Table of Content

25 November 2025, Volume 0 Issue 11

Previous Issue   

Comprehensive prevention and control of geological hazard risks in transmission lines: a UAV-based multi-scale fusion approach

ZHONG Qin, LIU Yi, WANG Shenli, SHI Yi, HE Xiangkui, MA Li, YANG Xiaodong, ZHANG Di

2025, 0(11):  1-7.  doi:10.13474/j.cnki.11-2246.2025.1101

Asbtract ( )   HTML ( )   PDF (6064KB) ( )  

References | Related Articles | Metrics

The application of unmanned aerial vehicles (UAVs) in the comprehensive prevention and control of geological hazard risks for transmission lines still faces challenges including fragmented operational scenarios,insufficient data fusion,and lack of systematic methodologies.This study focuses on mountainous transmission lines threatened by geological hazards such as collapses and landslides.Targeting multi-scale scenarios encompassing macro-regional,meso-key,and micro-local dimensions,we fully utilize various photogrammetric techniques (vertical,oblique,and close-range imaging) combined with multi-scale scenario fusion methods.A technical framework and implementation process for UAV-based multi-scale fusion in geological hazard management are proposed,covering field operations,product generation,and practical applications.Key technical approaches including multi-mode field operation combinations and multi-scale indoor fusion modeling are elucidated.The proposed methodology was applied to a high-risk transmission line section in the Three Gorges Reservoir area of western Hubei province.Results demonstrate its effectiveness in supporting rapid construction of real-scene 3D platforms while fulfilling comprehensive requirements for hazard identification/assessment,mitigation design,and deformation monitoring,thereby verifying its practical applicability.

Interpretation of surface deformation of UHV transmission channel and quantitative evaluation of tower damage

LI Chunyi, GUO Yaxing, DING Laizhong, CUI Ximin, ZHENG Yuesong

2025, 0(11):  8-14.  doi:10.13474/j.cnki.11-2246.2025.1102

Asbtract ( )   HTML ( )   PDF (5729KB) ( )  

References | Related Articles | Metrics

In order to detect the surface deformation under the UHV transmission channel and quantitatively evaluate the damage of the tower,this paper uses 163 Sentinel-1A remote sensing images as the data source,and uses MT-InSAR technology to understand the surface deformation of the Lingshao line and the Shanwu line UHV transmission channel crossing Sanmenxia city,Henan province.The calculation model of surface cumulative settlement at any time is constructed,and the time benchmarks of MT-InSAR and leveling monitoring data are unified.The fusion of MT-InSAR and leveling monitoring data is realized by Kriging interpolation algorithm.A surface fitting parameter model considering underground mining space is created,and the calculation parameters of surface subsidence are inverted.According to the maximum inclination index of the surface,the stability of the transmission line tower is quantitatively evaluated.The results show that MT-InSAR technology can effectively detect the abnormal surface deformation area of UHV transmission channel,and the settlement and tilt deformation of transmission tower are closely related to underground mining of coal mine.The fusion of MT-InSAR and leveling data can effectively invert the calculation parameters of surface subsidence.The LS1130 tower in the coal mining subsidence area exceeds the allowable value of the maximum tilt of the surface and needs to be reconstructed; the maximum tilt deformation of the four towers of LS1132,LS1134,SW1006 and SW1008 is between 1.45 and 2.3 mm/m,resulting in a large tilt,which should be strengthened and repaired in time ; the maximum tilt deformation of the remaining towers is below 1 mm/m.The research results can provide safety warning for transmission lines,and also provide theoretical and technical basis for the evaluation and treatment of tower damage in coal mining subsidence area.

Disturbance analysis of forest aboveground biomass in power transmission corridors based on multi-source remote sensing data

LIU Haibo, ZHANG Su, RONG Jingguo, CHANG Jianyang, YANG Zhou

2025, 0(11):  15-21,39.  doi:10.13474/j.cnki.11-2246.2025.1103

Asbtract ( )   HTML ( )   PDF (5985KB) ( )  

References | Related Articles | Metrics

Aboveground biomass (AGB) of forests along transmission corridors is crucial for ecosystem stability and regional carbon cycling,yet quantitative assessments of potential spatial disturbances caused by transmission corridors remain limited.This study focuses on two typical transmission corridors in Fuling district,Chongqing,integrating multi-source remote sensing data including global ecosystem dynamics investigation (GEDI) spaceborne LiDAR,airborne AGB measurements,Sentinel-2 imagery,and shuttle radar topography mission (SRTM) data.A framework is developed for footprint-scale AGB retrieval and regional-scale mapping.Buffer and vertical profile analyses are further applied to quantify the impact of transmission corridors on forest AGB at both landscape and local scales.The footprint-scale AGB retrieval achieves a root mean square error (RMSE) of 19.36 Mg/hm²,and the regional-scale AGB map has an RMSE of 2.04 Mg/hm².Differences in AGB between inner and outer buffer zones are minimal,and fluctuations along vertical profiles are less than 5 Mg/hm².Overall,transmission corridors do not cause significant disturbances to forest AGB.Observed spatial variations in AGB are mainly attributable to natural heterogeneity.

Real-time and post-production methods of orthophoto maps of power transmission channels based on tight coupling of GPS and monocular SLAM

CHEN Jingchuan, BAO Ruoyu, LUO Chaowei, LIU Xuedan, DENG Baichuan, LEI Ming

2025, 0(11):  22-26,33.  doi:10.13474/j.cnki.11-2246.2025.1104

Asbtract ( )   PDF (2601KB) ( )  

References | Related Articles | Metrics

When encountering emergencies such as floods,fires,and mudslides,quickly identifying the disaster situation in the transmission channel is the primary task of the power grid's natural disaster emergency response.However,the traditional orthophoto image generation technology cannot meet the real-time requirements of emergency response.The existing technology has relatively few studies on the real-time generation of orthophoto images for long strip transmission channels,and lacks a systematic solution.This study proposes a monocular SLAM technology based on GPS tight coupling.By adding real geographic information scale and pose to the SLAM system,and using local and global optimization improved algorithms,the cumulative error problem in the traditional stitching algorithm is solved.Using dynamic Delaunay triangulation and orthorectification algorithms,and targeting the long strip characteristics of transmission lines,this paper adopts a real-time and post-event orthophoto generation strategy to optimize the video stitching process and ensure that the generated images have real-time and high resolution.The ablation experiment shows that the model generated by the systematic solution proposed in this paper effectively reduces the edge spikes,holes and image distortion in the orthophoto image,and the effect is better than a single SLAM model.Compared with the traditional Pix4D software,under the same conditions,the efficiency of post-orthophoto generation is increased by 34.64 times,and the integrity of orthophoto is increased by 70.59 percentage points.

Alteration information extraction from hyperspectral(GF-5)and multispectral images in the Gangdise metallogenic belt

LIU Xingyu, QIU Chunxia

2025, 0(11):  27-33.  doi:10.13474/j.cnki.11-2246.2025.1105

Asbtract ( )   PDF (19780KB) ( )  

References | Related Articles | Metrics

Wall rock alteration is a significant indicator of hydrothermal mineralization and provides crucial information for mineral exploration.With the development of remote sensing technology,multispectral imagery has been widely used in mineral identification and lithological mapping,while hyperspectral imagery has attracted considerable attention due to its superior capability in ground object recognition.The Chinese independently developed GF-5 satellite integrates the dual advantages of wide-swath and high spectral resolution technology,however,its capability for alteration mineral identification in remote areas of the Tibetan Plateau still needs validation.This study,focused on the Gangdise metallogenic belt,compares the effectiveness of GF-5 and ASTER imagery in detecting alteration minerals.Four mixed pixel decomposition methods integrated with SAM classification are applied to GF-5 data,enabling the successful identification of characteristic alteration mineral assemblages such as chlorite and muscovite.Meanwhile,the improved Crosta technique was employed to process ASTER imagery,extracting three types of alteration:iron staining,Mg-OH alteration,and Al-OH alteration.Ground drill hole validation showed that the overall identification accuracies of GF-5 and ASTER data were 0.875 and 0.625 respectively,demonstrating that GF-5 imagery can achieve high-precision mineral information identification in remote areas of the Tibetan Plateau,offering significant advantages and application prospects in mineral exploration.

Few-shot hyperspectral image classification with depth feature fusion

QIN Jinchun, PEI Hang, LIU Bing, YU Anzhu, CHEN Junming, FAN Junyi

2025, 0(11):  34-39.  doi:10.13474/j.cnki.11-2246.2025.1106

Asbtract ( )   PDF (3570KB) ( )  

References | Related Articles | Metrics

A depth feature extraction method for hyperspectral image classification is proposed to address the small sample problem.The proposed method first utilizes a pre-trained base model to extract depth maps from hyperspectral images as prior information,which is then fused with spectral information for classification.To fully exploit the rich spectral information in hyperspectral images,a sliding window approach is employed to extract multiple depth maps along the spectral dimension,which are then stacked to form depth features.The method is based on the concept of multi-source remote sensing image fusion but does not require precisely registered multi-source remote sensing images,offering a plug-and-play advantage.Extensive classification experiments on three hyperspectral image datasets validate the effectiveness of the method.

An improved YOLOv11-based algorithm for interchange bridge recognition in remote sensing imagery

HUANG Chuanshu, LI Jiatian, YANG Kun

2025, 0(11):  40-46.  doi:10.13474/j.cnki.11-2246.2025.1107

Asbtract ( )   PDF (7709KB) ( )  

References | Related Articles | Metrics

In response to the challenges of large target scale differences and complex backgrounds leading to low detection accuracy of overpasses in remote sensing images,this paper proposes a solution based on the YOLOv11 framework.The approach introduces hypergraph computing to explore the intrinsic relationships between cross-layer features,and presents a parallel processing channel selection module based on a gating mechanism that dynamically selects and strengthens task-relevant features,enhancing the model's ability to focus on key information.Additionally,a learnable scaling factor is embedded in the regression part to construct a regression-optimized detection head,improving the adaptive ability of bounding box predictions and enhancing network performance.Experimental results show that,on the Dior dataset for the overpass category,the proposed algorithm achieves mAP＿50 and mAP＿50:95 of 80.5%and 53.1%,respectively,outperforming the comparison algorithms,effectively improving the detection accuracy and robustness of overpass targets in complex backgrounds.

UAV-HSfM modeling method and its empirical analysis in the complex mountainous areas of central Yunnan Plateau

ZHANG Chi, GAN Shu, YUAN Xiping, LUO Weidong, MA Chong, LI Yi

2025, 0(11):  47-52,61.  doi:10.13474/j.cnki.11-2246.2025.1108

Asbtract ( )   PDF (24165KB) ( )  

References | Related Articles | Metrics

As a key technique within 3D reconstruction technology,the structure from motion (SfM) algorithm enables 3D scene reconstruction when integrated with UAV imagery.To address the issues of error accumulation in traditional incremental SfM(ISfM) and the high computational complexity of global SfM (GSfM) optimization in drone-based 3D reconstruction,this paper proposes a hybrid drone modeling method (UAV-HSfM). This approach first estimates camera rotations using a global algorithm and then computes camera positions through an incremental algorithm,thereby optimizing the 3D reconstruction process.The study was conducted in a mountainous test area in the central Yunnan Plateau.The results demonstrate that:①The proposed method achieves more stable feature extraction;②The sparse point cloud exhibits both completeness in quantity and geometric accuracy; ③The elevation root mean square error (RMSE) of 0.056 m shows significant improvement,with a more concentrated error distribution.In conclusion,the UAV-HSfM method effectively combines the advantages of GSfM and ISfM,significantly enhancing the accuracy of 3D reconstruction in complex mountainous terrains.

A review on the development of distributed SLAM technology for multiple unmanned platforms

XIONG Chao, WU Meng, DUAN Xuzhe, ZHAO Pengcheng, LU Chuanfang, XU Jian

2025, 0(11):  53-61.  doi:10.13474/j.cnki.11-2246.2025.1109

Asbtract ( )   PDF (1113KB) ( )  

References | Related Articles | Metrics

In response to the key issues such as low collaborative positioning accuracy,environmental perception conflicts,and limited communication bandwidth in the distributed simultaneous localization and mapping (SLAM) technology for multiple unmanned platforms,this paper adopts a research strategy combining literature analysis and method comparison,and conducts research based on the current domestic and foreign research status.It systematically sorts out the core challenges of distributed SLAM in aspects such as time synchronization,data association,and topological optimization.Secondly,it conducts a comparative analysis of representative domestic and foreign datasets for multiple unmanned platforms from the dimensions of platform type,sensor configuration,and scene complexity,revealing the deficiencies of existing datasets in terms of heterogeneous system adaptability.Finally,it looks forward to the development trend of the distributed SLAM technology for multiple unmanned platforms and points out future research directions of multi-modal data fusion and lightweight map construction.The research results provide theoretical support and data reference for the collaborative operation of multiple unmanned platforms and have guiding significance for improving the autonomous collaboration ability in complex environments.

Future landslide susceptibility monitoring and assessment in the eastern Qilian Mountains

XU Chen, BAO Shuai, LIU Mengmeng

2025, 0(11):  62-69.  doi:10.13474/j.cnki.11-2246.2025.1110

Asbtract ( )   PDF (38662KB) ( )  

References | Related Articles | Metrics

To address the difficulty of quantifying future climate impacts in static landslide assessments,this study proposes a dynamic evaluation method integrating multi-source factors with climate scenarios.Taking the eastern section of the Qilian Mountains as the study area,we firstly construct an indicator system by integrating multi-source data such as topography and geology.Multiple machine learning models are then compared to select the optimal algorithm.Finally,we couple scenarios from the sixth coupled model intercomparison project (CMIP6) (SSP1-2.6,SSP2-4.5,SSP5-8.5) to predict landslide susceptibility in different future periods.The results show that the random forest model performs best.In the baseline period (1991—2020),medium-high susceptibility areas were concentrated in low-slope valleys.Under the SSP5-8.5 scenario,the proportion of high and very high susceptibility areas is projected to increase to 25.4% by the end of the 21 century,whereas SSP1-2.6 could limit it to around 20%.This study overcomes the limitations of static assessments and reveals that,under high-emission pathways,climate change will significantly exacerbate landslide risks in the eastern Qilian Mountains.Emission reduction is therefore crucial.

Evaluation of landslide susceptibility by integrating time-series InSAR and PSO-SVM models

SHANG Xuewei, YANG Rui, CHEN Yuhan, GAO Fufang

2025, 0(11):  70-77,123.  doi:10.13474/j.cnki.11-2246.2025.1111

Asbtract ( )   PDF (10983KB) ( )  

References | Related Articles | Metrics

To address the issues of outdated landslide inventories and the lack of dynamic factors in the landslide susceptibility assessment process,this study takes Midu county,Yunnan province,as a case study.Surface deformation information from ascending and descending orbits was retrieved using the SBAS-InSAR technique.Combined with high-resolution optical imagery and slope data,landslides were identified to update the existing landslide inventory.In addition,based on 13 static conditioning factors such as slope and aspect,surface deformation rate was introduced as a dynamic factor,and a PSO-SVM model was employed to conduct landslide susceptibility assessment.The results indicate that:①The monitoring results from ascending and descending orbit in the study area exhibit significant differences both in deformation rates and in the spatial extent of deformation.These discrepancies are primarily attributed to variations in satellite orbit geometry as well as local topographic characteristics;②Integrating ascending and descending orbit datasets effectively overcomes the limitations of single-orbit observations,thereby improving the comprehensiveness and accuracy of landslide identification;③Incorporating dynamic factors yields higher assessment accuracy,with AUC,precision,recall,and F1-score reaching 0.886,0.847,0.851,and 0.855,respectively.This approach provides valuable reference for landslide prevention and early warning in Midu county.

An instance segmentation method for mining-induced ground fissures in shallow coal seams using UAV imagery

SUN Bin, ZHANG Ruiling, REN Chenfeng, LIN Yunhao, SUN Chao, LIU Yihan, LIU Mengjie, YUAN Debao, XU Zhihua

2025, 0(11):  78-83.  doi:10.13474/j.cnki.11-2246.2025.1112

Asbtract ( )   PDF (9843KB) ( )  

References | Related Articles | Metrics

Ground fissures induced by shallow coal mining significantly damage the ecological environment of mining areas.Timely detection and landfill treatment can prevent secondary hazards such as spontaneous combustion of residual coal and water inrush during rainy seasons.This paper proposes an improved YOLOv8 model for instance segmentation of ground fissures from UAV imagery.Firstly,the backbone of YOLOv8 is replaced with a pyramid vision transformer (PVT) to enhance multi-scale and high-resolution feature learning for dense fissures,improving geometric recognition capabilities.Then,UAV images from the Huangyuchuan Coal Mine in Inner Mongolia were processed to create the HYCdata dataset for model training.Experiments demonstrate that the modified YOLOv8 outperforms the original model,achieving a mAP_0.5 of 74.3%,providing an effective solution for automatic segmentation of widespread mining-induced fissures.

Land cover classification in multi-modal remote sensing images using dual attention and multi-branch losses

YU Xiaowei, ZHENG Yadong, LIANG Li

2025, 0(11):  84-90,153.  doi:10.13474/j.cnki.11-2246.2025.1113

Asbtract ( )   PDF (3589KB) ( )  

References | Related Articles | Metrics

Existing land cover classification methods still face numerous challenges in feature extraction and fusion quality.This paper proposes a multimodal image classification method named dual-attentive triple-branch fusionNet (DATF-Net),which integrate dual-attention collaboration and multi-branch joint loss.A cross-correlation feature enhancement strategy is adopted,and a channel-spatial attention collaboration mechanism is introduced to achieve comprehensive fusion of complementary multimodal features.The consistency of different branch decisions is ensured under the joint constraint of dice loss and cross-entropy loss.Ablation and comparative experiments are conducted on the Dongying multi-modal image dataset.The results demonstrate that the dual-attention collaboration mechanism and multi-branch joint loss function both contribute to improving land cover classification accuracy.Compared with other methods,DATF-Net achieves optimal performance in precision across various land cover categories and multiple overall classification metrics.Notably,The OA and FWIoU of DATF-Net outperform the second-best method (VFesuNet) by 7.9% and 12.66% respectively.The proposed method effectively mitigate the speckle noise interference in SAR images,enhance the coherence of classification boundaries,and improve classification accuracy and robustness in complex urban scenarios.

Spatio-temporal trends and driving mechanisms of vegetation NPP in Shaanxi province from 2000 to 2023

CHANG Dee, WEI Haixia, CHEN Liyan

2025, 0(11):  91-98.  doi:10.13474/j.cnki.11-2246.2025.1114

Asbtract ( )   PDF (7827KB) ( )  

References | Related Articles | Metrics

Quantitatively analyzing the spatio-temporal variations and driving mechanisms of vegetation net primary productivity (NPP) in Shaanxi province based on long-term remote sensing data is crucial for assessing regional ecosystem stability and elucidating carbon cycle dynamics.Utilizing MODIS NPP products and multi-source data from 2000 to 2023,this study integrated Sen's trend analysis,the Hurst index,partial correlation analysis,residual analysis,and the geodetector method to analyze the spatio-temporal dynamics and driving factors of vegetation NPP in Shaanxi province at the pixel scale.NPP showed a significant increasing trend from 2000 to 2023,with a growth rate of 8.19 gC·m^-2·a^-1.The spatial distribution of NPP showed higher values in the south and lower values in the north; 97.63% of the area exhibited increasing trends.The Hurst index indicated that the improving trend would persist in 99.15% of the region.The spatial heterogeneity of precipitation and temperature impacts was significant (positive synergy in Northern Shaanxi,temperature dominance in Guanzhong/Southern Shaanxi). Human activities promoted vegetation restoration in 76.26%of the area.Evapotranspiration (ET),precipitation,and landform type are the dominant driving factors.Among their interactions,ET∩precipitation and ET ∩ landform exhibite the strongest explanatory power.The strong interaction effects of land use/population density ∩ ET (q>83%) and∩precipitation (q>74%) indicate that changes in vegetation NPP result from the deep coupling of natural and socioeconomic factors.

Robot dense RGB-D SLAM algorithm based on 3D Gaussian primitive scene representation

ZHANG Guo, WEI Ling, HE Shengxi

2025, 0(11):  99-103.  doi:10.13474/j.cnki.11-2246.2025.1115

Asbtract ( )   PDF (1225KB) ( )  

References | Related Articles | Metrics

Dense simultaneous localization and mapping is a crucial technology in robotics.Recent research on 3D Gaussian splatting has demonstrated that high-quality scene reconstruction and real-time rendering can be achieved using multiple cameras in different poses.Against this backdrop,this paper introduces 3D Gaussian splatting into SLAM,representing the scene using 3D Gaussian primitives and implementing a dense visual SLAM algorithm using RGB-D cameras.This algorithm overcomes the limitations of previous radiance field-based representations,particularly in terms of fast rendering and optimization,the ability to recognize previously mapped regions,and structured map expansion by adding more Gaussians.Extensive experimental results demonstrate that the proposed dense RGB-D SLAM algorithm improves performance by up to 2 times compared to existing methods in terms of camera pose estimation,map construction,and new view synthesis.

Dynamic water level responses and drought-flood risk assessment at Yangtze River Datong station from GNSS-IR

CHEN Zilong, JIN Shuanggen

2025, 0(11):  104-109.  doi:10.13474/j.cnki.11-2246.2025.1116

Asbtract ( )   HTML ( )   PDF (2093KB) ( )  

References | Related Articles | Metrics

Affected by the global climate change and human activities,the dynamics of river water levels have become increasingly complex and varied.Drought and flood disasters pose significant threats to the ecological safety and socio-economic development of river basins.Nowadays,GNSS interferometric reflectometry (GNSS-IR) provide a unique means to invert water levels and assess drought-flood risk.This study uses ground-based GNSS multi-path reflectometry data to invert water level data at the Yangtze River Datong station from August 2023 to July 2024,and analyzes water level dynamic response characteristics and drought-flood risk assessment in conjunction with hydrological and meteorological factors.The results show that during the flood season,water level dynamics are primarily driven by rainfall,exhibiting trends similar to changes in runoff,indicating that heavy rainfall events directly elevate water levels.Long-term water level changes exhibit clear seasonal and dynamic response patterns,with correlation coefficients of 0.73,0.74,and 0.89 for temperature,precipitation,and runoff,respectively.Additionally,based on the standardized water level index (SWI),eight drought and flood events of varying risk levels were identified during the study period.Three of these events reached moderate drought/flood risk,all concentrated in the summer of 2023,with daily average water level changes exceeding 0.1 m/d.The analysis of water level dynamics and the assessment of drought-flood risk demonstrate the performance and application potential of GNSS-IR technology in river water level monitoring.

Automatic monitoring and early warning system and application of geological disaster in mountainous expressway of western Guangdong

XU Hua, CHEN Longwang, HAN Fuqing, WANG Weili, WANG Guozhi

2025, 0(11):  110-117,163.  doi:10.13474/j.cnki.11-2246.2025.1117

Asbtract ( )   PDF (10493KB) ( )  

References | Related Articles | Metrics

As expressways extend into mountainous regions,a large number of high and steep slopes are formed through excavation.This has led to frequent geological disasters such as collapses,landslides,and debris flows,posing severe threats to highway safety.To enable dynamic early warning through multi-source monitoring of geological hazards on expressway slopes in the Western Guangdong mountainous areas and ensure safe highway operation,a warning indicator system is established.This system addresses the complex geological conditions and concentrated typhoon and rainstorms characteristic of the Western Guangdong mountainous areas,focusing on slope displacement,displacement rate,structural stress,and rainfall as monitoring parameters.Based on the deformation characteristics of weathered granite residual soil slopes,a four-level early warning system (blue,yellow,orange,and red) and corresponding thresholds are proposed.Integrating Beidou surface deformation monitoring,deep horizontal displacement monitoring,anchor cable stress monitoring,distributed optical fiber monitoring,and rainfall monitoring,an automated monitoring and early warning system for highway slopes is developed.Analysis of typical slope monitoring data reveales that during the monitoring period,the extreme values for displacement and rate from Beidou surface deformation monitoring are -99 mm and 7.40 mm/d,respectively.For deep horizontal displacement monitoring,the extreme values for displacement and rate are 55.386 mm and 3.83 mm/d,respectively.The single-day extreme rainfall is 458.1 mm.All these parameters reach the red warning level,and warnings are successfully issued.The implementation of corresponding measures preventes further slope deformation.This demonstrates that the developed system can timely reflect slope deformation characteristics and development stages,quickly and accurately identify potential geological hazards,achieve multi-level and multi-dimensional early warning for highways,and effectively improve the accuracy of slope geological disaster warnings.This system can serve as a valuable reference for geological disaster monitoring and early warning of highway slopes in China.

Inspection of hole defects in underwater anti-ship steel boxed cofferdam of water crossing bridges

LIU Chengcai, XIE Xiaowang, HU Jian, ZHANG Yiqing, YAN Jing, ZHU Yanjie

2025, 0(11):  118-123.  doi:10.13474/j.cnki.11-2246.2025.1118

Asbtract ( )   PDF (7419KB) ( )  

References | Related Articles | Metrics

Due to the poor accessibility in deep and turbid water environments,traditional underwater inspection methods (e.g.,manual diving and underwater photography) struggle to detect corrosion-induced hole defects in underwater steel structures.To advance the development of bridge underwater inspection technology,this study proposes an automated detection method for corrosion voids in underwater anti-collision steel cofferdams based on 3D sonar point cloud modeling.The proposed method integrates second-nearest-neighbor spacing statistical features with an Alpha Shape algorithm to construct an adaptive Alpha Shape-based edge detection model for point clouds.Subsequently,a polygon decomposition technique is applied to segment individual voids from the identified edge point clouds,thereby achieving automated recognition and geometric quantification of corrosion voids in underwater steel cofferdams.Experimental validation through underwater measurements demonstrates that the proposed method achieves an average accuracy of 76.2% in hole defect assessment.Furthermore,the method is successfully applied to inspect thin-walled steel plates on the main pier of a Yangtze River bridge,detecting a total void area of 0.542 m².This research provides a novel technical framework and methodological reference for the digital and intelligent inspection of underwater infrastructure.

Real-time localization and high-fidelity mapping algorithm for unmanned vehicles driven by 3D Gaussian splatting technology

LU Zhiqiang, PANG Qingkai, WEI Jian

2025, 0(11):  124-128.  doi:10.13474/j.cnki.11-2246.2025.1119

Asbtract ( )   PDF (1392KB) ( )  

References | Related Articles | Metrics

To address the high computational cost and low efficiency of combining neural rendering with SLAM for camera localization and high-fidelity reconstruction,this paper proposes a SLAM algorithm for unmanned vehicles based on 3D Gaussian splatting(3D GS). This algorithm improves localization efficiency through explicit geometric representation and integrates implicit representations to learn illumination and texture to achieve high-precision reconstruction of complex scenes.This method employs a multi-level training strategy based on a Gaussian pyramid to enhance the ability to capture multi-scale details.Experimental results demonstrate excellent performance on multiple datasets,with a 30% improvement in PSNR on the Replica dataset,validating its efficiency and reconstruction quality.

Research on carbon sink effectiveness monitoring in ecological protection and restoration: a case study of the integrated protection and restoration project in the Dongting Lake region

QUAN Sixiang, ZHANG Tai, ZHANG Ya, YIN Ziqiang, WANG Zhenxiang

2025, 0(11):  129-133,145.  doi:10.13474/j.cnki.11-2246.2025.1120

Asbtract ( )   PDF (2176KB) ( )  

References | Related Articles | Metrics

To scientifically evaluate the implementation effects of ecological protection and restoration projects,it is crucial to study monitoring techniques for the carbon sink effectiveness of ecological protection and restoration.This paper takes the integrated protection and restoration project in the Dongting Lake region as the research object,and constructs a “space-air-ground” integrated carbon sink monitoring method driven by remote sensing data.It systematically assesses the vegetation carbon storage and soil carbon storage in 2023 and 2024,as well as the carbon sink capacity before and after the restoration of typical sample plots,and analyzes the spatial distribution characteristics.The results show that:①The vegetation carbon storage increased from 2.332 4 million tons to 6.881 4 million tons,and the soil carbon storage increased from 9.407 8 million tons to 13.166 1 million tons.②The vegetation carbon storage and soil carbon storage show a spatial distribution pattern of being higher in the west and lower in the east.③The annual carbon sink capacity of the typical sample plot increased from 669.78 kg CO₂ to 1 231.36 kg CO₂ before and after restoration.This paper can provide technical reference for the carbon sink effect monitoring of large-scale ecological restoration projects and provide scientific basis for the ecological protection and restoration of the Yangtze River Economic Belt.

Research and practice on safety compliance technology for online update of autonomous driving maps

MA Xiaolong, MA Zhaoting, ZHAO Yuanchun, FANG Chiyu, FEI Wenkai, YAN Chunli, WANG Yueming, SHI Yihui

2025, 0(11):  134-139.  doi:10.13474/j.cnki.11-2246.2025.1121

Asbtract ( )   HTML ( )   PDF (2970KB) ( )  

References | Related Articles | Metrics

With the continuous expansion of the application fields of autonomous driving maps,the business models and management mechanisms for their data collection,transmission,update,and distribution are facing new transformations and challenges.On the premise of ensuring the security of geographic information,how to achieve efficient collaboration between crowdsourced collection and online update of autonomous driving map data has become a key problem that needs to be urgently solved in this field under the new situation.This paper analyzes the security and compliance requirements of crowdsourced collection and online update of map data,as well as the current state of research in this area.It proposes a safety and compliance technology for online update of autonomous driving maps that integrates the confidential processing of geographic information and commercial cryptographic protection.Through the exploration and practice of national-level scientific research project,it is verified that this technology can meet the performance requirements of related businesses of crowdsourced collection and online update,and effectively improve the security protection ability of autonomous driving maps and related geographic information data.

DEM-aided pure integer programming phase unwrapping algorithm for dual baseline InSAR

HUO Hui, LI Geshuang, MIAO Changwei, KONG Lingpeng

2025, 0(11):  140-145.  doi:10.13474/j.cnki.11-2246.2025.1122

Asbtract ( )   PDF (5515KB) ( )  

References | Related Articles | Metrics

To solve the phase unwrapping (PU) problem in areas with spectral aliasing and abrupt topographic changes,the problem of solving the integer cycle number of dual-baseline InSAR phase unwrapping is transformed into a pure integer programming (PIP) problem,and a DEM-assisted pure integer programming phase unwrapping algorithm is proposed.Firstly,a PIP model with the intercept on the vertical axis as the objective function and a ray as the constraint condition is constructed.Then,the branch and bound algorithm of operations research theory is used to solve the optimal integer solution of ambiguity numbers.Finally,on the premise that the true misunwrapping point is extracted by DEM,and the PU is completed by replacing the ambiguity number with the highest frequency in the square window.The feasibility,effectiveness and universality of the proposed algorithm are proved through the comparative experiments of the simulation data and real data with the branch-cut method,the minimum cost flow method,the Chinese remainder theorem algorithm and the clustering analysis algorithm.The algorithm has good unwrapping capability in phase under-sampling areas and terrain mutation area,and weakens the requirement for the mutual prime of the interferogram baselines.

Automatic calculation method of material source volume for debris flow channels

XU Yingjie, DONG Xiujun, DENG Bo, QIU Chengyue

2025, 0(11):  146-153.  doi:10.13474/j.cnki.11-2246.2025.1123

Asbtract ( )   PDF (14012KB) ( )  

References | Related Articles | Metrics

The calculation of debris flow material source volume is crucial for disaster risk assessment and the formulation of emergency response measures.Traditional volume measurement methods often rely on ground surveys or simple remote sensing analyses,which are time-consuming and lack sufficient accuracy,making them unsuitable for rapid response demands.This study proposes an automatic calculation method for debris flow source volume based on high-precision airborne LiDAR data.Firstly,digital elevation models and orthophotos are imported into Earth Survey software to obtain the basic vector files of debris flow channels.Secondly,B-spline curves of the material sources in debris flow channels are constructed using these files,and the predicted basal surface of the debris flow material sources is generated by fitting the basal surface.Finally,the source volume of the debris flow channels is calculated using calculus methods and compared with two-phase DEM differencing data to validate the accuracy.Using Youyi village in Luding county,Sichuan province as the study area,results show that for the 21 interpreted debris flow channels,9 are selected for validation,achieving an average accuracy of over 80%.This method provides robust support for disaster risk management and the development of emergency response plans.

Hole detection and filling in road surface point clouds from vehicle-mounted LiDAR

CHEN Jianping, SHI Jian, DAI Xiangxi, ZHANG Qinyu, HAN Wenquan

2025, 0(11):  154-158.  doi:10.13474/j.cnki.11-2246.2025.1124

Asbtract ( )   PDF (3143KB) ( )  

References | Related Articles | Metrics

Vehicle-mounted LiDAR scans of urban roads often contain holes in the road surface point cloud due to occlusions by vehicles,pedestrians,and roadside objects.These gaps degrade the completeness and accuracy of downstream 3D modeling and spatial analysis.We propose an automated pipeline for hole detection and repair in road surface point clouds.Firstly,noise and non-road points are removed via filtering and clustering techniques to isolate the road surface.Next,a multi-scale Alpha Shape algorithm extracts the 2D road boundary and identifies boundary breaks caused by occlusion; NURBS curves then restore a continuous boundary.Finally,the area within the repaired boundary is partitioned into a regular grid,empty cells are clustered to locate holes,and a quadratic surface is fitted to surrounding points to interpolate and fill each hole.Experiments on ten diverse urban road segments demonstrate that our method achieves 96.5% hole-filling success for common occluders,produces smooth,seamless transitions with the original data,and preserves critical geometric features such as curb and corner shapes.The proposed approach reliably repairs small-scale occlusions in road surface point clouds and is broadly applicable to most urban driving scenarios.

Rapid measurement of track alignment based on inertial navigation with multi-source information constraints

WANG Xinming, FANG Bole, LIU Jianjian

2025, 0(11):  159-163.  doi:10.13474/j.cnki.11-2246.2025.1125

Asbtract ( )   PDF (2120KB) ( )  

References | Related Articles | Metrics

In response to the problems of low efficiency,dependence on CPⅢ control network,poor environmental adaptability,and error divergence in traditional track measurement technology,this paper proposes a rapid measurement of track alignment based on inertial navigation with multi-source information constraints.Integrating multiple sensors such as BeiDou,INS,gauge sensors,odometry,etc.,a time synchronization mechanism for each sensor is designed.Data acquisition software is developed.BeiDou fixed point coordinate fitting,track alignment information constraint,and 3D reconstruction algorithm are constructed.Multi source information fusion and inertial navigation error suppression are achieved.Tests have shown that the absolute measurement accuracy of the track can reach 10 mm,the relative measurement accuracy of short wave is better than 0.3 mm,and the long wave is better than 3 mm.This system breaks through the bottleneck of traditional technology,meets the requirements of precise track measurement,and provides a new solution for the intelligent development of track measurement.

Prediction of settlement interval of high-speed railway subgrade based on composite neural network GRNN-BP

LI Dewei, ZHANG Sheng, SUN Tong, HE Quanpeng

2025, 0(11):  164-169.  doi:10.13474/j.cnki.11-2246.2025.1126

Asbtract ( )   PDF (2546KB) ( )  

References | Related Articles | Metrics

To solve the problem of low prediction accuracy caused by lack of subgrade settlement data,a prediction model of settlement interval of high-speed railway subgrade combined with PS-InSAR technology,generative adversarial network(GAN) data expansion and GRNN-BP composite neural network is proposed.Firstly,PS-InSAR technology is used to obtain the settlement value of the subgrade,analyze the correlation between environmental factors and the settlement of the subgrade,and build the original sample set.Secondly,the network connection layer is set up to connect the two kinds of networks,and the advantages of GRNN and BP neural network are used to form a composite neural network.Finally,GAN is used to expand the data set,and the expanded data set is input into GRNN-BP to predict the settlement of high-speed railway subgrade.The experimental results show that the prediction accuracy of the model can be improved effectively by inputting the expanded data samples into the composite neural network and training it.GRNN-BP can not only provide high-precision point prediction results,but also construct clear and reliable prediction intervals.Compared with the other four models,GRNN-BP has more reliable prediction results for the settlement of high-speed railway subgrade.

Analysis of changes in ecosystem service value in Guangzhou using GEE land use classification

ZHAO Xiaoyang, QIU Yongkang, LIU Yang, DUAN Peng, FU Leyi

2025, 0(11):  170-175.  doi:10.13474/j.cnki.11-2246.2025.1127

Asbtract ( )   PDF (6381KB) ( )  

References | Related Articles | Metrics

Quantifying the ecosystem service value (ESV) of different land use types can provide important basis for urban ecological resource protection and national spatial planning.This article takes Guangzhou city as the research area,uses the GEE platform remote sensing dataset and random forest algorithm for land use classification,combines the equivalent factor method to calculate the spatial distribution of ESV in the study area,and evaluates the impact of human activity intensity on ESV changes.The results show that: ①From 2014 to 2024,the growth rate of construction land area has reached 5.93%.The increased area was mainly converted from cultivated land,unused land,and grassland,while the land area of other land types showed a decreasing trend except for forest land.② The ESV in Guangzhou mainly comes from forests and water bodies,accounting for over 95% of the total.From 2014 to 2024,the total ESV in the study area decreased by about 112 million yuan,and the areas where ESV decreased were mainly located around the built-up areas of Guangzhou,mainly due to urban expansion.③ The spatial distribution of human activity intensity in Guangzhou city shows a characteristic of low intensity in the north and high intensity in the middle and west,and there is a negative correlation between human activity intensity and ESV.With the increase of human activity,ESV shows a downward trend.

3D Gaussian splatter modeling technology for multi-source surveying and mapping geographic information data fusion

XU Zhengbing, FENG Guozheng, ZOU Yang, YE Fei

2025, 0(11):  176-180.  doi:10.13474/j.cnki.11-2246.2025.1128

Asbtract ( )   PDF (3841KB) ( )  

References | Related Articles | Metrics

To address the issues faced by existing 3D modeling methods,such as limited modeling parameters,low efficiency,high labor costs,and lack of details.Firstly,by integrating oblique photography 3D models,refined laser point clouds,and geographic information semantic technology,we break through the bottleneck of data fusion under complex occlusions (vegetation,buildings),providing multi-type,full-coverage elements and reliable data sources such as textures,geometry,and point clouds for modeling.Then,the 3D Gaussian splash (3DGS) modeling method is applied,utilizing anisotropic Gaussian ellipsoids to construct scenes,combined with optimization and fast rasterization techniques.And high-quality and efficient reconstruction of full-element real-world models under large elevation differences is achieved,with model accuracy reaching the centimeter level.The research shows that the 3DGS method based on multi-source surveying and mapping geographic information fusion can effectively solve the 3D modeling challenges in complex environments with high occlusions and large elevation differences,providing a new paradigm for high-quality and efficient 3D reconstruction.

Application of BDS interferometric reflectometry in river water level monitoring

HUANG Xiaoming, ZHU Nan, FEI Xinlong, CHENG Yongjuan, CHEN Xugang, ZHANG Zhixuan, PAN Ming

2025, 0(11):  181-184.  doi:10.13474/j.cnki.11-2246.2025.1129

Asbtract ( )   PDF (3161KB) ( )  

References | Related Articles | Metrics

This study investigates the applicability of BeiDou Navigation Satellite System Interferometric Reflectometry (BDS-IR) technology for river water level monitoring,aiming to address the limitations of traditional methods,such as insufficient spatio-temporal coverage and high construction and operational costs.A dual-frequency signal (B1I/B2I) interference physical model is established,integrated with a multi-path effect characteristic parameter inversion algorithm.Dynamic water level monitoring experiments are conducted in typical reaches of the middle Yangtze River,and the results are validated against data from hydrometric stations.The validation shows that the root mean square error (RMSE) between the BDS-IR derived water levels and the reference data is 1.13 cm,with a correlation coefficient of 0.97,achieving centimeter-level inversion accuracy.The findings demonstrate that BDS-IR technology offers a high-precision and low-cost technical solution for watershed flood control,disaster mitigation,and intelligent water resource management.