Search

This paper takes smart cartography in the context of artificial intelligence as the starting point, analyzes the impact of space-time big data on cartography, points to three possible directions for the evolu‐ tion of the map, explains the connotation of smart cartography and its characteristics of crowd-sourced da‐ ta, user participation, model coupling and visual analysis, discusses in depth the explicit and implicit knowl‐ edge representation of smart cartography, proposes the idea of establishing a map knowledge graph, and outlines a platform development paradigm based on low-code technology. [ABSTRACT FROM AUTHOR]

Objectives: Geological hazard points and hidden danger points are the data basis for geological hazard evaluation, while the existing records of geological hazard points have poor timeliness and are incomplete. To solve this problem, the deformation information obtained by multi-temporal interferometric synthetic aperture radar (InSAR) was integrated into the geological hazard evaluation model. And we explore how to make better use of the deformation information. Methods: The greater the deformation level, the greater the possibility of geological hazards. This paper not only takes the deformation points obtained by multi-temporal InSAR as the geological hazard points/hidden danger points, but also integrates the deformation level information obtained by multi-temporal InSAR as an evaluation factor into the susceptibility evaluation model, making full use of the effective deformation information obtained by multi-temporal InSAR. And the evaluation model adopts the coupling model based on information value model and the analytic hierarchy process model to obtain the susceptibility evaluation and zoning of the geological hazards in Baiyin City, Gansu Province,China. Results: Through the verification of the existing geological disaster point data, it is found that the partitions in this paper are in good agreement with the existing geological hazard points distribution.In the designated extremely high-prone areas, there are nearly 8 geological disaster points within 10 km2 , while less than one in the extremely low-prone areas. Conclusion: The multi-temporal InSAR deformation information added to the geological hazard evaluation model greatly improves the timeliness and quantity of records of geological hazard points/hidden points. However, only one kind of synthetic aperture radar data cannot completely identify all geological hazard points/hidden danger points, due to the limitations of incidence angle and microwave wavelength. In the futher work, we will focus on the combination of multiple deformation monitoring technologies to jointly monitor surface deformation, such as multi-sensor and multi-track InSAR technology, airborne light laser detection and ranging and high-resolution optical remote sensing. [ABSTRACT FROM AUTHOR]

Objectives: Convolutional neural network (CNN) was widely used in landslide susceptibility assessment because of its powerful feature extraction capability. However, with the demand for scene diversification and high accuracy, the algorithm of CNN was constantly improved. The practice of improving accuracy by deepening the network hierarchy or combining other models tended to increase the number of model parameters and the amount of calculation greatly, which led to the difficulty of model training or the overfitting of results, thus limiting its practical application. We want to solve the above problems from the model structure. of feature maps, one-dimensional CNN（1D-CNN） and two-dimensional CNN（2D-CNN） were connected to maintain network depth, limit model parameters, and reduce computation. The parameters sharing of multidimensional convolution kernels was used to capture the deep coupling features of different dimensions and different landslide factors, so as to make full use of features and avoid overfitting. Taking Sedongpu gully in Xizang,China as the experimental area, this paper selected 11 kinds of landslide influencing factors to analyze the landslide susceptibility. Results: The results show that the multi-dimensional CNN coupling structure had the same training efficiency as the shallow 2D-CNN with fewer parameters due to the reduced computational effort. While compared with the deep 2D-CNN with the approximate number of parameters, the training time of the proposed method was significantly reduced and the training cost was lower. In addition, the coupled model had a stronger feature learning ability than the independent 1D-CNN and 2D-CNN, and therefore obtained higher model accuracy. Under each confusion matrix metric of the testing data, the coupled model received higher scores, and thus obtained more reliable landslide susceptibility assessment results. Conclusions: The multidimensional CNN coupling model proposed in this paper is a reliable method applicable to landslide susceptibility assessment. This study provides new theoretical guidance and technical support for further landslide hazard monitoring and prevention [ABSTRACT FROM AUTHOR]

Objectives: The conventional group sparse representation (GSR) model encounters challenges in precisely estimating the sparsity of individual image groups, because the influence of noise results in distortion in the restoration of original image group. Methods: This paper introduces a strategy denominated as the group sparse residual constraint. This strategy leverages non-local self-similarity priors to derive precise estimate of sparse coefficients for each original image group. Subsequently, the sparse coefficients of corresponding degraded image groups are constrained to approximate this estimate. Furthermore, the intricate nature of noise in area-array satellite images (e. g., Luojia3-01 satellite images), surpasses the simplicity of additive white Gaussian noise (AWGN). Consequently, conventional GSR model tailored for AWGN exhibits a reduction in efficacy when deployed in real image denoising scenarios. To address this, this paper introduces two distinct weight matrices into the data fidelity term and regularization, respectively, with the intent of characterizing the statistical features inherent in both images and noise. This paper adopts the alternating direction method of multipliers for optimizing the proposed image denoising task. Results: The effectiveness of the proposed model is validated by the experiments using simulated data and real images obtained from the Luojia3-01 satellite. In the experiments of simulated data, the proposed model outperforms other methods in removing AWGN and spatially variant noise. In the experiments of real images, the entropy of denoised images by the proposed model shows improvements of 2.03%, 1.18%, 1.26%, 1.24%, and 2.10%, compared to the block-matching and 3-D filtering method, multi-channel weighted nuclear norm minimization method, nonlocally centralized sparse representation method, low-rankness guided group sparse representation method, and trilateral weighted sparse coding method, respectively. Conclusions: The proposed method exhibits good performance in processing area-array satellite images with low signal-to-noise ratios and complex noise structures, significantly enhancing the quality of the image. By effectively reducing noise while preserving crucial details and edges, it provides a reliable solution for enhancing the usability and interpretability of area-array satellite images, particularly in challenging conditions. [ABSTRACT FROM AUTHOR]

Objectives: The dynamic and high spatial resolution data obtained by optical video satellites provide a new technical means for the dynamic observation of the earth. As a new generation of intelligent mapping remote sensing scientific test satellite, Luojia3-01 satellite was launched in early 2023, and it could acquire high-definition color video by gazing mode. However, the original video frame rate is low, only 6 frame per-second. To further improve the fluency and reduce the stuttering effect of visual perception, this paper carries out related research on intelligent video frame interpolation (VFI) of Luojia3-01 satellite video. Methods: First, a satellite video image stabilization method based on the perspective transformation model is proposed after analyzing the imaging errors during the gaze mode. Second, satellite video is greatly different from daily video, and there is no available satellite VFI dataset at present. This paper builds a VFI dataset termed as Luojia3_VFISet based on the stabilized satellite videos, which covers different scenes. Finally, based on the flow-agnostic video representation (FLAVR) VFI network without optical flow module, the FLAVR_Plus VFI network is proposed to further improve the interpolation effect for satellite video by in⁃ troducing the feature coding information of different scales from the encoder into the decoding process. Results: The experimental results show that the interpolation of FLAVR_Plus VFI network raises the peak signal to noise ratio (PSNR) and structural similarity (SSIM) to 35.544 6 dB and 0.917 9, respectively, and it enhances the quality of synthesized intermediate frames with hardly-observed artifacts under different scenarios. Compared with other methods, the proposed network can improve the PSNR by 0.5% to 7.2% and the SSIM by 0.5% to 8.7%, respectively. Conclusions: In this paper, the application of satellite VFI is studied by taking Luojia3-01 satellite as an example. The proposed Luojia3_VFISet dataset is conducive to the development of related research. The proposed FLAVR_Plus VFI network can effectively improve the fluency of the Loujia3-01 satellite video by generating interframes without artifacts and provide more interframe information for subsequent applications. [ABSTRACT FROM AUTHOR]

Objectives: Individual driving destination prediction has important value in location-based services such as personalized services recommendation and intelligent transportation. However, most of the existing deep learning methods construct travel features on the basis of densely-sampled trajectory points, which generate data redundancy but only achieve limited information gain, and in turn introduce negative impacts on model training. Road intersection sequence can simplify the expression of driving trajectories restricted on the roads, and reduce the model training cost. Meanwhile, transfer preference between intersections and current movement mode imply the spatiotemporal correlation between road intersections and destinations, and are capable to represent individual travel intention to some extent, but few researches have applied them to destination prediction. Methods: This paper proposes a novel individual driving destination prediction method that takes intersection transfer preference (ITP) and current movement mode (CMM) into account, named ITPCMM. First, it constructs input features in units of each intersection, and adopts graph attention network to learn the individual-level transfer coefficients between intersections in different time slots. Then, it uses long short-term memory (LSTM) to capture the long-term dependence of transfer preferences. Second, ITP-CMM constructs time cyclic encoding features and driving features to learn current move mode by using two-layered LSTM. Third, intersection transfer preferences and current move mode are fused and weighted through feature intersection and attention mechanisms respectively. And the stacked residual networks are introduced to output prediction results. Results and Conclusions: Based on the trajectory data of twelve private car drivers in Shenzhen City, China for the whole year of 2019, this paper has conducted a group of experiments to verify the feasibility of the proposed method ITP-CMM. We verify the effectiveness of the proposed method by carrying out ablation experiments and comparing the prediction accuracy and stability of ITP-CMM with four baselines, including hidden Markov model, LSTM, distance neighboring dependencies model and attentionaware LSTM for real-time driving destination prediction by considering location semantics and location importance trajectory points. Besides, we demonstrate the value of transfer preference in capturing the spatial correlation between intersections through visual analytics, and explore the impact of the number of transfer intersections on the prediction accuracy. [ABSTRACT FROM AUTHOR]

Objectives：In mid- to late- April 2024, an extreme heavy rainfall event occurred in Shaoguan City, Guangdong Province, inducing a large number of landslides in Jiangwan Town, Shaoguan. People lost connection with the outside world for nearly 36 hours, which aroused widespread social concern. Rapidly and accurately identifying the basic characteristics of landslides, development and distribution patterns and formation conditions is crucial for disaster emergency decision-making and risk elimination and disposal. Methods: Using the post-disaster optical remote sensing images and combining with deep learning model, the rainfall-induced landslides in Jiangwan Town, Shaoguan, were quickly and automatically identified. Results:After manually calibration, a total of 1 192 landslides were deciphered, with a total area of about 3.14 km². The scale of the landslides was dominated by small and medium-sized landslides, which were mainly distributed as an aggregated belt along the river in the northeast-southwest direction, with a significant characteristic of concentrated occurrence. Spatial statistical analysis showed that the landslides were mainly distributed on concave slopes with slopes of 10°-30° in the range of 200-300 m elevation. Further quantitative analysis of the geomorphic controlling factors of landslides using the random forest model and SHAP theory reveals that different topographic and geomorphic factors have different degrees of nonlinear effects on landslide formation, and that multiple factors such as elevation, slope, and catchment conditions are coupled to jointly control the formation of landslides. Conclusions: This paper highlights the great advantage of deep learning-based intelligent identification and analysis technology in the emergency investigation and formation conditions analysis of landslide disasters, which can provide important technical support for the rapid assessment of disaster losses and risk identification. [ABSTRACT FROM AUTHOR]

Objectives: Flood disaster is a very destructive natural disaster. The main reasons for its generation include heavy rainfall, storm surge and dam break. When flood disaster occurs in populated areas such as cities and towns, the flood will directly threaten the safety of life and property of residents. Also it will cause paralysis of land and underground transportation, interruption of water and electricity transportation. In the process of flood rescue, quick and accurate identification of flooded roads is conducive to planning appropriate personnel transfer and material transportation routes, and reducing subsequent losses caused by floods. Aiming at the problem that roads in the flood disaster scenario cannot be automatically and correctly identified, this paper proposes an end-to-end flooded road detection method based on deep learning. Methods: The proposed method uses a three-branch encoder-decoder structure and uses strip convolution, in which the efficient extraction of linear features is realized. And the coordination dual attention mechanism can effectively guide the network, and realize the identification of road areas. The method can effectively utilize the historical optical remote sensing images before the disaster and the real-time optical remote sensing during the disaster. The image is applied to detect the flooded and non-flooded roads in the disaster-stricken area, and a comparative experiment is carried out on the self-built dataset. Results: The experimental results show that the precison and recall rate are 0.838 1 and 0.666 8 on pre-disaster road, 0.796 6 and 0.607 4 on post-disaster road, 0.780 0 and 0.661 4 on affected road respectively. Conclusions: The proposed method has achieved the goal of automatically identifying the flooded roads in the flood-stricken area. The ability of detecting flooded roads and non-flooded roads can provide strong support for flood disaster rescue and reduce losses of life and property caused by flood disasters. [ABSTRACT FROM AUTHOR]

Objectives: The landslides disaster is one of the most important factors for the construction of major infrastructure in the western China. How to effectively and reliably carry out wide-area landslide susceptibility prediction has always been a frontier difficulty in domestic and foreign studies. However, the present data-driven methods for prediction of landslide susceptibility in large-scale and complex scenarios still face two major issues: (1) The difference between various landslide inducing environments in a wide range of scenarios, would cause the difficulty to applying a single model to account for multiple landslide phenomenon; (2) small sample problem: Complex environmental tasks require models with large capacity and strong representative power, but there is a lack of sufficient landslide samples in practice. Methods: In response to the above problems, this paper takes Qijiang and Fuling District of Chongqing City, China as an example, proposes a local prediction strategy, and introduces the idea of meta-training an intermediate representation suited to be generalized, that can be adapted to the landslide sensitivity prediction task corresponding to the current local area, with only a very small number of samples in the local area, and number of iterations. Thus, the two issues mentioned can be well settled. Results: The proposed method is different from traditional methods such as support vector machines, multilayer perceptrons, and random forests, which require a large number of samples and gradient iterations to train the supervised model. Instead, only a small sample is required to fine-tune the intermediate model, which still improves the global accuracy by 1%-5%, the precision by 1%-3%, the F1-score by 0.5%-6%, and the recall rate is close to the highest level of other methods. Conclusions: The meta-learning paradigm enables few-shot adaptation of landslide susceptibility prediction model with superior statistical performance. [ABSTRACT FROM AUTHOR]

Objects: Mountain peak extraction technology determines the accuracy of the hill position classification and the efficiency of automatic classification of micro-landform. Because of the limitations of elevation and contour lines, such as loss of local mountain peaks and incomplete removal of false mountain peaks, other landform factors could be used to express the distribution feature around the mountain peaks. Methods: Based on the uniform distribution feature of the aspect around the mountain peaks, this paper proposes an efficient mountain peak extraction model, in which the aspect is centered on the mountain peak and gradually increases clockwise. Moreover, according to the ridge line fitting method and the recursive thought of depth first search algorithm, the false mountain peaks in the extraction results are removed, while the real mountain peaks are retained. Considering the negative influences of the terrain fragmentation of the entity digital elevation model(DEM), this paper experiments and analyzes with simulated DEM and entity DEM respectively. Results:The results show that the method overcomes the uncertainty of subjective threshold based on closed contour lines. The mountain peak extraction accuracy of simulated DEM can achieve 100% due to its continuity and smoothness. Considering the terrain fragmentation of entity DEM will make it ill-posed, we improve it by adjusting the aspect distribution constraint condition and obtain average extraction accuracy of 96.1%. Conclusions: An efficient mountain peak extraction model is proposed based on the digital terrain analysis technology and the uniform distribution feature of the aspect around the mountain peaks, and the topographic feature points are extracted from the perspective of terrain geometry. Compared with the traditional mountain peak extraction methods, the proposed method is relatively accurate and simple, and can effectively reduce the false mountain. [ABSTRACT FROM AUTHOR]

Objectives: The traditional methods of identifying urban features use spatial and statistical algorithms to extract analysis indicators, but feature evaluation indicators are very subjective. Street view images contain visual information of the city and can be used to identify urban features. Methods: Taking Qingdao, China as an example, this paper proposes a multi-scale semantic segmentation model, named MS-DeepLabV3+, based on street view images. The proposed model adds full feature extraction channels in the encoding process to aggregate multi-scale features, and adds multi-scale feature extraction channels in the decoding process to effectively capture low-level features. And convolutional block attention module and efficient channel attention modules focusing on key features are introduced to improve the accuracy of semantic segmentation of street views. The mean intersection over union, accuracy and recall of the proposed model have been increased by 3.47%, 2.37% and 3.96%, respectively. We build a multi-dimensional feature vector of the city in six dimensions, including environment dimension, facility convenience dimension, economic affluence dimension, transportation dimension, urban safety dimension and urban synthesis dimension. Based on the semantic segmentation results of the street view images, the data are combined with the point-of-interest data and residential land use data. At the plot scale, we extract the feature vectors and calculate the values in six dimensions to characterize the urban features of each urban area in Qingdao. This paper uses the Grad-CAM method for interpretable analysis of semantic segmentation models and the feature attribution SHAP method to mine the intrinsic drivers of multi-dimensional features in cities. Results: Different urban areas have different feature vectors, and the feature vectors of different urban areas have the advantages in specific dimensions. Conclusions: The above analysis helps optimize the multi-dimensional features in urban space for the planning and construction of cities. [ABSTRACT FROM AUTHOR]

Objectives: In recent years, a great breakthrough has been made in the text generation image problem based on generative adversarial networks (GAN). It can generate corresponding images based on the semantic information of the text, and has great application value. However, the current generated image results usually lack specific texture details, and often have problems such as collapsed modes and lack of diversity. Methods: This paper proposes a determinant point process for generative adversarial networks(GAN-DPP) to improve the quality of the generated samples, and uses two baseline models, Stack-GAN++ and ControlGAN, to implement GAN-DPP. During the training, it uses determinantal point process kernel to model the diversity of real data and synthetic data and encourages the generator to generate diversity data similar to the real data through penalty loss. It improves the clarity and diversity of generated samples, and reduces problems such as mode collapse. No extra calculations were added during training. Results: This paper compares the generated results through indicators. For the inception score, a high value indicates that the image clarity and diversity have improved. On the Oxford-102 dataset, the score of GAN-DPP-S is increased by 3.1% compared with StackGAN++, and the score of GAN-DPPC is 3.4% higher than that of ControlGAN. For the CUB dataset, the score of GAN-DPP-S increased by 8.2%, and the score of GAN-DPP-C increased by 1.9%. For the Fréchet Inception Distance score, the lower the value, the better the quality of image generation. On the Oxford-102 dataset, the score of GANDPP-S is reduced by 11.1%, and the score of GAN-DPP-C is reduced by 11.2%. For the CUB dataset, the score of GAN-DPP-S is reduced by 6.4%, and the score of GAN-DPP-C is reduced by 3.1%. Conclusions: The qualitative and quantitative comparative experiments prove that the proposed GAN-DPP method improves the performance of the generative confrontation network model. The image texture details generated by the model are more abundant, and the diversity is significantly improved. [ABSTRACT FROM AUTHOR]

Objectives: The quality of network structure is the key to obtain excellent classification performance. Most of existing deep models are designed by the trial-and-error method. When the same model processes different hyperspectral images (HSIs), different network hyperparameters need to be set artificially to adapt to the data characteristics of different images, greatly increasing the difficulty of network design. Aiming at the problem of deep learning model design in hyperspectral image classification, this paper proposes an automatic network structure search method for HSI classification. Methods: In the proposed method, the differentiable structure search technology is first used to search the network structure on the source HSI data sets, then the deep network model is constructed by stacking cells, and finally the classification performance is evaluated on the target HSIs. It should be pointed out that, the proposed method only conducts one search on the source HSI data sets, and the obtained deep model can be applied to other target HSI classification tasks, which can effectively improve the utilization rate of the model. To improve the generalization ability and classification accuracy of the model, multi-source and multi-resolution HSIs are employed to construct the precollected data set, and the partial channel connection operation is introduced to further improve the search efficiency. Results: Three HSIs inculding Houston, Chikusei and Pavia Center are selected as source data. Four HSIs including University of Pavia, Indian Pines, Salinas and Houston 2018 are selected as the target HSIs. The source data is used to determine the network model and the target HSIs are used to evaluate the classification performance. Experimental results show that the proposed method can automatically search out a universal deep network model suitable for HSI classification task, and this model can achieve better classification performance than the conventional deep learning models. The overall classification accuracies of University of Pavia, Indian Pines, Salinas and Houston 2018 are 98.15%, 98.74%, 97.30% and 74.47%, respectively. Conclusions: Compared with other deep models, the model in this paper is automatically determined by differentiable architecture search algorithm according to the data characteristics of HSIs, effectively improving the classification accuracy of target HSIs while avoiding complex network design. Extensive experiments verify the applicability and effectiveness of the proposed method in HSI classification. [ABSTRACT FROM AUTHOR]

Objectives: Curve simplification is of importance in automated map generalization; nevertheless, the Douglas-Peucker (DP) algorithm popularly used in map generalization is not automatic, because a key parameter called distance tolerance (ε) must be given by experienced cartographers and needs to be input before execution of the algorithm. Methods: To solve the problem, this paper proposed a method to automatically calculate ε and by which the automation of the DP algorithm is achieved. The method consists of the following steps: (1) A formula is constructed by the Hausdorff distance for calculating the similarity degree (Ssim) between a curve at a larger scale and its simplified counterpart at a smaller scale. (2) 15 linear rivers are selected, and each of them is manually simplified to get their counterparts at seven different scales. The Ssim of each original river and each of its simplified counterpart at a smaller scale can be obtained using the formula constructed by the Hausdorff distance, and 15×7=105 coordinate pairs consisting of (S,Ssim) can be got, and a function between Ssim and S are constructed by the curve fitting using the coordinates. (3) In the meanwhile, the 15 rivers are simplified using a number of ε, and the Ssim of each original river and each of its simplified counterpart at a smaller scale can be calculated using the formula constructed by the Hausdorff distance. In this way, a number of coordinate pairs (ε,Ssim) are got, and a function between ε and Ssim is constructed by the curve fitting. (4) By the function between Ssim and S and that between ε and Ssim, a formula between ε and S can be deducted. Using the formula ε can be calculated automatically, because in a map generalization task S is usually known. After this step, automation of the DP algorithm is achieved. Results: The experiment results show that (1) The proposed DP algorithm can automatically simplify the rivers in a specific geographical area to get the results at different scales; and (2) the resulting river curves generated by the proposed DP algorithm have a high degree of similarity with the ones made by experienced cartographers. Their average similarity degree is 0.927. Conclusion: The proposed DP algorithm can simplify curve features on maps automatically, and the results are highly intelligent and credible. Although only river data is tested in this paper, the principle of the proposed method can be extended to other linear features on maps. Our future work will be on improving the accuracy of the proposed DP algorithm using more river data so that the algorithm can be used in practical map generalization engineering. [ABSTRACT FROM AUTHOR]

Objectives In recent years, many deep learning-based classification models have been proposed. Because the size of histopathology images is extremely large, current methods usually divide it into many small patches of the same size and then build the classification model. The models first extract the features of patches and then use pooling methods etc. to integrate patch-level features into slide-level features for classification, in which the extraction of patch-level features and independent with the integration and classification process, making it impossible to adapt the feature extractor with the feedback of classification results. Therefore, they can not obtain the feature which is most useful for the whole slide image (WSI) classification. The objective of the paper is to solve the problems faced by above methods.Methods We propose an end-to-end attention pooling method based on a patch-scoring model. First, we build a patch scoring model to score each patch with the multiple instance learning methods, selecting patches based on the scores. Second, we utilize an attention-pooling module to integrate the features of selected patches to build a classification model. Besides, according to the character of the histopathology image, we propose to use a customized squared average function to normalize the attention weights to focus more on the high scored patches in positive WSI, making the classification results more interpretable.Results On the public available dataset CAMELYON16, the proposed method achieves F1 score at 0.644. And it achieves F1 score at 0.593 at the public available BACH dataset. When the new normalization function is used, the average attention weights difference between positive patches and negative patches is 0.017, which is much higher than when softmax function was used.Conclusions We can conclude from the experiments that the end-to-end attention pooling method is a better method for breast histopathology image classification. The proposed normalization function can make the attention module more interpretable. [ABSTRACT FROM AUTHOR]

Objectives: High-resolution agile optical satellites can continuously observe specific regions or targets, thereby realizing applications such as regional monitoring and target tracking. In order to obtain stable and high-quality video data, continuous observation data need to be geographically corrected and stabilized. This process requires a large amount of calculation, but the on-board computing and storage device cannot meet the demand. Therefore, this process can only be completed afterwards by on-ground system, resulting in a significant delay in information acquisition. Methods: This paper proposes an object-spaceconsistency-based real-time stabilization approach for Luojia3-01 satellite video data, which is based on data characteristics and computing capability of embedded on-board hardware. The proposed approach employs real-time geometric positioning to extract and correct the region of interest (ROI) in real-time during satellite imaging. Additionally, it employs the consistency of geographical information in real-time align adjacent image frames to ensure the relative accuracy between sequence frames. On the basis, it generates a sequence of video frames of ROI with geographical encoding in real-time through the construction of a parallel pipeline of algorithms. Results: The experiments are conducted by the actual data of Luojia3-01 satellite with the frame rate of 6 frame per second, and the proposed approach can realize the faster processing based on the embedded on-board processing hardware. Furthermore, the obtained video frame sequence with geographic encoding and the inter-frame stabilized accuracy is better than 0.328 pixels. Conclusions: The proposed approach can meet the requirements of on-board real-time service. [ABSTRACT FROM AUTHOR]

Objectives: A real-time and robust 3D dynamic object perception module is a key part of autonomous driving system. Methods: This paper fuses monocular camera and light detection and ranging (LiDAR) to detect 3D objects. First, we use convolutional neural network to detect 2D bounding boxes and generate 3D frustum region of interest (ROI) according to the geometric projection relation between camera and LiDAR. Then, we cluster the point cloud in the frustum ROI and fit the 3D bounding box of the objects. After detecting 3D objects, we reidentify the objects between adjacent frames by appearance features and Hungarian algorithm, and then propose a tracker management model based on a quad-state machine. Finally, a novel prediction model is proposed, which leverages lane lines to constrain vehicle trajectories. Results: The experimental results show that in the stage of target detection, the accuracy and recall of the proposed algorithm can reach 92.5% and 86.7%, respectively. The root mean square error of the proposed trajectory prediction algorithm is smaller than that of the existing algorithms on the simulation datasets including straight line, arc and spiral curves. The whole algorithm only takes approximately 25 ms, which meets the real-time requirements. Conclusions: The proposed algorithm is effective and efficient, and has a good performance in different lane lines. [ABSTRACT FROM AUTHOR]

Objectives: The side-scan sonar shipwreck detection method based on the YOLOv3 model has the problems of high miss-alarm rate of small targets, heavy model weight, and slow detection speed that fails to meet real-time requirements. Methods: This paper introduces the YOLOv5 algorithm and proposes an improved YOLOv5 model according to the characteristics of the side-scan sonar shipwreck dateset. We test YOLOv5a, YOLOv5b, YOLOv5c, YOLOv5d, YOLOv5s, YOLOv5m, YOLOv5l and YOLOv5x under the basic framework of YOLOv5 with eight different depth and width model structures. Then we choose the best structure by using genetic algorithm and K-means algorithm to optimize the detection frame, and to improve the loss function through complete intersection over union. Results: The results show that under the different range of intersection over union as 0.5 and 0.5-0.95,the average precisions of the improved YOLOv5a model are increased by about 0.3% and 0.6% than that of the original model,respectively. Compared with the YOLOv3 model, the average precisions of the improved YOLOv5a model are increased by 4.2% and 6.1%, respectively, and the detection speed reaches 426 frames per second which is almost doubled that of YOLOv3. Conclusions: The proposed method is more conducive to practical applications and engineering deployment. [ABSTRACT FROM AUTHOR]

Objectives: Facing the needs of digital twins for the whole life cycle of railway, the problems exist in independent expression, decentralized management and difficult in-depth integration of the continuously changing railway geographic and geologic data for a long time. Methods: This paper takes railway geographic and geologic elements as an inseparable whole, and proposes an integrated expression model of the three-domain association for railway geographic and geologic data, which includes ontology domain, change domain and state domain. First, it takes spatial-temporal change as the center and defines the change participants (elements), drivers (events) and presentation modes (processes). Then, it defines a multi-level spatial-temporal change association relationship between state level, entity level and process level. Finally, it constructs a knowledge graph of railway geologic data with multi-domain association, and finely portrays the semantic association relationship between three aspects of geological elements, events and processes. Results and Conclusions: The importance and potential of the model in typical tunnel construction applications are analyzed. [ABSTRACT FROM AUTHOR]

Objectives: Under the traditional satellite observation data processing operation mode, outlier detection methods in gravity gradiometry have the problems of insufficient accuracy, low efficiency, and so on. Methods: This paper constructs an intelligent outlier detection method, which combines variational autoencoder and gated recurrent unit by the significant characteristics of variational autoencoders in multimodal data integration analysis. First, on the basis of the origin and propertier of satellite gravity gradionmetry outliers, the gravity gradiometry dataset with outlier are simulated. Second, the network model captures the effective features of dataset by variational autoencoder, makes predictions on dataset by combining with gated recurrent unit, and automatically finds optimal convergence of the loss function by designing adaptive moment estimation as optimizer. Finally, the tested training model is applied to actual gravity observation data of civilian gravity satellites. Results: The accuracy of the proposed model in outlier detection reaches more than 98%, with a good detection effect on both discrete and regional gross errors. Conclusions: The trained network can quickly and accurately construct the sample features of the simulated gravity dataset，achieve fast and efficient gross error detection capabilities，and be effectively applied in data preprocessing of autonomous satellite gravity missions. [ABSTRACT FROM AUTHOR]

Objectives: Satellite on-orbit autonomous mission planning requires high-precision orbit prediction data to calculate the feasible imaging time windows and imaging posture requirements for the intended observation targets. This process involves integrating these factors with other constraint conditions to accomplish the mission planning. Methods: In the context of limited computational resources onboard satellites, this paper proposes a novel method to reduce resource consumption in orbit prediction process and meet the demands of autonomous mission planning. First, based on two-line element (TLE) and a simplified general perturbation model, the proposed method utilizes multiple sets of orbital data as input and introduces a new measurement metric. Then, it iteratively generates TLE by rationally setting data weights and progressively adjusting them to get the results of orbit prediction. Finally, the total error for orbit prediction and the initial value error of imaging parameters for simulating observation targets are calculated for quantitative evaluation. Results: Experimental results demonstrate that the proposed method requires only a small amount of input data to achieve a 72-hour orbit prediction with the accuracy better than 4 km and the average computation time of 12.76 s. Compared to high precision orbit propagator model, the proposed method provides higher prediction accuracy and faster execution speed in long-period orbit prediction. The calculated start imaging time error for simulated observation targets is less than 0.2 s, and the initial imaging posture errors in the three-axis directions are all better than 0.03°, which is lower than the attitude pointing accuracy requirement for the imaging process of camera onboard Luojia3-01 satellite. Conclusions: The proposed method offers high prediction accuracy and demands fewer temporal and spatial resources, making it of great significance for satellite on-orbit autonomous mission planning [ABSTRACT FROM AUTHOR]

Forecasting trends in sustainable sustainability levels in metropolitan areas serves as the foundation for developing development policies that are tailored to local conditions. Sustainability evaluation studies based on a multidimensional comprehensive index system fall short of highlighting the variability, systemic, and dynamic nature of metropolitan area development. To better understand the role of multidimensional scenarios in guiding the sustainable development of metropolitan areas, a greater emphasis on suburban spatial units and spatial heterogeneity characteristics is required. Methods: We simulate the trends of sustainable development levels of 16 municipal districts in Beijing, China from 2011 to 2030 by combining the system dynamics model and the shared socioeconomic pathways, and thus condense the metropolitan area's sustainable development model and compare the advantages and shortcomings of each development scenario under different models. Results: The findings show that: (1) Beijing's municipal districts perform differently in economic, social, and environmental subsystems, and there are three types of typical districts: Synergistic development, ecologically biased development, and economic constraint. (2) Shared socioeconomic pathway (SSP) 1 is the most desirable urban sustainable development scenario among the four scenarios of SSP1, SSP2 (BAU (business-as-usual)), SSP3, and SSP5. (3) In order to achieve the transition from the historical trajectory to the SSP1 scenario, each district type should pursue a distinct development strategy. To improve the quality of coordinated development, synergistic development type zones should use social security and employment expenditures to compensate for the shortcomings of social development. Ecologically biased zones should reduce energy consumption and increase environmental protection investment to maintain environmental advantages, and economic constraint type zones can maintain high gross domestic product growth rates in the short term to get out of the development dilemma. Conclusions: The findings of this paper highlight the complex cause-and-effect relationships between urban economic, social, and environmental systems and offer a framework for system evaluation and simulation technology for the sustainable development of metropolitan areas. This framework will serve as a source of regulations and a foundation for management of urban refinement. [ABSTRACT FROM AUTHOR]

Objectives: Disaster event evolutionary graph can comprehensively express the development process of disaster, information of sub-events and multiple event relationships, providing knowledge services for disaster analysis. The problems in the construction of disaster event evolutionary graph include the lack of Chinese corpus containing spatiotemporal relationship, imprecise event extraction due to fuzzy boundaries of Chinese vocabulary, and difficulty in recognizing implied event relationships. Methods: This paper proposes a method for constructing disaster event evolutionary graph based on spatiotemporal relationship. First, a vocabulary enhancement event extraction model based on naming regularity (VENR) is designed to discern entity name patterns, and determine event information boundaries and types. Second, a relation recognition model by fusing attention and gateddilated convolution (AGDC) is proposed to obtain multi-dimensional text features, thereby uncovering potential event relationships. Additionally, a Chinese disaster corpus with spatiotemporal relationship (CDCSTR) is established. Results: We experiment on CDCSTR and Chinese emergency corpus (CEC), the results show that the F1 scores of VENR are 88.59% and 78.49%, respectively on CDCSTR and CEC. Compared with the existing methods, AGDC improves the performance of four tasks on CDCSTR, especially excelling in spatial relationship recognition, achieving a performance lead of over 3.08%. Conclusions: The Chengdu-Leshan expressway tailgating accident event evolutionary graph is built for verification, enabling the effective expression of event evolution and spatiotemporal features in real disaster scene, consequently supporting emergency responses to disaster. [ABSTRACT FROM AUTHOR]

Objectives: Recognition of building group patterns is an important part of the building generalization. Efficient methods for the recognition of building group patterns can effectively improve the quality of automated map generalization. Traditional recognition methods mainly include geometric methods and traditional machine learning methods, which are limited by the complex rule definitions and huge feature engineering. The graph convolution neural network (GCN) can overcome the limitations of traditional methods to some extent and has been successfully applied to the pattern analysis of buildings. However, there are few methods to recognize multiple building group patterns using GCN, and the existing GCN methods are almost based on spectral graph convolution, which do not consider local spatial information fully. Methods: As the spatial GCN is more efficient than the spectral GCN, this paper introduces a spatial GCN method DGCNN to recognize three building group patterns, including linear pattern, grid pattern and irregular pattern. The first step is to cluster the buildings into different groups, then some indices are selected to construct the feature vector of each group according to the Gestalt principles. Second, Delaunay triangulation (DT) and minimum spanning tree (MST) are chosen to construct graph structures of building groups. Finally, graph structures are used as input for the DGCNN model, and building group patterns are obtained after training. Results: The experiment selects the urban areas in Shanghai and we compare the accuracies with traditional spectral GCN method. The results show that the accuracy for the DGCNN model in the train set can reach 97.06%, and 95% for the test set. And the accuracies for the spectral GCN model in the train set is 90.90%, and 86.05% in the test set. Conclusions: The proposed method overcomes the limitations of spectral GCN, and it does not need graph Fourier transform. It can improve the recognition accuracy significantly and faces less difficulty in explaining the process of feature extraction. Hence it is an effective method for the recognition of building group patterns. [ABSTRACT FROM AUTHOR]

Objectives: In recent years, the climate change in Xinjiang region, China is complex. It is of great significance to study the terrestrial water storage (TWS) and its load deformation for the accurate maintenance of the regional reference frame. Methods: The satellite gravity data GRACE (gravity recovery and climate experiment) was used to invert the changes of land water storage in Xinjiang region from 2010 to 2014, and the independent component analysis was used to decompose it into spatiotemporal models, and the spatial characteristics of multiple time scales were analyzed. On this basis, the correction on the first-order of gravity potential coefficient and load Love number is carried out, the correction of the non-tidal atmospheric and ocean model (GAC) from GRACE is introduced to obtain the vertical change of the crust caused by the migration of the land water load. The GRACE inversion results are corrected by scale factor method derived from WaterGAP global hydrology model (WGHM). The influence of deformation is analyzed by introducing monthly precipitation data from global precipitation climatology project. And the deformation displacements of the 13 continuously operating reference system stations in the survey area are quantitatively compared. The correlation between TWS load signal of each station and the annual vertical displacement of global navigation satellite system (GNSS) is analyzed. Results: The results show that TWS in Xinjiang region presents multi-time scale characteristics, with obvious annual variation and long-term trend. The annual change is particularly significant near the western Pamirs Plateau, and the trend term is reflected in the long-term decrease year by year. The signal is relatively strong in the western Urumqi and Tianshan Mountains. In general, the change of land water load causes the vertical displacement of the crust in the study area, and the time series shows a small fluctuation with an amplitude of -1.5-1.5 mm, which is consistent with the deformation results of WGHM model. In addition to the obvious time lag, the time series of TWS load deformation and precipitation change are consistent with the overall trend. After GAC correction, the consistency between GRACE results and GNSS vertical displacement was enhanced, and the correlation coefficient and weighted root mean square (WRMS) were significantly improved. The correlation coefficient of each station was 0.63-0.91, and WRMS was 16.18%-58.97%. Conclusions: The research method in this paper can reveal more detailed hydrological signal characteristics in Xinjiang region, and the research results can provide important technical support for accurate maintenance of regional reference frame. [ABSTRACT FROM AUTHOR]

Objectives: As an important grain-producing area and the most populous plain area in China, the groundwater resources in the North China Plain are seriously exploited for agricultural irrigation and domestic water use, and it is of great ecological significance and socio-economic value to accurately monitor the changes of groundwater storage in the region and continuously recharge and restore the groundwater. Methods: The gravity satellite data were used to invert the groundwater storage changes in the North China Plain, and compared and analyzed with the hydrological model, precipitation information, South-North water transfer project and groundwater well information, and proposed a method to integrate shallow groundwater storage changes and deep groundwater changes to obtain the contribution of total groundwater. Re⁃ sults: The results show that: (1) Both gravity satellite and global land data assimilation system model estimated water storage changes in the North China Plain exhibit obvious seasonal characteristics. (2) The correlation coefficients between the gravity satellite inversions and the shallow groundwater well estimations are consistent in terms of interannual variability, multi-year trend variability and seasonal variability, with correlation coefficient greater than 0.8. (3) Considering the contribution of deep groundwater, the estimated results of integrating shallow and deep groundwater well information and the gravity satellite inversions are in better agreement than those using only shallow groundwater, and indicate that shallow groundwater is the main variation of groundwater storage change. (4) From 2003 to 2017, groundwater in the North China Plain is in a long-term deficit state, with a gravity satellite estimated groundwater deficit rate of-1.3± 0.6 cm/a. From 2018 to 2020, the gravity satellite estimated groundwater deficit rate is −1.9±0.7 cm/a, which is almost consistent with the total groundwater deficit rate of -1.8±0.8 cm/a during the same period taking into account the deep groundwater; the groundwater trend in 2021 is increasing, and the growth rate of gravity satellite inversion is 12.7±1.8 cm/a, which is more consistent with the total ground-water well estimation in the same period. The correlation coefficient is 0.98, and the weighted fused total groundwater storage change is more consistent than that estimated by using only shallow groundwater. (5) Precipitation and North-South water transfer project have a direct impact on water storage changes in the North China Plain, making a significant contribution to the mitigation of groundwater storage deficits and groundwater restoration. Conclusions: The research method in this paper provides a new way to accurately observe groundwater storage changes in the North China Plain, and provides reliable verification of gravity satellite monitoring results. [ABSTRACT FROM AUTHOR]

Objectives: Global navigation satellite system (GNSS) interferometric reflectometry (GNSS-IR) is a new passive remote sensing technique for determining surface environment parameters, which places an important part in the inversion of earth's surface properties, such as soil moisture monitoring, snow parameter retrieval, and vegetation remote sensing, etc. GNSS-IR offers several benefits over the traditional soil moisture inversion approach, including all-weather capability, high temporal precision, and cheap cost. Methods: Considering the fact that the existing soil moisture inversion algorithms only utilize one single feature of GNSS reflected signal and from the perspective of increasing data availability, this paper proposes a GNSS-IR soil moisture inversion approach that integrates multi-type feature data by utilizing phase, amplitude, and frequency extracted GNSS signals reflected by soil. The main work is to effectively filter all available features extracted from the original GNSS signal-to-noise ratio observations. The feasibility and effect of the suggested method are compared and evaluated using three machine learning models, including least squares support vector machine (LSSVM), random forest (RF), and back propagation neural network (BPNN). Results: Comparing the inversion effects of above three models, BPNN model has the best inversion effect, followed by RF model, and LSSVM model is the worst. The results show that the correlation coefficients between the reference value and soil moisture inversed by the multi-feature fusion method LSSVM, RF, and BPNN models are 0.830, 0.953, and 0.980, respectively, and the corresponding root mean square errors are 0.045, 0.035 and 0.032 cm³ /cm³ Conclusions: Compared with the single feature inversion method, both the accuracy and correlation coefficient of soil moisture inversion increase significantly. The results demonstrated that the proposed method has higher inversion accuracy and reliability than the single feature inversion method. [ABSTRACT FROM AUTHOR]

Soil moisture is an important parameter to measure the material and energy exchange between soil and atmosphere. It is a key environmental factor in the fields of hydrology, meteorology, agriculture and so on. Remote sensing inversion and model simulation are two basic means to obtain global soil moisture data. Remote sensing inversion can obtain soil moisture data with relatively high resolution, but there are often time-space gaps while land surface model can simulate spatiotemporal continuous data, but the spatial resolution is often coarse. Both two data cannot meet the fine observation of soil moisture. Therefore, it is very important to develop an effective method to obtain high-resolution, spatiotemporal seamless soil moisture data. Methods: Focusing on their complementary advantages, we propose a global seamless soil moisture data generation method integrating microwave remote sensing and model simulation. Specifically, for the vacant area of 9 km soil moisture data of soil moisture active passive satellite, GLDAS Noah 0.25° model assimilation data is introduced to establish the spatiotemporal fusion model between them. Through this model, Noah data is downscaled to 9 km to fill the vacant area. As for some areas that have not been filled, Noah data is used for fitting, interpolation and filling, and the residual correction is further carried out based on Poisson equation method, and then the spatiotemporal seamless 9 km daily soil moisture data is generated. Results: We evaluated the experimental results by simulation experiment evaluation, site data evaluation and triple collocation evaluation. Experiments show that the global seamless soil moisture data produced in this paper performs well in all evaluation methods. Conclusions: This method can effectively combine the spatial resolution advantage of remote sensing observation with the spatiotemporal continuity advantage of model simulation, provide spatiotemporal seamless global soil moisture data, and better meet the needs of global scale water cycle monitoring and water resources management. [ABSTRACT FROM AUTHOR]

Aiming at the problem of image registration difficulties caused by the significant differences of multi-source remote sensing images due to the sensor type, temporal phase and illumination conditions, this paper proposes a multisource image registration method guided by first-order Gaussian steerable filters. Methods: First, based on the geometric reference information of the image, the overlapping area of the reference image and the image to be aligned in the image space is calculated, the overlapping area is uniformly partitioned with the reference image surface as the reference, the corresponding image block to be aligned is calculated by the rational function model and the digital elevation model, and the affine transformation model is established to geometrically correct the image block to be aligned, so as to realize the coarse registration between the local images. Second, for feature detection, anti-cluster features from accelerated segmentation test improved by the chunking uniformity checking strategy are constructed to obtain a large number of uniformly distributed feature points. And for feature description, a set of multi-scale, multi-directional first-order Gaussian steerable filters are constructed to convolve the image, and by pooling the convolution results to achieve feature dimensionality reduction, a multi-source consistent feature description is constructed. Finally, feature matching is performed based on the nearest-neighbor principle, and high-precision correspondences are obtained by eliminating mismatches. And bundle adjustment is further performed based on the rational function model to calibrate the rational polynomial coefficients of the images to be aligned and geometrically correct the images, so as to realize the fine registration of the images. Results and Conclusions: Experimental results using multiple pairs of satellite multisource images show that the accuracy of the proposed method is better than 1 pixel on multitemporal optical data, optical-infrared data, and 1.5 pixels on optical-synthetic aperture radar data, and the computational efficiency is more than doubled compared to the existing similar methods. [ABSTRACT FROM AUTHOR]

Objectives: Spaceborne global navigation satellite system-reflectometry （GNSS-R） has been demonstrated to be effective in retrieving global sea surface wind speed, and many missions have been launched in orbit or planned. “Jilin-1” Kuanfu-01B satellite, as a commercial remote sensing satellite of China, equipped a GNSS-R instrument to observe Earth's physical parameters. This paper aims to preliminarily analyze and demonstrate the effectiveness and feasibility of GNSS-R data from “Jilin-1” Kuanfu01B satellite for retrieving sea surface wind speed. Methods: Three basic observables sensitive to wind speed, namely the delay-Doppler map (DDM) peak signal-to-noise ratio (SNR), the leading-edge slope and trailing-edge slope of the normalized delay waveform, are extracted from the DDMs of “Jilin-1” Kuanfu-01B satellite, and are used to demonstrate and assess the potential in retrieving wind speed. Three basic observables have obvious inversely proportional relationships with wind speed; therefore, an empirical exponential function is used to develop the geographic model function (GMF) between the basic observables and wind speed. Furthermore, the root mean square error (RMSE) between the retrieved wind speeds and European Centre for Medium-Range Weather Forecasts (ECMWF) data are obtained. Results: The RMSE of the retrieved wind speed using DDM peak SNR, leading-edge slope and trailingedge slope are 3.22 m/s, 3.67 m/s and 3.49 m/s without data quality control and data calibration. When SNR is controlled to be larger than −6 dB, the RMSE of the retrieved wind speed using the DDM peak SNR reduces to be within 2 m/s, and the useful data are 78.11 % of that without data quality control. For leading-edge slope and trailing-edge slope, when SNR threshold is −3 dB, the RMSE of retrieved wind speed is also decreased to be within 2 m/s, and the useful data reduce to 53.74 %. Conclusions: The results show that the GNSS-R data from “Jilin-1” Kuanfu-01B satellite can be used to measure sea surface wind speed. The data used in the paper are uncalibrated, therefore, the retrieved performances are not optimal. It is a key work to calibrate the influence of transmitted power, transmission gain, and propagation path on the DDM. [ABSTRACT FROM AUTHOR]

As the most dynamic component of terrestrial water cycling, soil moisture plays a pivotal role in the formation, transformation, and consumption of surface water. Simultaneously, it serves as a crucial parameter influencing hydrological processes, vegetation status, and climatic conditions. Therefore, highprecision, high spatiotemporal resolution soil moisture data holds significant importance across various fields, including agriculture, forestry, and meteorology. Presently, traditional methods such as the timeconsuming and labor-intensive drying-weighing technique are inadequate for large-scale monitoring demands, highlighting the advantages of remote sensing methods. However, optical sensors are susceptible to cloud cover and vegetation obscuration, while microwave remote sensing technology faces challenges in balancing spatial and temporal resolutions. The space-borne global navigation satellite system-reflectometry (GNSS R) technology, characterized by short revisit cycles and pseudo-random sampling, presents a new opportunity for soil moisture retrieval. This paper begins by introducing the fundamental theory of soil moisture retrieval using GNSS-R and analyzes error sources during the retrieval process. This includes the scattering and attenuation effects of surface factors such as vegetation and surface roughness on reflected signals, as well as the impact of water bodies on reflectivity. Existing error correction methods and their limitations are summarized, and potential avenues for improvement are discussed. Subsequently, the research progress in recent years regarding space-borne GNSS-R soil moisture retrieval is analyzed from three perspectives: (1) The development status of space-borne GNSS-R constellation. (2)An overview of algorithms for space-borne GNSS-R soil moisture retrieval, including empirical models, semi-empirical models, and machine learning methods. (3) Advancements and directions in integrating GNSS-R with other data for soil moisture retrieval. In conclusion, based on the scattering mechanism of reflected signals, the selection of auxiliary data, and the utilization of incident angle information, this paper discusses the technical challenges faced in current soil moisture retrieval using space-borne GNSS-R technology and provides insights into future research directions. [ABSTRACT FROM AUTHOR]

Objectives: Digital surface model is of great significance in the fields of real-life 3D modeling, smart city construction, natural resources management, geoscience research, and hydrology and water resources management. However, dense matching, as a core step in generating digital surface models, is prone to matching failures in regions with a lack of texture, disparity gap and inconsistent illumination. The sparse point cloud data with high accuracy and extensive coverage after aerial triangulation, which can be used as a priori information to improve the accuracy of dense matching results. Methods: First, this paper proposes a sparse point cloud guidance (SPCG) method for generating digital surface models of aerial images. The method aims to constrain the dense matching of images using sparse point cloud encrypted by aerial triangulation. The sparse point cloud guidance first selects stereo image pairs with good geometric configurations, high overlap, and extensive coverage. Then, the number of sparse points is extended by using the closest proximity clustering and pyramid propagation methods. Additionally, the matching cost of the extended points is optimized by using the improved Gaussian function to enhance the accuracy of the dense matching results. Finally, the sparse point cloud is fused with the dense matching point cloud to generate the digital surface model. Results: Experiments on simulated stereo images and real aerial stereo images show that the optimized semi-global matching by the SPCG method in this paper significantly improves the matching accuracy of the original semi-global matching algorithm and outperforms the semi-global matching optimized by the Gaussian method and the deep learning method, pyramid stereo matching network. Numerically, the percentages of disparity maps generated by semi-global matching with greater than 1, 2, or 3 pixels difference from the true disparities are 46.72%, 32.83%, or 27.32%, respectively, whereas the SPCG method decreases by 7.67%, 9.75%, or 10.28%, respectively, compared to the former. The experimental results of the multiview aerial images show that the SPCG method accurately generates the digital surface model of the whole survey area, and it is better than the digital surface model generated by the superior SURE software in both qualitative and quantitative aspects. Conclusions: Compared to the original dense matching, sparse point cloud-guided dense matching improves the matching accuracy in difficult matching regions such as weak textures, repetitive textures and depth discontinuities. In turn, high precision and high density point clouds are generated. A complete digital surface model is generated by the fusion of the densely matched point clouds. [ABSTRACT FROM AUTHOR]

Based on the application in the field of photogrammetry, this paper retrospects the binocular image dense matching method. First, the two categories of local dense image matching and global dense image matching methods are concisely compared, and the advantages, disadvantages, and main challenges of each method are pointed out. Then, the building occlusion problem that is difficult to deal with in dense matching is analyzed. The common building occlusion phenomenon in aerial photography is divided into five types, and the pertinence of the existing occlusion detection and filling algorithms is expounded, which provides technical ideas for solving the bottleneck problem in dense image matching. Finally, the development trend of dense image matching of binocular images is prospected. This paper can help readers fully understand the traditional binocular image dense matching technology, and it will be beneficial to the research of dense image matching based on deep learning. [ABSTRACT FROM AUTHOR]

High-precision global navigation satellite system (GNSS) positioning technology, as a hot spot in the field of satellite navigation, has expanded the in-depth development of the location-based service area, and has been widely used in smart cities, intelligent transportation, autonomous driving and other fields. First, this paper presents the current status and development of GNSS positioning technologies.Then, it explains the common basic theories such as integer ambiguity fixation, functional model and stochastic model fusion involved in multi-frequency multi-GNSS high-precision positioning, as well as the key techniques for improving GNSS positioning performance such as the non-model error processing, the cycle slip detection and data interruption repair. Based on the innovations of basic theories and key techniques, this paper introduces four new technologies for multi-frequency multi-GNSS high-precision positioning applications developed by Tongji University GNSS team in recent years. Finally, we remark the future development trend and challenges of GNSS-based high-precision positioning. [ABSTRACT FROM AUTHOR]

The deformation response models of earthquake cycle play a crucial role in studying the entire dynamic process of earthquakes, from inception, nucleation and rupture, to postseismic adjustments, as well as in exploring the rheological properties of Earth' s crust and mantle. They have significant implications in earthquake prediction, forecasting, seismic hazard mitigation and risk assessment. First, this paper introduces the characteristics of deformation during the interseismic phase of the earthquake cycle. It categorizes and summarizes the corresponding models from the perspectives of dimensionality and elastic-visc-oelastic models, and it elaborates on the characteristics, limitations and historical evolution of simulation methods for coseismic deformation, with a particular focus on mainstream approaches for determining the geometry of coseismic faults and selecting smoothing schemes. Then, this paper discusses the spatiotemporal evolution features of postseismic deformation, and categorizes them according to deformation mechanisms, including poroelastic rebound, viscoelastic relaxation and postseismic afterslip. It also outlines the characteristics and limitations of simulation methods for deformation phenomena and provides a classification summary of current mainstream postseismic combined models. Furthermore, this paper explores the opportunities and challenges faced by earthquake cycle deformation simulation methods in the rapidly accumulating data environment. It emphasizes the importance of three-dimensional viscoelastic models, considering the viscoelastic relaxation effects of Earth's crust and mantle. Finally, this paper offers prospects for future research directions, including single-fault earthquake cycle deformation simulation, postseismic mechanisms, dynamic modeling and interpretation of possible earthquake cycle deformation phenomena. [ABSTRACT FROM AUTHOR]

The “cloud control” paradigm, grounded on geocoded data such as digital orthophoto map (DOM) and digital elevation model (DEM), is gradually maturing for applications in medium and small-scale mapping. However, its essence of obtaining control points from the object space inherently carries errors within the DOM and DEM products, subsequently hindering its applicability in large-scale mapping scenarios. Constructing constraints based on original images with known accurate orientation parameters and adjusted object space tie points represents a theoretically superior cloud control paradigm. This paper proposes developing this image-based cloud control paradigm into a standardized survey and mapping product, termed digital control photo (DCP), aiming to better ensure positioning accuracy while reducing redundant labor and computational power consumption in revisit aerial survey tasks. This paper also introduces the application of DCP in calibration of large-format aerial cameras, geometric positioning of unmanned aerial vehicle images and satellite images, and suggests several recommendations for promoting the establishment of a DCP application system. [ABSTRACT FROM AUTHOR]

Objectives: In 2015, the Mw 7.8 earthquake in Nepal caused landslides and seriously impacted Nepal and its surrounding areas. Methods: In this paper, the Sindhupalchowk area, which is close to the epicenter of the earthquake, is selected as the study area. First, based on the L-band ALOS-2 and C-band Sentinel-1A synthetic aperture radar data, the detection and identification of post-earthquake landslides are carried out by using stacking-InSAR(interferometric synthetic aperture radar) technology. Second, the two-dimensional time series deformation characteristics of a typical landslide are obtained by combining the ascending and descending orbit data and multi-dimensional small baseline subset-InSAR technology. Results: Combined with optical images, 14 hidden points of landslide are delineated. The main deformation of the typical landslide occurs in the horizontal east-west direction, and the maximum deformation rate is −69 mm/a. Conclusions: By analyzing the trend and periodic term deformation signals in time series of the typical landslide, it is found that the earthquake has a significant acceleration effect on the landslide movement, and the increase of rainfall makes the landslide displacement show a periodic change from August to November every year. This paper can provide a reference for post-earthquake landslide monitoring re⁃ search. [ABSTRACT FROM AUTHOR]

Objectives: Reconstruction of ice flow velocity on the Antarctic Ice Sheet plays a critical role in estimating mass balance and contribution to global sea level rise. However, due to the low availability of optical satellite images before the 1990s, the time span of image pairs ranges from several years to more than ten years, leading to velocity overestimations. In this study, we correct the overestimated ice flow velocity based on the Lagrangian framework. After correction, we analyze the distribution of the overestimation values in the East Antarctic Ice Sheet (EAIS) according to the space and magnitude. Methods: We formulate a set of correction rules, including filtering points of ice flow velocity before correction and whether to retain results after correction. As a transition zone between grounded ice and ice shelf, ice flow velocity near the grounding line is one of the most important parameters of the ice sheet mass balance. In this paper, the overestimations of ice flow velocity are mainly studied from three aspects: First from different regions including inland region, grounding region and ice shelf region, then from different ocean sectors including Weddell Sea, West Indian Ocean, East Indian Ocean and Ross Sea, finally from different glaciers/ice shelves. Results: The average overestimation of ice flow velocity in EAIS is about 28 m/a. Both the grounding and ice shelf regions have larger overestimations than the inland region. The peak correction appears at ±25 km near the grounding line, and the maximum correction exceeds 300 m/a. As for ocean sectors, the overestimation of velocity is the largest in the East Indian Ocean. The average corrections of the Holmes Glacier, Ninnis Glacier, Publications Ice Shelf, and Totten Glacier in this sector exceed 50 m/a, much larger than that of EAIS. Conclusions: This paper shows that only a very small proportion of ice velocity points can meet the correction rules. The points show differences in regional distribution and they are concentrated mainly in the fast-flowing areas such as glaciers and ice shelves. Especially at the grounding zone and ice front areas of ice shelves, the ice flow velocity can be greatly overestimated. By analyzing the corrections and ice flow velocities in different ocean sectors, we found that the spatial acceleration trend of velocity in the East Indian Ocean is obvious and may be related to the interaction between the ocean and the ice shelves. The distribution of corrections varies among different ice shelves and glaciers. For example, Amery Ice Shelf has a larger correction at the confluence of three upstream glaciers, while Shirase Glacier has a larger correction in the inland region. The correction results in this paper have been applied to historical ice flow velocity products for East Antarctica from 1960s to 1980s. The overestimation correction of ice velocity is important for reducing the uncertainty in estimation and forecasting the contribution of Antarctic mass loss to the global sea level rise. [ABSTRACT FROM AUTHOR]

Multi-epoch interferometric synthetic aperture radar (InSAR) is a highly effective technique for monitoring deformation in urban areas. However, interpreting InSAR deformation can be challenging due to various factors, including inherent geometric imaging distortion, the intricate structure and deformation properties of targets in urban scenes, and the multiple scattering of microwave signals between objects in urban scenes. This paper discusses the challenges involved in interpreting time-series InSAR deformation: (1) Precisely identifying the location of deformation signals and linking them to their corresponding objects, i. e., determining where the deformation signal occurs, (2) understanding the mechanisms and factors that cause the detected deformation signals, i.e., determining what the deformation signal represents, (3) establishing the connection among the detected deformation signals, the deformation events, and the scattering mechanisms. We suggest a parametric framework to improve the accurate interpretation of InSAR deformation. This framework includes several factors, including kinematic characteristics (deformation rate, cumulative deformation, deformation gradient, and deformation model), geometric parameters (position, size, structure, orientation, and roughness), semantic information (land cover type, terrain morphology, texture, and auxiliary information on natural and anthropogenic disturbance) and physical properties (scattering mechanism, penetrability, extensibility, conductivity, and thermal conductivity). Our approach aims to enhance the representation of coherent points for a better understanding of InSAR deformation. This paper offers a comprehensive overview of the advancements achieved in extracting parameters of InSAR coherent points and interpreting deformation based on geometric parameters, semantic information, and physical properties. High-precision 3D positioning is crucial for InSAR fine monitoring in urban areas. It helps determine the source of deformation signals and facilitates the analysis of deformation mechanisms. Semantic information, such as 3D models, high-resolution optical images, laser point cloud data, and land use data, can aid in interpreting InSAR deformation. By combining InSAR deformation data with a deep learning approach, there is an opportunity to interpret deformations effectively. In urban environments, the scattering mechanism of ground objects is complex. Multiple scattering signals can provide effective observations of deformation and information about the target's size. However, combining the scattering mechanism of synthetic aperture radar signals to carry out parameter inversion and deformation mechanism interpretation of urban target terrain remains a challenge. The framework, which considers the geometric parameters, semantic information, and physical attributes of InSAR coherent points, will be crucial for deformation interpretation and mechanism cognition. This framework will enable fine deformation monitoring, intelligent recognition, and application in future urban scenes. [ABSTRACT FROM AUTHOR]

Objectives: The Mw 7.8 earthquake occurred in Turkey on February 6th, 2023, which is a rare powerful earthquake in recent years. Building collapse detection is an important part of post-earthquake damage assessment and emergency rescue. Facing the emergency rescue and post-earthquake disaster assessment, this paper aims to comprehensively utilize the intelligent interpretation technology of satellite remote sensing, coupled with high-resolution optical daily and night-time light remote sensing images, and to efficiently assess the collapsed buildings in the extreme disaster areas with low time and labor costs. Methods: In this paper, we present a multi-level rapid localization scheme for disaster-affected regions. First, the proposed approach involves the utilization of low- and medium-resolution night-time light data to estimate the extent of extreme disaster areas. Subsequently, high-resolution optical remote sensing imagery is employed to extract the collapsed buildings. By training a semi-supervised deep learning network with a small number of manually collected samples, the scheme automatically identifies collapse areas, thereby facilitating the precise localization of collapsed buildings. A high-resolution architecture network, HRNet, is employed in this study. HRNet has the advantages in both sematic and spatial feature representation, and can capture the collapsed buildings with muti-scale characteristics. Results: Six cities along the border between the Arab Plate and the Eurasian Plate are severely affected, from Adelman to Kahraman Marrash, and then to Antakia. A total of 2 377 collapsed buildings are detected in the extreme disaster areas. Except for Nurdaji and Antakia cities, where the proportion of collapsed buildings exceeded 2%, the proportion of collapsed buildings in the other four cities is close to 1%. After manual verification, the recall rate of the intelligent interpretation plan is between 74% and 93%. Conclusions: The efficiency and effectiveness of this approach are demonstrated through its application in assessing the building collapse situation after the Mw 7.8 earthquake in Turkey. The proposed work can serve as a timely reference for macro decision-making in post-earthquake emergency rescue operations. [ABSTRACT FROM AUTHOR]

Objectives: SSS (sea surface salinity) is an important physical and chemical indicator describing ocean state, and is also one of the most critical variables in global water cycle. In order to achieve largescale and continuous observation of the ocean salinity, SMOS (soil moisture and ocean salinity) mission is launched by European Space Agency to access brightness temperature for SSS retrieval. However, researches have pointed out that the accuracy of SMOS ocean salinity products cannot fully achieve scientific expect. Thus, a novel and effective method is desired for improving SMOS salinity. Methods: This paper selects study area in the Southeast Pacific, where the region is relatively less affected by radio frequency interference. The downloaded array for real-time geostrophic oceanography SSS data, SMOS level2 ocean salinity products and the Auxiliary data in the whole year of 2018 are matched both temporally and spatially. Based on the remote sensing mechanism and radiation transfer theory, five variables, including the first Stokes parameter (TH+TV), SST, UN10, VN10 and Ω are selected as sensitive factors. The ocean salinity retrieval model based on random forest is first established, and GridSearchCV(GS) algorithm is applied to optimize superparameters of the initial model, which is able to improve SSS accuracy further. Results: The experiment results show that： (1) For objective measurement, the mean absolute error (MAE) and root mean square error (RMSE) are adopted to compare the accuracy of the produced SSS from two established models and the SMOS products, in terms of Argo in-situ SSS as reference. According to experiments, MAEs of basic RF(random forest) model and GS‐RF model are both 0.08, which is evidently lower than SMOS SSS product's 0.66. Similarly, RMSEs of two models are also remarkable lower than SMOS's 0.93, while GS RF model is 0.14, indicating its slight advantage over RF model, which is 0.15. (2) All SSS retrieved from RF model and GS RF model distribute concentratedly around 36 psu, while SMOS ocean salinity product occasionally presents abnormal SSS that beyond 40 psu or below 30 psu, which is inappropriate. (3) In the testing set, error of SMOS SSS can reach −12.9 psu and 6 psu, while errors of two established models are no more than ±1.3 psu, noticeably lower than SMOS product. Meanwhile, GS‐RF model also present slight advantage than basic RF model on error measurement. Conclusions: Through experiments and analyses, we found that it is feasible to establish sea surface salinity retrieval methods based on random forest model, by deliberately selecting SSS sensitive factors, and advancedly optimizing super-parameters of model, especially with GridSearchCV algorithm. This paper has achieved reliable and satisfactory results, which have greatly improved the accuracy of ocean salinity data compared with SMOS products, thus benefit for relevant marine studies. [ABSTRACT FROM AUTHOR]

Objectives: Vegetation plays an important role in ecological environment monitoring. Studying vegetation cover change can provide reference for regional ecological environment protection. The region of Altay in Xinjiang belongs to the ecological function zone of water conservation mountain grassland, with rich natural resources and beautiful scenery. However， the ecological environmental problems were gradually emerging under the national development strategies. Therefore, monitoring the dynamics of vegetation change in the area with multi-source remote sensing data is necessary to explore the relationship between the economic development and ecological environment protection. Methods: This paper uses multisource remote sensing long time series data with breaks for additive seasonal and trend, topographical position and geostatistical analysis methods to monitor the vegetation cover dynamics in Altay of Xinjiang, China during 2000—2019. Results: This paper processed large number of multisource remote sensing data along with human activity data and showed that during the study period, the number of vegetation breakpoints increased year by year from 2003 to 2009, and then gradually decreased after 2009. Meanwhile, seven types of vegetation breakpoint trends were detected, among them, the more number of breakpoints from degradation to growth type was identified, and the types of disturbance degradation and continuous degradation were less. From the interannual change curve, the dynamic change of vegetation coverage showed a trend of first decreasing and then increasing, that is, the vegetation coverage showed a degradation trend from 2000 to 2008 and a significant improvement trend from 2008 to 2019, and the degradation was greater than the improvement. Meanwhile, the east slope with an elevation of more than 900 m and a slope of more than 15° (from north to east and from south to east) is the dominant topographic location of vegetation degradation type, accounting for 62.4%. Conclusions: The protection of ecological environment in Altay of Xinjiang still needs to be further strengthened and effective protection measures should be taken. [ABSTRACT FROM AUTHOR]

Objectives: It is difficult for the traditional traffic information acquisition method to obtain the real-time dynamic traffic flow information of large scope and full coverage. With urbanization, the increasing number of supertall buildings makes the city skyline a favorable platform for earth observation. This paper studies the urban remote sensing video production and traffic information extraction method using the urban skyline as the observation platform. Methods: First, the earth observation data shooting is carried out in the super-tall buildings. Then, the original oblique observation data is corrected by orthography, which is further fused with the satellite image to generate a large range of remote sensing video data of the city. Finally, we train a deep learning model and use it for vehicle classification and identification. The number and density of vehicles in the area are calculated based on the identification results. Results: This paper carries out data collection and processing analysis on the sightseeing floor of Ping, an Building in Shenzhen. Results show that the proposed method can produce low-cost, long-time, large-scale and high-quality urban remote sensing video. The vehicle detection based on the remote sensing video have high accuracy. Conclusions: The proposed framework can be used to effectively monitor the regional traffic flow and sever for the smart city management. Furthermore, based on its capability of long-time, high-resolution, full-coverage and real-time remote sensing, this method can also be applied to many other urban management fields, such as urban disaster emergency response, crowd monitoring, and construction site monitoring. [ABSTRACT FROM AUTHOR]

Objectives: Global navigation satellite system (GNSS) coordinate time series provide important data support for the study of crustal movement and deformation, and plate tectonics. Due to the noise caused by various external factors, the GNSS coordinate time series cannot reflect the real motion information of the station well. To effectively reduce the noise in the GNSS time series, we adopted a noise-reduction method combining genetic algorithm (GA) and variational mode decomposition (VMD-GA-VMD). Methods: First, the genetic algorithm was used to optimize VMD parameters, and the envelope entropy of the input signal was used as the fitness function of the genetic algorithm to find the optimal VMD parameter combination suitable for the signal. According to the optimized parameters, the signal was decomposed by VMD to obtain a series of modal components. Then we calculated the multi-scale permutation entropy (MPE) of each component and then regarded the MPE as the criterion of the noise component. Finally, according to the MPE, the noise components were identified and removed, and the remaining components were reconstructed to obtain the noise-reduced signal. In this paper, the noise reduction effect of GAVMD was analyzed through the example of noise reduction of analog signal and observation data, and compared with wavelet denoising (WD) and empirical mode decomposition (EMD) methods. Results: The experiment results show that: (1) The noise reduction results from the analog signals show that WD and EMD have the incomplete and excessive troubles on the noise reduction, respectively. However, GA-VMD can effectively eliminate noise and retain effective signals. From the evaluation index, compared with WD and EMD, the signal-to-noise ratio were increased by 5.18 dB and 2.91 dB, the correlation coefficient by 0.05 and 0.02, respectively, when using GA-VMD. (2) For the complex observation, we used the noise and velocity uncertainty as accuracy indicators to evaluate the noise reduction effects of the three methods. The results show that WD can only extract a part of the white noise, while EMD and GA-VMD can completely remove the white noise. GA-VMD can reduce the flicker noise to the range of 0 to 6 mm/a0.25. For the velocity uncertainty, the average gain rates of GA-VMD relative to the WD and EMD are 69% and 15.33%, respectively. GA-VMD has an average correction rate of 79.89% and 84.46% for the velocity uncertainty and flicker noise of GNSS coordinate time series. Conclusions: Therefore, GA-VMD is the most effective one among the three noise reduction methods, which can effectively reduce the noise in the GNSS time series and improve its accuracy. However, in this paper, we only discussed the effect of GA on VMD parameter optimization without comparing it with other methods. Hence, it will be the key for studying the advantages and shortcomings of those optimization algorithms in the selection of VMD, and improving the accuracy on the GNSS time series in the future. [ABSTRACT FROM AUTHOR]

Objectives: Millimeter wave radar has been widely used in automotive industry and other fields, but its application is mainly limited to the environmental perception of obstacles or specific tasks. At present, there is little research on the application of millimeter wave radar in the field of navigation and positioning. Methods: This paper first studies the raw data processing principle of millimeter wave radar, and then designs an indoor ego localization method which only depends on a low‐cost millimeter wave radar. The process mainly includes extracting centroid feature points using density based spatial clustering of applications with noise (DBSCAN) algorithm, matching centroid feature point pairs through nearest neighbor criterion, constructing nonlinear optimization function and solving positioning results using levenberg marquardt method. Results and Conclusions: Experiments show that indoor navigation and positioning can be solved in real time by using a low‐cost millimeter wave radar. Under static conditions, the average horizontal positioning accuracy can reach sub centimeter level (mean value is 0.82 cm and standard deviation is 0.47 cm). Under dynamic conditions, the absolute trajectory error can reach 0.66 m and the average heading angle error can reach 4.58°, which shows the feasibility of ego‐localization of low‐cost millimeter wave radar. Finally, this paper discusses the problems and feasible research ideas of low cost millimeter wave radar in navigation and positioning. [ABSTRACT FROM AUTHOR]

The Yangtze River Basin is an important part of China's ecology and social economy, and the land use/cover status is experiencing rapid changes. As the development of remote sensing technology, the Internet of things and social media, we can obtain huge, multi-dimensional relevant data of land use/cover, the process of land use/cover, driving factors and ecological response of research, the development of the Yangtze river basin planning and ecological protection provides a decision-making basis. First, this paper reviews the current research on land use/cover change in the Yangtze river basin, and then expounds the re⁃ search methods and conclusions in terms of the driving mechanism and ecological effects of land use/cover change. Finally, this paper discusses the difficulties and challenges faced by existing researches from the differences of land use/cover products, regional driving factors, complexity of ecological environmental change process and comprehensive evaluation of ecological service value and the possible directions of future researches. [ABSTRACT FROM AUTHOR]

Objectives: Gauss-Markov model, the model with condition equations and Gauss-Helmert model are three classical adjustment models. Gauss-Helmert model can be seen as the combined form of the other two models. They are usually distinguished by the number of unknown parameters. However, this classification method cannot reflect the essential relationship between them. In addition, the connections between Gauss-Markov model and the conditional model has been intensively investigated, while that of Gauss-Helmert model has been ignored. Methods: Some equivalent objective functions of the adjust⁃ ment of the least-square in the context of Gauss-Markov model and the conditional model are proposed based on the projection theory and the dual optimization technique. Based on the theory of matrix analysis, this paper realizes the conversion between Gauss-Markov model and Gauss-Helmert model according to partial orthogonality of coefficient matrices, thus completing the structure of these models and revealing the mathematical relationships. Results: The least-square adjustment within Gauss-Markov model and the model with condition equations are explained geometrically by employing the linear space and projector. The equivalence of the objective functions of these two models are also expounded. Combining the Pythagorean theorem and duality theory, we extend them to four equivalent objective functions and illustrate the construction principle. The transformation between Gauss-Markov model and Gauss-Helmert model are achieved by the so-called partial-orthogonality. Conclusions: The geometric reasoning of the least-squares adjustment is analyzed once more. Two objection functions are newly proposed, and the geometric relation and connotation between objective functions are revealed. Gauss-Helmert model is not only the combined model in terms of the model appearance, these three adjustment models can actually be mutually trans⁃ formed. As a result, the understanding of the mathematical nature of adjustment models and the corresponding estimation in this paper has theoretical and practical value. [ABSTRACT FROM AUTHOR]

Objectives: In view of the low accuracy of flow forecast in tidal reach due to the combined influence of runoff and tide, combining the ability of variational mode decomposition (VMD) to process non-stationary signals and the advantages of back propagation (BP) neural network to deal with nonlinear problems, this paper proposes a new method for discharge prediction in tidal reach based on the combination model of tidal current harmonic analysis and VMD-BP neural network. Methods: First, the harmonic analysis method is used to separate the tidal discharge and residual discharge from the original discharge. Second, a BP neural network is constructed based on error backpropagation algorithm. Then, the BP neural network is used to simulate and train the power flow data and residual current data processed by VMD. Finally, the output tidal discharge and residual discharge components are superimposed and reconstructed to obtain the final discharge prediction results. Results: Experiments performed at the Xuliujing section of the Yangtze Estuary show that the discharge prediction accuracy of BP neural network method is better than that of traditional harmonic analysis method, with the root mean square error (RMSE) is decreased by about 3 400 m³ /s, and the relative standard accuracy (RSD) is increased by about 6%. In addition, compared with the traditional method, RMSE of the proposed method in this paper is decreased by about 5 000 m³ /s, and the RSD is increased by about 9%. Conclusions: The combined model based on harmonic analysis and VMD-BP neural network can effectively improve the prediction accuracy of discharge in tidal reach. At the same time, it also provides a new idea for the discharge prediction in the waters with variable flow conditions. [ABSTRACT FROM AUTHOR]

Objectives: High definition map (HDM) is the “sensor” for automated driving (AD), which integrates the real-time data collected by various sensors and the prior data collected in the early stage, and serves the application of AD, including providing the necessary prior data and the corresponding over-the-horizon perception, calibration positioning, dynamic planning and decision-making and control. However, there still exists a gap between the supply of HDM data and the demand for AD knowledge, including difficulties in data retrieval due to the large volume of data and challenges in semantic understanding and intelligent decision-making due to weak data correlation. Therefore, how to balance the data supply of HDM and the knowledge demand of AD is the main goal of this paper. Methods: Knowledge graph (KG) is a representation of knowledge in a graph structure to describe entities and their relationships, involving entity extraction and relationship extraction, so that it can make the AD with interpretable, understandable and inferential. This means that KG can serve as an alternative and explicit simulation of the human mind and map cognition that driverless vehicles are missing. Therefore, we introduce KG on the basis of HDM data, and propose a framework of HDM-KG-driving task, so as to support the transformation of HDM data to AD knowledge. The construction of HDM-KG adopts the top-down method, that is, the pattern layer is first followed by the data layer. As the conceptual hierarchy of KG, pattern layer defines the concept, attribute and relation of map ontology, and it can explicitly describe the indirect information and implicit correlation information of map domain from the perspective of traffic. The data layer is an instantiation of the pattern layer, a way to populate the pattern layer through instance matching. Results: To verify the validity of the proposed method, a virtual simulation dataset based on OpenStreetMap data is constructed and converted into two formats. The first is the OpenDRIVE format, which is used to build the KG of static data, including the static knowledge of roads, lanes, intersections, road signs and road markings. The other is the data format of CARLA AD-simulator, which is used for AD simulation and the construction of dynamic real-time KG, including the dynamic knowledge of self-driving cars and vehicles in front of them, pedestrians and other traffic participants. Through the application of the resulting HDM-KG in semantic query, knowledge reasoning and local decision planning, the results show that the HDM-KG explicitly expresses the semantic relationship, enhances the application of HDM in AD, and can further improve AD by connecting additional knowledge such as safety requirements, traffic rules or scene background. Conclusions: HDM-KG can structurally and explicitly express the traffic semantics contained in map objects and the semantic relations between objects, traffic rule constraints, traffic scenes and other knowledge, and realize the three-layer evolution of data-information-knowledge of data sources-HDM-KG, providing support for semantic relation query, logical reasoning and motion planning required by driving tasks. In addition, the introduction of KG eliminates the need for driving tasks to perceive and predict objects and semantic relationships from data collected by sensors, which are explicitly defined by HDM-KG, reducing computational latency and improving safety and reliability. The top-down method of constructing KG will cost more labor and time, and the subsequent research will carry out the bottom-up method to automatically fill the KG with multi-source data. [ABSTRACT FROM AUTHOR]

Objectives: To analyze spatial disparities in urban living service facilities based on supply and demand service levels, we not only detect the spatial distribution characteristics of facilities, but also gain indepth insights into the underlying causes of these differences. Methods: This paper specifically focuses on Jianghan district in Wuhan city and conducts a case study by using the food market as a representative facility. The analysis involves constructing spatial disparity in accessibility of services (SDAS) index to assess the disparities between the supply and demand of life service facilities. Subsequently, SDAS is optimized by an improved gravity model based on the road traffic network and income levels of residents. Results: The results show that most community units in Jianghan district exhibit a high level of disparity between supply and demand, with notable differences observed in Wangjiadun central business district. The service capacity of life service facilities around communities significantly influences SDAS value under various conditions. The multi-mode case provides a more comprehensive reflection of the spatial differences between the facility accessibility and the needs of community units, offering a nuanced view of the actual situation. Conclusions: The results contribute to shaping city structure and facilitating optimal allocation of life service facility resources. It can provide scientific basis for layout optimization of public facilities in urban area. [ABSTRACT FROM AUTHOR]