Table of Content

25 February 2024, Volume 0 Issue 2

Previous Issue    Next Issue

Study of soil salinity remote sensing inversion method integrating crop type

ZHANG Shengnan, LU Miao, WEN Caiyun, SONG Yingqiang, KANG Lu, SHEN Junhui, YANG Minzhi

2024, 0(2):  1-7.  doi:10.13474/j.cnki.11-2246.2024.0201

Asbtract ( )   HTML ( )   PDF (2954KB) ( )  

References | Related Articles | Metrics

In coastal plain regions, soil salinity serves as one of the abiotic stressors limiting crop growth. The content of soil salinity is a critical determinant for crop cultivation. The variety of crops grown can indirectly indicate the extent of soil salinization. Therefore, this paper proposes an integrated approach for the inversion of soil salinity, incorporating crop type information. Based on Sentinel-2 MSI imagery in a typical coastal saline soil area in the Yellow River Delta. Firstly, crop type information was extracted using random forest classification and coded based on the OneHot method. Then, by integrating crop type information, environmental covariate data, and ground-measured salinity data, the adaptive boosting decision tree (AB-DT) model is applied for soil salinity estimation. Finally, the accuracy of salinity estimation is compared with other machine learning methods, including support vector machines, random forests, K-nearest neighbors, and decision trees. The results indicated that ①Incorporating crop type information enhances the accuracy of soil salinity estimation models. Among all models, the AB-DT model with fused crop type variables achieves the highest modeling set R² of 0.86 and validation set R²of 0.61.②The inclusion of crop type information enable to correct misclassifications of salinity levels and yield sharper boundaries in soil salinity estimation results. In conclusion, the incorporation of crop type information improves the accuracy of soil salinity estimation, providing a more reliable basis for agricultural management and decision-making.

Temporal and spatial correlation analysis of land cover change and net ecosystem carbon exchange in globally representative river basins

CHEN Lei, MA Ying, ZHA Fengli, LIN Yanzhen, SHI Guangbin

2024, 0(2):  8-13,31.  doi:10.13474/j.cnki.11-2246.2024.0202

Asbtract ( )   HTML ( )   PDF (6081KB) ( )  

References | Related Articles | Metrics

With socio-economic development, the management objectives of river basins shift from early flood control, water supply, and navigation to comprehensive management focusing on resource utilization and ecological protection, playing a significant role in the global ecosystem carbon balance. As an important indicator for quantifying ecosystem carbon sequestration capacity, net ecosystem carbon exchange (NEE) exhibits significant differences in carbon source/sink capacity among different surface cover types. Exploring the temporal and spatial correlation between globally representative river basin surface cover types and NEE is of great significance for basin ecological management. Based on the global 30 m surface cover dataset (GlobeLand30) and global land net ecosystem carbon exchange data, this study analyzes the temporal and spatial characteristics of surface cover and NEE changes in 8 typical river basins worldwide from 2000 to 2020, and investigates the temporal and spatial correlation between surface cover type changes and NEE. The results show that: ① From 2000 to 2020, the area of cropland, bare, artificial surfaces, wetlands, water, and tundra increased significantly, while grassland increased first and then decreases, it showed an overall increasing trend. ② From 2000 to 2020, the NEE value of the basin decreased first and then increased with the overall decreasing trend and enhanced carbon sink capacity. The basin exhibited a carbon sink, but the mean NEE showed an overall increasing trend, indicating a weakening carbon sink capacity. ③ During 2000—2020, there was a significant negative correlation between grassland and NEE, where the relationship between cropland, bare, artificial surface, wetland and water with NEE was positive in 2010—2020. The change of land cover type had a significant impact on NEE. This research provides theoretical references for basin carbon neutrality regulation and spatial optimization control, and promotes coordinated emission reduction and high-quality development of river basins.

Knowledge modeling of spatio-temporal changes of land cover in the Amazon Basin

ZHU Xiuli, ZHU Xinzhou, LIU Wanzeng, CHENG Dayu, ZHANG Xiaoying, ZHANG Ye

2024, 0(2):  14-18.  doi:10.13474/j.cnki.11-2246.2024.0203

Asbtract ( )   HTML ( )   PDF (1437KB) ( )  

References | Related Articles | Metrics

The spatio-temporal changes of land cover in the Amazon Basin affect the stability of global climate and ecosystems. Currently, research on land cover in the Amazon Basin is extensive, with massive data and information, but lacks a systematic spatio-temporal knowledge system. In order to better understand the knowledge of spatio-temporal changes of land cover in the Amazon Basin, this article introduces ontology modeling theory and proposes a knowledge modeling method for spatio-temporal changes of land cover in the Amazon Basin. The design and construction of the land cover pattern layer in the Amazon Basin are completed using the Protégé tool.And the knowledge modeling is completed by extracting spatio-temporal change data, spatio-temporal change information, and landscape pattern index of the Amazon Basin, providing support for analysis and decision-making of the Amazon Basin.

Extracting urban impervious area from multi-source image fusion data assisted by global land cover data

HUO Jiating, ZHAO Zhan, ZHU Xiuli

2024, 0(2):  19-25.  doi:10.13474/j.cnki.11-2246.2024.0204

Asbtract ( )   PDF (5848KB) ( )  

References | Related Articles | Metrics

In this paper, an automatic extraction method of urban impervious surface area from multi-source image fusion data is proposed using GlobeLand30 data asauxiliary data. Firstly, an image fusion method based on band mapping and wavelet transform is proposed to fuse Sentinel-2 and GF-2 images to obtain fusion images with high spatial resolution and spectral resolution. The fusion image has rich spectral and spatial characteristics, which is conducive to improving the ability of distinguishing impervious and non-impervious surface in complex urban areas. Then, the category information of GlobeLand30 data is used to automatically to obtain the initial samples, a variety of ground index such as vegetation index, water index, and built-up area index are constructed to refine the initial samples. Finally, the optimized training samples are used to train the classifier with spectral and ground index features to achieve automatic and accurate extraction of urban impervious surface. In this paper, images of GF-2 and Sentinel-2 in Jinan city in 2019 are used as experimental data. With the help of GlobeLand30 global land cover data with different phases, resolutions and images, the overall accuracy of impervious surface extraction is better than 92%, which verifies the effectiveness of the proposed method.

Design and representation method of interactive knowledge atlas for global land cover spatiotemporal change

DAI Ruyu, TI Peng, MEI Yuting, WAN Fangyi, LI Zhilin, CHEN Jun, ZHU Xiuli, LIU Wanzeng

2024, 0(2):  26-31.  doi:10.13474/j.cnki.11-2246.2024.0205

Asbtract ( )   PDF (4726KB) ( )  

References | Related Articles | Metrics

Atlas is an efficient tool for presenting land cover knowledge. However, existing atlases only present knowledge in given regions so that it is difficult to meet users' personalized and research requirements for different interest regions so as to decrease the practicality of atlases. In adclition, a sheet in traditional atlases is generally divided into different regions containing different information so that the relations of information containing in knowledge are not easy to understand for users. In order to meet the personalized and knowledge-based service needs of map applications, this study designed and produced an interactive knowledge atlas based on GlobeLand30 data. Besides the atlas sheets to represent the spatial and temporal land cover changing from the global, continental, and national scales, the atlas can also support users to select interesting regions and generate related changing knowledge, and provided different knowledge visualization templates and the function of interaction between graphics and text, which can help users to understand the knowledge content better and support customized service.

Wetland digital twin technology and its application

CAO Xuyue, SUN Yonghua, WANG Yanzhao, WANG Yihan, CHENG Xinglu, ZHANG Wangkuan

2024, 0(2):  32-38.  doi:10.13474/j.cnki.11-2246.2024.0206

Asbtract ( )   HTML ( )   PDF (1314KB) ( )  

References | Related Articles | Metrics

Wetlands provide enormous ecological services function for nature, which is an essential component for maintaining regional and national ecological security and promoting sustainable development. Digital twin technology provides a new method and tool for wetland research, which can achieve comprehensive digital virtualization of wetlands, real-time visualization of the entire process and collaborative intelligence of management decisions. It not only helps to understand the operational mechanism of wetland ecosystems, but also promotes the intelligent process of wetland research. In this paper, the characteristics of wetlands and digital twin technology are combined to study and analyze the construction of wetland digital twin systems from four aspects: wetland information monitoring layer, wetland information processing layer, wetland model analysis layer, and wetland application interaction layer, so as to achieve a highly integrated wetland digital twin system with informatization, digitalization and intelligence. Then, the key technologies of establishing a digital twin system for wetlands are introduced, including real-time monitoring technology, high performance data processing technology, complex wetland models, and visual interaction technology. In addition, the application of digital twin technology in wetland research was explored from five aspects: dynamic monitoring of wetland resources, wetland classification, wetland ecosystem assessment, wetland tourism, wetland planning and management. What's more, the current existing problems in the construction of wetland digital twin were discussed, including the fusion of multi-source data, complex wetland modeling and optimization, and visualization of wetland multiple scenes. It is expected that relevant work can provide reference for the further development and implementation of digital twin in wetland research.

Review of the current research status of subway deformation monitoring

HEI Junmiao, WANG Li, LI Ang

2024, 0(2):  39-44,79.  doi:10.13474/j.cnki.11-2246.2024.0207

Asbtract ( )   PDF (1655KB) ( )  

References | Related Articles | Metrics

This article includes the deformation monitoring techniques during subway construction and operation respectively, and then sorts out the current research status of subway deformation monitoring data processing methods. It summarizes the existing problems in the current research of subway deformation monitoring, and prospects the development direction of subway deformation monitoring research, providing some ideas for the systematization and automation of subway deformation monitoring research.

Cover information extraction and precision analysis in Karst area based on feature optimization

LIAO Chaoming, YUN Ziheng, LUO Heng, WEI Yuanyuan, LING Ziyan, PAN Guiying

2024, 0(2):  45-50.  doi:10.13474/j.cnki.11-2246.2024.0208

Asbtract ( )   PDF (2114KB) ( )  

References | Related Articles | Metrics

Karst areas have complex and irregular geomorphological features, which makes the accuracy of land use classification low. In this paper, Shanglin county in Nanning city is taken as the study area, and 33 feature variables are extracted and seven feature combination schemes are designed by combining multi-source data to explore the role of adding topography, texture, red-edge index and radar features on the extraction of land classes in karst areas. Combining the random forest OOB data error and recursive feature elimination method for feature optimisation, meanwhile introducing the third national land survey data to compare with the optimised classification results in order to evaluate its accuracy and reliability. The results of the study indicate that: ①Among the seven classification schemes, the traditional spectral features plus index features have the lowest classification accuracy, and the addition of topographic, texture, red-edge index and radar features can improve the classification accuracy, among which the texture features bring the most significant effect. ②The number of feature dimensions is reduced from 33 to 23 through feature optimisation, so that the classification accuracy reaches the highest, with an overall accuracy of 0.909 8, the overall accuracy is 0.909 8, and the Kappa coefficient is 0.884 9, which also reduces the complexity of the model and improves the computational efficiency. ③The classification results after feature selection are compared with the "three-tone data", and the overall accuracy is 0.852 5, which is in line with the actual situation of the study area. The classification method based on feature selection proposed in this paper can provide technical support and theoretical reference for the extraction of cover information in karst areas.

Open-pit mining area extraction based on deep learning and object-oriented image analysis in the Weining Beishan area

LIU Li, LI Shiyao, WANG Run, LIU Shaoyu, SONG Yongfei, NIU Ruiqing

2024, 0(2):  51-57.  doi:10.13474/j.cnki.11-2246.2024.0209

Asbtract ( )   HTML ( )   PDF (16110KB) ( )  

References | Related Articles | Metrics

Weining Beishan is a key area for the restoration and management of mining ecological environment in Ningxia. Quickly and accurately extracting information of the open-pit mining area and monitoring the progress of ecological restoration have become important tasks in the mining management. This article proposes a method for extracting information on open-pit mining areas from domestically produced high-resolution remote sensing satellite images in the Weinin Beishan area, which combines deep learning and object-oriented analysis. The method first uses the U-Net model, which supports small-sample learning, to perform initial recognition of the open-pit mining area, and then combines object-oriented analysis with spatial analysis methods to extract the mining area boundary. The experimental results show that the accuracy of identifying the spatial location of open-pit mining areas is 0.71, and the average spatial range extraction accuracy is 0.78. Based on this, the paper identifies and analyzes the ecological restoration status of open-pit mines in the Weining Beishan area from 2019 to 2021. Among the identified 125 open-pit mining areas, 43.2% have been restored, including 44 filled and leveled pits, 6 areas that have been redeveloped, and 4 areas that have been artificially revegetated. The method proposed in this article can accurately extract vector boundaries of open-pit mining areas without the need for feature engineering and can provide technical reference for mining remote sensing monitoring in Ningxia.

Influence of guided filtering at different scales on the classification accuracy of multi-spectral remote sensing images

LÜ Qiang, LI Chaokui, XIE Mengyuan, LI Hao, CHEN Jun

2024, 0(2):  58-62.  doi:10.13474/j.cnki.11-2246.2024.0210

Asbtract ( )   PDF (1852KB) ( )  

References | Related Articles | Metrics

Due to the complex diversity of features, accurate identification of their classification accuracy is of great significance to remote sensing data processing. In order to improve the classification accuracy of multi-spectral remote sensing data based on Landsat 8 data, this paper proposes a method of fusing NDVI and NDBI with different scales to classify multi-spectral remote sensing images. Firstly, the first principal component of the multi-spectral data is extracted as the guide image, the original image is the input image, and the guide filter feature set with filter radii of 2, 4, 6 and 8 is extracted in turn. Then,the guided filtering feature set with different filtering radii is fused with the NDVI and NDBI features of the image, and the method of support vector machine is used to supervise the classification, so as to explore the influence of guided filtering of different scales on the classification accuracy of multi-spectral remote sensing images. The experimental results show that:①Guided filtering can better retain the edge features of the image while removing noise.②Guided filtering can improve the classification accuracy of multi-spectral remote sensing images, and the classification accuracy of different sizes of guided filtering radius images and original images has been improved to different degrees compared with the original image,the highest overall accuracy reaches 99.776 3%, and the Kappa coefficient is 0.997 1.③Guided filtering of different scales will obtain different classification results,and when the filter radii R=2, the classification accuracy of the image is the highest.

Landslide stability monitoring in southwest mountainous areas based on SBAS-InSAR technology

XING Mingze, ZUO Xiaoqing, ZHANG Jianming, HUANG Cheng, LI Yongfa, BU Jinwei, SHI Chao

2024, 0(2):  63-68.  doi:10.13474/j.cnki.11-2246.2024.0211

Asbtract ( )   PDF (6409KB) ( )  

References | Related Articles | Metrics

In this study, the small baseline subset time-series InSAR (SBAS-InSAR) technique is employed to monitor surface deformation in the northeastern part of Xuanwei, Yunnan province. Deformation results from January 2021 to June 2023 are obtained, and the deformation characteristics of the selected typical landslide are analyzed. GNSS monitoring data are collected and compared with the InSAR deformation monitoring for validation. Furthermore, the response of InSAR deformation to precipitation is analyzed. The research findings demonstrate the effectiveness of SBAS-InSAR in monitoring typical landslide in the southwestern mountainous regions. These results offer support for the early identification of landslide hazards and hold significant reference value for the stability monitoring of similar landslide disasters.

Housing data spatialization research based on remote sensing images for rapid loss assessment after earthquakes

ZHANG Ping, LI Bijun, LI Yin, ZHANG Yimei, Temuqile, LIU Ke, LI Zhijun

2024, 0(2):  69-73.  doi:10.13474/j.cnki.11-2246.2024.0212

Asbtract ( )   PDF (1777KB) ( )  

References | Related Articles | Metrics

The convolutional neural network method can efficiently extract housing vector data from high-resolution remote sensing images, quickly obtain spatialization data of housing data, and improve the updating ability of earthquake emergency database. Based on the contour-guided and local structure-aware encoder-decoder network(CGSANet) model and the equal scale grid sampling method on the basis of partition, this paper establishes the spatialization model of housing construction area and housing structure types, and achieves spatialization of multi-type housing data in complex regional backgrounds. Taking Huangmei county as the study area, the model of housing data spatialization(1 km×1 km) is constructed, and the ability to identify housing data of different structural types is achieved. The model of housing data spatialization constructed can be used to update the earthquake emergency database, and is of great significance for improving the accuracy and timeliness of housing data.

Analysis of local surface subsidence characteristics in Tianjin based on InSAR technology

ZHANG Qian, MA Yue, ZHOU Hongyue, YAN Shiyong

2024, 0(2):  74-79.  doi:10.13474/j.cnki.11-2246.2024.0213

Asbtract ( )   PDF (39597KB) ( )  

References | Related Articles | Metrics

Tianjin is one of the areas in our country with the most severe ground subsidence. This article is based on the integrated distributed scatterer interferometry (DS-InSAR) technology, which processed 58 scenes of Sentinel-1A data from January 2021 to June 2023. It obtained the latest surface deformation characteristics in the southern region of Tianjin. By combining information on land use, hydrogeology, and other factors, a typical subsidence analysis was conducted. The results are as follows:①There is significant variation in the distribution of ground subsidence in Tianjin, with obvious uneven subsidence characteristics. The southwestern region is the most severely affected area, with a maximum subsidence rate of 85.2 mm/a. ②The ground subsidence in Tianjin is closely related to excessive groundwater extraction, increased surface loading, and geological structures. This study provides data support and decision-making basis for geological disaster prevention and control in Tianjin.

Algorithm of obtaining sea ice concentration using polarization features from fully polarimetric SAR data

CHEN Xingzhe, XIE Tao, WANG Minghua, ZHANG Xuehong, LI Jian, BAI Shuying

2024, 0(2):  80-84,89.  doi:10.13474/j.cnki.11-2246.2024.0214

Asbtract ( )   PDF (6049KB) ( )  

References | Related Articles | Metrics

This paper proposes an algorithm to obtain sea ice concetration(SIC) from fully polarimetric SAR data based on polarization features. Firstly, multilookprocess and filtering are performed on the fully polarimetric SAR data to obtain the coherence matrix and covariance matrix. Secondly, a number of polarization features are obtained through the coherence matrix and covariance matrix, and the correlation and redundancy analysis of these polarization features is carried out to construct the optimal feature space.Then, put the optimal feature space as input into the neural network classifier to obtain the SIC result. Finally, extract the sea ice concentration according to the SIC result. In this paper, two fully polarimetric Radarsat-2 images in the southern waters of Labrador are used to obtain the SIC. Compared with the commercial the SIC product of ASI-3125, the algorithm results of this paper are basically consistent with the distribution trend of the SIC product of ASI-3125,and generally slightly larger than the SIC product of ASI-3125. The standard deviation distributions are 3.46% and 6.82%, indicating that the use of high-resolution fully polarimetric SAR data has advantages in monitoring small-sized broken sea ice in the marginal area.

Analysis of the applicability of PointNet++ deep learning model for semantic segmentation of typical elements of urban as-built mapping

HUANG Yinghua, DONG Zhenchuan, LI Hao, CHEN Zhuang, LIU Changrui, ZHANG Xianzhou

2024, 0(2):  85-89.  doi:10.13474/j.cnki.11-2246.2024.0215

Asbtract ( )   PDF (3379KB) ( )  

References | Related Articles | Metrics

The traditional methods for processing urban completion mapping point cloud scene data obtained by 3D laser scanner have several limitations and cannot meet the demand for efficient processing of products in the information society. In this paper, we analyze the demand for classification of urban completion mapping point cloud scenes and study the automated processing of point cloud scenes using a deep learning network model. Firstly, we preprocess the input urban completion mapping data to achieve point cloud downsampling, denoising, and ground point and non-ground point segmentation. Secondly, manually labels five regional scenes with millimeter-level labels and performs data augmentation techniques. And finally tests the semantic segmentation performance and effect of the PointNet++ network in urban completion mapping point cloud scenes. The test results show that the PointNet++ network can achieve the semantic segmentation of laser point clouds in urban completion mapping point cloud scenes with a small number of samples, and the overall mIoU reaches 73.06%, meeting the demand for semantic automatic segmentation of urban completion mapping point clouds and offering a new approach to processing urban completion mapping point cloud data.

LiDAR-based non-destructive detection algorithm of tower crane verticality

ZHOU Mingduan, QIN Yuhan, ZHANG Wenyao, XU Xiang, ZHOU Qinghui

2024, 0(2):  90-94.  doi:10.13474/j.cnki.11-2246.2024.0216

Asbtract ( )   PDF (2222KB) ( )  

References | Related Articles | Metrics

Aiming at the many disadvantages of traditional theodolite measurement method,a novel LiDAR-based non-destructive detection algorithm of tower crane verticality is proposed. The point cloud data of the tower crane is obtained by using LiDAR,and the effective point cloud data of the standard section of the tower body is obtained through the pre-processing of point cloud alignment,stitching,de-noising and de-redundancy. The standard node cloud cross-section slice segmentation scheme is designed for the tower body,and the Marching Square algorithm is used to extract the cross-section slice polygon contour lines and determine the vertices of the contour lines and their corresponding centroids. The least squares parameter estimation method is used to fit the linear equation of the center point space,the direction vector of the tower axis line is determined,and vector operations are performed with the z-axis and x-axis of the station-center spatial coordinate system,respectively,and then the tilt angle,tilt azimuth and verticality parameters of the tower axis line are solved. The experimental results show that the tower crane verticality parameters obtained by the two kinds of segmentation scheme designed by the algorithm in this paper are 2.04‰ and 2.49‰ respectively,which are smaller than the results of the traditional theodolite measurement method of 3.02‰. The algorithm in this paper is effective,which can provide a kind of non-destructive detection algorithm for the high-precision monitoring of the verticality of tower crane.

3D model reconstruction and visualization of digital factory based on multi-source data fusion

DU Jiatao, WANG Fengyan, WANG Mingchang, WU Xiang, LIU Xingnan, MA Runze, AN Zhilei

2024, 0(2):  95-99.  doi:10.13474/j.cnki.11-2246.2024.0217

Asbtract ( )   PDF (11316KB) ( )  

References | Related Articles | Metrics

Providing current, detailed, and accurate spatial and geographic information data is crucial for assisting intelligent decision-making in areas such as enterprise safety production, emergency response command, and accident investigation. This article focuses on the key technology of virtual construction of digital factories, and uses a combination of macroscopic modeling with airborne oblique photogrammetry and fine modeling with local ground-based 3D laser scanning technology to construct a model of the factory area. With the support of the front-end Cesium framework and Web GL technology, combined with Microsoft's Web preview version of Bing maps service, users can visualize the “digital factory” 3D scene in their web browser, thereby enhancing the informationization level of enterprise safety production and preventing and controling major safety accidents.

Optimization of 3D scanning and inspection scheme for thin plate components based on TRIZ theory

WANG Zhanhui, FENG Chaojie, GUO Xiaofan, GAO Fei, DU Xiaopeng, WEI Xun

2024, 0(2):  100-106.  doi:10.13474/j.cnki.11-2246.2024.0218

Asbtract ( )   PDF (3488KB) ( )  

References | Related Articles | Metrics

To address the difficulties in overall data acquisition, low efficiency, and the inability to simulate real operating conditions in the process of three-dimensional scanning and inspection of thin plate components, researchers are currently optimizing the three-dimensional scanning and inspection scheme for thin plate components based on TRIZ innovation theory. After analyzing the existing issues, they employ methods such as causal chain analysis, nine-screen diagram, and technical contradictions, by integrating these methods with the actual scanning process of thin plate components, they are designing a multi-point flexible fixture to optimize the 3D scanning and inspection process. They validate the optimization process using automotive sheet metal parts as an example, providing a new, efficient, and accurate inspection method for quality assessment of thin plate components.

High-definition map data interaction mode for intelligent connected vehicle

ZHANG Chuang, YING Shen, WANG Runze, WANG Shuman, LI Lin

2024, 0(2):  107-112.  doi:10.13474/j.cnki.11-2246.2024.0219

Asbtract ( )   PDF (2847KB) ( )  

References | Related Articles | Metrics

With the development of autonomous driving technology, traditional map services can no longer meet the needs of autonomous vehicles with machines as the main body. A high-definition(HD) map with higher accuracy, more elements, and faster update frequency has become the key to autonomous vehicles. Analyzing the data interaction issues concerning HD maps in intelligent connected vehicle systems, this research examines relevant studies and reorganizes the HD map data in intelligent connected vehicles based on the data content and sources of HD maps. Furthermore, it elaborates on the content and interaction mechanism of HD map data exchange based on the “vehicle-road-cloud” communication framework, forming an applicable HD map data interaction model for intelligent connected vehicles. Finally, the designed data interaction model is validated in a simulation environment.

A deep monocular visual-inertial navigation algorithm with SuperGlue

LIU Yibo, WU Chuanwen, ZHOU Zongkun, CHEN Hua

2024, 0(2):  113-117.  doi:10.13474/j.cnki.11-2246.2024.0220

Asbtract ( )   PDF (3472KB) ( )  

References | Related Articles | Metrics

The deep learning method for images is an effective way to solve the problems of unstable feature extraction and tracking loss of traditional visual positioning algorithms in complex environments. In this paper, we propose a visual-inertial navigation algorithm based on VINS-Mono, which using SuperPoint to get feature points and track them by using SuperGlue. And evaluate it using Open-source dataset and real world experiments. Experimental results show that our algorithm has a significant improvement in positioning accuracy and stability compared with VINS-Mono, and the accuracy improvement can reach 26%.

Prelipitable water vapor inversion using BeiDou PPP-B2b service and analysis of heavy rainfall process in Hong Kong

CHENG Ankun, WANG Min, MENG Xin, JI Rui, SUN Shuang

2024, 0(2):  118-123.  doi:10.13474/j.cnki.11-2246.2024.0221

Asbtract ( )   PDF (4719KB) ( )  

References | Related Articles | Metrics

The precision single point positioning service (PPP-B2b service) of BDS-3 sends correction information to the user through the satellite B2b signal, which is free from the dependence on external communication, and provides a new technological way for real-time inversion of precipitable water vapor (PWV) in the atmosphere. In this paper, the accuracy of PWV inversion using PPP-B2b service is analysed experimentally, and the experimental results show that the difference between the inversion results using PPP-B2b service and those using the CODE hindcast precision ephemeris product is 3.70 mm in RMS and 3.61 mm in STD, and the differences with the ERA5 reanalysis data are 4.80 mm in RMS and 4.07 mm in STD, and the differences with the sounding results are 4.80 mm and 4.07 mm in STD, respectively. The RMS and STD of the difference with the ERA5 reanalysis data are 4.80 mm and 4.07 mm, and the RMS and STD of the difference with the sounding results are 3.75 mm and 7.16 mm, respectively, which verifies that the PWV inversion results using the PPP-B2b service have a good accuracy. Finally, by analysing the correlation between the actual precipitation and the PWV inversion results of the rainstorm process in Hong Kong, the feasibility of BDS PPP-B2b service for the short-term early warning of rainstorm disasters is preliminarily verified.

Classification method of concentrated contiguous arable land under non-uniform grid

LI Bo, ZHAO Rong, ZHANG Yu

2024, 0(2):  124-128.  doi:10.13474/j.cnki.11-2246.2024.0222

Asbtract ( )   PDF (1754KB) ( )  

References | Related Articles | Metrics

The fragmentation of arable land leads to high agricultural production cost and low efficiency. In order to improve the utilization efficiency of arable land, it is necessary to realize the centralized management of arable land. Aiming at the problem of data redundancy and grid scale selection when the uniform grid is used to express the arable land area, this paper proposes a method to determine the optimal starting and ending scales of the grid according to the average area of the patch and the area expression error, and constructs a non-uniform grid based on the quadtree. The analysis rules, evaluation index system and technical process of arable land centralized connection are established to realize the classification of arable land centralized connection under non-uniform grid. Taking Xuzhou city of Jiangsu province as the experimental area, the concentrated contiguous analysis of arable land is carried out. The results show that:①Compared with uniform grid, non-uniform grid reduces 43.8% data redundancy, realizes the minimum error expression of arable land area, and has stronger applicability.②The area of arable land in Xuzhou accounts for 88.6% of the total area of arable land, of which 50 % of the arable land is concentrated in the upper middle level.③In the contiguous area, 21.4% of the arable land is concentrated below the middle level, which should be used as the arable land to improve the overall level of the region.

Assessment and reflection on the operational application of Satellite-Derived Bathymetry in hydrography

WANG Zhao, BAI Tingying

2024, 0(2):  129-133,139.  doi:10.13474/j.cnki.11-2246.2024.0223

Asbtract ( )   PDF (1286KB) ( )  

References | Related Articles | Metrics

Bathymetry is the most important geographical element of hydrography, reflecting the geomorphology of the seabed and the navigability of maritime routes, and occupying an important position in hydrography. Satellite-derived bathymetry has become a potential complement to traditional acoustic bathymetry systems. Firstly, this paper summarizes the existing studies, and analyzes the elements considered in SDB from three aspects: reality factors, image factors, and bathymetry-derived models. Secondly, the applicability of SDB is evaluated against hydrographic accuracy and operationalization needs. The results show that:①The horizontal accuracy of SDB has met the requirements.②The relative vertical accuracy is around 20% for most multispectral data, which still does not meet the operationalized accuracy requirements.③There is already good practice in seafloor coverage using SDB. Finally, the problems are analyzed in terms of uncertainty controllability, bathymetry-derived model extrapolation, and multi-source data assimilation, and some thoughts and suggestions are given.

Point cloud construction and differential analysis based on airborne laser scanning and oblique image matching

TAN Jinshi, GAO Zhaozhong, YANG Minjing, ZU Weiguo, LIU Li

2024, 0(2):  134-139.  doi:10.13474/j.cnki.11-2246.2024.0224

Asbtract ( )   PDF (3559KB) ( )  

References | Related Articles | Metrics

Point clouds are an important part of 3D spatial data. Airborne laser scanning and tilt image matching are the two main point cloud construction techniques, which have commonalities and differences. The aim of this paper is to analyse the differences between airborne laser scanning and tilted image matching techniques for point cloud construction. Firstly, the specific methods of airborne laser scanning and tilted image matching point cloud construction are described, and then compare and analyse in terms of point cloud construction effect, data integrity, density, accuracy and vegetation penetration in conjunction with case studies. The results show that both techniques have good data integrity and can produce highly dense point clouds that far exceed the specification requirements, with comparable and high accuracy, but the laser scanning has loopholes in the occluded areas and distorted details in the blind local areas of image matching; the laser point cloud has good penetration and can achieve ground point cloud construction in different vegetation covered areas, while the image matching points have poor penetration. There is a lack in densely vegetated areas, can't even construct ground points. The results of the study provide a reference for the selection and optimization of subsequent point cloud construction methods, and have certain research and application value.

Identification of construction waste information with multiple features using object-oriented morphological operation

ZHANG Mengyuan, ZHAO Junhua, SUN Yumei, HAO Zongpeng

2024, 0(2):  140-143.  doi:10.13474/j.cnki.11-2246.2024.0225

Asbtract ( )   PDF (2944KB) ( )  

References | Related Articles | Metrics

The long-term storage and unscientific management of construction waste will cause various ecological and social problems, which will seriously affect the green and sustainable development of the city. In the research of solid waste information recognition, the difference of texture characteristics between buildings and construction waste is not considered, which may lead to confusion between them in the classification process. To solve this problem, mathematical morphology algorithm can be used to highlight the gray intensity characteristics of construction waste. Then the differences of morphological, spectral, geometric and texture characteristics of various ground objects are analyzed to realize object-oriented construction waste information extraction with multiple features. Taking Baohezhuang Village, Fangshan District, Beijing as an example, the experiment is conducted using WorldView-2 remote sensing image. The accuracy of construction waste extraction is evaluated by establishing confusion matrix and separability evaluation index. The overall accuracy is up to 96.6%, and the separation between construction waste and buildings is up to 1.000. The results show that this method can effectively solve the confusion problem between construction waste and buildings, and has reliable applicability in the information extraction of construction waste.

Production technology of city-level real scene 3D models based on oblique images and airborne LiDAR point clouds

ZHU Xuhe, LUO Ningxin, WANG Junyi

2024, 0(2):  144-147.  doi:10.13474/j.cnki.11-2246.2024.0226

Asbtract ( )   PDF (10687KB) ( )  

References | Related Articles | Metrics

With the advancement of real scene 3D modeling in China, the reconstruction of city-level real scene 3D models has become a key focus of research. However, there are challenges in the process of oblique photogrammetric modeling for urban areas, such as large amounts of data processing, complicated and diverse terrains and occlusion of viewpoints leading to model deformation. To address these issues, this paper summarizes a complete technical workflow for city-level real scene 3D model reconstruction by fusing oblique aerial images and airborne LiDAR point clouds. The proposed workflow is validated using data from Zhongshan city, and the results demonstrate that this workflow can effectively improve the efficiency and quality of city-level real scene 3D model reconstruction.

Research and application of 3D ground penetrating in road detection

ZHAO Zhen, HUANG Yong, FENG Kun

2024, 0(2):  148-152.  doi:10.13474/j.cnki.11-2246.2024.0227

Asbtract ( )   PDF (3075KB) ( )  

References | Related Articles | Metrics

In recent years,the development of urban construction and the development and utilization of underground space have brought convenience to people's lives, but to a certain extent, it has also led to the occurrence of urban road hazard, and the frequent occurrence of road collapse accidents has prompted us to conduct a comprehensive physical examination of urban roads.In this paper, the propagation characteristics of geo-radar in media such as pipe, culvert and cavity are numerically simulated based on GprMax, which is based on the principle of 3D finite-difference time-domain method.A preliminary forwardmodeling of radar images is established and the image characteristics of radar data bodies are analysed using the example of 3D ground-penetrating radar in road detection work in Tianjin area, while relevant experiences and practices in the detection of urban road subsurface hazardare are introduced,Which provide technical methods and guidance for road detection.

Overall denoising method for shield tunnel based on point cloud spatial expansion

SHEN Changbiao, XIA Yonghua, LIU Yong, WANG Dandan

2024, 0(2):  153-156.  doi:10.13474/j.cnki.11-2246.2024.0228

Asbtract ( )   PDF (3306KB) ( )  

References | Related Articles | Metrics

The denoising of point cloud data in shield tunneling is a prerequisite for subsequent tunnel structure analysis.This paper proposes an overall denoising method for shield tunnels based on the theory of point cloud spatial expansion and the cylindrical characteristics of shield tunnels. Firstly,the shield tunnel is spatially expanded to obtain a new topological structure,enabling the application of existing mature ground filtering algorithms for denoising.After direct pass filtering and slope filtering,the point cloud of the tunnel retains the complete lining structure while removing the ancillary facilities of tunnel.Experimental tests are conducted on both straight and curved tunnels, and the results demonstrate that this method achieves significant overall denoising of the tunnel with remarkable denoising effects.

Exploration of multi-source laser point cloud modeling in bridge maintenance

LIU Shuangchen, LI Shengfu, JIA Yang

2024, 0(2):  157-160,177.  doi:10.13474/j.cnki.11-2246.2024.0229

Asbtract ( )   PDF (4119KB) ( )  

References | Related Articles | Metrics

Under the influence of serious road load and secondary geological disasters, the daily maintenance of complex mountain highway bridges has brought great challenges. Using bridge 3D modeling technology to carry out online visualization, and installing monitoring equipment on key bridge sections, online monitoring of bridge health has become a new technical means. For roads and bridges that have been operating for a long time, the as-built design drawings are difficult to find and the real bridge model can not be reproduced through the drawings. Therefore, the ground-station laser and vehicle-mounted laser are combined to scan the bridges that have been operating for a long time, and the point cloud data obtained by the two methods are registered and filtered, and then the monomer modeling is carried out. Restore the real structure of the bridge. At the same time, using the highway industry BIM model standard, the components of the monomer bridge are split and coded to form a standardized bridge unit module. On this basis, GIS technology is used to develop expressway infrastructure platform, and bridge maintenance related operation modules are associated to realize online inspection and maintenance of existing expressway bridges.

Long-distance cross-sea height datum transfer method and verification in the Beibu Gulf engineering in Northern Guangxi

ZHONG Changhai, LIN Zile, HUANG Xin, KONG Jian

2024, 0(2):  161-164,182.  doi:10.13474/j.cnki.11-2246.2024.0230

Asbtract ( )   PDF (1702KB) ( )  

References | Related Articles | Metrics

The traditional methods for transferring height between land and sea benchmarking points are affected by factors such as the distance between islands and mainland, and the land-sea environment, leading to measurement inaccuracies. In this article, we propose a method using high-precision quasi-geoid surfaces for height transfer. We investigate error accumulation factors, provide a theoretical model for height transfer, and describe the specific technical process. In the Beibu Gulf area, this method is used to transfer the 1985 national height datum from the mainland to Weizhou Island and Xieyang Island, which are located 40 kilometers away from Beihai city. The elevation differences between GNSS leveling points after the transfer and the independent elevation differences from the second-order leveling measurement on Weizhou Island are all better than ±4.1 cm. This study provides practical reference for height transfer projects in offshore areas nationwide.

Research on the construction technology of 3D one-map of natural resources

ZHONG Yong, XIE Gangsheng, DUAN Wenzhou, ZOU Binwen

2024, 0(2):  165-169.  doi:10.13474/j.cnki.11-2246.2024.0231

Asbtract ( )   PDF (2212KB) ( )  

References | Related Articles | Metrics

The traditional 2D graphic technology is unable to meet the high-quality development needs of current natural resource management. And realistic 3D technology has become the key to solving this problem. In order to achieve 3D centralized management of various natural resources and provide basic support for unified investigation, monitoring, and evaluation of natural resources, based on big data, cloud computing, knowledge graph and other technologies, taking the Louxing district of Loudi city as an example, 3D one-map management platform for natural resources is designed and implemented to realize the management and utilization of natural resources in this paper. It explores its application in natural resource data management, data sharing, multidimensional mining and analysis, and all-round analysis and decision-making, providing reference experience for relevant work in various places.

Improved Mask RCNN method for shield tunnel leakage detection

WANG Jian, ZHENG Like, WU Binjie, QI Zhiyu

2024, 0(2):  170-177.  doi:10.13474/j.cnki.11-2246.2024.0232

Asbtract ( )   PDF (5777KB) ( )  

References | Related Articles | Metrics

Water leakage is an important characterization of the potential damage or defects of the shield tunnel structure. Rapid and accurate detection of the tunnel leakage site is of great significance to the safe operation and maintenance of the tunnel. However, most of the existing methods use optical images to detect the tunnel water leakage, but due to the tunnel space limitations and light conditions, it is difficult to obtain high-quality disease pictures. In this regard, a water leakage detection method based on terrestrial laser scanning point cloud and improved Mask RCNN is proposed. Firstly, the laser point cloud reflection intensity is corrected, and then the gray scale image is generated and the water leakage disease data set is established. Finally, the atrous convolution and deformation convolution are introduced in the Mask RCNN algorithm to realize the rapid detection of tunnel water leakage disease. The data collected in metro are used for verification. Experimental results show that, compared to the original algorithm and FCN algorithm, the detection accuracy of the proposed improved Mask RCNN algorithm is significantly improved, and it has a good performance in water leakage identification in shield tunnel.

Application analysis of UAV LiDAR system in urban rail transit construction

XU Huazhi, JIANG Wenting, CHEN Chunlei, ZHANG Yuntao, WANG Xiao

2024, 0(2):  178-182.  doi:10.13474/j.cnki.11-2246.2024.0233

Asbtract ( )   PDF (1476KB) ( )  

References | Related Articles | Metrics

In order to reduce the workload of field survey in urban rail transit construction and improve the efficiency and accuracy of mapping, this study is based on an actual project. The SF1650 six rotor UAV LiDAR system is selected to aerial photograph of the Jinan East Railway Station area. The obtained LiDAR point cloud and image data are used to make DEM and DOM, and then large-scale topographic maps and sectional maps of the area are made. Through field surveys, it has been verified that the accuracy of DEM and DOM fully meets the requirements for large-scale topographic maps in urban rail transit construction projects. Compared to traditional stereomapping, the accuracy and data utilization have been greatly improved, greatly reducing the workload of field surveys. The feasibility of the SF1650 UAV LiDAR system has been verified, providing a reference plan for future engineering applications.