Real-time updating and intelligent analysis method of geographic information based on cloud computing

Abstract

The present invention discloses a real-time update and intelligent analysis method of geographic information based on cloud computing, which relates to the field of data analysis technology, including collecting geographic information data through distributed sensors to generate labeled geographic information data; transmitting the labeled geographic information data to a cloud server and cleaning it to obtain cleaned geographic information data; extracting features from the cleaned geographic information data based on a deep learning model to form a geographic feature data set; comparing and analyzing the geographic feature data set with data in a historical geographic information database to generate a geographic information change report, and updating it in real time; establishing a geographic information early warning indicator system based on the geographic information change report, and pushing early warning information to a terminal device when it is detected that the geographic information change exceeds a preset threshold. The present invention realizes precise management and efficient analysis of geographic information data through multi-level and multi-stage abnormal data processing and feature extraction.

Description

Technical Field

[0001] The present invention relates to the field of data analysis technology, and in particular to a real-time updating and intelligent analysis method of geographic information based on cloud computing. Background Art

[0002] With the development of video surveillance technology, image transmission and recording systems have been widely used in various industries, especially in security monitoring, traffic management, industrial production and other fields. Traditional image transmission recorders are usually used to capture and record image data streams in real time and transmit them to storage media or cloud servers. However, with the dramatic increase in the amount of image data, how to efficiently process these large amounts of image data has become an urgent problem to be solved. In order to improve data processing efficiency, the technology of image transmission recorders has been continuously developed, adding functions such as compression coding, image differential analysis, and event-driven data capture. These technologies can effectively reduce data storage requirements, optimize the use of network bandwidth, and improve the real-time performance of data processing.

[0003] However, with the rapid growth of data scale and the complexity of application scenarios, how to efficiently process abnormal data and extract valuable features from it has become an urgent problem to be solved. In traditional geographic information processing methods, the cleaning of abnormal data usually relies on simple rule settings or manual intervention, such as screening and filtering data based on fixed thresholds. However, this method is difficult to deal with high-noise, nonlinear abnormal data that appear in complex and changeable actual environments. Especially in distributed data collection, due to the influence of various factors such as equipment environment and sensor accuracy, the proportion of abnormal data is high, which directly affects the reliability and effectiveness of subsequent data analysis.

[0004] In terms of feature extraction, existing technologies mostly rely on traditional statistical methods or shallow learning models. These methods have certain advantages when dealing with small-scale, low-complexity data sets, but they show obvious limitations when faced with massive geographic information data and multidimensional characteristics. Specifically, existing methods are difficult to fully utilize the potential complex correlation structure between data and cannot accurately identify high-value features from noise. In addition, traditional feature extraction methods have weak perception of change patterns, especially in rapidly changing geographical environments, which often leads to inaccurate identification results of changing areas, thereby affecting the overall performance of real-time updates and analysis. These problems significantly limit the application of geographic information technology in high-real-time and high-reliability scenarios such as emergency management and environmental monitoring. Summary of the invention

[0005] In view of the problems existing in the prior art, the present invention proposes a real-time updating and intelligent analysis method of geographic information based on cloud computing.

[0006] Therefore, the present invention provides a real-time updating and intelligent analysis method of geographic information based on cloud computing, which can solve the problems mentioned in the background technology.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0008] In a first aspect, the present invention provides a method for real-time updating and intelligent analysis of geographic information based on cloud computing, which comprises:

[0009] Collecting geographic information data through distributed sensors, marking the geographic information data according to timestamps and geographic coordinates, and generating marked geographic information data;

[0010] The marked geographic information data is transmitted to a cloud server and cleaned to obtain cleaned geographic information data;

[0011] Extracting features from the cleaned geographic information data based on a deep learning model to form a geographic feature data set;

[0012] Compare and analyze the geographic feature data set with the data in the historical geographic information database, identify the geographic information change area, generate a geographic information change report, and update it in real time;

[0013] Establishing a geographic information early warning indicator system based on the geographic information change report, and pushing early warning information to the terminal device when it is detected that the geographic information change exceeds a preset threshold;

[0014] The deep learning model includes a terrain feature extraction branch, a vegetation coverage analysis branch and a hydrological information processing branch;

[0015] The features extracted by the terrain feature extraction branch include slope, aspect, curvature, roughness and landform type;

[0016] The features extracted by the vegetation coverage analysis branch include vegetation density, growth status, coverage type and seasonal changes;

[0017] The features extracted by the hydrological information processing branch include runoff, catchment area, water system connectivity and flow direction distribution.

[0018] As a preferred solution of the real-time updating and intelligent analysis method of geographic information based on cloud computing of the present invention, wherein: the geographic information data includes surface elevation data, soil moisture data, vegetation coverage data and geological stratification data;

[0019] The distributed sensors include a surface morphology sensor, a soil moisture sensor, a vegetation cover sensor and a geological structure sensor.

[0020] As a preferred solution of the cloud computing-based real-time updating and intelligent analysis method of geographic information of the present invention, obtaining the cleaned geographic information data includes:

[0021] Verifying the integrity of the marked geographic information data, and if the integrity verification of the marked geographic information data fails, triggering an automatic retransmission mechanism;

[0022] Performing data cleaning on the verified marked geographic information data to obtain cleaned geographic information data;

[0023] The data cleaning includes outlier detection and elimination, duplicate data removal, data completion processing, data standardization processing and noise smoothing processing;

[0024] The outlier detection and elimination is determined based on the calculation of the Mahalanobis distance, and the calculation of the Mahalanobis distance is shown in the following formula: ;

[0025] in: ; ;in, The data point to be detected The normalized distance to the center of the data distribution, is the data point to be detected, is the mean vector in the time window, is the covariance matrix, is the current time, is the reference time point, is the spatial distance, and To adjust the parameters.

[0026] As a preferred solution of the real-time updating and intelligent analysis method of geographic information based on cloud computing of the present invention, the backbone network of the terrain feature extraction branch includes multiple residual block groups, and the calculation of the residual blocks in the residual block groups can be expressed by the following formula: ;

[0027] in: ; ; ;in, For the The weight matrix of each branch is used for feature transformation. is the feature transformation function, is the attention weight function, is the activation function, ReLU is used, BN is the batch normalization operation, is a 3×3×3 three-dimensional convolution operation, is a 1×1×1 three-dimensional convolution operation, is the global average pooling operation, , is the weight matrix of the fully connected layer, , is the bias term, is the characteristic modulation function.

[0028] As a preferred solution of the real-time updating and intelligent analysis method of geographic information based on cloud computing of the present invention, the vegetation coverage analysis branch includes a feature fusion mechanism, which can be expressed by the following formula: ;

[0029] in, For the The feature map of the layer, For the The feature map of the layer, For the The feature transformation function of the layer, For the The upsampling operator of the layer, For the The upsampling operator of the layer, For the The upsampling operator of the layer, with learnable parameters, is the network level index, , [·] is the feature concatenation operation.

[0030] As a preferred solution of the real-time updating and intelligent analysis method of geographic information based on cloud computing of the present invention, the hydrological information processing branch can be expressed by the following formula: ;

[0031] in: ; ;in, is the output feature matrix, To add a self-connected adjacency matrix, dimension [N, N], is the degree matrix, a diagonal matrix, is the input node feature matrix, dimension [N, C], is the learnable parameter matrix, dimension [C, F], is the time modulation factor, is the time attenuation coefficient, is the periodic modulation intensity, is the periodic modulation frequency, is the Hadamard product.

[0032] As a preferred solution of the real-time updating and intelligent analysis method of geographic information based on cloud computing of the present invention, wherein: the geographic information change report includes a comprehensive analysis of the spatial distribution characteristics of the changed area, the quantitative characteristics of the changed elements and the change trend;

[0033] The spatial distribution characteristics of the change areas include the spatial coordinates and range boundaries of the key change areas, and a hierarchical expression method is used to display the spatial distribution of the change intensity, and the area and proportion of each type of change area are counted; the specific location information of the key change areas is listed in a table, including the center coordinates, change area and main change type, and the spatial agglomeration characteristics of the changes are displayed through a heat map;

[0034] The quantitative characteristics of the changing elements include the quantitative characteristics of terrain elements, vegetation elements and hydrological elements respectively; the quantitative characteristics of the terrain elements include statistical data of elevation change, slope change amplitude, and degree of change in landform morphology; the quantitative characteristics of the vegetation elements include the vegetation coverage change rate, the conversion area matrix between different vegetation types, and the distribution characteristics of areas with abnormal growth conditions; the quantitative characteristics of the hydrological elements include the increase and decrease of water body area, the change value of river network density, and the degree of change in water system connectivity.

[0035] In a second aspect, the present invention provides a real-time updating and intelligent analysis system of geographic information based on cloud computing, which comprises:

[0036] A data collection module, used to collect geographic information data through distributed sensors, mark the geographic information data according to timestamps and geographic coordinates, and generate marked geographic information data;

[0037] A preprocessing module, used for transmitting the marked geographic information data to a cloud server and cleaning it to obtain cleaned geographic information data;

[0038] A feature extraction module, used to extract features from the cleaned geographic information data based on a deep learning model to form a geographic feature data set;

[0039] An analysis module is used to compare and analyze the geographic feature data set with the data in the historical geographic information database, identify the geographic information change area, generate a geographic information change report, and update it in real time;

[0040] The early warning module is used to establish a geographic information early warning indicator system based on the geographic information change report, and push early warning information to the terminal device when it is detected that the geographic information change exceeds a preset threshold.

[0041] The beneficial effect of the present invention is that through multi-level and multi-stage abnormal data processing and feature extraction, accurate management and efficient analysis of geographic information data are achieved. In the data processing stage, an improved sliding window Mahalanobis distance algorithm is used to accurately detect outliers, and the dual spatiotemporal weight function is combined to effectively capture the spatiotemporal anomaly features in geographic information data, greatly improving the accuracy of anomaly detection and reducing the false alarm rate. In the feature extraction stage, a multi-branch network structure based on deep learning extracts diversified and fine-grained feature information for different dimensions such as terrain, vegetation, and hydrology, such as the slope and curvature of the terrain, the density and health of the vegetation, and the flow and connectivity of the hydrology, thereby forming a comprehensive geographic feature data set. In addition, the scheme deeply analyzes the changing spatiotemporal characteristics and correlations through historical data comparison, and finally provides a set of highly reliable geographic information early warning indicator systems that can dynamically respond to changing trends and provide a scientific basis for geographic environment monitoring and management. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Among them:

[0043] Figure 1 The flowchart is a method for real-time updating and intelligent analysis of geographic information based on cloud computing. DETAILED DESCRIPTION

[0044] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0045] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0046] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0047] The present invention is described in detail with reference to schematic diagrams. When describing the embodiments of the present invention, for the sake of convenience, the cross-sectional diagrams showing the device structure will not be partially enlarged according to the general scale, and the schematic diagrams are only examples, which should not limit the scope of protection of the present invention. In addition, in actual production, the three-dimensional dimensions of length, width and depth should be included.

[0048] At the same time, in the description of the present invention, it should be noted that the directions or positional relationships indicated by the terms "upper, lower, inner and outer" are based on the directions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as limiting the present invention. In addition, the terms "first, second or third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0049] In the present invention, unless otherwise clearly specified and limited, the terms "install, connect, connect" should be understood in a broad sense, for example: it can be a fixed connection, a detachable connection or an integral connection; it can also be a mechanical connection, an electrical connection or a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0050] Example 1: Reference Figure 1 , which is the first embodiment of the present invention, and provides a real-time updating and intelligent analysis method of geographic information based on cloud computing, including:

[0051] S1: Collect geographic information data through distributed sensors, mark the geographic information data according to timestamps and geographic coordinates, and generate marked geographic information data.

[0052] A plurality of distributed sensor nodes are arranged in a geographical area, and the distributed sensor nodes include a surface morphology sensor, a soil moisture sensor, a vegetation coverage sensor and a geological structure sensor; the distributed sensor nodes transmit data via a wireless communication network, and each sensor node has a unique device identification code; the distributed sensor collects geographic information data at a preset sampling frequency, and the sampling frequency can be dynamically adjusted according to the rate of environmental change, and the sampling frequency ranges from 10 seconds to 1 hour; the collected geographic information data include surface elevation data, soil moisture content data, vegetation coverage data and geological stratification data; the collected geographic information data is associated with the standard time of the collection moment to generate timestamp information; at the same time, the GPS module built into each sensor node is used to obtain the latitude and longitude information and altitude information of the collection location to form geographic coordinate information; the timestamp information and geographic coordinate information are encapsulated into a data packet of the geographic information data to form tagged geographic information data, and the tagged geographic information data is stored using a unified data format standard.

[0053] S2: The marked geographic information data is transmitted to a cloud server, and the cloud server cleans the marked geographic information data according to a preset data cleaning rule to obtain cleaned geographic information data.

[0054] The marked geographic information data is transmitted to the cloud server via the 5G network. The cloud server adopts a dual-machine hot backup architecture and is configured with a distributed storage architecture and a load balancing module. The distributed storage architecture adopts the HDFS file system, the data block size is set to 128MB, and the replication factor is set to 3. The load balancing module adopts a consistent hashing algorithm to evenly distribute data requests to each server node.

[0055] After receiving the marked geographic information data, the cloud server first performs a data packet integrity check, and uses a 32-bit CRC check algorithm to identify errors or lost data in the data transmission process. Data packets that fail the check will trigger an automatic retransmission mechanism, and the upper limit of the number of retransmissions is set to 3 times;

[0056] For the verified and marked geographic information data, the cloud server processes it according to the preset data cleaning rules, including outlier detection and elimination, duplicate data removal, data completion, data standardization and noise smoothing.

[0057] Preferably, outlier detection and elimination uses an improved Mahalanobis distance algorithm based on a sliding window to identify outliers, the window size is set to 24 hours, and abnormal data points with a confidence level greater than 99.9% are eliminated; specifically, as shown in the following formula: ;

[0058] in: ; ;in, The data point to be detected The normalized distance to the center of the data distribution, is the data point to be detected, is the mean vector in the time window, is the covariance matrix, is the current time, is the reference time point, is the spatial distance, and To adjust the parameters.

[0059] The improved Mahalanobis distance algorithm is better by introducing a dual spatiotemporal weight function , the time decay term With spatial weight Combined with the above, accurate anomaly detection of geographic information data is achieved. In the calculation process, the covariance matrix is ​​dynamically updated using a sliding window. , so that the algorithm can accurately characterize the spatiotemporal anomaly characteristics of geographic information data while maintaining the original advantages of Mahalanobis distance, with a detection accuracy of more than 95% and a false alarm rate reduced to less than 5%, which is suitable for the identification of sudden and gradual anomalies in geographic information data.

[0060] If necessary, Calculated outlier criterion:

[0061] when When , it is judged as an abnormal value;

[0062] The degrees of freedom are The chi-square distribution of Critical value at the significance level;

[0063] is the data dimension. In this embodiment =4;

[0064] α is set to 0.001, corresponding to a confidence level of 99.9%.

[0065] S3: Perform feature extraction on the cleaned geographic information data based on a deep learning model to extract terrain features, vegetation coverage and hydrological information to form a geographic feature data set.

[0066] The cleaned geographic information data is subjected to feature extraction based on a deep learning model, wherein the deep learning model adopts a multi-branch parallel network structure, including a terrain feature extraction branch, a vegetation coverage analysis branch, and a hydrological information processing branch; the network structure of the deep learning model is as follows:

[0067] The terrain feature extraction branch adopts an improved 3D-ResNeXt structure. Its input layer receives rasterized surface elevation data. The grid unit size is 5 meters × 5 meters. The backbone network contains 4 residual block groups, each of which contains 6 basic units. Each basic unit adopts a combination of "1×1 convolution-3×3 convolution-1×1 convolution". An attention mechanism is added to the residual connection, and the attention weight is calculated by the sigmoid function. The final output features are: slope, aspect, curvature, roughness and landform type.

[0068] More specifically, the residual block calculation in the backbone network of the terrain feature extraction branch can be expressed as follows: ;

[0069] in: ; ; ;in, For the The weight matrix of each branch is used for feature transformation. is the feature transformation function, is the attention weight function, is the activation function, ReLU is used, BN is the batch normalization operation, is a 3×3×3 three-dimensional convolution operation, is a 1×1×1 three-dimensional convolution operation, is the global average pooling operation, , is the weight matrix of the fully connected layer, , is the bias term, is the characteristic modulation function.

[0070] The vegetation coverage analysis branch adopts a densely connected U-Net++ structure. The encoder part contains 5 layers of deeply separable convolutional layers, and the decoder part uses skip connections and progressive feature fusion. The output of each encoding layer passes through a spatial pyramid pooling module, so that the feature map scale is gradually downsampled from 1 / 2 to 1 / 32. Finally, the output features are obtained: vegetation density, growth status, coverage type and seasonal changes.

[0071] More specifically, the feature fusion mechanism of the vegetation coverage analysis branch can be expressed as follows: ;

[0072] in, For the The feature map of the layer, For the The feature map of the layer, For the The feature transformation function of the layer, For the The upsampling operator of the layer, For the The upsampling operator of the layer, For the The upsampling operator of the layer, with learnable parameters, is the network level index, , [·] is the feature concatenation operation.

[0073] The hydrological information processing branch adopts a graph neural network structure, in which nodes represent water body units and edges represent water flow connections. A gated graph convolutional layer is used to extract local hydrological features, and a global attention mechanism is applied to capture long-range dependencies. The obtained time series information is then integrated to obtain output features: runoff, catchment area, water system connectivity, and flow direction distribution.

[0074] More specifically, the feature extraction of the hydrological information processing branch can be expressed as follows: ;

[0075] in: ; ;in, is the output feature matrix, To add a self-connected adjacency matrix, dimension [N, N], is the degree matrix, a diagonal matrix, is the input node feature matrix, dimension [N, C], is the learnable parameter matrix, dimension [C, F], is the time modulation factor, is the time attenuation coefficient, is the periodic modulation intensity, is the periodic modulation frequency, is the Hadamard product.

[0076] Furthermore, a geographic feature data set is formed according to the features of the three different dimensions obtained. The geographic feature data set adopts a hierarchical storage structure, including: a basic feature layer (storing terrain features such as elevation, slope, aspect, curvature and landform type, vegetation features such as coverage, type, leaf area index and biomass, and hydrological features such as runoff, catchment area and river network density); a statistical feature layer (containing statistics of each element such as mean, variance, spatial distribution characteristics, etc.); and an associated feature layer (recording the relationship between elements such as terrain-vegetation correlation, hydrological-geomorphological coupling, etc.).

[0077] S4: Compare and analyze the geographic feature data set with the data in the historical geographic information database, identify the geographic information change area, and generate a geographic information change report.

[0078] Spatially align the newly generated geographic feature data with the historical data to unify the projection coordinate system and resolution. Resample the elevation data and ensure that the spatial positions of vegetation cover and hydrological features are accurately matched. Pair the data at the same time interval based on the timestamp.

[0079] The terrain changes, vegetation changes and hydrological element changes are monitored separately. Specifically, by calculating the elevation difference map, the terrain undulation change area is identified, the changes in slope and slope direction are further analyzed, the landform changes are extracted, and finally the change in local terrain undulation is counted.

[0080] Specifically, first, at the terrain level, by calculating the elevation difference ΔH (marking the change area when |ΔH|>0.5 meters), analyzing the slope change ΔS (dividing the change into a drastic change of |ΔS|>15°, a significant change of 5°<|ΔS| ≤ 15°, and a slight change of 2°<|ΔS| ≤ 5°), the landform morphological change was evaluated (when the surface roughness changed and the boundary displacement was >10 meters, it was recorded as a change area). At the vegetation level, the normalized difference vegetation index ΔNDVI was calculated (significant decrease ΔNDVI<-0.2, significant increase ΔNDVI>0.2, general change |ΔNDVI| ∈ [0.1, 0.2]), and the leaf area index change ΔLAI was analyzed (marked as an anomaly when the LAI decreased by >30% and lasted for >1 season). At the hydrological level, the area change rate ΔA = (A current - A history) / A history was calculated, the river network density change ΔD was evaluated and the areas exceeding the threshold were identified.

[0081] It should be noted that elevation difference, slope change, etc. are all calculated based on the results of deep learning feature extraction, for example:

[0082] Changes in river network density Feature extraction based on improved 3D-ResNeXt residual block: ;

[0083] in: ; Through multi-factor cross-validation, the reliability of the changes is confirmed, such as the mutual confirmation of terrain changes and vegetation cover changes, and the correlation analysis between water system changes and terrain transformation. Based on the standard deviation of the change amount, the thresholds of short-term changes (differences between two consecutive periods), cumulative changes (total differences with the base period) and change trends (directions and rates of changes in multiple consecutive periods) are established, and finally a geographic information change report is formed, which includes:

[0084] Spatial distribution characteristics of the change area. The spatial distribution characteristics of the change area include the spatial coordinates and range boundaries of the key change area. The spatial distribution of the change intensity is displayed in a hierarchical expression (spatial distribution map of significant change area, medium change area, and slight change area), and the area and proportion of each type of change area are counted. The specific location information of the key change area is listed in a table, including key information such as the center coordinates, change area, and main change type. At the same time, the spatial agglomeration characteristics of the change are displayed through a heat map.

[0085] Quantitative characteristics of changing elements. For terrain elements, provide statistical data on elevation change, slope change, and degree of landform change; for vegetation elements, provide the vegetation coverage change rate, the conversion area matrix between different vegetation types, and the distribution characteristics of abnormal growth areas; for hydrological elements, provide statistics on the increase and decrease of water body area, river network density change, and degree of change in water system connectivity. All statistical indicators include basic statistics such as mean, standard deviation, maximum, and minimum.

[0086] Comprehensive analysis of change trends. Identify the dominant direction and pattern of changes in each element, evaluate the continuity and stability of changes, and predict short-term change trends. Emphasize the spatial correlation of changes, such as the impact of terrain transformation on vegetation distribution, the effect of water system changes on surrounding landforms, etc. For areas with major changes, describe in detail the temporal characteristics and spatial evolution laws of the change process to provide a basis for subsequent monitoring and management.

[0087] S5: Establishing a geographic information early warning indicator system according to the geographic information change report, and pushing early warning information to the terminal device when it is detected that the geographic information change exceeds a preset threshold.

[0088] The early warning indicator system adopts a multi-level classification structure, including basic monitoring indicators and statistical analysis indicators. Basic monitoring indicators include terrain change indicators (such as surface subsidence rate, landslide volume, erosion depth), vegetation change indicators (such as vegetation coverage reduction rate, biomass loss, vegetation health), and hydrological change indicators (such as river channel shrinkage, water pollution level, flow anomalies). Statistical analysis indicators include statistical characteristics such as spatial aggregation of change areas, temporal continuity index and element association intensity.

[0089] The warning level is divided into five levels, namely, extremely serious (red), serious (orange), major (yellow), general (blue) and warning (green). Different threshold systems are set for different warning levels. The threshold is determined by comprehensively considering the historical change rules and numerical simulation results, and dynamically adjusted according to regional characteristics. For example, for areas prone to geological disasters, the warning threshold of terrain changes is lowered; for ecologically fragile areas, the sensitivity of vegetation changes is increased; for water resource shortage areas, the monitoring of hydrological anomalies is strengthened.

[0090] The early warning trigger mechanism adopts multi-dimensional judgment criteria. Single indicator over-threshold trigger requires a key indicator to exceed the preset threshold; multi-indicator combination trigger requires multiple related indicators to meet the early warning conditions at the same time; trend analysis trigger is based on the extrapolated prediction results of the change trend. At the same time, an early warning upgrade mechanism is established, and the early warning level is automatically upgraded when the degree of change is monitored to continue to deteriorate or the scope of impact continues to expand.

[0091] For example, in this embodiment, the warning trigger mechanism includes but is not limited to the following rules:

[0092] Single indicator direct warning conditions:

[0093] If the surface displacement rate exceeds 5cm / day, a yellow warning will be triggered; if the surface displacement rate exceeds 10cm / day, an orange warning will be triggered; if the surface displacement rate exceeds 20cm / day, a red warning will be triggered; if the vegetation coverage decreases by more than 10% within 24 hours, a blue warning will be triggered; if the vegetation coverage decreases by more than 20% within 24 hours, a yellow warning will be triggered; if the water pollution index exceeds the regional standard by 1.5 times, a yellow warning will be triggered; if the surface settlement rate exceeds 3cm / month, a blue warning will be triggered.

[0094] Multi-indicator combination warning conditions:

[0095] If the width of surface cracks increases by more than 2cm and the groundwater level drops by more than 1 meter, a yellow warning will be triggered; if the slope displacement exceeds 3cm and the rainfall exceeds 50mm / day, an orange warning will be triggered; if the vegetation coverage decreases by 15% and the surface temperature rises by 5℃, a yellow warning will be triggered; if the river channel shrinkage exceeds 10% and the water pollution index increases by 50%, an orange warning will be triggered.

[0096] Spatial aggregation warning conditions:

[0097] If the area of ​​the changed region exceeds 1 square kilometer and shows significant spatial correlation, a yellow warning will be triggered; if the continuous change area exceeds 5 square kilometers, an orange warning will be triggered; if the abnormal area shows a trend of spreading and the rate exceeds 0.5 square kilometers / day, a red warning will be triggered.

[0098] Warning upgrade conditions:

[0099] If the blue warning continues for 72 hours without improvement, it will be upgraded to a yellow warning; if the yellow warning area expands by 50%, it will be upgraded to an orange warning; if new danger signs appear and the original warning level is yellow, it will be upgraded to an orange warning; if multiple yellow warning areas are spatially correlated, it will be upgraded to an orange warning.

[0100] Warning downgrade conditions:

[0101] If the monitoring indicators return to normal and remain stable for 72 hours, the warning level will be lowered to one level; if the affected area is reduced by 70% and there are no new abnormalities, the warning level will be lowered to one level; if there are no abnormal fluctuations within 30 consecutive days of monitoring, the warning will be lifted.

[0102] Special area warning conditions:

[0103] If the surface displacement in areas prone to geological disasters exceeds 2cm / day, a yellow warning will be triggered; if the vegetation loss in ecologically fragile areas exceeds 5%, a yellow warning will be triggered; if the water quality index in water source protection areas exceeds the standard by 20%, an orange warning will be triggered; if the ground subsidence in densely populated areas exceeds 1cm / month, a yellow warning will be triggered.

[0104] The push method adopts a multi-channel collaborative push mechanism. Real-time push is carried out through the mobile terminal APP to ensure that front-line staff can obtain warning information in the first time, and then it is regularly updated through the Web platform to facilitate managers to query historical records and analyze change patterns. Finally, in order to ensure the reliability of information transmission, traditional methods such as SMS and email are used as supplementary channels. In terms of push frequency, differentiated strategies are adopted for different levels of warnings. High-level warnings are pushed in real time and tracked regularly, and low-level warnings are summarized and pushed regularly.

[0105] Furthermore, this embodiment also provides a real-time update and intelligent analysis system of geographic information based on cloud computing, including:

[0106] A data collection module, used to collect geographic information data through distributed sensors, mark the geographic information data according to timestamps and geographic coordinates, and generate marked geographic information data;

[0107] A preprocessing module, used for transmitting the marked geographic information data to a cloud server and cleaning it to obtain cleaned geographic information data;

[0108] A feature extraction module, used to extract features from the cleaned geographic information data based on a deep learning model to form a geographic feature data set;

[0109] An analysis module is used to compare and analyze the geographic feature data set with the data in the historical geographic information database, identify the geographic information change area, generate a geographic information change report, and update it in real time;

[0110] The early warning module is used to establish a geographic information early warning indicator system based on the geographic information change report, and push early warning information to the terminal device when it is detected that the geographic information change exceeds a preset threshold.

[0111] This embodiment also provides a computer device, which is suitable for the real-time update and intelligent analysis method of geographic information based on cloud computing, including a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement the real-time update and intelligent analysis method of geographic information based on cloud computing as proposed in the above embodiment.

[0112] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0113] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the method for implementing real-time updating and intelligent analysis of geographic information based on cloud computing as proposed in the above embodiment is implemented.

[0114] The storage medium proposed in this embodiment and the data storage method proposed in the above embodiment belong to the same inventive concept. The technical details not fully described in this embodiment can be found in the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0115] Embodiment 2: This is the second embodiment of the present invention. This embodiment provides a real-time update and intelligent analysis method of geographic information based on cloud computing. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0116] In order to verify the actual effect of the geographic information data acquisition and processing system based on distributed sensors, a mountainous area (simulated area) was selected as the experimental object. The area includes complex surface topography, vegetation coverage and hydrological system, and the experimental range is about 50 square kilometers. The experimental steps are as follows:

[0117] 80 distributed sensor nodes are deployed in the experimental area, with a distribution density of 1.6 / km², including surface morphology sensors, soil moisture sensors, vegetation cover sensors and geological structure sensors. Each sensor node has a unique device identification code and is connected to the central server through a 5G communication network.

[0118] The sampling frequency of the sensor is set to 30 seconds, and data is recorded at high resolution, including surface elevation, soil moisture, vegetation coverage, and geological stratification data. The GPS module of each node provides accurate latitude, longitude, and altitude information, and generates a timestamp combined with the standard time of the collection moment to form a tagged geographic information data packet.

[0119] The collected geographic information data is transmitted to the cloud server through the 5G network. The cloud server is configured with HDFS distributed storage architecture, with a data block size of 128MB and a replication factor of 3. After the data is received, a 32-bit CRC check is first performed. Data packets that fail the check trigger automatic retransmission to ensure data integrity.

[0120] Data cleaning includes outlier detection, duplicate data removal, data completion, standardization and noise smoothing. Among them, outlier detection uses an improved Mahalanobis distance algorithm based on a sliding window with a window size of 24 hours, and removes abnormal data points with a confidence level greater than 99.9%.

[0121] The cleaned geographic information data is used to extract features through a deep learning model. The improved 3D-ResNeXt structure is used to extract terrain features (such as slope, aspect, etc.), the U-Net++ structure is used to analyze vegetation cover (such as cover type and seasonal changes), and the graph neural network is used to process hydrological information (such as runoff and water system connectivity). The resulting geographic feature data set adopts a hierarchical storage structure, which contains basic features, statistical features, and associated features. The generated geographic feature data is compared and analyzed with the regional historical database to identify areas of terrain change, vegetation change, and hydrological change, and a geographic information change report is generated. The key change areas are displayed through spatial heat maps, and the change trends are analyzed using time series.

[0122] A multi-level early warning indicator system is established based on the change report, and single indicator and combined indicator thresholds are set. When the geographic information changes exceed the set value, the early warning signal is triggered and the graded early warning information is pushed to the terminal. The specific experimental data is shown in the following table:

[0123] Table 1 Experimental data table

[0124] Parameter name Terrain elevation difference (m) Slope change (°) Change in vegetation coverage (%) Changes in Vegetation Health Index Change in runoff volume (m³ / s) Change in river network density (km / km²) Time (hours) Region 1 0.8 18 -12 -0.15 3.2 -0.6 12 Area 2 1.2 22 -5 -0.08 1.8 -0.4 24 Area 3 0.5 10 -18 -0.22 4.1 -0.9 48 Region 4 1.5 15 -7 -0.12 2.5 -0.3 36 Area 5 0.9 12 -10 -0.17 3.0 -0.5 72 Area 6 0.7 8 -14 -0.19 3.7 -0.7 60

[0125] The above data clearly show the superiority and innovation of the method of the present invention in detecting changes in geographic information. The following key points can be analyzed from the data of area 1 to area 6:

[0126] Through the improved Mahalanobis distance algorithm, the system can effectively identify abnormal changes in different areas. For example, the vegetation coverage in area 3 decreased by 18%, with high detection accuracy and low noise interference rate. In areas 2 and 4, the slope changes reached 22° and 15° respectively, and combined with the changes in the vegetation health index, the impact of terrain changes on vegetation growth can be inferred, which is usually difficult to quantify multi-dimensional data simultaneously in traditional methods.

[0127] Moreover, the system can not only detect the changes of a single element, but also comprehensively analyze the relationship between multiple geographical elements. For example, the change in runoff in area 5 is highly correlated with the reduction in river network density, indicating that the change in hydrological characteristics is caused by the reduction in vegetation.

[0128] Compared with the existing technology, the present invention has obvious advantages in the diversity of geographic information collection, the accuracy of data cleaning, deep learning optimization of feature extraction and comprehensive change analysis, providing reliable technical support for geographic information monitoring and early warning. The final generated change report shows the specific distribution of the change area and the main change types, and proposes an accurate early warning strategy, providing a comprehensive scientific basis for decision makers.

[0129] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for real-time updating and intelligent analysis of geographic information based on cloud computing, characterized in that: include: Collecting geographic information data through distributed sensors, marking the geographic information data according to timestamps and geographic coordinates, and generating marked geographic information data; The marked geographic information data is transmitted to a cloud server and cleaned to obtain cleaned geographic information data; Extracting features from the cleaned geographic information data based on a deep learning model to form a geographic feature data set; Compare and analyze the geographic feature data set with the data in the historical geographic information database, identify the geographic information change area, generate a geographic information change report, and update it in real time; Establishing a geographic information early warning indicator system based on the geographic information change report, and pushing early warning information to the terminal device when it is detected that the geographic information change exceeds a preset threshold; The deep learning model includes a terrain feature extraction branch, a vegetation coverage analysis branch and a hydrological information processing branch; The features extracted by the terrain feature extraction branch include slope, aspect, curvature, roughness and landform type; The features extracted by the vegetation coverage analysis branch include vegetation density, growth status, coverage type and seasonal changes; The features extracted by the hydrological information processing branch include runoff, catchment area, water system connectivity and flow direction distribution; The cleaned geographic information data includes: Verifying the integrity of the marked geographic information data, and if the integrity verification of the marked geographic information data fails, triggering an automatic retransmission mechanism; Performing data cleaning on the verified marked geographic information data to obtain cleaned geographic information data; The data cleaning includes outlier detection and elimination, duplicate data removal, data completion processing, data standardization processing and noise smoothing processing; The outlier detection and elimination is determined based on the calculation of the Mahalanobis distance, and the calculation of the Mahalanobis distance is shown in the following formula: ， in: ， in, The data point to be detected The normalized distance to the center of the data distribution, is the data point to be detected, is the mean vector in the time window, is the covariance matrix, is the current time, is the reference time point, is the spatial distance, and To adjust the parameters; The backbone network of the terrain feature extraction branch includes a plurality of residual block groups, and the calculation of the residual blocks in the residual block groups can be expressed by the following formula: ， in: ， in, For the The weight matrix of each branch is used for feature transformation. is the feature transformation function, is the attention weight function, is the activation function, ReLU is used, BN is the batch normalization operation, is a 3×3×3 three-dimensional convolution operation, is a 1×1×1 three-dimensional convolution operation, is the global average pooling operation, , is the weight matrix of the fully connected layer, , is the bias term, is the characteristic modulation function.

2. The method for real-time updating and intelligent analysis of geographic information based on cloud computing according to claim 1, characterized in that: The geographic information data includes surface elevation data, soil moisture data, vegetation coverage data and geological stratification data; The distributed sensors include surface morphology sensors, soil moisture sensors, vegetation cover sensors and geological structure sensors.

3. The method for real-time updating and intelligent analysis of geographic information based on cloud computing as claimed in claim 2, characterized in that: The vegetation coverage analysis branch includes a feature fusion mechanism, which can be expressed by the following formula: ， in, For the The feature map of the layer, For the The feature map of the layer, For the The feature transformation function of the layer, For the The upsampling operator of the layer, For the The upsampling operator of the layer, For the The upsampling operator of the layer, with learnable parameters, is the network level index, ∈ [1, ], [·] are feature concatenation operations.

4. The method for real-time updating and intelligent analysis of geographic information based on cloud computing as claimed in claim 3, characterized in that: The hydrological information processing branch can be expressed by the following formula: ， in, ， in, is the output feature matrix, To add a self-connected adjacency matrix, dimension [N, N], is the degree matrix, a diagonal matrix, is the input node feature matrix, dimension [N, C], is the learnable parameter matrix, dimension [C, F], is the time modulation factor, is the time attenuation coefficient, is the periodic modulation intensity, is the periodic modulation frequency, is the Hadamard product.

5. The method for real-time updating and intelligent analysis of geographic information based on cloud computing as claimed in claim 4, characterized in that: The geographic information change report includes the spatial distribution characteristics of the changed area, the quantitative characteristics of the changed elements and a comprehensive analysis of the change trend; The spatial distribution characteristics of the change areas include the spatial coordinates and range boundaries of the key change areas, and a hierarchical expression method is used to display the spatial distribution of the change intensity, and the area and proportion of each type of change area are counted; the specific location information of the key change areas is listed in a table, including the center coordinates, change area and change type, and the spatial agglomeration characteristics of the changes are displayed through a heat map; The quantitative characteristics of the changing elements include the quantitative characteristics of terrain elements, vegetation elements and hydrological elements respectively; the quantitative characteristics of the terrain elements include statistical data of elevation change, slope change amplitude, and degree of change in landform morphology; the quantitative characteristics of the vegetation elements include the vegetation coverage change rate, the conversion area matrix between different vegetation types, and the distribution characteristics of areas with abnormal growth conditions; the quantitative characteristics of the hydrological elements include the increase and decrease of water body area, the change value of river network density, and the degree of change in water system connectivity.

6. A cloud computing-based geographic information real-time update and intelligent analysis system, based on the cloud computing-based geographic information real-time update and intelligent analysis method according to any one of claims 1 to 5, characterized in that: include: A data collection module, used to collect geographic information data through distributed sensors, mark the geographic information data according to timestamps and geographic coordinates, and generate marked geographic information data; A preprocessing module, used for transmitting the marked geographic information data to a cloud server and cleaning it to obtain cleaned geographic information data; A feature extraction module, used to extract features from the cleaned geographic information data based on a deep learning model to form a geographic feature data set; An analysis module is used to compare and analyze the geographic feature data set with the data in the historical geographic information database, identify the geographic information change area, generate a geographic information change report, and update it in real time; The early warning module is used to establish a geographic information early warning indicator system based on the geographic information change report, and push early warning information to the terminal device when it is detected that the geographic information change exceeds a preset threshold.

Images