegetation distribution is extracted from field measurements; the parameterized vegetation reflection model is generated by considering the surface roughness and slope, and vegetation canopy reflection characteristics; then, the waveform simulator is developed based on the echo theory for satellite lidars. The simulated waveforms based on the field measurement data in Greater Khingan Mountains and the GLAS echo waveforms have good consistency with R2 equal to 0.91. The smaller footprint diameter is beneficial to retrievethe vegetation information that locates at the terrain with a large slope.This research is of reference to the lidar system design for the developingsatellite lidar system of our country.
Image super-resolution reconstruction is the method that uses one or several low-resolution images to reconstruct a high-resolution image. Sparse representation has been widely used in single image super-resolution reconstruction. However, the contents can vary significantly across different patches in a single image, and the fixed dictionaries, which common super-resolution algorithms based on sparse representation often used, cannot suit for every patch. This paper presents a novel approach for single image super-resolution based on sparse representation, which trains the dictionary with external database and the input low-resolution image itself. With the nonlocal similar patches extracted from the input image, the dictionary is updated by on-line dictionary learning method to ensure that the new dictionary is suitable for every patch in the image. Extensive experiments on natural images and remote sensing images show that the method with on-line dictionary learning achieves better results than those of the state-of-the-art algorithms in terms of both objective and visual evaluations.
In automatic map compilation, spatial conflicts among various spatial objects in maps occur at one time. The displacement of these objects will influence their proximity objects, so the collaborative displacement of them is needed. In this paper, we try to put forward a collaborative method for solving spatial conflict among various spatial objects after identifying spatial proximity conflict automatically, based on cartographic rules which are common in topographic map compilation. Rules related to map feature displacement are formalized with help of a parameter table. Firstly, the possible conflict area is recognized automatically based on constrained Delaunay triangles, then regions in which map features could be shifted are identified, and defined as operational displacement zone. In this zone, associated line networks for displacement propagation is established. Map objects in this zone are shifted collaboratively based on associated line networks and Beam displacement model of energy minimization, considering cartographic rules. At last, the effectiveness and availability of this given method in this paper is verified by means of the experiment.
The development of 3D space planning and management strongly require 3D data to support volumetric representation of 3D space. This paper aims to construct 3D closed building from building information with CityGML LoD3 data and provides the framework and workflow that how to transform and construct 3D closed buildings from CityGML LoD3 data step by step. The semantic objects between 3D property unit and building information are quite different and the 3D space they describe and represent are thorough different. According to the requirement of the geometry of 3D cadastral objects, semantic relationships and object correspondences between CityGML LoD3 and 3D closed building are calibrated. Based on these, the paper gives the methods to extract the geometry data from CityGML and recombines the data to enclose the 3D closed building that satisfy certain 3D geometric rules. The approach in this paper develops a space transformation idea way to construction 3D property unit from building information and can make up the deficiency of the traditional acquisition methods for 3D cadastre.
Motivated by the idea of total least squared method, a total error-based multiquadric method (MQ-T) has been developed to decrease the effect of both horizontal and vertical errors inherent in sample points on surface modeling. Two examples including a numerical test and a real-world example were, respectively, employed to test the robustness of MQ-T to sample errors. The numerical test indicates that when sample points are only subject to vertical errors, MQ-T has a similar performance to MQ. When sample points are subject to horizontal errors, MQ-T is more accurate than MQ. In the real-world example, MQ-T was used to construct DEMs with sample points collected by a total station instrument, and its accuracy was compared with those of classical interpolation methods including inverse distance weighting, ordinary Kriging (Kriging) and ANUDEM. Results indicate that with the decrease of sample density, the interpolation accuracies of all methods become lower. Regardless of sample density, MQ-T is always more accurate than the other methods. Yet, compared with Kriging, MQ-T has a peak-cutting problem.
Multi-holed plane object, as one of the abstracts of the real world, mainly represent geographic objects having more than one interior boundary, such as areas that contain a few lakes, or lakes with islands. To realize the matching between these spatial objects, the paper proposed a model of similarity measurement on multi-holed regions, with several restrictions being taken into account. In this model, the multi-holed plane object was viewed as a micro-spatial-scene, where holes and direction between holes playing roles of spatial objects and spatial distribution relations respectively. Taking into the direction between holes and the shape of holes account, Fourier descriptor was utilized to describe the shape of holes and Feature Matrix of direction was applied to represent the distribution relationship between holes, then the process of measuring similarity would be transformed into a constraint satisfaction problem (CSP). Association graph containing nodes and edges could be adopted to represent the matching solutions of CSP. In the paper, A case study of Urmia, a lake in Iran, is given to illustrate the whole process of similarity measurement among the shapes of the lake in different years, and the result of experiment is presented to be simple and applicable.
Similarity measurement is the key of Geography Information System (GIS), it is widely applied to spatial retrieval, spatial information integration and spatial data mining, etc. By considering the scale difference and studying the spatial semantics of spatial scenes, this paper build a formalized description model of spatial scene. According to the description model, multi-scale spatial scenes can be abstracted to feature matrices that contain the essential features. With feature matrices, we establish the initial probability matrix of spatial scenes, and iteratively update the matrix by relaxation labeling approach until it convergences to a global minimum value. After that the matched objects between two spatial scenes can be found. Accordingly, we calculate the spatial scenes similarity. In the experiment stage, Wuhan residential region data was adopted. We analyze the accuracy and total time spent of spatial scene matching process under different neighborhood searching radius. The experimental result confirms that spatial scene matching based on relaxation labeling approach has a high accuracy.
A network construction method based on temporal-spatial influence is proposed. Then a crime transmission network based on the impact strength of nodes is constructed based on the method. The characteristic parameters of complex networks are introduced. The concept of degree, average degree, clustering coefficient and analysis of criminal network are carried out. The results show that degree of the nodes are related to the future crime rate, which can be used in crime forecasting; Distribution of node's degree has no scaling property. Burglary can also occur within areas rarely be infringed. And the degree of the node is closely related to the crime rate future. Therefore, even if the crime rate is relatively low, the district should also pay attention to the changes of node's degree; Coefficient of crime aggregation has predictability to the future crime rate. Higher aggregation coefficient means the state of crime may change in the future. The spatial-temporal characteristic of important nodes in crime prediction was analyzed at last.
Association rules mining of event sequences aims to discover interesting patterns of different neighboring events and plays an important role in understanding their mutual relationship. However, for most existing methods, the distribution characters of events in the sequences are usually ignored and selecting proper thresholds is really a tough task, which brings about the problems of redundant results or interesting rules missing. Thus, new measuring indexes were defined and a context-based method for multiple event sequences mining was proposed. Results of both the simulated experiment and practical cases emphasized that the proposed method could effectively reduce the redundancy in the results in comparison with the classic MOWCATL method. Moreover, there was good consistency between the measuring indexes, which eases the selection of generated rules. Finally, the proposed method was applied to mine association rules between and PM2.5 concentration and several meteorological factors. Results indicated that the most associated meteorological factor with PM2.5 concentration was the humidity and an eligible environment for high PM2.5 concentration were high humidity, low temperature and weak winds.
Aiming at the problem that the single rule cellular automata image encryption is easy to be attacked by the plain text, this paper analyzes the limitation of the key space, and proposes a new encryption algorithm based on the high order reversible cellular automata. Through the analysis of the reversible cellular automaton characteristics, we construct the rules of higher order reversible cellular automata, combined with results of raster map after quadtree decomposition, which replacing the traditional method of all pixels of multiple cycle iterative encryption. Encrypt the raster map under the premise of not increasing cellular automata structure complexity. Experiments show that the proposed method is of high key space and high encryption efficiency. It can effectively resist differential attacks and clear text attacks, and it is also suitable for real time image encryption based on the integrity of map data.
Because of the area distortion of Gauss projection, it is necessary to calculate the map patch area on Earth ellipsoid with high area precision in some applications.In this paper, the area calculation method of ellipsoid patch is studied and improved, and the improved algorithm is analyzed empirically. The results show that the approximating definite integrals by the improved Rectangular Rule has higher computational efficiency than by the conventional rectangular rule. The improved rectangle method can replace the middle layer algorithm and the bottom algorithm in ellipsoid patch area calculation method and simplify the calculating process. With the help of decimal number type variables in C # language, a high accuracy value of the ellipsoid trapezoidal area is obtained using the improved algorithm. The ellipsoid area of large trapezoidal block can be obtained, and the ellipsoid area of arbitrary polygon can be easily solved also by the improved algorithm.
The process of the typical algorithm of hierarchical agglomerative clustering (HAC) reflects the multi-scale property of studying data, which is crucial in the research of Geography, Cartography and Remote Sensing. However, the typical algorithm is inefficient and costs too much memory space. In this paper, considering the transitivity of distances among points, the studying data are divided into equilateral orthogonal grids to avoid redundant computation; moreover, the feasibility of the proposed algorithm is proved in theory and extended to N-dimensional space. The proposed algorithm follows the same single-link clustering rule as the typical algorithm; and therefore, it generates the same multi-scale clustering series as the typical algorithm. Since there is no distance-matrix, the proposed algorithm costs much less memory space. Experimental results show that although without any manual intervention and distance-matrix, the proposed algorithm obviously improves the efficiency of HAC. However, the advantage decreases as the extending of the dimension number.
A linear feature Morphing method is proposed based on the fact that the spatial characteristic of linear element is represented by bent structure. First, for linear features at different scales, by construction and classification of Constrained Delaunay Triangulation, we can build the multiway trees to express the curves' bend hierarchical structure. Then, by the matching of multiway trees, we can get the bends corresponding relationship between two linear features at different scales. By the importance evaluation of matching bends, we can divide the linear element into different line segments. Last, different strategies are implemented for different type of bends. For the corresponding line segments, linear interpolation operation is adopted to exaggerate the small bend or shrink the big bend; for the no-matching segments, deletion operation is implemented. Experimental results show that the proposed Morphing method for linear features can meet the map generalization requirements, and keep the balance of curvature and quantity of bends, which finally achieving smooth and gradient continuous generalization of linear features.
Change detection is an essential step in spatial data updating. By analyzing existing matching methods, this paper proposes a method for single-and dual-carriageway roads matching considering dual-carriageway road characteristics to adapt the incremental updating of dual-carriageway roads, it is able to extract change information utilizing this method. For the completeness of dual-carriageway roads, the method selects polygons composed of dual-carriageway roads as matching objects, and designs a calculation model about overall single-and dual-carriageway roads matching indicator based on analyzing orientation, length and location relation of old single-carriageway road and polygon. Finally, change information is extracted with the matching relation confirmed by overall matching indicator. Through comparative experiments and analysis, it is proved that this the method can realize changing information extraction in incremental updating of urban dual-carriageway roads, and is really of practical significance.
Due to the drift of yaw, low accuracy and accumulative erroring in the procedure of using smartphone to realize Pedestrain dead reckoning algorithm, a map-aided KF-PF multi-filter algorithm is used to optimize PDR algorithm. Based on the traditional PDR algorithm, a Kalman Filter, fusing output of gyroscope and cartographic information primarily, is used to get the orientation, then using the map-matching particle filter to process the route results. The experimental results show that the flexibility of indoor positioning is improved, in the meanwhile, the stability and preciseness of the positioning results is enhanced and the algorithm can eliminate the error of the drift of yaw. Compared with the traditional particle filter, the map-matching particle filter can decrease the number of particles and computation burden effectively, which makes the possibility for realizing the real-time indoor localization.