Intelligent multi-mode earth surface change monitoring method and system

Abstract

The invention discloses an intelligent multi-modal earth surface change monitoring method and system. The method comprises the steps of multi-source data acquisition, data preprocessing, data fusion, AI analysis, real-time monitoring and early warning, three-dimensional modeling, user customization service and the like. Multi-modal data such as high-resolution satellite images, unmanned aerial vehicle aerial photography, ground LiDAR data and environmental sensor data are fused, and intelligent analysis is performed by using a deep learning algorithm, so that comprehensive, efficient and accurate monitoring of earth surface changes is realized. According to the method, a scientific basis can be provided for the fields of urban planning, disaster early warning, natural resource management and the like, user customized services are supported, and personalized requirements of different users are met.

Description

Technical Field

[0001] The present invention relates to the field of geographic information technology, and particularly to an intelligent multi-modal surface change monitoring method and system, which realizes efficient and accurate monitoring of surface changes by integrating multiple data sources and using artificial intelligence technology. Background Art

[0002] In the field of geographic information monitoring, surface change monitoring is of great significance for urban planning, disaster warning, natural resource management, etc. Traditional surface change monitoring methods mainly rely on a single data source, such as satellite images or ground sensors, and there are problems such as limited monitoring accuracy and inability to comprehensively reflect surface changes.

[0003] Monitoring methods based on a single data source are difficult to obtain multi-dimensional information of the surface. For example, although satellite images can provide large-scale surface coverage information, they may not be able to accurately capture some local and subtle changes; while ground sensors can provide high-precision local data, but the coverage range is limited. In addition, traditional methods are less efficient in data processing and analysis, often requiring a large amount of manual intervention, and it is difficult to achieve real-time monitoring and warning.

[0004] With the development of technologies such as remote sensing technology, artificial intelligence, and the Internet of Things, multi-modal data fusion and intelligent analysis have become the development trend of surface change monitoring. However, existing multi-modal monitoring methods still have deficiencies in aspects such as the accuracy and efficiency of data fusion and the accuracy of intelligent analysis. Summary of the Invention

[0005] The present invention provides an intelligent multi-modal surface change monitoring method, which realizes comprehensive, efficient, and accurate monitoring of surface changes by integrating multi-modal data such as high-resolution satellite images, unmanned aerial vehicle aerial photography, ground LiDAR data, and environmental sensor data, and using deep learning algorithms for intelligent analysis, providing a scientific basis for fields such as urban planning, disaster warning, and natural resource management. The specific steps are as follows:

[0006] Multi-source data collection: Collect multi-modal data of surface images, terrain, and environmental parameters;

[0007] Data preprocessing: Perform denoising, calibration, and enhancement processing on the collected original multi-modal data;

[0008] Data fusion: Use spatial registration algorithms and multi-scale fusion technologies to fuse data from different data sources to generate a comprehensive data set;

[0009] AI analysis: Use deep learning algorithms to analyze the comprehensive data set and automatically identify surface change features;

[0010] Real-time monitoring and early warning: Combining Internet of Things technology, it conducts real-time monitoring of surface changes. When specific surface changes are detected, the early warning system is triggered;

[0011] Three-dimensional modeling: Using multi-modal data to construct a high-precision three-dimensional terrain model and supporting time series analysis to display the dynamic process of surface changes;

[0012] User-customized services: According to user needs, generate customized monitoring reports, provide decision-making support services, and at the same time support modular configuration for users to select different functional modules as needed.

[0013] Preferably, in the multi-source data acquisition step, according to the terrain complexity, weather conditions and accuracy requirements of the monitoring area, the acquisition frequency and resolution of satellite images are dynamically adjusted, where the acquisition frequency ranges from 1 time per day to 1 time per week, and the resolution is from 0.1 meter to 1 meter.

[0014] Preferably, in the data fusion step, the ICP algorithm is used for spatial registration, and the number of iterations is dynamically adjusted according to the terrain complexity.

[0015] Preferably, the multi-scale fusion technology adopts the wavelet transform method, and the number of decomposition layers is 2 - 5 layers.

[0016] Preferably, in the AI analysis step, the deep learning algorithm includes the collaborative work of a convolutional neural network and a generative adversarial network, where the convolutional neural network is used for preliminary feature extraction and classification, and the generative adversarial network is used for enhancing feature recognition and anomaly detection.

[0017] Preferably, the convolutional neural network adopts the ResNet-50 architecture, the loss function is the weighted cross-entropy loss, and an L1 regularization term is added, with the regularization coefficient being 0.01;

[0018] The generator of the generative adversarial network generates simulated data based on the features extracted by the convolutional neural network, and the discriminator optimizes the model parameters through the adversarial mechanism.

[0019] Preferably, in the real-time monitoring and early warning step, edge computing technology is used for preliminary processing at the data acquisition end to reduce data transmission and processing delays.

[0020] Preferably, in the three-dimensional modeling step, terrain modeling is achieved through point cloud generation, triangular mesh construction and texture mapping, and lightweight rendering technology is used to support Web-side interaction.

[0021] Preferably, in the user-customized service step, the supported functional modules include the urban planning module, the disaster management module and the agricultural monitoring module.

[0022] The present invention also provides an intelligent multi-modal surface change monitoring system, which is applied to the monitoring of surface transformation and includes:

[0023] A data acquisition module that dynamically predicts the optimal acquisition parameters of the monitoring area through machine learning algorithms, including high-resolution satellite images, UAV aerial photography, ground LiDAR data, and environmental sensor data.

[0024] A data preprocessing module for denoising, correcting, and enhancing the original data;

[0025] A data fusion module for integrating multi-source data through spatial registration algorithms and multi-scale fusion technologies to generate a comprehensive data set;

[0026] An AI analysis module for automatically identifying surface change features through deep learning algorithms;

[0027] A real-time monitoring and early warning module that combines Internet of Things technology to achieve real-time data processing and early warning functions;

[0028] A 3D modeling module for constructing a dynamic 3D terrain model and supporting time series analysis;

[0029] A cloud platform and big data processing module for storing, analyzing, processing, and providing interactive map services for the data during the monitoring process;

[0030] A user customization service module for providing modular configuration and customized monitoring reports.

[0031] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0032] (1) The present invention integrates various remote sensing data and ground sensor data deeply for the first time. By fusing multi-modal data, multi-dimensional information of the surface can be obtained, making up for the deficiencies of traditional single data source monitoring and improving the comprehensiveness and accuracy of monitoring.

[0033] (2) Using deep learning algorithms for intelligent analysis to automatically identify surface change features reduces manual intervention and improves the analysis efficiency and accuracy.

[0034] (3) The real-time monitoring and early warning function can timely detect abnormal situations in surface changes and provide support for disaster early warning and emergency response.

[0035] (4) The 3D modeling and time series analysis functions can intuitively display the dynamic process of surface changes and provide strong decision-making basis for fields such as urban planning and natural resource management.

[0036] (4) The user customization service meets the personalized needs of different users and improves the practicality and flexibility of the system. Brief Description of the Drawings

[0037] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the embodiments will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can be obtained based on these drawings.

[0038] Figure 1 is the architecture diagram of the entire monitoring system of the present invention;

[0039] Figure 2 is the fusion flow chart of multi-source data in the embodiments of the present invention;

[0040] Figure 3 is the AI-driven image analysis flow chart in the embodiments of the present invention;

[0041] Figure 4 is the three-dimensional dynamic modeling schematic diagram in the embodiments of the present invention;

[0042] Figure 5 is the schematic diagram of the user-customized service interface in the embodiments of the present invention;

[0043] Figure 6 is the real-time monitoring and early warning flow chart in the embodiments of the present invention;

[0044] Figure 7 is the architecture diagram of the cloud platform and big data processing in the embodiments of the present invention;

[0045] Figure 8 is the schematic diagram of the method proposed by the present invention in different application scenarios. Detailed Embodiments

[0046] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be noted that, without conflict, the following embodiments and the features in the embodiments can be combined with each other; and all other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present disclosure without creative efforts belong to the scope of protection of the present disclosure.

[0047] This embodiment provides an intelligent multi-modal surface change monitoring method, as Figure 1-7 shown, specifically including the following steps:

[0048] (1) Collect multi-modal data of surface images, terrain, and environmental parameters

[0049] Specifically, high-resolution satellite images with a resolution up to sub-meter level are obtained through commercial satellites to comprehensively cover the monitoring area. An unmanned aerial vehicle (UAV) is used to carry a multispectral camera for low-altitude photography to obtain high-resolution and wide-coverage image data. A ground laser scanner is used to collect ground LiDAR data to achieve fine topographic measurement of local areas. An environmental sensor network for temperature, humidity, rainfall, soil moisture, etc. is deployed to monitor environmental parameters in real time, providing multi-source data support for subsequent analysis. The multi-source data collected by the data acquisition module is transmitted to the data preprocessing module through a standardized data interface. The data interface follows [specific data interface standards, such as JSON-RPC, etc.] to ensure the stability and compatibility of data transmission. During data transmission, the data is encapsulated in a unified data format (such as GeoTIFF format for image data, LAS format for LiDAR data, etc.) for subsequent processing. At the same time, to ensure the accuracy of data transmission, [data verification methods, such as CRC verification, etc.] are used to verify the data, and if the verification fails, the data is retransmitted.

[0050] (2) Denoise, correct, and enhance the collected original multimodal data

[0051] Specifically, the collected original data is denoised, corrected, and enhanced. Radiometric correction and atmospheric correction algorithms are used to improve the accuracy of image data, remove noise interference in the data, correct the geometric and radiometric errors of the data, and improve the data quality, providing a reliable data basis for subsequent processing.

[0052] (3) Use spatial registration algorithms and multi-scale fusion techniques to fuse data from different data sources to generate a comprehensive dataset

[0053] Spatial registration algorithms (such as the ICP algorithm) are used to eliminate coordinate deviations between different data sources. The ICP algorithm is used for spatial registration, and the number of iterations is dynamically adjusted according to the terrain slope (formula: N = 50 + 4×(θ / 15°), where θ is the slope). The error in the plain area is < 0.1 m, and the error in the mountainous area is < 0.15 m (test data source: Reports of 10 typical regions of the National Administration of Surveying, Mapping and Geoinformation, No. CGIS-2023-001). Combining multi-scale fusion techniques, experiments comparing different decomposition levels (1 - 5 levels) show that when the decomposition level is 3, the peak signal-to-noise ratio (PSNR) of data fusion reaches 45 dB. A time synchronization algorithm (such as the Kalman filter) is introduced to solve the problem of timestamp alignment of multi-source data. An adaptive registration parameter adjustment strategy is proposed to dynamically optimize the number of iterations and convergence threshold of the ICP algorithm according to the characteristics of different data sources. A data quality assessment matrix is established, including: resolution error tolerance ≤ 0.05 m (according to GB / T 13989-2012); time synchronization threshold ≤ 30 s (UTC time calibration); spectral band matching degree ≥ 95%.

[0054] (4) Analyze the comprehensive dataset using deep learning algorithms to automatically identify surface change features

[0055] Apply deep learning algorithms (such as Convolutional Neural Network CNN and Generative Adversarial Network GAN) to automatically identify surface change features such as land cover change, urban expansion, vegetation degradation, and water body change. The model adopts the ResNet-50 architecture, the loss function is weighted cross-entropy loss, the weights are allocated according to the reciprocal of the class sample size, and the L1 regularization coefficient is 0.01. The ResNet-50 model is trained based on the publicly available dataset WHU Building Dataset (containing 20,000 annotated images). The learning rate of 0.001 is determined through gradient descent convergence testing, and the accuracy of the validation set is 98.5% (compared with 82.3% of the traditional threshold segmentation method). Support the incremental learning mechanism, automatically update the model parameters according to the newly collected data, and continuously optimize the AI model to adapt to new surface change features. Establish a dual-channel feedback mechanism. When the user annotation error > 2%, trigger model fine-tuning (learning rate 0.0001). When the environmental parameter mutation > 15%, initiate transfer learning (freeze the first 3 layers of the feature layer). Increase the test data for the model performance under extreme weather conditions, such as the recognition accuracy in scenarios of cloud cover, haze, rain, snow, etc. Explore the application of transfer learning and use the pre-trained model to quickly adapt to new scenarios.

[0056] (5) Combine Internet of Things technology to monitor surface changes in real time and trigger the warning system when specific surface changes are detected

[0057] Combine Internet of Things technology to achieve real-time or near-real-time data processing capabilities. Use edge computing technology to perform preliminary processing at the data collection end to reduce data transmission and processing latency.

[0058] When specific surface changes (such as changes related to disasters like landslides and floods) are detected, immediately trigger the warning system. Provide specific latency evaluation data, such as the average time from data collection to warning trigger is 3 minutes. In the case of receiving 1 million sensor data per second, the system latency can still be maintained within 3 minutes.

[0059] Requirements for distributed edge computing nodes:

[0060] Computing unit: 4-way GPU (computing power ≥ 20 TFLOPS)

[0061] Storage bandwidth ≥ 10 GB / s

[0062] Positioning module: Support Beidou-3 / GPS / Galileo (timing accuracy ≤ 1 μs).

[0063] (6) Construct a high-precision 3D terrain model using multi-modal data, support time series analysis, and display the dynamic process of surface changes.

[0064] Construct a high-precision 3D terrain model using multi-modal data, and achieve precise terrain modeling through key technologies such as point cloud generation, triangulation network construction, and texture mapping.

[0065] Support time series analysis and display the dynamic process of surface changes. Adopt lightweight rendering technology to ensure efficient loading and interaction of the model on the Web side.

[0066] Support access to WebGL 2.0 standard and VR / AR devices

[0067] Time axis comparison function (synchronously play ≥ 3 phase data)

[0068] LOD classification: dynamically switch between 1:500 and 1:5000

[0069] Define the time resolution index, for example, update the terrain model once per hour.

[0070] Propose a lightweight rendering algorithm based on GPU acceleration to improve the loading speed on the Web side.

[0071] (8) Store data in the cloud, and use a distributed computing framework (such as Hadoop, Spark) to support fast analysis and access of large-scale data. Provide a Web-based interactive map service for convenient user access and use. Integrate data visualization tools to support multi-dimensional data analysis and display.

[0072] (9) Generate customized monitoring reports according to user needs, provide decision support services, and at the same time support modular configuration for users to select different function modules as needed.

[0073] Generate customized monitoring reports according to user needs and provide decision support services.

[0074] Support modular configuration, and users can select different function modules according to their needs, such as urban planning module, disaster management module, agricultural monitoring module, etc.

[0075] Design a unified functional module interface specification to support rapid integration and switching of functional modules in different fields.

[0076] Support RESTful API interface, the request format is JSON, and the response format supports JSON and XML. Developers can access specific function modules, such as urban planning module or disaster management module, through GET / POST methods.

[0077] In addition, such as Figure 8As shown, the above method proposed by the present invention is applied to different scenarios to verify its actual effects, and specific application examples are as follows:

[0078] Example 1: Urban expansion monitoring

[0079] Objective: To monitor the annual expansion of a certain city.

[0080] Method:

[0081] (1) Use high-resolution satellite images (resolution 0.5 meters) and UAV aerial photography data (resolution 0.1 meters) to obtain surface cover information.

[0082] (2) The AI analysis module adopts the ResNet-50 model, with training parameters: learning rate 0.001, number of training epochs 100, and L1 regularization coefficient 0.01. The basis for model parameter selection: Determine the learning rate 0.001 and L1 regularization coefficient 0.01 through cross-validation (5-fold), the ratio of training set / validation set is 8:2, and data augmentation includes random rotation (±15°) and brightness adjustment (±20%).

[0083] (3) The 3D modeling module generates point clouds through LiDAR data, constructs a terrain model using the Delaunay triangulation algorithm, and uses UAV images for texture mapping.

[0084] Result:

[0085] (1) Generate an annual urban expansion report: The area of newly added construction areas is 12.5 square kilometers, a 18% increase compared to the previous year.

[0086] (2) Recognition accuracy: The recognition accuracy of the method of the present invention for newly added buildings reaches 98.5% (the traditional method is 82.3%), and the efficiency is increased by 16.2%.

[0087] (3) Data source: The data in this embodiment is based on the urban monitoring project of the Natural Resources Bureau of a certain city in 2023, and the monitoring area is 200 square kilometers.

[0088] (4) Application effect: By analyzing the expansion direction, the planning department optimized the traffic network layout, and it is expected to save 15% of land resources.

[0089] Example 2: Landslide warning

[0090] Objective: To monitor and warn of landslide risks in a certain mountainous area.

[0091] Method:

[0092] (1) Deploy ground LiDAR (monitoring radius 500 meters, acquisition frequency 10 minutes / time) and rainfall sensors (accuracy 0.1 mm).

[0093] (2) The real-time monitoring module sets the terrain change threshold at 0.1 meters and the rainfall warning threshold at 50 mm / 24 hours.

[0094] (3) The 3D modeling module dynamically updates the terrain model, and slope analysis combines with geological data to evaluate risks.

[0095] Results:

[0096] (1) Early warning timeliness: The system issues a landslide warning 12 hours in advance (data from a geological disaster monitoring project in a certain province in 2022).

[0097] (2) Disaster losses: Evacuate people in time and take protective measures, reducing disaster losses by 75% (no warning by traditional methods).

[0098] (3) Risk assessment: Through 3D model analysis, it is determined that the area of the area to be reinforced is 3.2 square kilometers, and the recurrence probability of landslides is reduced by 60% after formulating a drainage plan.

[0099] (4) Comparative experiment: The traditional method uses image analysis based on threshold segmentation (test data set: 2,000 samples, source: a monitoring project of a city's natural resources bureau in 2023). The recognition accuracy of the present invention on the same data set is 98.5% (82.3% for the traditional method).

[0100] Example 3: Monitoring of agricultural pests and diseases

[0101] Objective: Monitor the growth status of crops and the occurrence of pests and diseases in a certain farmland.

[0102] Methods:

[0103] (1) UAV multispectral image acquisition (once a week, bands include visible light and near-infrared), combined with a temperature and humidity sensor network (accuracy ±2%).

[0104] (2) The AI analysis module uses a fine-tuned ResNet-50 model, with a learning rate of 0.0001 and 50 training epochs. Data augmentation includes rotation and scaling.

[0105] (3) Establish a correlation model between pests and diseases and meteorological conditions (when the temperature is 25 - 30°C and the humidity is 70 - 80%, the probability of pests and diseases is 85%).

[0106] Results:

[0107] (1) Early warning effect: Detect signs of pests and diseases 7 days in advance and guide farmers to use biological control (release natural enemy insects).

[0108] (2) Yield and environmental protection: The yield loss of crops is reduced by 22% (35% loss by traditional methods), and the amount of pesticide used is reduced by 30%.

[0109] (3) Data source: The data in this embodiment is based on the actual measurement results of a certain agricultural experimental base in 2023, with a monitoring area of 500 mu.

[0110] (4) Model verification: The correlation model between temperature, humidity and the probability of pests and diseases is verified by the Pearson correlation coefficient (r = 0.89). The test data is based on the actual measurement results of a certain agricultural experimental base in 2023 (sample size: 500 mu, monitoring period: 6 months).

[0111] This embodiment also proposes an intelligent multi-modal surface change monitoring system, including:

[0112] Data acquisition module, which dynamically predicts the optimal acquisition parameters of the monitoring area through machine learning algorithms, including high-resolution satellite images, UAV aerial photography, ground LiDAR data, and environmental sensor data. The acquisition frequency and resolution of satellite images are dynamically adjusted according to the terrain complexity, weather conditions, and accuracy requirements of the monitoring area, where the acquisition frequency ranges from 1 time per day to 1 time per week, and the resolution is from 0.1 meter to 1 meter.

[0113] Data preprocessing module, used to perform denoising, correction, and enhancement processing on the original data;

[0114] Data fusion module, used to integrate multi-source data through spatial registration algorithms and multi-scale fusion technologies to generate a comprehensive data set;

[0115] AI analysis module, used to automatically identify surface change features through deep learning algorithms;

[0116] Real-time monitoring and early warning module, which combines Internet of Things technology to achieve real-time data processing and early warning functions;

[0117] 3D modeling module, used to build a dynamic 3D terrain model and support time series analysis;

[0118] Cloud platform and big data processing module, used to store, analyze, process, and provide interactive map services for the data during the monitoring process;

[0119] User customization service module, used to provide modular configuration and customized monitoring reports. The supported functional modules include urban planning module, disaster management module, and agricultural monitoring module.

[0120] The above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present invention should be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. An intelligent multi-modal surface change monitoring method, characterized in that It includes the following steps: Multi-source data collection: Collect multi-modal data such as images, terrain, and environmental parameters of the ground surface; Data preprocessing: Denoise, correct, and enhance the collected original multi-modal data; Data fusion: Use spatial registration algorithms and multi-scale fusion techniques to fuse data from different data sources to generate a comprehensive dataset; AI analysis: Use deep learning algorithms to analyze the comprehensive dataset and automatically identify ground surface change features; Real-time monitoring and early warning: Combine Internet of Things technology to conduct real-time monitoring of ground surface changes. When specific ground surface changes are detected, trigger the early warning system; 3D modeling: Use multi-modal data to build a high-precision 3D terrain model and support time series analysis to display the dynamic process of ground surface changes; User customization service: Generate customized monitoring reports according to user needs, provide decision-making support services, and at the same time support modular configuration for users to select different function modules as needed.

2. The intelligent multimodal surface change monitoring method according to claim 1, characterized in that, In the multi-source data collection step, according to the terrain complexity, weather conditions, and accuracy requirements of the monitoring area, dynamically adjust the acquisition frequency and resolution of satellite images, where the acquisition frequency range is from 1 time per day to 1 time per week, and the resolution is from 0.1 meter to 1 meter.

3. An intelligent multi-modal ground change monitoring method according to claim 2, characterized in that, In the data fusion step, the ICP algorithm is used for spatial registration, and the number of iterations is dynamically adjusted according to the terrain complexity.

4. An intelligent multi-modal ground change monitoring method according to claim 3, characterized in that, The multi-scale fusion technology uses the wavelet transform method, and the number of decomposition layers is 2 - 5 layers.

5. An intelligent multi-modal surface change monitoring method according to claim 1, characterized in that, In the AI analysis step, the deep learning algorithm includes the collaborative work of a convolutional neural network and a generative adversarial network. Among them, the convolutional neural network is used for preliminary feature extraction and classification, and the generative adversarial network is used for enhancing feature recognition and anomaly detection.

6. The intelligent multi-modal surface change monitoring method according to claim 5, characterized in that The convolutional neural network adopts the ResNet-50 architecture, the loss function is weighted cross-entropy loss, and an L1 regularization term is added, with a regularization coefficient of 0.01; The generator of the generative adversarial network generates simulated data based on the features extracted by the convolutional neural network, and the discriminator optimizes the model parameters through the adversarial mechanism.

7. An intelligent multi-modal surface change monitoring method according to claim 1, characterized in that, In the real-time monitoring and early warning step, edge computing technology is used for preliminary processing at the data collection end to reduce data transmission and processing delays.

8. The intelligent multimodal surface change monitoring method according to claim 1, characterized in that In the 3D modeling step, terrain modeling is realized through point cloud generation, triangular mesh construction, and texture mapping, and lightweight rendering technology is used to support Web-side interaction.

9. The intelligent multi-modal ground change monitoring method according to claim 1, characterized in that In the user customization service step, the supported function modules include the urban planning module, the disaster management module, and the agricultural monitoring module.

10. An intelligent multi-modal surface change monitoring system, applied to the method according to any one of claims 1-9, characterized in that, It includes: A data collection module that dynamically predicts the best collection parameters of the monitoring area through machine learning algorithms, including high-resolution satellite images, unmanned aerial vehicle aerial photography, ground LiDAR data, and environmental sensor data; A data preprocessing module for denoising, correcting, and enhancing the original data; A data fusion module for integrating multi-source data through spatial registration algorithms and multi-scale fusion techniques to generate a comprehensive dataset; An AI analysis module for automatically identifying ground surface change features through deep learning algorithms; A real-time monitoring and early warning module that combines Internet of Things technology to achieve real-time data processing and early warning functions; A 3D modeling module for building a dynamic 3D terrain model and supporting time series analysis; Cloud platform and big data processing module, which are used for storing, analyzing, processing data during the monitoring process and providing interactive map services; User customization service module, which is used to provide modular configuration and customized monitoring reports.

Images