Top Read Articles

    Published in last 1 year |  In last 2 years |  In last 3 years |  All
    Please wait a minute...
    For Selected: Toggle Thumbnails
    Multi-modal remote sensing large foundation models: current research status and future prospect
    Yongjun ZHANG, Yansheng LI, Bo DANG, Kang WU, Xin GUO, Jian WANG, Jingdong CHEN, Ming YANG
    Acta Geodaetica et Cartographica Sinica    2024, 53 (10): 1942-1954.   DOI: 10.11947/j.AGCS.2024.20240019.
    Abstract1327)   HTML134)    PDF(pc) (10357KB)(1226)       Save

    The increasing remote sensing capabilities for Earth observation have eased the access to abundant data and enabled the emergence and development of remote sensing foundation models (RSFMs). Designing distinct deep neural networks and optimizing for different data and task types require substantial development efforts and prohibitively high computational resources. In order to address these issues, researchers in the remote sensing field have shifted their focus to the study of RSFMs and presented many dedicated designed unified models. To enhance the generalizability and interpretability of RSFMs, the integration of extensive geographic knowledge has been recognized as a pivotal/key approach. While existing works have explored or incorporated geographic knowledge into the architecture design or pre-training methods of RSFMs, there lacks of a comprehensive survey to review the current status of geographic knowledge-guided RSFMs. Therefore, this paper starts with summarizing and categorizing large-scale pre-training datasets and then provides an overview of the research progress in this field. Subsequently, we introduce intelligent interpretation algorithms for remote sensing imagery guided by geographic knowledge, along with advancements in the exploration and utilization of geographic knowledge specifically tailored for RSFMs. Finally, several future research prospects are outlined to tackle the persisting challenges in this field, aiming to shed light on future investigations into RSFMs.

    Table and Figures | Reference | Related Articles | Metrics
    Research progress and trend of intelligent remote sensing large model
    Qin YAN, Haiyan GU, Yi YANG, Haitao LI, Hengtong SHEN, Shiqi LIU
    Acta Geodaetica et Cartographica Sinica    2024, 53 (10): 1967-1980.   DOI: 10.11947/j.AGCS.2024.20240053.
    Abstract884)   HTML84)    PDF(pc) (14566KB)(764)       Save

    AI large models, with their advantages in generalization, universality, and high accuracy, have become the cornerstone of various AI applications such as computer vision, natural language processing. Based on the analysis of the development process, value, and challenges of AI large models, this article first discusses the research progress of remote sensing large models from three perspectives: data, model, and downstream tasks. At the data level, there is a transition from single modality to multi-modality; at the model level, there is a shift from small models to large models; and at the downstream task level, there is a development from single-task to multi-task. Next, the article explores three key development directions for remote sensing large models: multi-modal remote sensing large models, interpretable remote sensing large models, and reinforcement learning from human feedback(RLHF). Furthermore, it realizes a construction approach for remote sensing large models, namely “construction of unlabeled dataset-self-supervised model learning-downstream transfer application”. Technical experiments have been conducted to validate the significant advantages of remote sensing large models. Finally, the article concludes and provides prospects, emphasizing the need to focus on application tasks and combine theoretical methods, engineering technology, and iterative applications to achieve low-cost training, efficient and fast inference, lightweight deployment, and engineering-based applications for remote sensing large models.

    Table and Figures | Reference | Related Articles | Metrics
    Large language model-driven GIS analysis: methods, applications, and prospects
    Huayi WU, Zhangxiao SHEN, Shuyang HOU, Jianyuan LIANG, Anqi ZHAO, Haoyue JIAO, Zhipeng GUI, Xuefeng GUAN
    Acta Geodaetica et Cartographica Sinica    2025, 54 (4): 621-635.   DOI: 10.11947/j.AGCS.2025.20240468
    Abstract680)   HTML402)    PDF(pc) (3633KB)(1035)       Save

    The rapid development of large language models (LLMs) provide a new approach for GIS analysis, leading to the large language model-driven GIS analysis technical architecture (LLM4GIS). Based on the latest research up to October 2024, this paper reviews the evolution of GIS analysis and summarizes the LLM4GIS technical architecture from 3 aspects: application modes, datasets and evaluation methods. It also summarizes the research progress of LLM in GIS analysis tasks such as knowledge question-answering, knowledge extraction, spatiotemporal reasoning, and analyzing and modeling. Finally, the paper prospects the future research directions of GIS4LLM in 5 aspects: collaborative understanding of multimodal spatio-temporal data, balancing generalization with depth, enhancing interpretability and credibility, transitioning to embodied intelligence and edge intelligence, and the development of intelligent and universal GIS analysis. This paper provides inspiration for achieving mutual empowerment between LLM4GIS and GIS4LLM.

    Table and Figures | Reference | Related Articles | Metrics
    Autonomous situatedness map representation: a theoretical discussion on intelligent cartography in the era of large models
    Zhilin LI, Zhu XU, Li SHEN, Jingzhong LI, Tian LAN, Jicheng WANG, Tingting ZHAO, Tinghua AI, Peng TI, Wanzeng LIU, Jun CHEN
    Acta Geodaetica et Cartographica Sinica    2024, 53 (11): 2043-2052.   DOI: 10.11947/j. AGCS.2024.20240222.
    Abstract538)   HTML83)    PDF(pc) (2933KB)(672)       Save

    Making mapping system automatically conducting map design and production through intelligent techniques has always been the goal pursued by the cartographic community and the frontier research direction of the International Cartographic Association. Since the 1980s, artificial intelligence has been applied in cartography, gradually solving the automation problems of some processes and improving the production efficiency of map making. However, the level of automation in key steps such as map design is still extremely low, which cannot meet the “customized” and “ubiquitous” mapping demand in the information age. Fortunately, since 2023, artificial intelligence technology represented by large language models such as GPT-4 and Gemini has made breakthroughs and achieved “quasi-general artificial intelligence”, which shows strong language comprehension, reasoning and expression ability. This paper explores the use of large models to improve the intelligence level of map making systems, aiming to establish a new generation of intelligent mapping theory and method system. This paper first analyzes the bottleneck problems of the existing digital mapping system and points out the necessity of establishing a new generation of intelligent mapping technology; then it analyzes the nature and capabilities of large models and demonstrates the sufficiency of establishing such a new generation; then it further analyzes the possibility and methods of combining them, proposes an intelligent mapping framework in the era of large models (e.g. situatedness map representation); finally, it discusses the key technical issues of situatedness map representation: “autonomous consciousness of mapping context”, “autonomous design and production of maps” and “autonomous human-computer interaction in situatedness ”.

    Table and Figures | Reference | Related Articles | Metrics
    The transformation of the scientific concept of GIS: from Map-based GIS to Space-oriented GIS
    Renzhong GUO, Yebin CHEN, Zhigang ZHAO, Ding MA, Biao HE, Weixi WANG, Wuyang HONG, Minmin LI
    Acta Geodaetica et Cartographica Sinica    2024, 53 (10): 1853-1862.   DOI: 10.11947/j.AGCS.2024.20240152.
    Abstract488)   HTML91)    PDF(pc) (12116KB)(571)       Save

    From the 1960s to the present, GIS has undergone a development journey of over 60 years, during which its connotations have evolved from geographic information system (GISystem) to geographic information science (GIScience), and further to geographic information service (GIService). During this period, GIS research has primarily been based on the two-dimensional abstract expression logic of cartography (Map-based GIS), achieving abstract analysis and representation of the real world. However, with the continuous emergence of new technologies and new demands such as 3D real scene, digital twins, and city information modeling (CIM), the original two-dimensional logic of GIS is facing challenges in the collection, processing, and fusion of multi-source heterogeneous spatiotemporal big data, the representation of complex spatiotemporal dynamic processes, and the mining of potential spatiotemporal patterns. How to adjust the scientific positioning of GIS to adapt to the multi-type, multi-level, and multi-role needs of spatial object expression and analysis in the digital society has become an important issue that GIS development urgently needs to consider. From a methodological perspective, we deeply analyzes the bottlenecks of Map-based GIS in spatial representation, spatial analysis, and comprehensive application. Furthermore, based on the development needs of GIS in the new era, we propose a scientific concept transformation model from Map-based GIS to Space-oriented GIS, integrating the logical thinking of Map-based GIS from the perspectives of theoretical foundations, management models, visualization forms, and functional positioning, and innovative applications under the transformation of GIS scientific concepts. This research aims to provide reference ideas for the development of GIS in the new era.

    Table and Figures | Reference | Related Articles | Metrics
    Global registration method for multi-station point clouds based on the bundle adjustment method
    Qingzhou MAO, Mengxuan XIA, Qingquan LI, Jing ZHU, Tingli FAN
    Acta Geodaetica et Cartographica Sinica    2024, 53 (9): 1663-1670.   DOI: 10.11947/j.AGCS.2024.20240075
    Abstract482)   HTML87)    PDF(pc) (8299KB)(427)       Save

    In the extraction of parallelism and warping data for large-scale, high-density ice-making pipes, issues such as low detection accuracy and incomplete data coverage are prevalent. This paper proposes a method based on bundle adjustment for regional networks, utilizing a 3D laser acquisition approach with multi-prism target spheres. The target sphere centers are extracted using a radius-constrained random sample consensus (RANSAC) sphere fitting method. By applying coordinate transformations between the absolute coordinates of the station point cloud origins, the relative coordinates of the target centers, and the absolute coordinates of the target centers, a global solution for the positions and orientations of all stations is achieved using bundle adjustment. The method was validated using scan data from the National Speed Skating Oval. The results show that after point cloud matching, the internal consistency accuracy is 2.6 mm, and the external consistency accuracy is 1.9 mm, demonstrating higher acquisition accuracy compared to existing methods.

    Table and Figures | Reference | Related Articles | Metrics
    Positioning performance analysis and evaluation for standalone BDS receivers
    Chuang SHI, Chenlong DENG, Lei FAN, Fu ZHENG, Tao ZHANG, Yuan TIAN, Guifei JING, Jie MA
    Acta Geodaetica et Cartographica Sinica    2025, 54 (1): 1-13.   DOI: 10.11947/j.AGCS.2025.20240127
    Abstract467)   HTML56)    PDF(pc) (5525KB)(441)       Save

    China's BeiDou navigation satellite system (BDS) has completed its global constellation establishment and began to provide positioning, navigation, and timing (PNT) services to global users. Based on the early principle of multi-system compatibility and interoperability, in the current market all the mainstream GNSS receivers support multi-system satellite signal reception. In order to improve the autonomy and security of the BDS PNT services, the government departments have issued opinions on accelerating the research, development and promotion, utilization of homemade standalone BDS positioning terminals. Since the standalone BDS receiver can no longer rely on the guidance of other system's signals during signal acquisition, its hardware and positioning performance may be changed, thus it is urgent to evaluate the navigation and positioning performance of the homemade standalone BDS receiver. In this paper, the M300 Pro standalone BDS receiver is selected to carry out a series of test and evaluation experiments, and the hardware performance of the receiver such as time to first fix (TTFF), signal quality and observation noise are evaluated first. Then the positioning performance such as station coordinate estimation, single point positioning (SPP), precise point positioning (PPP), static baseline solution and real-time kinematic (RTK) positioning are analyzed and discussed by using the self-developed BDS precise data processing software platform named space Geodetic spatio-temporal data analysis and research software (GSTAR). The experimental results show that the cold TTFF of the selected standalone BDS receiver is lower than 40 s, the average ratio of the intact observation data is more than 95%, and standard deviations of pseudorange and carrier phase measurement noise are 0.051 7 m and 0.003 4 cycles, respectively, which is basically consistent with the hardware performance of multi-GNSS receivers at home and abroad. Using the selected standalone BDS receiver, the single-day solution precision of the horizontal directions of station coordinates is 3.5 mm and the up direction is 9.9 mm; the precision of 2.208 m in horizon and 2.502 m in vertical can be realized with single-epoch pseudorange SPP; the precision of horizontal directions of kinematic PPP with ambiguity resolution (PPP-AR) is better than 3 cm and the up direction is better than 5 cm; the convergence time of the PPP-AR is better than 27 min; the repeatability accuracy of single-day solution for baselines shorter than 20 km is better than 0.7 cm in the horizontal direction and 1.8 cm in the vertical direction, and the RTK positioning accuracy for short baselines will not exceed 3 cm in the horizontal direction and 5 cm in the vertical direction. The homemade standalone BDS receiver has initially possessed the ability to independently provide reliable high-precision positioning services.

    Table and Figures | Reference | Related Articles | Metrics
    Road extraction networks fusing multiscale and edge features
    Genyun SUN, Chao SUN, Aizhu ZHANG
    Acta Geodaetica et Cartographica Sinica    2024, 53 (12): 2233-2243.   DOI: 10.11947/j.AGCS.2024.20230291
    Abstract438)   HTML79)    PDF(pc) (6974KB)(391)       Save

    Extracting roads using remote sensing images is of great significance to urban development. However, due to factors such as variable scale of roads and easy to be obscured, it leads to problems such as road miss detection and incomplete edges. To address the above problems, this paper proposes a network (MeD-Net) for road extraction from remote sensing images integrating multi-scale features and focusing on edge detail features. MeD-Net consists of two parts: road segmentation and edge extraction. The road segmentation network uses multi-scale deep feature processing (MDFP) module to extract multi-scale features taking into account global and local information, and is trained using group normalization optimization model after convolution. The edge extraction network uses detail-guided fusion algorithms to enhance the detail information of deep edge features and uses attention mechanisms for feature fusion. To verify the algorithm performance, this paper conducts experiments using the Massachusetts road dataset and the GF-2 road dataset in Qingdao area. The experiments show that MeD-Net achieves the highest accuracy in both datasets in terms of intersection-over-union ratio and F1 value, and is able to extract roads at different scales and maintain road edges more completely.

    Table and Figures | Reference | Related Articles | Metrics
    3D Gaussian radiation field modeling for real-scene bridges
    Wei MA, Qiang TU, Jianping PAN, Lidu ZHAO, Wei TU, Qingquan LI
    Acta Geodaetica et Cartographica Sinica    2024, 53 (9): 1694-1705.   DOI: 10.11947/j.AGCS.2024.20240071
    Abstract423)   HTML58)    PDF(pc) (25350KB)(447)       Save

    Realistic 3D modeling and digital twins have become essential foundations for bridge operation and management. However, given the complex geometric structures of bridges, current 3D modeling methods face issues such as large amounts of raw data collection, low modeling efficiency, and missing or deformed model details. In response to these challenges, this paper investigates a bridge realistic 3D reconstruction method based on 3D Gaussian radiance fields. This method utilizes 3D Gaussian functions to construct a Gaussian radiance field from sparse point clouds generated by captured images. Adaptive optimization of radiance field parameters is performed based on stochastic gradient descent, and real-time visualization of the 3D model is achieved through differentiable rasterization, resulting in high-quality bridge 3D reconstruction and rendering. The study explores the impact of different image resolutions and various parameter changes on bridge modeling. Comparisons with traditional methods are made to provide theoretical and technical support for further bridge applications, promoting efficient and accurate realistic 3D reconstruction of complex bridge structures.

    Table and Figures | Reference | Related Articles | Metrics
    Six geographic application paradigms of big data
    Lun WU, Yuanqiao HOU, Yu LIU
    Acta Geodaetica et Cartographica Sinica    2024, 53 (8): 1465-1479.   DOI: 10.11947/j.AGCS.2024.20230199
    Abstract415)   HTML73)    PDF(pc) (16210KB)(499)       Save

    With the advent of the big data era, multi-source big data is on the rise, leading to the integration of data-driven research paradigms with geography. Geospatial big data based on individual behavior offers observations of massive individual behavior patterns, thereby achieving “from people to places” social perception and supporting various applications such as urban management, transportation, and public health. This article delineates six application paradigms focusing on geospatial big data from an application perspective, ranging from describing spatio-temporal distributions at a low level to optimizing spatial decision-making at a high level. The first direction involves a simple characterization of the spatio-temporal features of geographic phenomena and elements, while the second to fourth directions focus on exploring the rules and mechanisms behind spatio-temporal distribution characteristics. The last two directions provide support at the decision-making level. Furthermore, this article highlights issues in data acquisition, analysis methods, and application goals in big data applications.

    Table and Figures | Reference | Related Articles | Metrics
    Prospects for the development of deep space navigation technologies
    Gang LI, Fang XIE, Yifan WU, Jun LU, Shuren GUO, Yingchun LIU, Chengeng SU, Xiaoheng YANG
    Acta Geodaetica et Cartographica Sinica    2025, 54 (3): 397-409.   DOI: 10.11947/j.AGCS.2025.20230493
    Abstract411)   HTML73)    PDF(pc) (2825KB)(370)       Save

    Deep space navigation is a fundamental element ensuring the efficient implementation of all kinds of deep space missions. With the substantial increase in the number of cislunar and Mars missions, as well as the great extension in distance of interplanetary missions, the entire process of spacecraft cruise, flyby, orbiting, descent landing, and surface activities pose increasingly higher demands for navigation. The traditional methods, such as deep space network, are hard to fulfill the requirements for multi-task, real-time and autonomous navigation. The main characteristics of deep space missions are analyzed, such as increasing in number, focusing on Moon and Mars, transforming from exploration to development and extending in distance. The paper summarizes three typical navigation requirements for cislunar transfer space, the outer space and surface of planets (like Moon and Mars), and ultra-far deep space. Then the paper studies the characteristics and trends of currently in-use, available and potential technologies. Moreover, the suggestions for the development of deep space navigation technology are proposed, including continuously improving the performance of in-use technologies, actively promoting the experimental deployment of available technologies, and strengthening the fundamental research of potential technologies. This study can provide reference for the technical development in the field of deep space navigation.

    Table and Figures | Reference | Related Articles | Metrics
    Key technologies for spaceborne SAR payload of LuTan-1 satellite system
    Yunkai DENG, Yu WANG, Kaiyu LIU, Naiming OU, Dacheng LIU, Heng ZHANG, Jili WANG
    Acta Geodaetica et Cartographica Sinica    2024, 53 (10): 1881-1895.   DOI: 10.11947/j.AGCS.2024.20230263.
    Abstract400)   HTML29)    PDF(pc) (20110KB)(403)       Save

    LuTan-1 (referred as LT-1) is China's first civil synthetic aperture radar (SAR) satellite mission to monitor the ground deformation with high precision by differential interferometry technology. The LT-1A and LT-1B have been success-fully launched on January 26 and February 27, 2022, respectively. The data acquisition schedule of LT-1 mission is divided into two stages, which corresponding to two specific orbit configurations. In the first stage, two satellites fly in a compact formation to get the digital elevation model (DEM) using the bistatic InSAR strip mode. In the second stage, both satellites fly in a common reference orbit with 180° separation. The revisit time of the individual satellite is 8 days, and it can be reduced to 4 days with two satellites. LT-1 satellite constellation can stably obtain time series data, so that we can monitor the ground deformation with high precision. Moreover, the multi-mode polarimetric payload will be utilized to obtain single-pass multi-polarimetric InSAR and hybrid polarimetric SAR data for forestry, land resource surveys, disaster monitoring, etc. In this paper, the key technologies of the LT-1 SAR payload, including phase synchronization, ambiguity suppression and system calibration, are systematically described and analyzed.The maximum resolution of LT-1 is 3 m, and the maximum swath width is 400 km, respectively. The azimuth ambiguity-to-signal ratio (AASR) of the interference wave position is better than -20 dB.The performance is partially demonstrated by ground testing and on-orbit actual measurement data.

    Table and Figures | Reference | Related Articles | Metrics
    Analysis of heavy rainstorm in Beijing in 2023 based on GNSS observations
    Fei YANG, Yingying WANG, Zhicai LI, Boyao YU, Junli WU, Yunchang CAO, Shu ZHANG
    Acta Geodaetica et Cartographica Sinica    2025, 54 (1): 14-25.   DOI: 10.11947/j.AGCS.2025.20230548
    Abstract389)   HTML89)    PDF(pc) (9638KB)(467)       Save

    By the end of July 2023, Beijing and its surrounding areas were severely impacted by an extreme rainstorm, which was the result of a combination of typhoon “Doksuri” “Khanun”, and geographic factors. The precipitable water vapor (PWV) is one of the key factors influencing rainfall, to explore its relationship with rainfall in different process of the rainstorm is of great significance for further establishment of a rainstorm warning model. In this study, 34 GNSS stations, 34 meteorological stations, 1 radiosonde station and ERA5 datasets in and around Beijing were utilized, the GNSS-PWV data with high accuracy from July 25th, 2023 to August 1st, 2023 were obtained using GAMIT 10.71. An improved interpolation algorithm has been proposed to retrieve gridded PWV data with a high spatiotemporal resolution. Then, the accuracy of the GNSS-PWV was evaluated from multiple perspectives using the radiosonde and ERA5 data as references. Finally, the relationship between the PWV variation and extreme rainfall and the relationship between tropospheric delay gradient and rainfall trend are analyzed from the perspective of time and space by the combination of the rainfall data from the meteorological stations. Results showed that the correlation coefficient between GNSS-PWV and RS-PWV was up to 0.99, the root mean square error (RMSE) and bias were about 0.52 mm and -0.52 mm, respectively. In the comparison with ERA5 data, the RMSE of GNSS-PWV is less than 6 mm and the bias range is -4~1.5 mm, the gridded PWV has a RMSE of about 4 mm and a bias about 1 mm. The spatiotemporal analysis shows that the PWV increases sharply before the occurrence of this rainstorm, keeps increasing during the rainstorm, and could not dissipate immediately after the end of the rainstorm. This phenomenon is related to the co-influence of “Doksuri” and “Khanun”. In addition, the tropospheric delay gradient at each station shows a consistent northeastward direction, which is consistent with the transport trend of PWV high value from southwest to northeast in space, and is consistent with the actual precipitation route.

    Table and Figures | Reference | Related Articles | Metrics
    Digital twin and spatio-temporal intelligence of geospatial information system
    Yuanxi YANG
    Acta Geodaetica et Cartographica Sinica    2025, 54 (2): 213-220.   DOI: 10.11947/j.AGCS.2025.20240515
    Abstract385)   HTML54)    PDF(pc) (1079KB)(436)       Save

    The digital twin system of geospatial information is an important support system for geospatial information service and an important foundation for the development of intelligent society. The digital twin system of geospatial information has more special requirements in terms of accuracy, systemization and reliability, compared to other industrial digital twin systems. This paper describes the basic rules of the establishment of the geospatial digital twin system from perception, description and mapping to statistics, prediction and deduction. It is emphasized that the perception of geographic entities should be accurate, the space-time reference should be consistent, the attribute description should be correct, the historical information should be dependable, the mapping relationship should be complete, the statistic trend should be systematic, the variation prediction should be rigorous, and the auxiliary decision making should be scientific. The related research topics of the geospatial digital twins are generally classified. The problems to be paid attention are listed. Finally, the relationship between the geospatial digital twin and spatio-temporal intelligence is discussed. The basic process and key technologies in the construction of geospatial digital intelligence system are pointed out.

    Table and Figures | Reference | Related Articles | Metrics
    Analysis of InSAR time-series deformation monitoring accuracy of domestic satellite
    Bing XU, Yan ZHU, Zhiwei LI, Huiwei YI, Miaowen HU, Qi CHEN, Kun HAN, Xun DU
    Acta Geodaetica et Cartographica Sinica    2024, 53 (10): 1930-1941.   DOI: 10.11947/j.AGCS.2024.20230572.
    Abstract380)   HTML33)    PDF(pc) (10153KB)(418)       Save

    The successful launch of the Lutan-1 satellite group (LT-1) has achieved the development of China's L-band interferometric SAR satellite from scratch. For obstacle avoidance, a small part of the spatial baselines of LT-1 satellite was long, but the length of the baseline has been controlled to within 400 meters after the orbit adjustment. In order to verify the availability and accuracy of LT-1 satellite data, this article takes the Datong mining area in Shanxi Province as an example and obtains 25 LT-1 strip pattern image data from December 23, 2022 to May 20, 2023, respectively, for SBAS-InSAR and PS-InSAR data processing. By comparing and analyzing the deformation monitoring results of time-series InSAR and GPS stations in the light of sight, the standard deviations of the two are 5.7 mm/a (SBAS-InSAR) and 3.4 mm/a (PS-InSAR), respectively. The root mean square error of the time series is less than 5 mm, indicating high consistency. The research has shown that domestically produced LT-1 satellites have high-precision deformation monitoring capabilities, providing reliable data assurance for domestic terrain surveying and deformation monitoring.

    Table and Figures | Reference | Related Articles | Metrics
    Large models enabling intelligent photogrammetry: status, challenges and prospects
    Mi WANG, Xu CHENG, Jun PAN, Yingdong PI, Jing XIAO
    Acta Geodaetica et Cartographica Sinica    2024, 53 (10): 1955-1966.   DOI: 10.11947/j.AGCS.2024.20240068.
    Abstract368)   HTML37)    PDF(pc) (9351KB)(479)       Save

    Developed from deep learning and transfer learning techniques, large models leverage vast training datasets and immense parameter capacities to create scale effects, thus inspiring the model's emergent capabilities and demonstrating strong generalization and adaptability in numerous downstream tasks. Large models, represented by ChatGPT and SAM, signify the arrival of the era of general artificial intelligence, providing new theories and techniques for the automation and intelligence of Earth's spatial information processing. To further explore the methods and pathways for large models to empower the field of photogrammetry, this paper reviews the basic problems and mission tasks in the field of photogrammetry, summarizes the research achievements of deep learning methods in intelligent photogrammetric processing, analyzes the advantages and limitations of supervised pre-training methods aimed at specific tasks; Besides, we elaborates on the characteristics and research progress of general artificial intelligence large models, focusing on the generalizability of large models in basic visual tasks and the potential in three-dimensional representation; Finally, this paper explores the current challenges and future trends of large model technologies in the field of photogrammetry, from the perspectives of training data, model fine-tuning strategies, and heterogeneous multi-modal data fusion strategies.

    Table and Figures | Reference | Related Articles | Metrics
    Multi-star tracker angular velocity reconstruction method considering temperature effect correction
    Danyi HU, Yunlong WU, Yun XIAO, Yue QIU, Xiaohui WU, Yulong ZHONG
    Acta Geodaetica et Cartographica Sinica    2024, 53 (9): 1748-1760.   DOI: 10.11947/j.AGCS.2024.20240093
    Abstract363)   HTML47)    PDF(pc) (10506KB)(236)       Save

    High-precision satellite attitude control is an important data preprocessing aspect of satellite gravity mission operation. The key payload star trackers onboard the gravity field and steady-state ocean circulation explorer (GOCE) satellite inevitably experience temperature variations in its low orbit, leading to inter-boresight angles (IBA) deviations ranging from 2 to 14 arcseconds, directly impacting the accuracy of satellite attitude. Quantitative analysis of temperature effects on satellite attitude and precise determination of satellite angular velocities are essential steps in the satellite data preprocessing workflow, directly influencing the accuracy of high-precision gravity gradient component reconstruction. In this study, based on the characteristics of the GOCE satellite mission, we develop a temperature effect correction method for joint attitude quaternion reconstruction using multiple star trackers. This method involves constructing a linear function of temperature-related relative attitude offsets between star trackers, establishing a weighted matrix considering the precision differences among sensor axes, and obtaining the optimal quaternion reconstruction of attitude velocities based on the principle of least squares. Additionally, in the original attitude data processing, we propose a logarithmic quaternion Hermite hypersurface interpolation method for data optimization. The research results demonstrate that the corrected attitude quaternions calculated from star tracker data exhibit no significant deviation when compared with reference frame information. Moreover, after temperature effect correction, the noise level of angular velocity for each tracker axis significantly decreases by approximately two orders of magnitude, achieving an accuracy of 10-10 rad·s-1 and significantly improving the precision of velocity reconstruction. Additionally, the angular velocity accuracy of each tracker axis maintains good consistency.The power spectral density of the gravity gradient trace calculated based on this method shows a more significant improvement in the whole frequency domain.

    Table and Figures | Reference | Related Articles | Metrics
    Weakly supervised building change detection integrating multi-scale feature fusion and spatial refinement for high resolution remote sensing images
    Xin YAN, Li SHEN, Junjie PAN, Yanshuai DAI, Jicheng WANG, Xiaoli ZHENG, Zhi-lin LI
    Acta Geodaetica et Cartographica Sinica    2024, 53 (8): 1586-1597.   DOI: 10.11947/j.AGCS.2024.20230118
    Abstract361)   HTML45)    PDF(pc) (20867KB)(293)       Save

    To alleviate the heavy dependence of deep learning methods on large-scale high-cost pixel-level annotations, in this paper, we propose a novel weakly supervised method, named MDF-LSR-Net, for high-resolution remote sensing building change detection. Specifically, the proposed method first designs a multi-scale difference feature aggregation module to make better use of multi-scale difference features to generate change heatmaps. Then, by utilizing the local spatial consistency of the low-level fused difference features, MDF-LSR-Net presents a local spatial refinement module to enhance the integrity and accuracy of change regions in heatmaps. Finally, the change detection model is trained based on the high-quality change heatmaps. Experimental results on publicly available datasets, including WHU and LEVIR, demonstrate that our proposed method can obtain more integral and accurate change heatmaps, leading to significantly improved detection performance of the final change detection model. The final model has achieved over 65% points in IOU and over 79% points in F1 on the WHU dataset.

    Table and Figures | Reference | Related Articles | Metrics
    Rapid single point positioning enhancement service and application based on urban CORS
    Yang LIU, Guang YANG, Xiaohui CHENG, Xiao ZHANG
    Acta Geodaetica et Cartographica Sinica    2024, 53 (9): 1706-1714.   DOI: 10.11947/j.AGCS.2024.20240077
    Abstract338)   HTML36)    PDF(pc) (8322KB)(222)       Save

    In the realm of smart city development, the scalability, privacy, and heightened reliability of CORS (continuously operating reference station) location services have emerged as pivotal focal points. This paper presents an innovative urban CORS augmented positioning service leveraging PPP-RTK (precise point positioning-real-time kinematic) technology. It delineates the formulaic methodology for achieving enhanced precision single-point positioning within urban environments, elucidating the estimable parameters essential for urban CORS augmented positioning and their practical implementation on client platforms. Through integration with real-time data sourced from the Guangzhou CORS network, the efficacy of the PPP-RTK service is rigorously evaluated. Test results demonstrate that the initial epoch of the PPP-RTK client plane rapidly converges to centimeter-level accuracy, with vertical elevation convergence achieved within approximately seven epochs. Furthermore, the performance of the enhanced positioning parameter SSR2OSR (state space representation to observation space representation) in PPP-RTK is found to be comparable to that of short baseline RTK solutions, thus substantiating its capacity to cater to the monitoring requirements of a substantial user base. On-board experiments exhibit superior 3D accuracy, surpassing lane-level precision by a margin of less than 0.5 meters, underscoring the capability of PPP-RTK positioning to fulfill the stringent reliability criteria essential for various positioning applications.

    Table and Figures | Reference | Related Articles | Metrics
    In-orbit application parameters test and analysis of L-band differential interferometric SAR satellite constellation
    Xinming TANG, Tao LI, Xiang ZHANG, Xiaoqing ZHOU, Jing LU, Xuefei ZHANG
    Acta Geodaetica et Cartographica Sinica    2024, 53 (10): 1863-1872.   DOI: 10.11947/j.AGCS.2024.20230240.
    Abstract333)   HTML43)    PDF(pc) (17127KB)(353)       Save

    After one year in-orbit test of the China's first L-Band differential interferometric SAR (L-SAR) satellite, which is also named as LuTan-1, we finally conclude that the satellite has reached the pre-defined accuracies. In this paper, we analyzed the key technologies that related to the parameters of the quantitative applications. Three kinds of parameters belonging the long-time geolocation capacity, the digital surface model (DSM) accuracy, as well as the deformation products accuracy are introduced. Results show that the geolocation error of L-SAR is within 3.9 m. We obtained the DSM calibration and validation data in Henan and Jiangsu Provinces. DSM accuracies before interferometric calibration, after calibration and after least square estimation are 13.2, 6.3 and 2.9 m, respectively. We have defined three kinds of deformation products, i. e., deformation products generated using differential InSAR, stacking and multi-temporal InSAR. Accuracies of the three deformation products obtained in Datong of Shanxi Province are 2.7 mm, 8.6 mm/a and 3.7 mm, respectively. Meanwhile, 90% of the strict regression orbit radius was tested to be smaller than 323 m. The components of the decoherence sources were discussed. Final coherence before filtering is better than 0.7 when ignoring the temporal decoherence. The in-orbit test result is encouraging and the satellite will be widely used after its delivery from satellite producers to the users.

    Table and Figures | Reference | Related Articles | Metrics