CN118171918A - Air-space-ground landslide monitoring method and system based on multi-source data - Google Patents
Air-space-ground landslide monitoring method and system based on multi-source data Download PDFInfo
- Publication number
- CN118171918A CN118171918A CN202410591712.9A CN202410591712A CN118171918A CN 118171918 A CN118171918 A CN 118171918A CN 202410591712 A CN202410591712 A CN 202410591712A CN 118171918 A CN118171918 A CN 118171918A
- Authority
- CN
- China
- Prior art keywords
- data
- monitoring
- landslide
- space
- foundation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 207
- 238000000034 method Methods 0.000 title claims abstract description 20
- 238000012545 processing Methods 0.000 claims abstract description 55
- 230000007613 environmental effect Effects 0.000 claims abstract description 26
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 21
- 239000013598 vector Substances 0.000 claims description 44
- 230000000007 visual effect Effects 0.000 claims description 27
- 230000011218 segmentation Effects 0.000 claims description 15
- 238000000605 extraction Methods 0.000 claims description 13
- 238000012876 topography Methods 0.000 claims description 10
- 238000004458 analytical method Methods 0.000 claims description 8
- 238000006073 displacement reaction Methods 0.000 claims description 7
- 230000004044 response Effects 0.000 claims description 7
- 239000002689 soil Substances 0.000 claims description 6
- 238000012935 Averaging Methods 0.000 claims description 5
- 230000004913 activation Effects 0.000 claims description 5
- 230000008859 change Effects 0.000 claims description 5
- 238000009792 diffusion process Methods 0.000 claims description 5
- 238000001914 filtration Methods 0.000 claims description 5
- 238000012806 monitoring device Methods 0.000 claims description 5
- 239000013049 sediment Substances 0.000 claims description 5
- 238000013507 mapping Methods 0.000 claims description 4
- 230000008901 benefit Effects 0.000 abstract description 2
- 230000006870 function Effects 0.000 description 8
- 238000005516 engineering process Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 238000001514 detection method Methods 0.000 description 2
- 230000002265 prevention Effects 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000003628 erosive effect Effects 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Landscapes
- Pit Excavations, Shoring, Fill Or Stabilisation Of Slopes (AREA)
- Testing Or Calibration Of Command Recording Devices (AREA)
Abstract
The invention relates to the technical field of landslide monitoring, and particularly discloses an aerospace-earth landslide monitoring method and system based on multi-source data. The system comprises a data acquisition layer, a data processing layer and an application service layer; the air-based monitoring data, the space-based monitoring data and the foundation monitoring data are processed under different scales, the advantages of the air-based monitoring means are fully exerted, the intelligent judgment of landslide hazard is carried out on the foundation monitoring data by utilizing the artificial intelligent convolutional neural network model, the regional video monitoring data and the influence of the environmental data on landslide judgment are considered in the foundation monitoring data, the accuracy of landslide monitoring early warning is further improved, and powerful support is provided for landslide judgment by workers.
Description
Technical Field
The invention belongs to the technical field of landslide monitoring, and particularly relates to an aerospace-earth landslide monitoring method and system based on multi-source data.
Background
Landslide hazard is one of serious geological disasters, and dynamic monitoring and early warning of landslide are key to disaster prevention and reduction. The initial deformation of landslide in time is smaller, and the middle-late deformation is larger; spatially representing the spatial patterns of trailing edge tension cracks, leading edge bulge, and side edge shear strain. The parameters available for monitoring mainly include stress, deformation, water parameters, acoustic emission, etc., wherein the most widely used is deformation. Therefore, the detection of landslide needs to comprehensively consider the characteristics of time and space evolution, and the multi-parameter monitoring is combined to carry out comprehensive detection.
At present, in landslide monitoring and early warning, a plurality of monitoring modes exist, because the defects of a single monitoring mode are obvious, the technology of 'empty', 'sky', 'earth' is advantageous and insufficient, the satellite remote sensing image can realize large-scale coverage and can provide a large amount of basic data, but the main reflection of the satellite remote sensing image is a plane image, the elevation precision of a monitored area cannot be accurately reflected, and the image is greatly affected by cloud layer interference and the like. The unmanned aerial vehicle measures that the operation time, place are flexible, and data reality is good, the precision is high, but can't carry out more careful soil water content, measure of parameter such as surface erosion volume. The ground monitoring often uses a sensor, only punctiform local parts can be monitored, and although the data is accurate, especially the pertinence is strongest in the manual investigation, the whole area is difficult to develop due to time and effort waste or extremely large data volume. Therefore, in recent years, the three technologies are usually complemented and comprehensively utilized, but with the development of big data and artificial intelligence technologies, the space is greatly improved in the aspects of system construction and algorithm of space-earth integrated monitoring, and the invention aims to provide a space-earth landslide monitoring method and system based on multi-source data.
Disclosure of Invention
The invention aims to overcome the problems in the prior art and provide an air-space-ground landslide monitoring method and system based on multi-source data.
In order to achieve the above object, the present invention provides the following technical solutions:
The sky-land landslide monitoring method based on the multi-source data is characterized by comprising the following steps of:
Acquiring space-based monitoring data, space-based monitoring data and foundation monitoring data; the space-based monitoring data are GNSS monitoring data, the space-based monitoring data are unmanned aerial vehicle aerial image data, and the ground-based monitoring data are ground real-time video data and environment monitoring data;
Analyzing and processing the space-based monitoring data, positioning a large-scale landslide risk area, and starting day-based data acquisition equipment of the corresponding area; analyzing and processing the space-based monitoring data to extract three-dimensional point cloud data and model three-dimensional data of the slope surface, positioning a mesoscale landslide risk area, and starting foundation data acquisition equipment of the corresponding area; analyzing and processing the foundation monitoring data to realize real-time video data feature extraction and environment data feature extraction, and judging landslide hazard by adopting a pre-trained artificial convolutional neural network model;
And setting an early warning rule and an early warning response according to the monitoring result and the landslide hazard degree, and realizing visual display of the monitoring data.
Further, the analyzing and processing the space-based monitoring data further includes:
Setting n observation time points, m monitoring points and 1 datum point, counting the earth height difference of each monitoring point and the datum point on each observation time point, taking the m earth height differences measured for the first time as initial values, subtracting the initial values of the corresponding monitoring points from the earth height differences on n-1 observation time points to obtain settlement data, and transmitting the settlement data into a visual monitoring unit to realize visual monitoring; and analyzing the settlement amount data, and determining an area with the settlement amount larger than a preset threshold value as a large-scale landslide risk area.
Further, the analyzing and processing the day-based monitoring data further comprises:
In the large-scale landslide risk area, starting unmanned aerial vehicle aerial photographing equipment to collect multi-angle unmanned aerial vehicle image data; extracting GPS coordinates and camera gestures in the multi-angle unmanned aerial vehicle image data, generating sparse point cloud data of a target area, and obtaining dense point cloud data through feature point matching, reconstruction, spatial point diffusion and filtering; and carrying out three-dimensional modeling and space-time evolution analysis on the ground surface of the large-scale landslide risk area according to the dense point cloud data, and determining the area with the ground surface displacement deformation exceeding a preset threshold value within a preset time range as a mesoscale landslide risk area.
Further, the analyzing and processing the foundation monitoring data further comprises:
in the mesoscale landslide risk area, starting a foundation video monitoring device to acquire real-time video data, and transmitting the real-time video data to a visual monitoring unit for realizing manual monitoring; extracting a key frame image set in the real-time video data, carrying out target identification and segmentation on the key frame image set, carrying out gridding treatment on a landslide area in the key frame image set after the target segmentation, and extracting coordinates and characteristic points of each grid point; integrating grid point coordinates and feature points in a landslide region in the key frame image set after the target segmentation to form a time sequence feature vector;
Collecting environmental data of a preset time range in the mesoscale landslide risk area, wherein the environmental data comprise rainfall, soil humidity, gradient, slope direction, plane curvature, section curvature, topography humidity index, sediment transport index and topography roughness index; each item in the environmental data is respectively subjected to averaging treatment and then spliced to form an environmental data feature vector;
Normalizing the time sequence feature vector and the environment feature vector, fusing the time sequence feature vector and the environment feature vector into foundation data features, inputting the foundation data features into a pre-trained artificial convolutional neural network model, and outputting a probability vector of landslide hazard degree;
The pre-trained artificial convolutional neural network model comprises an input layer, an output layer, 2 hidden layers and 3 convolutional layers, relu is adopted as an activation function, and cross entropy is adopted as a loss function.
Further, setting an early warning rule and an early warning response according to the monitoring result and the landslide hazard degree, and further comprising:
And establishing a mapping relation between a monitoring result and landslide hazard degree, wherein the hazard degree is divided into three gears of high, medium and low, and the monitoring result comprises an empty base, a space base and a foundation monitoring result.
Further, realize visual show to monitoring data, further include:
Setting parameters to be monitored, graphically drawing discrete monitoring data to reflect time change characteristics, and triggering an alarm to remind workers of paying attention when certain monitoring data is higher than a preset value.
The invention also realizes the following technical scheme:
the air-sky landslide monitoring system based on the multi-source data is characterized by comprising the following functional modules:
Data acquisition layer: the method comprises the steps of acquiring space-based monitoring data, space-based monitoring data and foundation monitoring data; the space-based monitoring data are GNSS monitoring data, the space-based monitoring data are unmanned aerial vehicle aerial image data, and the ground-based monitoring data are ground real-time video data and environment monitoring data;
Data processing layer: the system comprises an empty base data processing unit, a data acquisition unit and a data processing unit, wherein the empty base data processing unit is used for analyzing and processing the empty base monitoring data, positioning a large-scale landslide risk area and starting a day base data acquisition device of the corresponding area; the day-based data processing unit is used for analyzing and processing the day-based monitoring data, extracting three-dimensional point cloud data and modeling three-dimensional data of the slope surface, positioning a mesoscale landslide risk area and starting foundation data acquisition equipment of the corresponding area; the foundation data processing unit is used for analyzing and processing the foundation monitoring data, realizing real-time video data feature extraction and environment data feature extraction, and realizing landslide hazard judgment by adopting a pre-trained artificial convolutional neural network model;
application service layer: the system comprises an early warning unit, a monitoring unit and a control unit, wherein the early warning unit is used for setting early warning rules and early warning responses according to monitoring results and landslide hazard degrees; and the visual monitoring unit is used for realizing visual display of the monitored data.
Further, the space-based data processing unit further includes:
Setting n observation time points, m monitoring points and 1 datum point, counting the earth height difference of each monitoring point and the datum point on each observation time point, taking the m earth height differences measured for the first time as initial values, subtracting the initial values of the corresponding monitoring points from the earth height differences on n-1 observation time points to obtain settlement data, and transmitting the settlement data into a visual monitoring unit to realize visual monitoring; and analyzing the settlement amount data, and determining an area with the settlement amount larger than a preset threshold value as a large-scale landslide risk area.
Further, the day data processing unit further includes:
In the large-scale landslide risk area, starting unmanned aerial vehicle aerial photographing equipment to collect multi-angle unmanned aerial vehicle image data; extracting GPS coordinates and camera gestures in the multi-angle unmanned aerial vehicle image data, generating sparse point cloud data of a target area, and obtaining dense point cloud data through feature point matching, reconstruction, spatial point diffusion and filtering; and carrying out three-dimensional modeling and space-time evolution analysis on the ground surface of the large-scale landslide risk area according to the dense point cloud data, and determining the area with the ground surface displacement deformation exceeding a preset threshold value within a preset time range as a mesoscale landslide risk area.
Further, the foundation data processing unit further includes:
in the mesoscale landslide risk area, starting a foundation video monitoring device to acquire real-time video data, and transmitting the real-time video data to a visual monitoring unit for realizing manual monitoring; extracting a key frame image set in the real-time video data, carrying out target identification and segmentation on the key frame image set, carrying out gridding treatment on a landslide area in the key frame image set after the target segmentation, and extracting coordinates and characteristic points of each grid point; integrating grid point coordinates and feature points in a landslide region in the key frame image set after the target segmentation to form a time sequence feature vector;
Collecting environmental data of a preset time range in the mesoscale landslide risk area, wherein the environmental data comprise rainfall, soil humidity, gradient, slope direction, plane curvature, section curvature, topography humidity index, sediment transport index and topography roughness index; each item in the environmental data is respectively subjected to averaging treatment and then spliced to form an environmental data feature vector;
Normalizing the time sequence feature vector and the environment feature vector, fusing the time sequence feature vector and the environment feature vector into foundation data features, inputting the foundation data features into a pre-trained artificial convolutional neural network model, and outputting a probability vector of landslide hazard degree;
The pre-trained artificial convolutional neural network model comprises an input layer, an output layer, 2 hidden layers and 3 convolutional layers, relu is adopted as an activation function, and cross entropy is adopted as a loss function.
Further, the early warning unit further includes:
Establishing a mapping relation between a monitoring result and landslide hazard degree, wherein the hazard degree is divided into three steps of high, medium and low, and the monitoring result comprises an empty base, a space base and a foundation monitoring result;
further, the visual monitoring unit further includes:
Setting parameters to be monitored, graphically drawing discrete monitoring data to reflect time change characteristics, and triggering an alarm to remind workers of paying attention when certain monitoring data is higher than a preset value.
According to the application, by utilizing the combination of the space-sky-ground integrated landslide monitoring means, the landslide monitoring range is continuously reduced, the landslide monitoring precision is improved, and the advantage complementation of each monitoring method is realized.
The application realizes the processing and the fusion of the space-sky multi-source data, and effectively utilizes the characteristics of the data collected under different scales to monitor landslide.
According to the landslide risk judging method, the artificial intelligent convolutional neural network is used for judging the landslide risk of the foundation monitoring data, the input characteristics of the convolutional neural network are combined with the real-time monitoring video data and the environment data, and the accuracy of landslide risk judgment is further improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the embodiments will be briefly described below, it being understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and other related drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a functional flow chart of an air-to-ground landslide monitoring method based on multi-source data;
FIG. 2 is a functional block diagram of an air-to-ground landslide monitoring system based on multi-source data;
FIG. 3 is a functional flow diagram of a space-based data processing unit according to the present application;
FIG. 4 is a functional flow chart of the day-based data processing unit of the present application;
fig. 5 is a functional flowchart of the foundation data processing unit of the present application.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. The following description of at least one exemplary embodiment is merely exemplary in nature and is in no way intended to limit the invention, its application, or uses. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Example 1
As shown in fig. 1, the embodiment provides an aerospace-land landslide monitoring method based on multi-source data, which includes the following steps:
Acquiring space-based monitoring data, space-based monitoring data and foundation monitoring data; the space-based monitoring data are GNSS monitoring data, the space-based monitoring data are unmanned aerial vehicle aerial image data, and the ground-based monitoring data are ground real-time video data and environment monitoring data;
Analyzing and processing the space-based monitoring data, positioning a large-scale landslide risk area, and starting day-based data acquisition equipment of the corresponding area; analyzing and processing the space-based monitoring data to extract three-dimensional point cloud data and model three-dimensional data of the slope surface, positioning a mesoscale landslide risk area, and starting foundation data acquisition equipment of the corresponding area; analyzing and processing the foundation monitoring data to realize real-time video data feature extraction and environment data feature extraction, and judging landslide hazard by adopting a pre-trained artificial convolutional neural network model;
And setting an early warning rule and an early warning response according to the monitoring result and the landslide hazard degree, and realizing visual display of the monitoring data.
Example 2
As shown in fig. 2, the embodiment provides an aerospace-land landslide monitoring system based on multi-source data, which comprises the following functional modules:
Data acquisition layer: the method comprises the steps of acquiring space-based monitoring data, space-based monitoring data and foundation monitoring data; the space-based monitoring data are GNSS monitoring data, the space-based monitoring data are unmanned aerial vehicle aerial image data, and the ground-based monitoring data are ground real-time video data and environment monitoring data;
Data processing layer: the system comprises an empty base data processing unit, a data acquisition unit and a data processing unit, wherein the empty base data processing unit is used for analyzing and processing the empty base monitoring data, positioning a large-scale landslide risk area and starting a day base data acquisition device of the corresponding area; the day-based data processing unit is used for analyzing and processing the day-based monitoring data, extracting three-dimensional point cloud data and modeling three-dimensional data of the slope surface, positioning a mesoscale landslide risk area and starting foundation data acquisition equipment of the corresponding area; the foundation data processing unit is used for analyzing and processing the foundation monitoring data, realizing real-time video data feature extraction and environment data feature extraction, and realizing landslide hazard judgment by adopting a pre-trained artificial convolutional neural network model;
application service layer: the system comprises an early warning unit, a monitoring unit and a control unit, wherein the early warning unit is used for setting early warning rules and early warning responses according to monitoring results and landslide hazard degrees; and the visual monitoring unit is used for realizing visual display of the monitored data.
Example 3
As shown in fig. 3, the present embodiment provides an air-space-ground landslide monitoring system based on multi-source data, where the air-space-data processing unit specifically includes:
Setting n observation time points, m monitoring points and 1 datum point, counting the earth height difference of each monitoring point and the datum point on each observation time point, taking the m earth height differences measured for the first time as initial values, subtracting the initial values of the corresponding monitoring points from the earth height differences on n-1 observation time points to obtain settlement data, and transmitting the settlement data into a visual monitoring unit to realize visual monitoring; and analyzing the settlement amount data, and determining an area with the settlement amount larger than a preset threshold value as a large-scale landslide risk area.
GNSS is mainly used for monitoring surface displacement. Firstly, a base station is constructed, necessary equipment is installed, then a solar panel is placed, a pipeline is connected to a GNSS station pole, a storage battery is placed in the GNSS station pole, SIM (Subscriber Identity Module) cards are installed in the equipment, a GPRS antenna, the GNSS antenna and input voltage are connected, and finally, a professional technician inputs related data information into the equipment to establish a landslide monitoring change system, so that real-time monitoring is realized, and meanwhile, early warning and forecasting information is issued to a specific area by utilizing a multimedia mode such as a television, a short message, the Internet and the like to guide disaster prevention and reduction work.
In the invention, GNSS settlement displacement statistics is carried out, and GNSS monitoring feasibility demonstration, reference point stability analysis and GNSS measurement accuracy analysis are also required to ensure the reliability, stability and accuracy of GNSS monitoring data, thereby ensuring the accuracy of monitoring results.
Example 4
As shown in fig. 4, the present embodiment provides an air-space-ground landslide monitoring system based on multi-source data, where the day-base data processing unit specifically includes:
In the large-scale landslide risk area, starting unmanned aerial vehicle aerial photographing equipment to collect multi-angle unmanned aerial vehicle image data; extracting GPS coordinates and camera gestures in the multi-angle unmanned aerial vehicle image data, generating sparse point cloud data of a target area, and obtaining dense point cloud data through feature point matching, reconstruction, spatial point diffusion and filtering; and carrying out three-dimensional modeling and space-time evolution analysis on the ground surface of the large-scale landslide risk area according to the dense point cloud data, and determining the area with the ground surface displacement deformation exceeding a preset threshold value within a preset time range as a mesoscale landslide risk area.
When the large-scale landslide risk area is positioned by utilizing the air-based monitoring means, the application utilizes the unmanned aerial vehicle aerial photographing equipment to photograph the image of the large-scale area, and before photographing, the flight route, the flight angle, the flight height, the flight speed, the flight time and the aerial photographing view finding scheme of the unmanned aerial vehicle are required to be designed. According to the application, ground objects can be shot from multiple angles by utilizing the unmanned aerial vehicle inclination aerial photogrammetry technology, so that more accurate ground information can be obtained. When the image processing is carried out, the multi-angle image data can be spliced and fused, so that the overall feature extraction of the target is realized.
Three-dimensional modeling of the ground object target is realized through the extracted point cloud data, time data corresponding to the image is combined, space-time evolution analysis is carried out on a target area after the three-dimensional modeling, and an area with the spatial deformation of the target within a certain time range larger than a preset threshold value is positioned to be used as an area for next foundation monitoring.
Example 5
As shown in fig. 5, the present embodiment provides an aerospace-land landslide monitoring system based on multi-source data, where the foundation data processing unit specifically includes:
in the mesoscale landslide risk area, starting a foundation video monitoring device to acquire real-time video data, and transmitting the real-time video data to a visual monitoring unit for realizing manual monitoring; extracting a key frame image set in the real-time video data, carrying out target identification and segmentation on the key frame image set, carrying out gridding treatment on a landslide area in the key frame image set after the target segmentation, and extracting coordinates and characteristic points of each grid point; integrating grid point coordinates and feature points in a landslide region in the key frame image set after the target segmentation to form a time sequence feature vector;
Collecting environmental data of a preset time range in the mesoscale landslide risk area, wherein the environmental data comprise rainfall, soil humidity, gradient, slope direction, plane curvature, section curvature, topography humidity index, sediment transport index and topography roughness index; each item in the environmental data is respectively subjected to averaging treatment and then spliced to form an environmental data feature vector;
Normalizing the time sequence feature vector and the environment data feature vector, fusing the time sequence feature vector and the environment data feature vector into a foundation data feature vector, inputting the foundation data feature vector into a pre-trained artificial convolutional neural network model, and outputting a probability vector of landslide hazard degree;
The pre-trained artificial convolutional neural network model comprises an input layer, an output layer, 2 hidden layers and 3 convolutional layers, relu is adopted as an activation function, and cross entropy is adopted as a loss function.
After the mesoscale landslide risk area is positioned, corresponding foundation monitoring equipment is started to acquire foundation monitoring data. The inventor finds that the imaging data monitoring of the landslide area can reflect the landslide hazard degree of the area to a certain extent, but the landslide is only a phenomenon characteristic, and the landslide factors of a deeper level are related to the environmental factors, so that the environmental factors are also considered when the artificial intelligent convolutional neural network model is trained. Because of the numerous environmental factors, the application selects partial environmental factor variables to extract the environmental feature vectors.
The time sequence feature vector and the environment feature vector are expressed in the form of feature value sequences, and in order to unify dimensions, the time sequence feature vector and the environment feature vector are normalized and mapped into a [0,1] range, and then are spliced to form a final ground data feature vector which is input into a pre-trained artificial convolutional neural network model, and the output of the model is a probability value vector of the landslide hazard degree of a target area. When the model is trained, the training data are marked as three categories of high, medium and low landslide hazard degrees.
It should be appreciated by those skilled in the art that embodiments of the invention may be provided as a method, system, computer device, or computer-readable storage medium. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
The foregoing description of the preferred embodiment of the invention is not intended to limit the invention in any way, but rather to cover all modifications, equivalents, improvements and alternatives falling within the spirit and principles of the invention.
Claims (10)
1. The sky-land landslide monitoring method based on the multi-source data is characterized by comprising the following steps of:
Acquiring space-based monitoring data, space-based monitoring data and foundation monitoring data; the space-based monitoring data are GNSS monitoring data, the space-based monitoring data are unmanned aerial vehicle aerial image data, and the ground-based monitoring data are ground real-time video data and environment monitoring data;
Analyzing and processing the space-based monitoring data, positioning a large-scale landslide risk area, and starting day-based data acquisition equipment of the corresponding area; analyzing and processing the space-based monitoring data to extract three-dimensional point cloud data and model three-dimensional data of the slope surface, positioning a mesoscale landslide risk area, and starting foundation data acquisition equipment of the corresponding area; analyzing and processing the foundation monitoring data to realize real-time video data feature extraction and environment data feature extraction, and judging landslide hazard by adopting a pre-trained artificial convolutional neural network model;
And setting an early warning rule and an early warning response according to the monitoring result and the landslide hazard degree, and realizing visual display of the monitoring data.
2. The method for air-space-ground landslide monitoring based on multi-source data of claim 1, wherein analyzing the air-space-based monitoring data further comprises:
Setting n observation time points, m monitoring points and 1 datum point, counting the earth height difference of each monitoring point and the datum point on each observation time point, taking the m earth height differences measured for the first time as initial values, subtracting the initial values of the corresponding monitoring points from the earth height differences on n-1 observation time points to obtain settlement data, and transmitting the settlement data into a visual monitoring unit to realize visual monitoring; and analyzing the settlement amount data, and determining an area with the settlement amount larger than a preset threshold value as a large-scale landslide risk area.
3. The method for air-space landslide monitoring based on multi-source data of claim 2 wherein analyzing the space-based monitored data further comprises:
In the large-scale landslide risk area, starting unmanned aerial vehicle aerial photographing equipment to collect multi-angle unmanned aerial vehicle image data; extracting GPS coordinates and camera gestures in the multi-angle unmanned aerial vehicle image data, generating sparse point cloud data of a target area, and obtaining dense point cloud data through feature point matching, reconstruction, spatial point diffusion and filtering; and carrying out three-dimensional modeling and space-time evolution analysis on the ground surface of the large-scale landslide risk area according to the dense point cloud data, and determining the area with the ground surface displacement deformation exceeding a preset threshold value within a preset time range as a mesoscale landslide risk area.
4. A method of air-to-ground landslide monitoring based on multi-source data of claim 3 wherein analyzing the ground-based monitoring data further comprises:
in the mesoscale landslide risk area, starting a foundation video monitoring device to acquire real-time video data, and transmitting the real-time video data to a visual monitoring unit for realizing manual monitoring; extracting a key frame image set in the real-time video data, carrying out target identification and segmentation on the key frame image set, carrying out gridding treatment on a landslide area in the key frame image set after the target segmentation, and extracting coordinates and characteristic points of each grid point; integrating grid point coordinates and feature points in a landslide region in the key frame image set after the target segmentation to form a time sequence feature vector;
Collecting environmental data of a preset time range in the mesoscale landslide risk area, wherein the environmental data comprise rainfall, soil humidity, gradient, slope direction, plane curvature, section curvature, topography humidity index, sediment transport index and topography roughness index; each item in the environmental data is respectively subjected to averaging treatment and then spliced to form an environmental data feature vector;
Normalizing the time sequence feature vector and the environment feature vector, fusing the time sequence feature vector and the environment feature vector into foundation data features, inputting the foundation data features into a pre-trained artificial convolutional neural network model, and outputting a probability vector of landslide hazard degree;
the pre-trained artificial convolutional neural network model comprises an input layer, an output layer, 2 hidden layers and 3 convolutional layers, relu is adopted as an activation function, and cross entropy is adopted as a loss function period.
5. The method for monitoring the sky-land landslide based on multi-source data according to claim 4, wherein a mapping relation between a monitoring result and a landslide hazard level is established, wherein the hazard level is divided into three steps of high, medium and low, and the monitoring result comprises a sky-base monitoring result, a sky-base monitoring result and a foundation monitoring result; setting parameters to be monitored, graphically drawing discrete monitoring data to reflect time change characteristics, and triggering an alarm to remind workers of paying attention when certain monitoring data is higher than a preset value.
6. The air-sky landslide monitoring system based on the multi-source data is characterized by comprising the following functional modules:
Data acquisition layer: the method comprises the steps of acquiring space-based monitoring data, space-based monitoring data and foundation monitoring data; the space-based monitoring data are GNSS monitoring data, the space-based monitoring data are unmanned aerial vehicle aerial image data, and the ground-based monitoring data are ground real-time video data and environment monitoring data;
Data processing layer: the system comprises an empty base data processing unit, a data acquisition unit and a data processing unit, wherein the empty base data processing unit is used for analyzing and processing the empty base monitoring data, positioning a large-scale landslide risk area and starting a day base data acquisition device of the corresponding area; the day-based data processing unit is used for analyzing and processing the day-based monitoring data, extracting three-dimensional point cloud data and modeling three-dimensional data of the slope surface, positioning a mesoscale landslide risk area and starting foundation data acquisition equipment of the corresponding area; the foundation data processing unit is used for analyzing and processing the foundation monitoring data, realizing real-time video data feature extraction and environment data feature extraction, and realizing landslide hazard judgment by adopting a pre-trained artificial convolutional neural network model;
application service layer: the system comprises an early warning unit, a monitoring unit and a control unit, wherein the early warning unit is used for setting early warning rules and early warning responses according to monitoring results and landslide hazard degrees; and the visual monitoring unit is used for realizing visual display of the monitored data.
7. An aerospace ground landslide monitoring system of claim 6 and wherein the aerospace data processing unit further comprises: setting n observation time points, m monitoring points and 1 datum point, counting the earth height difference of each monitoring point and the datum point on each observation time point, taking the m earth height differences measured for the first time as initial values, subtracting the initial values of the corresponding monitoring points from the earth height differences on n-1 observation time points to obtain settlement data, and transmitting the settlement data into a visual monitoring unit to realize visual monitoring; and analyzing the settlement amount data, and determining an area with the settlement amount larger than a preset threshold value as a large-scale landslide risk area.
8. An aerospace landslide monitoring system of claim 7 wherein the aerospace data processing unit further comprises: in the large-scale landslide risk area, starting unmanned aerial vehicle aerial photographing equipment to collect multi-angle unmanned aerial vehicle image data; extracting GPS coordinates and camera gestures in the multi-angle unmanned aerial vehicle image data, generating sparse point cloud data of a target area, and obtaining dense point cloud data through feature point matching, reconstruction, spatial point diffusion and filtering; and carrying out three-dimensional modeling and space-time evolution analysis on the ground surface of the large-scale landslide risk area according to the dense point cloud data, and determining the area with the ground surface displacement deformation exceeding a preset threshold value within a preset time range as a mesoscale landslide risk area.
9. An aerospace landslide monitoring system based on multi-source data of claim 8 wherein the ground-based data processing unit further comprises:
in the mesoscale landslide risk area, starting a foundation video monitoring device to acquire real-time video data, and transmitting the real-time video data to a visual monitoring unit for realizing manual monitoring; extracting a key frame image set in the real-time video data, carrying out target identification and segmentation on the key frame image set, carrying out gridding treatment on a landslide area in the key frame image set after the target segmentation, and extracting coordinates and characteristic points of each grid point; integrating grid point coordinates and feature points in a landslide region in the key frame image set after the target segmentation to form a time sequence feature vector;
Collecting environmental data of a preset time range in the mesoscale landslide risk area, wherein the environmental data comprise rainfall, soil humidity, gradient, slope direction, plane curvature, section curvature, topography humidity index, sediment transport index and topography roughness index; each item in the environmental data is respectively subjected to averaging treatment and then spliced to form an environmental data feature vector;
Normalizing the time sequence feature vector and the environment feature vector, fusing the time sequence feature vector and the environment feature vector into foundation data features, inputting the foundation data features into a pre-trained artificial convolutional neural network model, and outputting a probability vector of landslide hazard degree;
The pre-trained artificial convolutional neural network model comprises an input layer, an output layer, 2 hidden layers and 3 convolutional layers, relu is adopted as an activation function, and cross entropy is adopted as a loss function.
10. The air-ground landslide monitoring system of claim 9 and wherein the pre-warning unit further comprises:
Establishing a mapping relation between a monitoring result and landslide hazard degree, wherein the hazard degree is divided into three steps of high, medium and low, and the monitoring result comprises an empty base, a space base and a foundation monitoring result;
The visual monitoring unit further comprises:
Setting parameters to be monitored, graphically drawing discrete monitoring data to reflect time change characteristics, and triggering an alarm to remind workers of paying attention when certain monitoring data is higher than a preset value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410591712.9A CN118171918B (en) | 2024-05-14 | Air-space-ground landslide monitoring method and system based on multi-source data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410591712.9A CN118171918B (en) | 2024-05-14 | Air-space-ground landslide monitoring method and system based on multi-source data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118171918A true CN118171918A (en) | 2024-06-11 |
| CN118171918B CN118171918B (en) | 2024-07-30 |
Family
ID=
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105184863A (en) * | 2015-07-23 | 2015-12-23 | 同济大学 | Unmanned aerial vehicle aerial photography sequence image-based slope three-dimension reconstruction method |
| CN111207728A (en) * | 2020-03-03 | 2020-05-29 | 王刚强 | Unmanned aerial vehicle measures equipment of side slope and foundation ditch excavation volume |
| KR102305468B1 (en) * | 2021-05-12 | 2021-09-30 | 씨티씨 주식회사 | Landslide distributed detection system based on deep learning |
| CN113887515A (en) * | 2021-10-28 | 2022-01-04 | 中国自然资源航空物探遥感中心 | Remote sensing landslide identification method and system based on convolutional neural network |
| CN116504032A (en) * | 2023-06-28 | 2023-07-28 | 湖南科技大学 | Landslide hazard monitoring and early warning method and system based on live-action three-dimension |
| CN116580532A (en) * | 2023-04-20 | 2023-08-11 | 西藏大学 | Geological disaster early warning system based on radar remote sensing technology |
| CN117058841A (en) * | 2023-06-20 | 2023-11-14 | 国能北电胜利能源有限公司 | Trinity slope monitoring, grading and early warning system for strip mine stope |
| CN117249862A (en) * | 2023-10-13 | 2023-12-19 | 中国地质大学（北京） | Method for monitoring space-ground integrated ecological environment of grassland strip mine |
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105184863A (en) * | 2015-07-23 | 2015-12-23 | 同济大学 | Unmanned aerial vehicle aerial photography sequence image-based slope three-dimension reconstruction method |
| CN111207728A (en) * | 2020-03-03 | 2020-05-29 | 王刚强 | Unmanned aerial vehicle measures equipment of side slope and foundation ditch excavation volume |
| KR102305468B1 (en) * | 2021-05-12 | 2021-09-30 | 씨티씨 주식회사 | Landslide distributed detection system based on deep learning |
| CN113887515A (en) * | 2021-10-28 | 2022-01-04 | 中国自然资源航空物探遥感中心 | Remote sensing landslide identification method and system based on convolutional neural network |
| CN116580532A (en) * | 2023-04-20 | 2023-08-11 | 西藏大学 | Geological disaster early warning system based on radar remote sensing technology |
| CN117058841A (en) * | 2023-06-20 | 2023-11-14 | 国能北电胜利能源有限公司 | Trinity slope monitoring, grading and early warning system for strip mine stope |
| CN116504032A (en) * | 2023-06-28 | 2023-07-28 | 湖南科技大学 | Landslide hazard monitoring and early warning method and system based on live-action three-dimension |
| CN117249862A (en) * | 2023-10-13 | 2023-12-19 | 中国地质大学（北京） | Method for monitoring space-ground integrated ecological environment of grassland strip mine |
Non-Patent Citations (3)
| Title |
|---|
| GUOSEN XU: "feature-based constraint deep CNN method for mapping rainfull-induced landslides in remote regions with mountainous terrain：an application to brazil", 《IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING》, vol. 15, 23 March 2022 (2022-03-23), pages 2644 - 2659 * |
| 朱祺夫: "基于TLS点云数据与卷积神经网络的滑坡变形分析与预测研究", 《中国优秀硕士学位论文全文数据库 基础科学辑》, no. 1, 15 January 2019 (2019-01-15), pages 008 - 175 * |
| 苏伟: "《软土地区高速铁路建造技术》", 30 June 2021, 中国铁道出版社, pages: 284 - 287 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112598881B (en) | Geological disaster monitoring method and device and computer equipment | |
| CN113611082B (en) | Unmanned aerial vehicle railway slope monitoring and early warning system and method | |
| CN111339826B (en) | Landslide unmanned aerial vehicle linear sensor network frame detecting system | |
| CN107479065B (en) | Forest gap three-dimensional structure measuring method based on laser radar | |
| Mason et al. | Data utilization in flood inundation modelling | |
| CN113869629A (en) | Laser point cloud-based power transmission line safety risk analysis, judgment and evaluation method | |
| CN113723403A (en) | Landslide monitoring method and device, electronic equipment and storage medium | |
| Gonçalves et al. | Monitoring Local Shoreline Changes by Integrating UASs, Airborne LiDAR, Historical Images and Orthophotos. | |
| CN115762067A (en) | Mountain landslide monitoring system based on fusion of laser point cloud and video data | |
| CN115267786A (en) | Resunet-GAN global radar inversion method and device fusing satellite observation brightness and elevation | |
| Karantanellis et al. | 3D hazard analysis and object-based characterization of landslide motion mechanism using UAV imagery | |
| CN114048944A (en) | Estimation method for people to be evacuated and houses to be damaged under rainstorm induced geological disaster | |
| CN116778329A (en) | Urban road underground shallow disease detection method, device, equipment and medium | |
| CN117076928A (en) | Bridge health state monitoring method, device and system and electronic equipment | |
| CN117152371B (en) | Three-dimensional topographic mapping method and system | |
| CN111881566B (en) | Landslide displacement detection method and device based on live-action simulation | |
| CN118171918B (en) | Air-space-ground landslide monitoring method and system based on multi-source data | |
| Taylor et al. | Evaluation of low-cost Raspberry Pi sensors for structure-from-motion reconstructions of glacier calving fronts | |
| CN107860375A (en) | A kind of landslide disaster volume rapid extracting method based on three-dimensional laser scanning technique | |
| CN116913046A (en) | Rainfall landslide risk early warning and monitoring method | |
| CN117373209A (en) | Interface type soil landslide monitoring and early warning method and device based on concave-convex body theory | |
| CN118171918A (en) | Air-space-ground landslide monitoring method and system based on multi-source data | |
| Comert et al. | Object based building extraction and building period estimation from unmanned aerial vehicle data | |
| Peitzsch et al. | Detecting snow depth changes in avalanche path starting zones using uninhabited aerial systems and structure from motion photogrammetry | |
| Yang et al. | A framework of linear sensor networks with unmanned aerial vehicle for rainfall-induced landslides detection |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |