Quick response method and system for geological disasters

Abstract

The invention discloses a geological disaster quick response method and system. According to the method, state value extraction is carried out on image data of geological disasters to obtain health state data of the geological disasters, a multi-dimensional data fusion model of a geological disaster target area is constructed, then fusion feature data of the geological disaster target area is obtained, the early warning state of the geological disasters is determined, and then the early warning state of the geological disasters is obtained based on a Django server framework. According to the method, the health state data of the geological disasters and the early warning state of the geological disasters are visualized in real time by means of Pyecarts, the visualization result of geological disaster danger early warning is obtained, and finally, real-time geological disaster danger early warning operation is executed based on the visualization result of geological disaster collision prevention. According to the invention, the health state data of the road surface can be rapidly and effectively obtained, and the health state data of the road surface and the early warning state of geological disasters can be rapidly and visually displayed, so that early warning operation which is more practical is adopted.

Description

Technical Field

[0001] The present invention relates to the field of early warning of geological disasters, and in particular to a method and system for rapid response to geological disasters. Background Art

[0002] At present, existing methods can combine multimodal fusion data for geological disaster warning, such as fusing camera sensor data in geological disaster target areas with satellite radar data in geological disaster target areas to obtain more accurate geological disaster detection results, which have certain timeliness and practicality. However, since existing methods cannot quickly and effectively obtain road health monitoring data, and cannot quickly and intuitively display road health monitoring data and slope fusion data, the geological disaster detection results obtained by them have certain limitations, which in turn leads to insufficient subsequent dangerous warnings, bringing more serious harm to pedestrians and vehicles on the geological disaster road surface and under the slope. Summary of the invention

[0003] In view of the above-mentioned deficiencies in the prior art, the present invention provides a method and system for rapid response to geological disasters.

[0004] In order to achieve the above-mentioned object of the invention, the technical solution adopted by the present invention is:

[0005] A method for rapid response to geological disasters, comprising the following steps:

[0006] S1. Determine the target monitoring area for geological disasters, collect road surface image data, slope surface image data and slope surface millimeter wave radar data in real time in the target monitoring area, extract state values ​​from the road surface image data to obtain road surface health status data, and perform multi-target feature extraction on the slope surface image data to obtain a slope surface feature map;

[0007] S2, constructing a slope multidimensional data fusion model, using the slope multidimensional data fusion model to fuse the slope millimeter wave radar data and the slope feature map in step S1, obtaining the slope fusion feature map of the target monitoring area, and determining the early warning status of the geological disaster according to the slope fusion feature map of the target monitoring area;

[0008] S3. Based on the Django server framework, Pyecharts is used to visualize the health status data of the road surface in step S1 and the warning status of the geological disaster in step S2 in real time to obtain the visualization results of the geological disaster warning;

[0009] S4. Based on the visualization result of the geological disaster warning in step S3, perform real-time geological disaster warning operation.

[0010] Furthermore, in step S1, the state value of the road surface image data is extracted to obtain the health status data of the road surface, specifically: a road surface health status detection model is designed with the road surface image data as input and the road surface health status data as output, the road surface health status data includes the road surface deformation value, the road surface inclination value and the road surface height change value, and the road surface health status detection model is used to extract the state value of the road surface image data to obtain the health status data of the road surface.

[0011] Further, in step S1, multi-target feature extraction is performed on the slope image data to obtain a slope feature map, including the following steps:

[0012] A1. Determine the geological disaster image features that are independent of each other and obey the Gaussian distribution, and construct the multidimensional feature space of geological disasters based on the geological disaster image features, which can be expressed as:

[0013] X=span(ξ1,…ξ K , …ξ N )

[0014] Among them: X is the multidimensional feature space of geological disasters, span is the spatial symbol, ξ1 is the geological disaster image feature of the first dimension, ξ K is the K-th dimension geological disaster image feature, ξ N is the geological disaster image feature of the Nth dimension;

[0015] A2. Based on the multidimensional feature space of geological disasters in step A1, the slope image data is converted into target data to be extracted, which is expressed as:

[0016]

[0017]

[0018] Where: f is the slope image, i is the serial number of the target to be extracted in the slope image, M is the number of targets to be extracted in the slope image, and f i is the i-th target to be extracted in the slope image, j is the serial number of the image feature dimension, N is the number of image feature dimensions, ξ j is the geological disaster image feature of the jth dimension, x j is the characteristic decomposition coefficient of the j-th dimension;

[0019] A3. Construct the imaging matrix of multi-target features, expressed as:

[0020]

[0021] in: is the imaging matrix of multi-target features, n is the sequence number of the feature coefficient, ω n is the nth characteristic coefficient;

[0022] A4. Based on the target data to be extracted converted in step A2, the imaging matrix of the multi-target features in step A3 is used to extract the multi-target features of the slope image data to obtain a slope feature map, which is expressed as:

[0023]

[0024] in: It is the slope characteristic map.

[0025] Further, in step S2, the slope multi-dimensional data fusion model includes an input layer, an encoder, a feature splicing layer and a decoder;

[0026] The input layer is used to splice the slope millimeter wave radar data and the slope feature map to obtain the input feature map, which is expressed as:

[0027]

[0028]

[0029] X rader ∈R H×W×1

[0030] Where: X input is the input feature map, concat is the concatenation operator, is the slope characteristic map, X rader is the slope millimeter-wave radar data, R is a placeholder, H is the height of the image, W is the width of the image, and 4 is the number of channels;

[0031] The encoder is used to perform convolution and pooling operations on the input feature map to obtain a compressed feature map, which is expressed as:

[0032]

[0033]

[0034] in: is the compressed feature map after the input feature map is convolved by the l-th layer encoder, l is the current number of layers, σ is the activation function, is the weight coefficient of the convolution of the encoder at layer l, is the compressed feature map after the input feature map is convolved and pooled by the (l-1)th layer encoder. is the bias coefficient of the convolution of the encoder at layer l, is the compressed feature map after the input feature map is pooled by the l-th layer encoder, and MaxPool is the maximum pooling operation;

[0035] The feature concatenation layer is used to concatenate the compressed feature map in the encoder with the upsampled feature map in the decoder to obtain the detail feature map, which is expressed as:

[0036]

[0037] in: is the detail feature map of the input feature map spliced ​​by the lth layer, is the feature map after the input feature map is upsampled by the l-th layer decoder, is the compressed feature map after the input feature map is convolved by the (Ll-1)th layer encoder, and L is the total number of encoder layers;

[0038] The decoder is used to upsample the compressed feature map and perform convolution on the detail feature map to output the slope fusion feature map of the target monitoring area, which is expressed as:

[0039]

[0040]

[0041]

[0042] Among them: UpSamplc is the upsampling operation, is the compressed feature map after the input feature map is pooled by the (Ll-1)th layer encoder, is the feature map after the input feature map is pooled by the l-th layer decoder, is the weight coefficient of the convolution of the l-th layer decoder, is the bias coefficient of the convolution of the l-th layer decoder, X output is the slope fusion feature map of the target monitoring area, W output is the weight coefficient of the last layer of decoder convolution, b output is the bias coefficient of the last decoder convolution.

[0043] Furthermore, in step S2, the warning status of the geological disaster is determined based on the slope fusion feature map of the target monitoring area, specifically: the disaster type, disaster area, disaster location, disaster movement direction and disaster movement speed of the target monitoring area are obtained based on the slope fusion feature map of the target monitoring area, and the warning status of the corresponding geological disaster is determined based on the disaster type, disaster area, disaster location, disaster movement direction and disaster movement speed of the target monitoring area.

[0044] Furthermore, in step S3, based on the Django server framework, Pyecharts is used to visualize the health status data of the road surface in step S1 and the warning status of the geological disasters in step S2 in real time to obtain the visualization results of the geological disaster warning, specifically: a visualization platform for geological disaster warning is built based on the Django server framework, and the visualization platform for geological disaster warning includes a data layer, an application layer and a presentation layer; the data layer has a built-in MySQL database for storing the health status data of the road surface and the warning status of the geological disasters; the application layer is used to process the request instructions of the presentation layer, determine whether to interact with the data layer according to the request instructions, and determine the method of data interaction; the presentation layer is used to transmit the request instructions, and visualize the health status data of the road surface and the warning status of the geological disasters received after the request instructions are passed in real time using Pyecharts to obtain the visualization results of the geological disaster warning.

[0045] Furthermore, when the application layer determines whether to interact with the data layer according to the request instruction, SQL injection protection is set, and the data layer is operated through Django's model and query set to interact with the data.

[0046] A geological disaster early warning system using the above method, comprising a data acquisition module, a geological disaster detection module, a data visualization module and a geological disaster early warning module;

[0047] The data acquisition module is used to determine the target monitoring area for geological disasters, collect road surface image data, slope image data and slope millimeter wave radar data in real time in the target monitoring area, extract state values ​​from the road surface image data to obtain road health status data, and perform multi-target feature extraction on the slope image data to obtain a slope feature map;

[0048] The geological disaster detection module is used to construct a slope multi-dimensional data fusion model, use the slope multi-dimensional data fusion model to fuse the slope millimeter wave radar data and the slope characteristic map, obtain the slope fusion characteristic map of the target monitoring area, and determine the early warning status of the geological disaster based on the slope fusion characteristic map of the target monitoring area;

[0049] The data visualization module is used to visualize the health status data of the road surface and the early warning status of geological disasters in real time based on the Django server framework using Pyecharts to obtain the visualization results of geological disaster early warnings;

[0050] The geological disaster warning module is used to perform real-time geological disaster warning operations based on the visualization results of geological disaster warning.

[0051] The present invention has the following beneficial effects:

[0052] (1) The present invention extracts multi-target features from slope image data to obtain a slope feature map, then fuses the slope millimeter-wave radar data and the slope feature map using a slope multi-dimensional data fusion model to obtain a slope fusion feature map of a target monitoring area, and determines the early warning state of a geological disaster based on the slope fusion feature map of the target monitoring area. This process integrates more fusion features of the target monitoring area, and can make the acquired early warning state of a geological disaster more accurate and in line with reality.

[0053] (2) The present invention is based on the Django server framework and uses Pyecharts to visualize the health status data of the road surface and the warning status of geological disasters in real time, obtain the visualization results of geological disaster warnings, and finally perform real-time geological disaster warning operations based on the visualization results of geological disaster warnings. It can quickly and effectively obtain the health status data of the road surface, and quickly and intuitively display the health status data of the road surface and the warning status of geological disasters, so as to take more practical warning operations;

[0054] (3) The present invention builds a visualization platform for geological disaster early warning based on the Django server framework. The visualization platform for geological disaster early warning includes a data layer, an application layer and a presentation layer. When the application layer determines whether to interact with the data layer according to the request instruction, SQL (Structured Query Language) injection protection is set. The data layer is operated through Django's model and query set to interact with the data, thereby avoiding directly splicing the input request instruction into the SQL statement, thereby reducing the risk of SQL injection;

[0055] (4) The present invention provides a geological disaster early warning system, including a data acquisition module, a geological disaster detection module, a data visualization module and a geological disaster early warning module; the data acquisition module can determine the target monitoring area of ​​geological disasters, collect road surface image data, slope image data and slope millimeter wave radar data of the target monitoring area in real time, extract the state value of the road surface image data to obtain the health state data of the road surface, and extract multi-target features of the slope image data to obtain the slope feature map; the geological disaster detection module can construct a slope multidimensional data fusion model, use the slope multidimensional data fusion model to fuse the slope millimeter wave radar data and the slope feature map, obtain the slope fusion feature map of the target monitoring area, and determine the early warning state of the geological disaster according to the slope fusion feature map of the target monitoring area; the data visualization module can use Pyecharts based on the Django server framework to visualize the health state data of the road surface and the early warning state of the geological disaster in real time to obtain the visualization results of the geological disaster early warning; the geological disaster early warning module can perform real-time geological disaster early warning operations according to the visualization results of the geological disaster early warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 It is a flow chart of a rapid response method for geological disasters;

[0057] Figure 2 This is a schematic diagram of the structure of an early warning system for geological disasters. DETAILED DESCRIPTION

[0058] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0059] like Figure 1 As shown, a method for rapid response to geological disasters includes steps S1-S4, which are as follows:

[0060] S1. Determine the target monitoring area for geological disasters, collect road surface image data, slope image data and slope millimeter wave radar data of the target monitoring area in real time, extract the state value of the road surface image data to obtain the health status data of the road surface, and extract multi-target features of the slope image data to obtain the slope feature map.

[0061] In an optional embodiment of the present invention, the present invention constructs a data acquisition module, through which pavement image data, slope image data and slope millimeter-wave radar data of the target monitoring area are collected in real time, state value extraction is performed on the pavement image data to obtain pavement health status data, and multi-target feature extraction is performed on the slope image data to obtain a slope feature map.

[0062] The present invention extracts state values ​​from road surface image data to obtain road surface health status data. Specifically, a road surface health status detection model is designed with road surface image data as input and road surface health status data as output. The road surface health status data includes road surface deformation value, road surface inclination value and road surface height change value. The road surface health status detection model is used to extract state values ​​from road surface image data to obtain road surface health status data.

[0063] The present invention sets corresponding weights for the road deformation value, the road inclination value and the road height change value, and then obtains the road health status value based on the road deformation value, the road inclination value and the road height change value and the corresponding weights. The road health status value sets four corresponding threshold intervals, which respectively represent no danger warning, low risk warning, medium risk warning and high risk warning.

[0064] The present invention adopts the open source SAM large model when designing the road health status detection model.

[0065] The present invention uses a data acquisition module to extract multi-target features from slope image data to obtain a slope feature map, including the following steps:

[0066] A1. Determine the geological disaster image features that are independent of each other and obey the Gaussian distribution, and construct the multidimensional feature space of geological disasters based on the geological disaster image features, which can be expressed as:

[0067] X=span(ξ1,…ξ K , …ξ N )

[0068] Among them: X is the multidimensional feature space of geological disasters, span is the spatial symbol, ξ1 is the geological disaster image feature of the first dimension, ξ K is the K-th dimension geological disaster image feature, ξ N is the geological disaster image feature of the Nth dimension.

[0069] A2. Based on the multidimensional feature space of geological disasters in step A1, the slope image data is converted into target data to be extracted, which is expressed as:

[0070]

[0071]

[0072] Where: f is the slope image, i is the serial number of the target to be extracted in the slope image, M is the number of targets to be extracted in the slope image, and f i is the i-th target to be extracted in the slope image, j is the serial number of the image feature dimension, N is the number of image feature dimensions, ξ j is the geological disaster image feature of the jth dimension, x j is the characteristic decomposition coefficient of the j-th dimension.

[0073] A3. Construct the imaging matrix of multi-target features, expressed as:

[0074]

[0075] in: is the imaging matrix of multi-target features, n is the sequence number of the feature coefficient, ω n is the nth characteristic coefficient.

[0076] A4. Based on the target data to be extracted converted in step A2, the imaging matrix of the multi-target features in step A3 is used to extract the multi-target features of the slope image data to obtain a slope feature map, which is expressed as:

[0077]

[0078] in: It is the slope characteristic map.

[0079] S2. Construct a slope multidimensional data fusion model, use the slope multidimensional data fusion model to fuse the slope millimeter wave radar data and the slope feature map in step S1, obtain the slope fusion feature map of the target monitoring area, and determine the warning status of geological disasters based on the slope fusion feature map of the target monitoring area.

[0080] In an optional embodiment of the present invention, the present invention constructs a geological disaster detection module, constructs a slope multidimensional data fusion model in the geological disaster detection module, uses the slope multidimensional data fusion model to fuse the slope millimeter wave radar data and the slope feature map, obtains the slope fusion feature map of the target monitoring area, and determines the warning status of the geological disaster based on the slope fusion feature map of the target monitoring area.

[0081] The present invention constructs a slope multi-dimensional data fusion model, which includes an input layer, an encoder, a feature splicing layer and a decoder.

[0082] The input layer is used to splice the slope millimeter wave radar data and the slope feature map to obtain the input feature map, which is expressed as:

[0083]

[0084]

[0085] X rader ∈R H×W×1

[0086] Where: X input is the input feature map, concat is the concatenation operator, is the slope characteristic map, X rader is the slope millimeter-wave radar data, R is a placeholder, H is the height of the image, W is the width of the image, and 4 is the number of channels.

[0087] The splicing calculation of the input layer in the present invention is based on time and space, that is, the slope millimeter wave radar data and the slope feature map are aligned by using time and space, and then spliced.

[0088] The encoder is used to perform convolution and pooling operations on the input feature map to obtain a compressed feature map, which is expressed as:

[0089]

[0090]

[0091] in: is the compressed feature map after the input feature map is convolved by the l-th layer encoder, l is the current number of layers, σ is the activation function, is the weight coefficient of the convolution of the encoder at layer l, is the compressed feature map after the input feature map is convolved and pooled by the (l-1)th layer encoder. is the bias coefficient of the convolution of the encoder at layer l, It is the compressed feature map after the input feature map is pooled by the l-th layer encoder, and MaxPool is the maximum pooling operation.

[0092] The feature concatenation layer is used to concatenate the compressed feature map in the encoder with the upsampled feature map in the decoder to obtain the detail feature map, which is expressed as:

[0093]

[0094] in: is the detail feature map of the input feature map spliced ​​by the lth layer, is the feature map after the input feature map is upsampled by the l-th layer decoder, It is the compressed feature map after the input feature map is convolved by the (Ll-1)th layer encoder, and L is the total number of encoder layers.

[0095] The decoder is used to upsample the compressed feature map and perform convolution on the detail feature map to output the slope fusion feature map of the target monitoring area, which is expressed as:

[0096]

[0097]

[0098]

[0099] Among them: UpSamplc is the upsampling operation, is the compressed feature map after the input feature map is pooled by the (Ll-1)th layer encoder, is the feature map after the input feature map is pooled by the l-th layer decoder, is the weight coefficient of the convolution of the l-th layer decoder, is the bias coefficient of the convolution of the l-th layer decoder, X output is the slope fusion feature map of the target monitoring area, W output is the weight coefficient of the last layer of decoder convolution, b output is the bias coefficient of the last decoder convolution.

[0100] The present invention determines the early warning status of geological disasters according to the slope fusion feature map of the target monitoring area, specifically: the disaster type, disaster area, disaster location, disaster movement direction and disaster movement speed of the target monitoring area are obtained according to the slope fusion feature map of the target monitoring area, and the early warning status of the corresponding geological disaster is determined according to the disaster type, disaster area, disaster location, disaster movement direction and disaster movement speed of the target monitoring area.

[0101] The early warning status of geological disasters is expressed as:

[0102] AS=(Le, Ty, Ar, Lo, Di, Ve)

[0103] Among them: AS is the warning status of geological disasters, Le is the disaster level, Ty is the disaster type, Ar is the disaster area, Lo is the disaster location, Di is the disaster movement direction, and Ve is the disaster movement speed.

[0104] The disaster levels in the present invention include no danger, low risk, medium risk and high risk. The disaster level is determined by the disaster area, disaster location, disaster movement direction and disaster movement speed of the disaster type according to actual experience. The disaster type includes landslide, mud-rock flow and collapse.

[0105] S3. Based on the Django server framework, Pyecharts is used to visualize the health status data of the road surface in step S1 and the warning status of the geological disaster in step S2 in real time to obtain the visualization results of the geological disaster warning.

[0106] In an optional embodiment of the present invention, the present invention constructs a data visualization module. The data visualization module is based on the Django server framework, and uses Pyecharts to visualize the health status data of the road surface and the warning status of geological disasters in real time, and obtains the visualization results of geological disaster warnings, specifically: a visualization platform for geological disaster warnings is built based on the Django server framework, and the visualization platform for geological disaster warnings includes a data layer, an application layer, and a presentation layer; the data layer has a built-in MySQL database for storing the health status data of the road surface and the warning status of geological disasters; the application layer is used to process the request instructions of the presentation layer, and determines whether to interact with the data layer according to the request instructions, and determines the mode of data interaction; the presentation layer is used to transmit the request instructions, and the health status data of the road surface and the warning status of geological disasters received after the request instructions are passed are visualized in real time using Pyecharts to obtain the visualization results of geological disaster warnings.

[0107] The request instructions of the presentation layer include HTTP (HyperText Transfer Protocol) request instructions and WebSocket (a full-duplex communication protocol) request instructions. In the present invention, for data with high real-time requirements, i.e., the early warning status of geological disasters, WebSocket request instructions are used to establish a persistent connection with the application layer to receive data push from the data layer in real time. In the present invention, for non-real-time demand data, i.e., road health status data, HTTP request instructions are periodically initiated. The present invention combines the use of HTTP request instructions and WebSocket request instructions to improve the efficiency and real-time performance of data transmission and reduce the consumption of network resources.

[0108] The present invention sets cross-site request forgery protection when transmitting request instructions at the presentation layer, and generates a random CSRF (Cross-site request forgery) token for the request instruction. The application layer checks whether the request instruction contains a valid CSRF token to verify whether the request instruction comes from a trusted source.

[0109] The present invention sets SQL injection protection when the application layer determines whether to interact with the data layer according to the request instruction, operates the data layer through the Django model and query set to interact with the data, avoids directly splicing the input request instruction into the SQL statement, and thus reduces the risk of SQL injection.

[0110] The present invention sets cross-site scripting attack protection when the presentation layer uses HTML pages for visualization, automatically performs HTML escape on data to prevent cross-site scripting attacks. When data is rendered to an HTML page, the HTML tags therein will be converted into HTML entities, thereby avoiding the execution of malicious scripts.

[0111] S4. Based on the visualization result of the geological disaster warning in step S3, perform real-time geological disaster warning operation.

[0112] In an optional embodiment of the present invention, the present invention constructs a geological disaster warning module. The geological disaster warning module performs real-time geological disaster warning operations based on the visualization results of geological disaster warnings. In the present invention, geological disaster risk warnings include no-risk warnings, low-risk warnings, medium-risk warnings, and high-risk warnings. High-risk warnings use both LED display screens and sound and light alarm devices to perform warning operations; medium-risk warnings use both LED display screens and light alarm devices to perform warning operations; and low-risk warnings use light alarm devices to perform warning operations.

[0113] like Figure 2As shown, a geological disaster early warning system using the above method includes a data acquisition module, a geological disaster detection module, a data visualization module and a geological disaster early warning module.

[0114] In an optional embodiment of the present invention, the data acquisition module is used to determine the target monitoring area for geological disasters, collect road surface image data, slope image data and slope millimeter wave radar data of the target monitoring area in real time, extract state values ​​of the road surface image data to obtain road health status data, and perform multi-target feature extraction on the slope image data to obtain a slope feature map.

[0115] In an optional embodiment of the present invention, the geological disaster detection module is used to construct a slope multidimensional data fusion model, use the slope multidimensional data fusion model to fuse the slope millimeter wave radar data and the slope feature map, obtain the slope fusion feature map of the target monitoring area, and determine the warning status of the geological disaster based on the slope fusion feature map of the target monitoring area.

[0116] In an optional embodiment of the present invention, the data visualization module is used to visualize the health status data of the road surface and the warning status of geological disasters in real time using Pyecharts based on the Django server framework to obtain visualization results of geological disaster warnings.

[0117] In an optional embodiment of the present invention, the geological disaster warning module is used to perform real-time geological disaster warning operations based on the visualization results of the geological disaster warning.

[0118] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0119] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0120] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0121] The present invention uses specific embodiments to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

[0122] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific variations and combinations that do not deviate from the essence of the present invention based on the technical revelations disclosed by the present invention, and these variations and combinations are still within the protection scope of the present invention.

Claims

1. A method for rapid response to geological disasters, It is characterized in that The following steps are involved: S1. Determine the target monitoring area for geological disasters, collect road surface image data, slope surface image data and slope surface millimeter wave radar data in real time in the target monitoring area, extract state values ​​from the road surface image data to obtain road surface health status data, and perform multi-target feature extraction on the slope surface image data to obtain a slope surface feature map; S2, constructing a slope multidimensional data fusion model, using the slope multidimensional data fusion model to fuse the slope millimeter wave radar data and the slope feature map in step S1, obtaining the slope fusion feature map of the target monitoring area, and determining the early warning status of the geological disaster according to the slope fusion feature map of the target monitoring area; S3. Based on the Django server framework, Pyecharts is used to visualize the health status data of the road surface in step S1 and the warning status of the geological disaster in step S2 in real time to obtain the visualization results of the geological disaster warning; S4. Based on the visualization result of the geological disaster warning in step S3, perform real-time geological disaster warning operation.

2. The rapid response method for geological disasters according to claim 1, It is characterized in that In step S1, the state value of the road surface image data is extracted to obtain the health status data of the road surface. Specifically, a road surface health status detection model is designed with the road surface image data as input and the road surface health status data as output. The road surface health status data includes road surface deformation value, road surface inclination value and road surface height change value. The road surface health status detection model is used to extract the state value of the road surface image data to obtain the health status data of the road surface.

3. The rapid response method for geological disasters according to claim 1, It is characterized in that In step S1, multi-objective feature extraction is performed on the slope image data to obtain a slope feature map, including the following steps: A1. Determine the geological disaster image features that are independent of each other and obey the Gaussian distribution, and construct the multidimensional feature space of geological disasters based on the geological disaster image features, which can be expressed as: X=span(ξ 1 ,…x K ,…x N ) Among them: X is the multidimensional characteristic space of geological disasters, span is the spatial symbol, ξ 1 is the first dimension of geological disaster image features, ξ K is the K-th dimension geological disaster image feature, ξ N is the geological disaster image feature of the nth dimension; A2. Based on the multidimensional feature space of geological disasters in step A1, the slope image data is converted into target data to be extracted, which is expressed as: Where: f is the slope image, i is the serial number of the target to be extracted in the slope image, M is the number of targets to be extracted in the slope image, and f i is the i-th target to be extracted in the slope image, j is the serial number of the image feature dimension, N is the number of image feature dimensions, ξ j is the geological disaster image feature of the jth dimension, x j is the characteristic decomposition coefficient of the j-th dimension; A3. Construct the imaging matrix of multi-target features, expressed as: in: is the imaging matrix of multi-target features, n is the sequence number of the feature coefficient, ω n is the nth characteristic coefficient; A4. Based on the target data to be extracted converted in step A2, the imaging matrix of the multi-target features in step A3 is used to extract the multi-target features of the slope image data to obtain a slope feature map, which is expressed as: in: It is the slope characteristic map.

4. The rapid response method for geological disasters according to claim 1, It is characterized in that In step S2, the slope multi-dimensional data fusion model includes an input layer, an encoder, a feature splicing layer and a decoder; The input layer is used to splice the slope millimeter wave radar data and the slope feature map to obtain the input feature map, which is expressed as: X rader ∈R H×W×1 Where: X input is the input feature map, concat is the concatenation operator, is the slope characteristic map, X rader is the slope millimeter-wave radar data, R is a placeholder, H is the height of the image, W is the width of the image, and 4 is the number of channels; The encoder is used to perform convolution and pooling operations on the input feature map to obtain a compressed feature map, which is expressed as: in: is the compressed feature map after the input feature map is convolved by the l-th layer encoder, l is the current number of layers, σ is the activation function, is the weight coefficient of the convolution of the encoder at layer l, is the compressed feature map after the input feature map is convolved and pooled by the (l-1)th layer encoder. is the bias coefficient of the convolution of the encoder at layer l, is the compressed feature map after the input feature map is pooled by the l-th layer encoder, and MaxPool is the maximum pooling operation; The feature concatenation layer is used to concatenate the compressed feature map in the encoder with the upsampled feature map in the decoder to obtain the detail feature map, which is expressed as: in: is the detail feature map of the input feature map spliced ​​by the lth layer, is the feature map after the input feature map is upsampled by the l-th layer decoder, is the compressed feature map after the input feature map is convolved by the (Ll-1)th layer encoder, where l is the total number of encoder layers; The decoder is used to upsample the compressed feature map and perform convolution on the detail feature map to output the slope fusion feature map of the target monitoring area, which is expressed as: Among them: UpSample is the upsampling operation, is the compressed feature map after the input feature map is pooled by the (Ll-1)th layer encoder, is the feature map after the input feature map is pooled by the l-th layer decoder, is the weight coefficient of the convolution of the l-th layer decoder, is the bias coefficient of the convolution of the l-th layer decoder, X output is the slope fusion feature map of the target monitoring area, W output is the weight coefficient of the last layer of decoder convolution, b output is the bias coefficient of the last decoder convolution.

5. The rapid response method for geological disasters according to claim 1, It is characterized in that In step S2, the warning status of the geological disaster is determined according to the slope fusion feature map of the target monitoring area, specifically: the disaster type, disaster area, disaster location, disaster movement direction and disaster movement speed of the target monitoring area are obtained according to the slope fusion feature map of the target monitoring area, and the warning status of the corresponding geological disaster is determined according to the disaster type, disaster area, disaster location, disaster movement direction and disaster movement speed of the target monitoring area.

6. The rapid response method for geological disasters according to claim 1, It is characterized in that In step S3, based on the Django server framework, Pyecharts is used to visualize the health status data of the road surface in step S1 and the warning status of the geological disasters in step S2 in real time to obtain the visualization results of the geological disaster warning, specifically: a visualization platform for geological disaster warning is built based on the Django server framework, and the visualization platform for geological disaster warning includes a data layer, an application layer and a presentation layer; the data layer has a built-in MySQL database for storing the health status data of the road surface and the warning status of the geological disasters; The application layer is used to process the request instructions of the presentation layer, determine whether to interact with the data layer based on the request instructions, and determine the method of data interaction; the presentation layer is used to transmit the request instructions, and use Pyecharts to visualize the road health status data and geological disaster warning status received after the request instructions are passed in real time to obtain the visualization results of geological disaster warning.

7. The method for rapid response to geological disasters according to claim 6, It is characterized in that When the application layer determines whether to interact with the data layer based on the request instruction, SQL injection protection is set up, and the data layer is operated through Django's models and query sets to interact with the data.

8. A geological disaster early warning system using the method described in any one of claims 1 to 7, It is characterized in that It includes data acquisition module, geological disaster detection module, data visualization module and geological disaster early warning module; The data acquisition module is used to determine the target monitoring area for geological disasters, collect road surface image data, slope image data and slope millimeter wave radar data in real time in the target monitoring area, extract state values ​​from the road surface image data to obtain road health status data, and perform multi-target feature extraction on the slope image data to obtain a slope feature map; The geological disaster detection module is used to construct a slope multi-dimensional data fusion model, use the slope multi-dimensional data fusion model to fuse the slope millimeter wave radar data and the slope characteristic map, obtain the slope fusion characteristic map of the target monitoring area, and determine the early warning status of the geological disaster based on the slope fusion characteristic map of the target monitoring area; The data visualization module is used to visualize the health status data of the road surface and the early warning status of geological disasters in real time based on the Django server framework using Pyecharts to obtain the visualization results of geological disaster early warnings; The geological disaster warning module is used to perform real-time geological disaster warning operations based on the visualization results of geological disaster warning.

Images