Table of Content

20 November 2019, Volume 48 Issue 11

Previous Issue    Next Issue

Review

Recent progress in taxi trajectory data mining

WU Huayi, HUANG Rui, YOU Lan, XIANG Longgang

2019, 48(11):  1341-1356.  doi:10.11947/j.AGCS.2019.20190210

Asbtract ( )   HTML   PDF (1439KB) ( )  

References | Related Articles | Metrics

The development of big data technology, internet of thing and precise positioning has promoted the progress of city perception. The increasing taxi trajectory data not only records the pathway of taxis, but also implies the real-time traffic status, the information of urban dwellers' travel rule, urban structure and potential social problems. It is of great significance to mine and analyze the taxi trajectory data for smart transportation, urban planning etc. This paper reviews the field of taxi trajectory data analysis and applications in the past ten years. From the perspective of research methodology, four categories are identified:spatial statistical, time series analysis, graph and network analysis, and machine learning. Each category is reviewed with its current research situation, advantages disadvantages. Later on, applications, hot topics and future trends of taxi trajectory analysis are summarized to four areas including traffic management, resources and environmental protection, city planning, and human mobility. Finally, the current challenges and the future research directions in the field of taxi trajectory data mining are proposed.

Photogrammetry and Remote Sensing

A progressive simplification method of navigation road map based on mesh model

GUO Qingsheng, LIU Yang, LI Meng, CHENG Xiaoxi, HE Jie, WANG Huihui, WEI Zhiwei

2019, 48(11):  1357-1368.  doi:10.11947/j.AGCS.2019.20180552

Asbtract ( )   HTML   PDF (9599KB) ( )  

References | Related Articles | Metrics

In this paper the graphic generalization of roads is transformed into linear graphic simplification based on mesh model when the navigation data precision is decreased. A progressive collaborative graphic generalization method for roads in the navigation data is proposed based on the mesh model, which can ensure that the single road and road networks after generalization are similar to the original high-precision data in spatial directional relationship. In the process of data handling, the navigational requirements are applied to constrain the progressive graphic simplification of the road. Besides, characteristics of road shape and the spatial direction relationship at the intersection of roads are maintained while visualizations of the roads in navigation is taken into consideration. The algorithm proposed in this paper has been verified with navigation data in actual production, which proves the method is both effective and practical.

Adaptive hierarchical spatio-temporal index construction method for vector data under peer-to-peer networks

WU Zheng, WU Pengda, LI Chengming

2019, 48(11):  1369-1379.  doi:10.11947/j.AGCS.2019.20190143

Asbtract ( )   HTML   PDF (1740KB) ( )  

References | Related Articles | Metrics

Spatio-temporal index is one of the key technologies for storage and management of spatio-temporal data. Index methods based on spatial filling curve (SFC) have drawn wide attention in recent year. However, the existing methods for the vector data mostly focus on the implementation of spatial index, which is difficult to take into account both the efficiency of time query and spatial query. For non-point elements (line elements and polygon elements), it is always difficult to determine the optimal index level. Therefore, this paper proposes an adaptive hierarchical spatio-temporal index construction method for vector data under peer-to-peer networks. Firstly, a joint coding of spatio-temporal information based on the combination strategy of partition key and sort key is proposed. Then, the spatio-temporal expression structure of point elements and non-point elements are designed. Finally, an adaptive multi-level tree is proposed to realize the spatio-temporal index (multi-level sphere 3, MLS3) based on the spatio-temporal characteristics of geographical entities. Experiments are carried out using actual data of trajectory (point elements), highway (line elements) and building (surface elements) data. By comparing with the XZ3 indexing algorithm proposed by GeoMesa, it is proved that the indexing method in this paper can effectively solve the problems of hierarchical division and spatio-temporal expression of non-point elements, and can effectively avoid storage hotspots while achieving efficient spatio-temporal retrieval.

Spatial-temporal clustering by fast search and find of density peaks

WANG Peixiao, ZHANG Hengcai, WANG Haibo, WU Sheng

2019, 48(11):  1380-1390.  doi:10.11947/j.AGCS.2019.20180538

Asbtract ( )   HTML   PDF (4838KB) ( )  

References | Related Articles | Metrics

Spatial-temporal clustering algorithm is the basic research topic of geographic spatial-temporal big data mining. In view of the problem that traditional CFSFDP clustering algorithm cannot be applied in spatio-temporal data mining, this paper proposes a spatio-temporal constraint algorithm called ST-CFSFDP(spatial-temporal clustering by fast search and find of density peaks). ST-CFSFDP adds time constraint on the basis of CFSFDP algorithm, and modifies the calculation strategy of sample attribute value, which not only solves the problem of multi-density peak of single cluster set in the original algorithm, but also can distinguish and identify clusters at the same location and at different times. In this paper, the simulated spatiotemporal data and real indoor location trajectory data were used for the experiment, the results show that the ST-CFSFDP algorithm has a recognition rate of 82.4% at a time threshold of 90 s and a distance threshold of 5 m,which is better than the classic ST-DBCSAN, ST-OPTICS and ST-AGNES algorithm increased by 5.2%, 4.2%, and 7.6%, respectively.

Decision tree model for extracting road intersection feature from vehicle trajectory data

WAN Zijian, LI Lianying, YANG Min, ZHOU Xiaodong

2019, 48(11):  1391-1403.  doi:10.11947/j.AGCS.2019.20190011

Asbtract ( )   HTML   PDF (18292KB) ( )  

References | Related Articles | Metrics

Crowd sourcing vehicle trajectory data imply the latest road network information. Therefore, studies on the extraction of road features from trajectory data provide the opportunity for efficient construction and renewal of road datasets. Since a road network is composed of road intersections and road segments, the extraction of road intersections plays an important role in road network generation. Due to the lack of accurate mechanisms for intersection extraction, problems such as omission and distortion of road intersections occur frequently. A method is proposed to identify and extract road intersections from vehicle trajectory data. Firstly, it is analyzed that the differences in shape and kinetic features between trajectories from intersection areas and non-intersection areas. Secondly, the decision tree method is employed to construct a trajectory segment classification model, which enables the extraction of lane-changing segments in intersection areas with the support of trajectory division model using a sliding window strategy. Thirdly, a method that based on Hausdorff distance is designed to cluster trajectory segments in intersection areas, and intersection structures are obtained by extracting the central lines of the trajectory segment clusters. Experiments on real-life trajectory datasets were implemented and results showed the effectiveness of the proposed method.

Vector geometry based method for the extraction of slope of aspect by using DEMs

HU Guanghui, XIONG Liyang, TANG Guoan

2019, 48(11):  1404-1414.  doi:10.11947/j.AGCS.2019.20180447

Asbtract ( )   HTML   PDF (19950KB) ( )  

References | Related Articles | Metrics

Aspect matrix, as the source data for calculating the slope of aspect (SOA), has the characteristic of directionality. Thus, misunderstanding and error would be produced if SOA was calculated by scalar method because its source data has the directional property. On a basis of the mathematical Gaussian surface and 5 m resolution DEM data of different loess landform sample areas, the mathematical vector method is proposed to calculate SOA with a full consideration of directional property of the aspect matrix. In this method, the original aspect matrix has been transformed into the polar coordinate system, and the vector geometric representation of aspect matrix could be achieved. Then the SOA is calculated on a basis of this vector transformed aspect data. In the end, a comparative analysis is conducted among the proposed method and traditional scalar methods. The results show that SOA calculated by the proposed method could effectively avoid the extreme error in the north direction and the inaccuracy when the aspect difference exceeds 180°. Meanwhile, a more reasonable SOA result could be achieved in the other main areas, and a more stable result can be obtained by using our method in different resolution DEMs. The proposed vector geometry method could help to provide a reference for accurate digital terrain analysis, and it is also an important practice to solve problem in DTA by mathematical vector geometry.

Fusion and visualization method of dynamic targets in surveillance video with geospatial information

ZHANG Xu, HAO Xiangyang, LI Jiansheng, LI Pengyue

2019, 48(11):  1415-1423.  doi:10.11947/j.AGCS.2019.20180572

Asbtract ( )   HTML   PDF (10512KB) ( )  

References | Related Articles | Metrics

The dynamic foreground target analysis of surveillance video is an important basis for security construction in safe cities and smart campuses. The dynamic attributes can be assigned to static geographic data by integrating surveillance video with geospatial data. For the integration of traditional surveillance video and geographic information data, only video data is projected into geospatial, which causes storage issue and difficult in understanding video content. The model that integrates geographic information and foreground dynamic targets, and integrates surveillance video with geographic information is proposed in this paper. And the dynamic foreground target and the tracking trajectory in the image are projected into the geospatial by the derived mapping model. The multi-layer fusion mode is utilized to distribute the video dynamic summary content in geospatial space.

Cartography and Geoinformation

Real-time BeiDou landslide monitoring technology of “light terminal plus industry cloud”

BAI Zhengwei, ZHANG Qin, HUANG Guanwen, JING Ce, WANG Jiaxing

2019, 48(11):  1424-1429.  doi:10.11947/j.AGCS.2019.20190167

Asbtract ( )   HTML   PDF (3102KB) ( )  

References | Related Articles | Metrics

Implementing the high-precision, real-time and three-dimensional deformation monitoring for the landslide area, which is the prerequisite for the accurate warning of landslide disasters. GNSS technology is currently the only way to directly obtain the three-dimensional vector deformation of landslide disaster surface, but GNSS has two problems of high-cost and low-reliability in large-scale landslide monitoring. The ideas of Internet of Things and the concept of "cloud platform plus monitoring terminal" are proposed in this article. Thus, we develop a real-time BeiDou/GNSS monitoring equipment with thousands RMB cost. The millimeter monitoring and warning cloud platform also are developed independently. This equipment successfully applied to real-time monitoring and early warning of Heifangtai landslide in Gansu Province. Cooperated with early warning system of the Chengdu University of Technology, we issued an accurate warning signal 40 minutes in advance to avoid casualties and property losses. The remote video surveillance installed on the landslide body recorded the whole process of landslide disaster for the first time.

The identification method of gross error detection failpoint in L₁-norm estimation

YAN Guangfeng, CEN Minyi

2019, 48(11):  1430-1438.  doi:10.11947/j.AGCS.2019.20180395

Asbtract ( )   HTML   PDF (1016KB) ( )  

References | Related Articles | Metrics

The gross errors reflected in the corresponding closure errors of conditional equations obtained by L₁-norm estimation are more significant than those in LS residuals, and thereby, the former estimation is helpful for the gross error detection and location. Unfortunately, there is a kind of observations which have the ability to detect and locate gross errors, no matter how large the gross error is, but it cannot be accurately located in L₁-norm estimation. For convenience of discussion, such observation is called as Robustness Failpoint in L₁-norm estimation, RFP-L₁ for short. Obviously, only to meet such a premise, the results of gross error detection based on L₁-norm estimation can be accurate and reliable, i.e., it is ensured that there is no RFP-L₁ in the surveying system, or whether the RFP-L₁ contains gross error can be judged accurately. And in this process, the accurate identification of RFP-L₁ is the basis for solving the problem. From conditional equation, the calculation formula of influence coefficient which reflect the extent of the influence of gross error on corresponding closure errors of conditional equations is derived, and the distinguish relation of judging whether an observation is RFP-L₁ or not according to the value of least influence coefficient is formulated. Furthermore, the numerical characteristic of design matrix that might contain RFP-L₁ is explored. And at the end, a method of identifying RFP-L₁ is put forward. The simulation results show that the least influence coefficient reflects the influence of gross errors on the objective function of L₁-norm estimation, and the least influence coefficient of normal observation and RFP-L₁ have a significant regularity equal to one and less than one respectively. In addition, it can be concluded that there is no RFP-L₁ in the surveying system with design matrix containing only±1 and 0.

Geodesy and Navigation

Coregistration scheme and error analysis of multi-mode SAR image based on geometric coregistration

WU Wenhao, ZHANG Lei, LI Tao, LONG Sichun, DUAN Meng, ZHOU Zhiwei, ZHU Chuanguang, JIANG Tingchen

2019, 48(11):  1439-1451.  doi:10.11947/j.AGCS.2019.20180440

Asbtract ( )   HTML   PDF (18556KB) ( )  

References | Related Articles | Metrics

With the development of radar imaging technology and the diversification of imaging modes, the accuracy of coregistration in the direction of azimuth required by SAR interferometry is different. At present, due to the strict theoretical model and the high accuracy, the coregistration scheme with the image information based on geometric coregistration under the condition of precise orbits become the first choice for interference processing. In this paper, TerraSAR-X images have been taken as an example to demonstrate the accuracy of interferometric coregistration and satellite orbit required by different imaging modes. Furthermore, theoretical analysis and experimental validation have also been applied. It is proved that under the condition of precise orbits, SAR interferometric processing of strip-mode image of TerraSAR-X and other satellites can be achieved by geometric registration. While, spotlight imaging mode requires image coherence or spectral diversity to further optimization of the geometric coregistration results. Since the ESD (Enhanced Spectral Diversity) offset estimation method offers the highest accuracy. It has been used to estimate the geometric coregistration errors of Sentinel-1 satellite images with different orbit and DEM conditions. The results suggest that geometric coregistration error is mainly caused by the tangential error of the satellite orbit, and the difference of geometric corregistration of three kinds of DEM is much less than 0.001 pixels, which can meet the needs of Sentinel-1 image interference registration.

Urban change detection by aerial remote sensing using combining features of pixel-depth-object

ZHAO Shengyin, AN Ru, ZHU Meiru

2019, 48(11):  1452-1463.  doi:10.11947/j.AGCS.2019.20180554

Asbtract ( )   HTML   PDF (21128KB) ( )  

References | Related Articles | Metrics

The selection and optimization of features play an important role in the recognition of remote sensing image. Usually, object-based methods cannot make full use of spectral information whereas pixel-based cannot take the advantage of spatial geometry information of remote sensing images. Therefore, a novel urban change detection method is proposed to combine pixel-scale and object-scale features from aerial remote sensing images. First, a feature space was established with spectrum, derivative index, texture, geometry, surface height and convolution neural network layers features. Second, a large number of important features were selected by using LightGBM (Light Gradient Boosting Machine) algorithm. Third, the selected features were used in the random forest classifier to produce the land cover maps from the aerial remote sensing images of Yixing City in 2012 and 2015. Finally, change matrix was used to detect urban change. The results show that the combination of pixel, depth, object features and LightGBM feature selection algorithm has the best recognition effect. Meanwhile, the average accuracy and Kappa coefficient of the proposed method are 88.50% and 0.86, which is 10.50%, 15.00% and 4.00% higher than that based on pixel, deep or object features recognition alone. And the accuracy of urban change detection is 87.50%. Hence, the proposed method is an effective method for urban change detection using aerial remote sensing images.

Linear feature detection for hyperspectral subpixel mapping

LIU Zhaoxin, ZHAO Liaoying, LI Xiaorun, CHEN Shuhan

2019, 48(11):  1464-1474.  doi:10.11947/j.AGCS.2019.20180221

Asbtract ( )   HTML   PDF (9267KB) ( )  

References | Related Articles | Metrics

The ignorance of the spatial structure of subpixel is one of the factors that influence the accuracy of hyperspectral subpixel mapping. In order to effectively solve this problem, a subpixel mapping algorithm is proposed that based on linear feature detection in mixed pixels. Firstly, the mixed pixels of typical linear feature classes are determined by spectral unmixing. Then, based on the maximum linear index method of the complete straight-line set, the linear features of remaining mixed pixels are determined. The template matching method is used in conjunction with the pixel attraction to determine the classes of linear subpixels. Finally, the subpixel categories of the remaining mixed pixels are iteratively determined based on the linear optimization method. The experimental results of real data and simulation data show that the proposed method can effectively improve the precision of subpixel mapping.