2019 Vol. 44, No. 10
NDVI (normalized difference vegetation index)is one of the most important vegetation index, which can reflect vegetation growth and coverage, and it is of great significance to perform real-time monitoring. In this paper, we first take advantage of the amplitude of single-to-noise generated from GPS reflected signal to calculate normalized amplitude, then the characteristics of time series, frequency spectrogram and correlation between the normalized amplitude and NDVI extracted from MODIS products are analyzed. At last, NDVI inversion models with linear regression and BP neural network based on GPS reflection signal are presented. The results show that there are significant annual and seasonal characteristics within the normalized amplitude calculated by GPS reflection signal and NDVI, the correlation coefficient of linear regression is about 0.7, RMSE is 0.05-0.09. Moreover, BP inversion model within consideration of soil moisture is superior to the linear inversion model, the correlation coefficient is increased by about 5%, RMSE is 0.03-0.09. It indicates that NDVI change inversion using GPS reflection signals is feasible, which will provide an alternative approach to monitoring NDVI with high temporal resolution and low cost.
The variation of tropopause serves as an indicator of global scale meteorological change. Although a lot of research on tropopause has been done, it has rarely been discussed in GPS meteorology. In fact, the tropopause shows strong correlation with weighted mean temperature and precipitable water vapor (PWV). The correlations suggest that the tropopause has been the main influencing factor, except surface temperature. Yao has proposed a formula that links the weighted mean temperature and tropopause. The formula is rearranged and the approximated quadratic functional relationship between tropopause and weighted mean temperature is derived. The following works focus on discussing the abscissa of extremum and the coefficient sign of quadratic function, using radiosonde data. The sign from one of the coefficients of quadratic function is plotted in China region. Combing the map of positive-negative signs and vertex height of abscissa of extremum with the seasonal changes of tropopause, it's easy to predict the influence of tropopause on weighted mean temperature. After discussing the relationship between tropopause and weighted mean temperature, its influence on PWV is easy to describe. The curve of PWV change will be sharp or mild when tropopause enhances or reduces the influence of surface temperature on weighted mean temperature.
LiDAR is an effective remote sensing method for monitoring the distribution of atmospheric particulate matter (PM), which can overcome the shortcoming of scattered distribution and failure to achieve regional monitor of the conventional ground monitoring stations. In order to determine the horizontal distribution and transportation of PMs in urban areas, a vertical observation LiDAR and a horizontal scanning LiDAR are applied to achieve the goal. A more accurate retrieval method is developed by combining the slope algorithm and Fernald algorithm. The transport, distribution and concentration are analyzed using three-dimensional LiDAR data. The results of the horizontal aerosol extinction coefficient of the LiDAR are compared with the PM concentration from the ground state-controlled station to analyze their correlation. The results show that the three-dimensional atmospheric LiDAR can effectively reveal the distribution of PMs in large urban areas, and has the advantages of wide coverage and high detection efficiency.
Very long baseline interferometry plays a critical role on the orbit determination of the lunar and deep space probes, as well as the radio science research. Based on a first-order approximation model of the probe's topocentric range-rate, this paper proposes a new narrow band tracking method for the weak signal received by an antenna with relatively small aperture. As long as the Doppler dynamics has been estimated from the signal received by a large antenna with high G/T value, the weak signal, received by the small antenna during the same time duration, could be Doppler compensated. The raw data acquired by Chinese Jiamusi station and European Space Agency's New Norcia station have been processed, and the results demonstrate the effectiveness of this method. Digital phase locked loop with 5 Hz noise bandwidth could be used to track the differential one-way ranging(DOR) tone received by New Norcia, after Doppler compensation by estimation results from Jiamusi carrier tracking, and the root mean square of the tracking phase errors for DOR tone is around 3.4 degree. This method could be used in the radio metric navigation for the future lunar and deep space missions, as well as the inter-agency cross-supports.
In traditional method, the model of elevation axis fitting is not rigor, which influences the accuracy and reliability of rotation center. To overcome this drawback, two new methods are proposed to get rotation center respectively based on elevation axes intersecting and spherical fitting. The trajectory of target points in different azimuths are fitted to acquire elevation axes which contain azimuth information, and the intersection of elevation axes is taken as rotation center, which is called as the elevation axes intersection method. This method is more reliable than the traditional method. The target points are fitted as a sphere to obtain a series of spherical centers, and their average is taken as the rotation center, which is named as the spherical fitting method. This method avoids the process of circle fitting and axes intersecting, which is more reasonable than the others. An uplink antenna array consists of three φ3 m antennas is taken as experimental object. Total station and industrial photogrammetric system are combined to get the coordinates of points on antenna. By employing the two new proposed methods, the antenna rotation centers are precisely acquired in a sub-millimeter level.
The BeiDou global navigation system was officially launched in 2017 and will adopt a new BeiDou global ionospheric delay correction model(BDGIM). Using high-precision grid ionospheric data released by CODE and the Slant ionospheric delay derived from dual-frequency observations as a reference, the accuracy of the BDGIM model is analyzed and evaluated, and compared with the BeiDou Klobuchar and GPS Klobuchar models. The results show that in the region of China, the BDGIM model and the BeiDou Klobuchar model are comparable in accuracy and superior to the GPS Klobuchar model. On a global scale, the accuracy of the BDGIM model is better than that of the BeiDou Klobuchar and the GPS Klobuchar models. Different ionospheric models are implemented for single point positioning. The results show that the BDGIM model has a 13% improvement over the positioning accuracy of the BeiDou Klobuchar model, and a 7%-10% improvement over the GPS Klobuchar model.
The BeiDou navigation satellite system have three attitude control modes-yaw-steering(YS) mode, orbit-normal(ON) mode and continuous yaw-steering(CYS) mode. The orbit accuracy and best orbit determination projects for different satellites were a hot topic at present in different time and different attitude control mode, which are researched and whose the dynamical mechanisms is analyzed in this paper. The results are as follows. For BeiDou-2 satellites, multi-satellite precise orbit determination by fixing satellite/station clock offset and extended CODE model (ECOM) 5 SRP should be the best project, user equivalent range errors (UERE) root mean square (RMS) in ON mode is 2.08 m, global satellite laser ranging (SLR) data fitting RMS is ablout 1 m. For BeiDou-3 satellites, multi-satellite precise orbit determination and ECOM 5 SRP should be the best project, UERE RMS in CYS mode is 1.22 m, SLR data fitting RMS is 0.23 m, which is comparable to that in YS mode. For GEO satellites near spring or autumn equinox, multi-satellite precise orbit determination by fixing satellite/station clock offset and ECOM 9 SRP should be the best project, UERE RMS is 0.72 m.
In recent years, the positioning function of smart devices, such as mobile phone and tablets, gets more and more popular, which is a great convenience to people's daily life. Now, the best horizontal accuracy of positioning that most ordinary smart devices can achieve is about 5 meters. In 2016, Google Inc. launched Android Nougat (7.0) operating system, supporting the compatible smartphone and tablets to obtain global navigation satellite system (GNSS) raw measurements, which makes it possible to compute pseudorange and carrier phase measurements to achieve precise handset positioning. The target of this paper is to evaluate the precision of different positioning methods on the smart devices. A Huawei P9 smartphone was used as experimental devices and set statically in good observation condition to collect GNSS raw measurements. We compute the positioning precise, analyze the accuracy and compare the accuracy with that of the NovAtel DL-V3-L1 receiver which was set nearby at the same time. Experimental results show that the Huawei P9 smartphone can achieve decimeter-level positioning accuracy through precise point positioning or static relative surveying.
Due to the inability to obtain ellipsoidal height data in the past, gravity anomaly is chosen as the boundary condition in the traditional third geodetic boundary value problem. The development of GNSS technology brings opportunities for the development of the second boundary value problem. The relatively mature third boundary value theory provides undoubtedly a good reference for the second boundary value problem. Therefore, this paper deals with how to use the Molodensky theory for the third boundary value problem to calculate the quasi-geoid for the second boundary value problem. In this paper, the relationship between the Hotine operator and gradient operator is deduced. Then the method of solving the second boundary value problem based on the Molodensky theory is presented. Experiments show that the accuracy of the quasi-geoid by this method is equivalent to that by the traditional Molodensky method in the third boundary value problem. Thus it is feasible to solve the second boundary value problem based on the Molodensky theory.
An automated method for aggregating depth contour based on triangulation is proposed. Firstly, an automatic method to detect aggregating contours is proposed, which doesn't need to reco-gnize the submarine topography features. Secondly, triangles to construct the bridge areas are filtered, and the bridge areas are expanded and related "holes" are filled. Next, the bend areas to generalize are detected. Then generalized areas are derived by merging the bride areas and generalizing bend areas, and then the aggregated result contours are achieved by conducting symmetric difference calculation on the boundary of generalized areas and the aggregating contours. Test results show that the method proposed in this paper can detect the aggregating contours correctly, and the aggregated results not only can guarantee navigation safety but also can achieve a legible and smooth depth contour.
This paper discusses the ill-posed nonlinear least squares problem, and proposes an adaptive relaxation algorithm based on the regularization method for stabilizing the nonlinear parameter estimation. The improved algorithm achieves the adaptive selection on the regularization parameter and iterative step by using an incremental geometric regularization parameter and the minimal residual criterion. The numerical convergence experiments of the method are performed. The results show that the numerical precision of our proposed method is better than that of the linearized adjustment estimation, and the convergence property is more efficient than the iterative Tikhonov regularization method.
To determine a more precise reference point (RP) coordinate of Very Long Baseline Interfero-metry (VLBI) antenna in a local surveying, the effects of azimuthal track deformations in four-wheels and pedestal-based radio telescopes are analyzed, and an azimuthal steel-track residual deformation correction model is firstly proposed and derived to correct the coordinates of target points (TPs), which are used to determine the RP position. Using the real measurements of TPs and the leveling data of the steel track, systematic effects in residual of TP coordinates are explained and theoretical TP coordinates are corrected so that the corresponding RP precision is significantly improved. The results show that applying azimuthal steel-track residual deformation correction and estimation models can improve the standard error of post-fit TP residuals by 50% and 65%, respectively. The RP precision can improve about 30%. Moreover, the proposed steel-track residual deformation estimation model can be well applied to monitor the azimuth steel-track deformation of four-wheels and pedestal-based antenna.
The reasonable selection of GNSS/leveling points is very important to GNSS height fitting, step-by-step elimination method is a good method for the optimization selection of joint-observation points. The traditional elimination method selects the GNSS/leveling points based on the minimum fitting error of height anomaly, which will easily lead to uneven joint-observation points. According to this view, this paper proposes to optimize the joint-observation point based on the area size of Thiessen polygons generated by the GNSS/leveling points, and on this basis to improve the traditional method, namely taking into account both the size of the height anomaly fitting error and the area size of polygons generated by Thiessen method (referred to as the synthesis method). The experimental results show that the synthesis method can improve the distribution of joint-observation points, and get the fitting results of height anomaly with high stability and accuracy.
The theoretical calculation model of magnetic anomaly of submarine pipeline is derived by Poisson formula method of gravity and magnetic field. For two known submarine gas pipelines, the forward simulation calculation and sea test verification are carried out. The forward results show that positive magnetic anomalies with a span of about 20.0 m and up to 329.8 nT and 362.4 nT respectively are produced on an observation plane of about 8.0 m from the submarine pipeline. The measured distribution spans of magnetic anomalies of submarine pipeline are consistent with the results of forward modeling, but the details of their distributions are different. Both positive and negative magnetic anomalies exist. By analyzing the distribution characteristics of the extreme points of magnetic anomalies, it is speculated that the thermal residual magnetism at the junctions of the pipeline may lead to the detailed characteristics of the measured magnetic anomalies.
The generation of initial seamline network is one key step in image mosaicking. The quality of generation has a massive impact on subsequent local optimization. In this paper, a novel method for generating initial seamline network is designed. The overlapping information of effective regions of images is encapsulated in the form of a bit array and assigns to the geometric elements such as points, lines and faces. The design contributes to excavate the spatial relation between overlapping areas and to facilitate the calculation of joint priorities of overlapping areas. The main consideration of seam line generation is no longer subject to the specific shape of overlapping areas, but only the degree of overlap. Therefore, it greatly improves the robustness of seam line generation. In addition, the simple bitwise operation reduces the complexity of algorithm and thus improves the computational efficiency. This method is compared with the area voronoi diagram with overlap method. The experiments show that the proposed method has advantages in robustness and efficiency of seam line generation.
The combination of camera and 2D laser-rangefinder has been widely used in the fields of surveying, pilotless and robot. Extrinsic calibration between the two sensors is a prerequisite for fusing the texture information from images and depth information from the laser. To tackle this problem, a method of calibrating the extrinsic parameters between a camera and a 2D laser-rangefinder is proposed. This method establishes three geometric constraints between the laser scanning line and the V-shaped chessboard plane, including point to plane constraint, line to plane constrain and point to line constraint. The extrinsic parameters can be solved and optimized by redundant geometric constrains which help mitigate the impacts of noises in the laser points and image data. Experiments show that the proposed algorithm achieves higher accuracy and robustness than previous methods. And a quantitative evaluation criteria of the extrinsic parameters is proposed by calculating the distance between the intersecting point of the laser lines and the V-shaped chessboard line.
To handle the issue of low accuracy performance of crowd abnormal behavior detection in video surveillance systems, an abnormal crowd behavior detection and location approach based on spatial-temporal cube is proposed in this paper. The optical flow method is first used to calculate the optical flow field of feature points which are obtained by the equidistant sampling method. Then, the velocity, orientation and orientation entropy of the feature points are obtained. And statistical histograms of the three parameters are mapped into the corresponding cubic space to extract the spatial-temporal cube feature for describing the spatial-temporal features. A blocking method is used to divide the image into several sub regions, and the spatial-temporal cubes of each sub region are calculated. Finally, a cascade classifier based on nearest neighbor classification and support vector machine is designed to detect and locate crowd abnormal behaviors. Experimental results show that the proposed method can effectively detect and locate abnormal crowd behaviors in videos.
In order to solve the problem of multi-domain adjacency in the modeling of complex orebodies, we implemented an adaptive meshing method for complex orebodies based on the Delaunay refinement method, which could generate multi-domain surface models and volume models for the finite element simulation in the later stage. The method relies on the constraint Delaunay triangulation to approximate the domain and the surface, and the Delaunay refinement to ensure the approximate accuracy of the domain and the grid quality of the cell. On the basis of constraint Delaunay triangulation, the multi-domain polyhedron method of prediction was constructed by separating the intermediate domain or fitting the intermediate domain by distance field. The method of adaptive meshing of multi-domain was implemented. Experimental tests of Beiminghe Iron Mine show that the method can generate a seamless and continuous multi-domain model with high-quality tetrahedral and triangular meshes by mesh optimization. And it can avoid many degraded and singular triangles by contour-matching. It is significant to apply this method in the reverse engineering, finite element analysis, scientific computing visualization and other fields.
Community detection is an important topic in spatial network researches. It can discover interesting spatial patterns, which help understand the spatial structures hidden in the networks. However, community detection in spatial networks is more difficult than traditional networks. Since it has to consider some other spatial correlations, e.g., spatial contiguity, geographic distance. This paper proposes a new spatial community detection algorithm-geographial weighted central node distance based Louvain method (GND-Louvain). It uses a geographical and network dual-constrain to measure the distance and to calculate the distances decay effects between spatial nodes and meta-communities. It also extends the famous fast unfolding community detection algorithm-Louvain, by using the distance-modularity. In addition, a merge order is defined during the GND-Louvain's optimization process to get high quality and stable community results. Comparative experiments are designed on five different community detection methods based on five different spatial constraints, which are ① no distance constraint, ② spatial contiguity edge constraint, ③ geometry median center distance, ④ degree center distance, and ⑤ degree weighted geometry median center distance. The experimental data comes from the Chinese railway lines. And results prove that the newly defined GND-Louvain algorithm can produce more accurate spatial communities than others. And the merge orders also ensure the high quality and stable results.
In order to meet the need of highly efficient processing of metro tunnel point cloud, this paper proposes a hybrid index model which combines the R-tree and grids.In global area, the metro tunnel point cloud is divided into grids which are organized by using R-tree method according to its spatial distribution characteristics. And a hybrid spatial index which combines the quadtree and octree is used to manage the point cloud in one grid. In order to improve the visualization effect of metro tunnel point cloud, we propose a method for constructing levels of cletail structure based on the area of the grids. In addition, single file storage method is used to store point cloud data. Experiments results show that the proposed method has advantages in the management and visualization of massive point cloud effect over traditional methods.
After the disaster, the increase of suspended matter in water and the high water content crops will lead to the low accuracy of conventional methods of water body information extraction. In response to this problem, a new method of extracting flood submerging area for farmland based on tasseled cap transformation is presented in this paper. First, the remote sensing images before and after disaster are pre-processed by radiometric calibration and atmospheric correction. Then, the tasseled cap transformation is performed based on the coefficients corresponding to the sensor to obtain the greenness component and the wetness component. Third, the wetness component is divided by the OTSU method, and combined with the greenness component to obtain the warterbody information. Finally, the spatial overlay analysis of the waterbody information and farmland vector data is carried out to extract the flood submerging area for farmland. Taking Yueyang City of Hunan Province and its vicinity as research area, the accuracy of the proposed method is evaluated both qualitatively and quantitatively to verify the validity and applicability of the method. The results shows that the boundary of flood submerging area for farmland is clear, the range is more accurate, and the producer's accuracy and the user's accuracy are 0.97 and 0.90, respectively. This paper provides references for agricultural risk evaluation and dynamic monitoring of flood disaster.
Current lake selection methods mostly use the form of selecting as a whole, and it is difficult for them to take into account the attribute characteristics, distribution characteristics and topological characteristics of the lake. By analyzing and imitating the cognitive behavior and process of artificial lakes selection, this paper proposes a lake selection method based on dynamic multi-scale clustering. Firstly, we set the area threshold to select the lakes with large area, then select the "isolated" lake through the buffer, then utilize the dynamic multi-scale clustering to the lake group to divide into areas with different density, decide the selection numbers by square root law and adopt the different selection strategy for the different areas, among whose lakes are selected according to the comprehensive evaluation of importance calculated by iterative principal component analysis in the lake group class with a large number of features until the number of lakes reaches the selection number. Experiments show that our proposed method maintains the morphological structure and density contrast of the lake group effectively, under the premise of considering the importance.
In this paper, we propose an individual income level inference method based on travel behavior of urban residents. Firstly, we define 13 residents' travel behavior features from 3 different aspects. These features are the mobility indicators based on trajectories, home-based travel characteristics, and activity chain patterns. Secondly, 80% of the daily travel diary data of residents in Guangzhou in 2013 is taken as the training sample and the remaining as the test data. Thirdly, the random forest method is trained and used to infer individual's income level. The experiment results show that our proposed method can obtain an overall accuracy of 80%. Among 13 travel behavior features, the home-based features (i.e. the mode of the travel distance from home during working time (9:00 AM to 18:00 PM)), the activity chain pattern, and travel distance and scope related features (i.e., the maximum distance between two successive activity points, and radius of gyration) have high feature importance. However, the features representing the spatial heterogeneity of the activity points have relatively low importance, such as the spatial diversity.