CN117474164A - Community earthquake disaster safety prediction method and terminal based on transfer learning - Google Patents
Community earthquake disaster safety prediction method and terminal based on transfer learning Download PDFInfo
- Publication number
- CN117474164A CN117474164A CN202311455130.XA CN202311455130A CN117474164A CN 117474164 A CN117474164 A CN 117474164A CN 202311455130 A CN202311455130 A CN 202311455130A CN 117474164 A CN117474164 A CN 117474164A
- Authority
- CN
- China
- Prior art keywords
- disaster
- remote sensing
- earthquake
- post
- sensing image
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 46
- 238000013526 transfer learning Methods 0.000 title claims abstract description 37
- 230000009467 reduction Effects 0.000 claims abstract description 41
- 238000011160 research Methods 0.000 claims abstract description 29
- 238000012545 processing Methods 0.000 claims abstract description 21
- 238000000611 regression analysis Methods 0.000 claims abstract description 16
- 238000003860 storage Methods 0.000 claims abstract description 13
- 238000013527 convolutional neural network Methods 0.000 claims description 20
- 238000013508 migration Methods 0.000 claims description 13
- 230000005012 migration Effects 0.000 claims description 13
- 238000002372 labelling Methods 0.000 claims description 7
- 238000012847 principal component analysis method Methods 0.000 claims description 4
- 238000007726 management method Methods 0.000 abstract description 9
- 230000002265 prevention Effects 0.000 abstract description 8
- 238000013468 resource allocation Methods 0.000 abstract description 6
- 238000004590 computer program Methods 0.000 description 6
- 230000008859 change Effects 0.000 description 4
- 238000011835 investigation Methods 0.000 description 4
- 238000010219 correlation analysis Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000000605 extraction Methods 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 238000013439 planning Methods 0.000 description 2
- 230000003449 preventive effect Effects 0.000 description 2
- 230000015556 catabolic process Effects 0.000 description 1
- 238000013145 classification model Methods 0.000 description 1
- 238000005520 cutting process Methods 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
- G06Q50/265—Personal security, identity or safety
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- Tourism & Hospitality (AREA)
- Multimedia (AREA)
- Artificial Intelligence (AREA)
- Medical Informatics (AREA)
- Development Economics (AREA)
- Software Systems (AREA)
- General Business, Economics & Management (AREA)
- Databases & Information Systems (AREA)
- Marketing (AREA)
- Computing Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Entrepreneurship & Innovation (AREA)
- Primary Health Care (AREA)
- Operations Research (AREA)
- Game Theory and Decision Science (AREA)
- Educational Administration (AREA)
- Computer Security & Cryptography (AREA)
- Quality & Reliability (AREA)
- Geophysics And Detection Of Objects (AREA)
- Alarm Systems (AREA)
Abstract
The invention discloses a community earthquake disaster safety prediction method, a terminal and a storage medium based on transfer learning, wherein the method comprises the following steps: acquiring a disaster front light index and a disaster rear light index of each space grid at night in a research area, and calculating a disaster front-disaster rear difference value of the light indexes of each space grid at night; acquiring remote sensing images of a research area, extracting characteristic elements of the remote sensing images in corresponding space grids according to pre-disaster-post-disaster differences of night light indexes of each space grid, and sorting according to the characteristic elements of the remote sensing images to obtain post-disaster statistical data; performing dimension reduction processing on the remote sensing image characteristic elements, and performing ridge regression analysis on the proportion of each remote sensing image characteristic element after dimension reduction and the post-disaster statistical data to obtain a seismic disaster safety prediction model; the invention provides a novel urban community space earthquake disaster safety prediction method based on remote sensing images and transfer learning, which can guide urban earthquake prevention and disaster reduction resource allocation and related space management and update work.
Description
Technical Field
The invention relates to the technical field of new generation information, in particular to a community earthquake disaster safety prediction method, a terminal and a storage medium based on migration learning.
Background
The earthquake disaster safety prediction is an important means for protecting the life safety of people and reducing the earthquake disaster loss, and the magnitude and potential influence of earthquake risks in specific areas can be known through the earthquake disaster safety prediction, so that decision basis is provided for related government departments and social organizations, and better support is provided for earthquake disaster prevention work.
However, the main basis of the conventional earthquake disaster safety prediction is expert experience and post-disaster statistical data. The former is relatively poor in objectivity and limited by cost, so that coverage is relatively weak, and the latter is difficult to guide space management before disaster, resource allocation and the like, so that preventive measures are relatively difficult to deploy in advance. Meanwhile, the application of the traditional remote sensing image technology in the earthquake disaster safety prediction is not mature enough, and the feature extraction of the existing remote sensing image often requires a large amount of manual labeling, so that a large amount of manual labeling samples are difficult to obtain, and the application of the remote sensing image in the earthquake disaster safety prediction is limited.
Accordingly, there is a need in the art for improvement.
Disclosure of Invention
The invention aims to solve the technical problems that a community earthquake disaster safety prediction method, a terminal and a storage medium based on transfer learning are provided for solving the technical problems that a large number of manual labeling samples are difficult to obtain in the traditional remote sensing image feature extraction.
The technical scheme adopted for solving the technical problems is as follows:
in a first aspect, the invention provides a community earthquake disaster safety prediction method based on transfer learning, which comprises the following steps:
acquiring a disaster front light index and a disaster rear light index of each space grid at night in a research area, and calculating a disaster front-disaster rear difference value of the light indexes of each space grid at night;
acquiring remote sensing images of a research area, extracting characteristic elements of the remote sensing images in corresponding space grids according to pre-disaster-post-disaster differences of night light indexes of each space grid, and sorting according to the characteristic elements of the remote sensing images to obtain post-disaster statistical data;
and performing dimension reduction processing on the remote sensing image characteristic elements, and performing ridge regression analysis on the proportion of each remote sensing image characteristic element after dimension reduction and the post-disaster statistical data to obtain a seismic disaster safety prediction model.
In one implementation manner, the acquiring the pre-disaster light index and the post-disaster light index of each space grid at night in the research area, and calculating the pre-disaster-post difference value of the light indexes of each space grid at night comprises:
acquiring global night light data of the previous year of the earthquake and the year of the earthquake in the research area;
determining a clipping region according to administrative division data of the research region, clipping the acquired night light data according to the preset space grid size, and generating night light data of the previous year and the earthquake year of each space grid earthquake;
and calculating the difference value of night light data of the previous year of earthquake and the year of earthquake of each space grid to obtain the disaster pre-disaster post difference value of night light indexes of each space grid.
In one implementation manner, after calculating the difference value of the night light data of the previous year of the earthquake and the year of the earthquake to obtain the disaster pre-disaster post difference value of the night light index of each space grid, the method further includes:
and dividing the degree of influence of the disasters of each space grid into pre-disaster-post-disaster differences of 2-10, 11-30 and more than 30 according to the pre-disaster-post-disaster differences of the night light indexes of each space grid.
In one implementation manner, the extracting the feature elements of the remote sensing image in the corresponding space grid according to the pre-disaster-post-disaster difference value of the night light index of each space grid includes:
identifying space grids with the difference exceeding 2 according to the pre-disaster-post-disaster difference value of the night light indexes of each space grid, and extracting remote sensing images in the corresponding space grids;
classifying and labeling the corresponding categories of the space grid remote sensing images according to the degree of influence of disasters on each space grid;
and identifying the remote sensing image characteristic elements corresponding to each space grid remote sensing image category by adopting a convolutional neural network.
In one implementation manner, the identifying, by using a convolutional neural network, the remote sensing image feature elements corresponding to each spatial grid remote sensing image category includes:
the proportion of each type of characteristic elements in the remote sensing image is identified by adopting a convolutional neural network, and the disaster influence degree of each space grid is predicted according to the proportion of each type of characteristic elements;
and extracting the characteristic elements of the remote sensing image according to the class of the degree of the influence of the disaster on the space grid.
In one implementation manner, the obtaining post-disaster statistical data according to the remote sensing image feature element arrangement includes:
and extracting the post-disaster statistical data according to the characteristic elements of the remote sensing images, and sorting according to administrative division to obtain the post-disaster statistical data of each area.
In one implementation manner, the performing dimension reduction processing on the remote sensing image feature elements, and performing ridge regression analysis on the proportion of each remote sensing image feature element after dimension reduction and the post-disaster statistical data to obtain a seismic disaster safety prediction model, where the ridge regression analysis includes:
performing dimension reduction processing on the remote sensing image characteristic elements to obtain dimension reduced remote sensing image characteristic elements;
and calculating the proportion of the feature elements of the remote sensing image after the dimension reduction, and carrying out ridge regression analysis on the feature elements and the post-disaster statistical data to obtain the earthquake disaster safety prediction model.
In one implementation manner, the performing the dimension reduction processing on the feature elements of the remote sensing image includes:
and performing dimension reduction processing on the remote sensing image characteristic elements by adopting a principal component analysis method, and calculating the average value of the dimension reduced remote sensing image characteristic elements.
In a second aspect, the present invention also provides a terminal, including: the system comprises a processor and a memory, wherein the memory stores a community earthquake disaster safety prediction program based on transfer learning, and the community earthquake disaster safety prediction program based on transfer learning is used for realizing the operation of the community earthquake disaster safety prediction method based on transfer learning according to the first aspect when being executed by the processor.
In a third aspect, the present invention also provides a computer-readable storage medium storing a community seismic disaster safety prediction program based on transfer learning, which when executed by a processor is configured to implement the operation of the community seismic disaster safety prediction method based on transfer learning according to the first aspect.
The technical scheme adopted by the invention has the following effects:
according to the invention, the disaster front light index and the disaster rear light index of each space grid at night in the research area are obtained, so that the disaster front-disaster rear difference value of the light indexes of each space grid at night can be calculated, and the area with higher urban earthquake safety risk can be known; acquiring remote sensing images of the research area, extracting and obtaining characteristic elements of the remote sensing images in the corresponding space grids according to the pre-disaster-post-disaster differences of the night light indexes of the space grids, and sorting and obtaining post-disaster statistical data according to the characteristic elements of the remote sensing images; finally, performing dimension reduction processing on the remote sensing image characteristic elements, and performing ridge regression analysis on the proportion of each remote sensing image characteristic element after dimension reduction and the post-disaster statistical data to obtain a seismic disaster safety prediction model; the invention provides a novel urban community space earthquake disaster safety prediction method based on remote sensing images and transfer learning, which effectively solves the problem of insufficient urban community earthquake disaster safety risk prediction means, and can also effectively guide the planning of related running plans of urban disaster prevention, disaster reduction and disaster relief and the proposal of community space management and update.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to the structures shown in these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a community seismic disaster safety prediction method based on transfer learning in one implementation of the invention.
Fig. 2 is a functional schematic of a terminal in one implementation of the invention.
The achievement of the objects, functional features and advantages of the present invention will be further described with reference to the accompanying drawings, in conjunction with the embodiments.
Detailed Description
The invention provides a community earthquake disaster safety prediction method and a terminal based on transfer learning. In order to make the objects, technical solutions and advantages of the present invention more clear and clear, the present invention will be further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
Exemplary method
The main basis of the conventional earthquake disaster safety prediction is expert experience and post-disaster statistical data. The former is relatively poor in objectivity and limited by cost, so that coverage is relatively weak, and the latter is difficult to guide space management before disaster, resource allocation and the like, so that preventive measures are relatively difficult to deploy in advance. Meanwhile, the application of the traditional remote sensing image technology in the earthquake disaster safety prediction is not mature enough, and the feature extraction of the existing remote sensing image often requires a large amount of manual labeling, so that a large amount of manual labeling samples are difficult to obtain, and the application of the remote sensing image in the earthquake disaster safety prediction is limited.
Aiming at the technical problems, the embodiment of the invention provides a community earthquake disaster safety prediction method based on transfer learning, which aims to provide a novel urban community space earthquake disaster safety prediction method based on remote sensing images and transfer learning, and can effectively guide urban earthquake prevention and disaster reduction resource allocation and related space management and updating work.
As shown in fig. 1, the embodiment of the invention provides a community earthquake disaster safety prediction method based on migration learning, which comprises the following steps:
step S100, acquiring a disaster front light index and a disaster rear light index of each space grid at night in the research area, and calculating a disaster front-disaster rear difference value of each space grid at night light index.
In this embodiment, the community earthquake disaster safety prediction method based on transfer learning is applied to a terminal, where the terminal includes but is not limited to: a computer, a mobile terminal and other devices; the terminal is provided with a training migration platform based on a community earthquake disaster safety prediction model of migration learning.
In the embodiment, a method for utilizing a convolutional neural network and transfer learning is provided, and a pre-trained 8-layer convolutional neural network model provided in Imagenet is utilized to extract the feature elements of the remote sensing images of the urban space of each security risk level; extracting each element of the community, calculating the proportion of each element, and carrying out regression analysis on the proportion and post-disaster statistical data to obtain a prediction model capable of predicting the safety of the earthquake disaster in the community space; in the embodiment, the remote sensing image data and transfer learning method is provided, and the intelligent model capable of predicting the safety of the community earthquake disasters is trained by extracting the community space elements with high safety risk and performing correlation analysis with post-disaster statistical data, so that the problem of insufficient prediction means of the safety risks of the community earthquake disasters in the city is effectively solved, and the establishment of related line plans of disaster prevention, disaster reduction and disaster relief in the city and the proposal of community space management and update can be effectively guided.
In this embodiment, urban space grid data is first produced, night light indexes before and after the disaster of each space grid are collected, and then a difference value between the night light indexes before and after the disaster is calculated so as to identify urban spaces with higher safety risks.
Specifically, in one implementation of the present embodiment, step S100 includes the steps of:
step S101, global night light data of the year before and the year after the earthquake in the research area are acquired.
In this embodiment, global night light data for the year immediately before and the year of the earthquake of the study area may be downloaded from a night light data official website. After night lamplight information of a study area before and after earthquake disaster is collected, the night lamplight data needs to be cut.
Step S102, determining a clipping area according to administrative division data of the research area, clipping the acquired night light data according to the preset space grid size, and generating night light data of the previous earthquake year and the previous earthquake year of each space grid.
In this embodiment, vector data of the investigation region can be usedAnd determining the cut area or determining the cut area as the research area according to the administrative boundary data of the research area. Wherein the preset space grid size can be set to be 1km 2 . Specifically, the ArcGIS software can be utilized to cut the downloaded data, and night light data of the previous year of the earthquake of the research area and the year of the earthquake of the research area after cutting is generated.
Step S103, calculating the difference value of night light data of the previous year of the earthquake and the year of the earthquake of each space grid, and obtaining the disaster pre-disaster post difference value of the night light index of each space grid so as to identify the activity change degree of each area before and after the earthquake of the research area.
In this embodiment, the difference between the night light data of the previous year of the earthquake of each space grid in the research area and the year of the earthquake of the research area is calculated, and the pre-disaster-post-disaster difference of the night light indexes of each space grid is obtained. Specifically, the difference value calculation can be performed on the night light data of the year before the earthquake of the investigation region and the year after the earthquake of the investigation region obtained in the step S102 through a map algebraic tool of an ArcToolbox in ArcGIS software, so as to generate light brightness change value data of each space grid of the investigation region, namely, the pre-disaster-post-disaster difference value of night light indexes of each space grid.
Specifically, in one implementation of the present embodiment, step S103 further includes the following steps:
step S104, dividing the degree of each space grid affected by the disaster into pre-disaster-post-disaster differences of 2-10, 11-30 and more than 30 according to the pre-disaster-post-disaster differences of the night light indexes of each space grid.
In this embodiment, the light degradation degree of each space grid can be classified into three types: descending by 2-10, 11-30 and more than 30, thereby classifying the disaster degree. According to the calculation result in step S103, the extent to which each area is affected by the disaster is classified into three types, i.e., pre-disaster-post-disaster differences 2-10, 11-30, and 30 or more.
In the embodiment, the area with higher urban earthquake safety risk can be better identified according to the change condition of the night light indexes before and after the disaster by manufacturing urban space grid data and collecting the night light indexes before and after the disaster of each grid.
As shown in fig. 1, in an implementation manner of the embodiment of the present invention, the community earthquake disaster safety prediction method based on migration learning further includes the following steps:
step S200, acquiring remote sensing images of a research area, extracting characteristic elements of the remote sensing images in the corresponding space grids according to pre-disaster-post-disaster differences of night light indexes of the space grids, and sorting according to the characteristic elements of the remote sensing images to obtain post-disaster statistical data.
In this embodiment, a remote sensing image of the area affected by the earthquake is obtained, and the remote sensing image may be obtained from a Google Static Maps API remote sensing image official website. And (3) extracting remote sensing images in the corresponding space grids as remote sensing image data related to disaster damage for the space grids with the difference value exceeding 2 between the front and the back of the disaster according to the difference value between the front and the back of the disaster of each space grid at night light index calculated in the step (S100). And identifying remote sensing image characteristic elements related to disaster damage by using a Convolutional Neural Network (CNN), and sorting related post-disaster statistical data.
Specifically, in one implementation of the present embodiment, step S200 includes the steps of:
step S201, identifying the space grids with the difference exceeding 2 according to the pre-disaster-post-disaster difference values of the night light indexes of the space grids, and extracting the remote sensing images in the corresponding space grids.
In this embodiment, the remote sensing image in the spatial grid with the pre-disaster-post-disaster difference exceeding 2 is extracted as the remote sensing image data related to disaster damage.
Step S202, the corresponding categories of the space grid remote sensing images are marked according to the degree of influence of disasters on the space grids.
In this embodiment, the degrees of disaster influence are divided into pre-disaster-post-disaster differences of 2-10, 11-30 and more than 30, and the categories of the remote sensing images corresponding to the spatial grids are marked according to the degrees of disaster influence of the three categories, so that the remote sensing image feature elements corresponding to the different categories are identified.
And step S203, identifying the remote sensing image characteristic elements corresponding to each space grid remote sensing image category by adopting a convolutional neural network.
In this embodiment, the feature elements of the remote sensing image related to the damage degree of the disaster are extracted by using the convolutional neural network model provided in Imagenet after the pre-training, the damage degree of the disaster is classified into three types of differences 2-10, 11-30 and more than 30 before the disaster and after the disaster by identifying the feature element information in the remote sensing image extracted in the step S200, and the migration learning task is regarded as a classification problem, and the feature elements of the remote sensing image related to the damage degree of the disaster are extracted by operating the classification model.
Specifically, in one implementation of the present embodiment, step S203 includes the steps of:
and step 203a, identifying the proportion of each type of characteristic elements in the remote sensing image by adopting a convolutional neural network, and predicting the disaster influence degree of each space grid according to the proportion of each type of characteristic elements.
In this embodiment, a Convolutional Neural Network (CNN) trained on Imagenet is adopted, and the task is regarded as a classification task for predicting the disaster influence degree of the region according to the proportion of each feature element in the remote sensing image by identifying the feature element corresponding to the remote sensing image category marked in step S202. 4000 remainder feature elements are empirically identifiable.
And step 203b, extracting the characteristic elements of the remote sensing image according to the degree of the disaster influence of the space grid.
In this embodiment, the feature elements of the remote sensing image in the urban space with security risks at each level are extracted by using the pre-trained 8-layer convolutional neural network model provided in Imagenet. And deleting the last layer (disaster damage degree classification layer) of the trained classification prediction neural network model to serve as a model for extracting the characteristic elements of each community.
And step S204, extracting the post-disaster statistical data according to the characteristic elements of the remote sensing images, and sorting according to administrative regions to obtain the post-disaster statistical data of each region.
In this embodiment, post-disaster statistics are sorted according to community administrative division. The post-disaster statistics include building damage rate, resident mortality, and the like.
In this embodiment, according to the pre-disaster-post-disaster difference value of the night light indexes of each space grid, the characteristic elements of the remote sensing images in the corresponding space grids are extracted by using the convolutional neural network, and the post-disaster statistical data are obtained by sorting the characteristic elements of the remote sensing images, so that the training of the prediction model can be realized by using the convolutional neural network to extract the space elements related to the earthquake safety risk in the remote sensing images and correlating with the post-disaster statistical data of the community. In this embodiment, the feature elements of the remote sensing image in the urban space with security risks at each level are extracted by using the pre-trained 8-layer convolutional neural network model provided in Imagenet. And deleting the last layer (disaster damage degree classification layer) of the trained classification prediction neural network model to serve as a model for extracting the characteristic elements of each community.
As shown in fig. 1, in an implementation manner of the embodiment of the present invention, the community earthquake disaster safety prediction method based on migration learning further includes the following steps:
and step S300, performing dimension reduction processing on the remote sensing image feature elements, and performing ridge regression analysis on the proportion of each dimension reduced remote sensing image feature element and the post-disaster statistical data to obtain a seismic disaster safety prediction model.
In this embodiment, after the required feature elements of the remote sensing image are extracted according to the trained model, a large number of feature elements of the remote sensing image are subjected to dimension reduction processing. And performing ridge regression analysis on the feature elements of the remote sensing image subjected to the dimension reduction treatment and post-disaster statistical data to obtain a prediction model capable of predicting the damage risk of the urban earthquake disaster.
Specifically, in one implementation of the present embodiment, step S300 includes the steps of:
step S301, performing dimension reduction processing on the remote sensing image feature elements to obtain the dimension reduced remote sensing image feature elements.
In this embodiment, the dimension reduction processing is performed on the feature elements of the remote sensing image, so that the calculated amount can be reduced, and the risk of overfitting can be reduced.
Specifically, in one implementation of the present embodiment, step S301 includes the steps of:
step S301a, performing dimension reduction processing on the remote sensing image characteristic elements by adopting a principal component analysis method, and calculating an average value of the dimension reduced remote sensing image characteristic elements.
In this embodiment, the principal component analysis method may be used to perform dimension reduction processing on the feature elements of the remote sensing image in python, so as to reduce the feature elements of the remote sensing image to 10-15 principal components, thereby reducing the operand and reducing the risk of overfitting. And calculating the average value of the feature elements after dimension reduction.
And step S302, calculating the proportion of the feature elements of the remote sensing images after the dimension reduction, and carrying out ridge regression analysis on the feature elements and the post-disaster statistical data to obtain the earthquake disaster safety prediction model.
In this embodiment, the regression model may be constructed using the ratio of each feature element as an independent variable and the post-disaster statistical data as an independent variable in python. Because the feature quantity extracted by the embodiment is more, the ridge regression can avoid the overfitting of the model by adding the regular term on the basis of square error, and further a prediction model capable of predicting the damage risk of the urban earthquake disaster is obtained.
In the embodiment, the feature elements of the remote sensing image are extracted, the dimension reduction processing is carried out, the proportion of the feature elements is calculated, regression analysis is carried out on the feature elements and post-disaster statistical data, and therefore a prediction model capable of predicting the safety of the earthquake disasters in the community space is obtained. Based on remote sensing image data and a migration learning method, intelligent models capable of predicting the safety of the community earthquake disasters are trained by extracting community space elements with high safety risk and performing correlation analysis with post-disaster statistical data. The method effectively solves the problem of insufficient prediction means of the earthquake disaster safety risk of the urban community, and can effectively guide the planning of related line plans of urban disaster prevention, disaster reduction and disaster relief and the proposal of community space management and update.
The following technical effects are achieved through the technical scheme:
according to the embodiment, the disaster front light index and the disaster rear light index of each space grid at night in the research area are obtained, and the disaster front-disaster rear difference value of each space grid at night light index is calculated; acquiring remote sensing images of a research area, extracting characteristic elements of the remote sensing images in corresponding space grids according to pre-disaster-post-disaster differences of night light indexes of each space grid, and sorting according to the characteristic elements of the remote sensing images to obtain post-disaster statistical data; performing dimension reduction processing on the remote sensing image characteristic elements, and performing ridge regression analysis on the proportion of each remote sensing image characteristic element after dimension reduction and the post-disaster statistical data to obtain a seismic disaster safety prediction model; according to the embodiment, a novel urban community space earthquake disaster safety prediction method based on remote sensing images and transfer learning is provided, areas with higher urban earthquake safety risks are marked according to the change condition of light indexes at night before and after the disaster, space elements related to the earthquake safety risks in the remote sensing images are extracted by utilizing a convolutional neural network through the transfer learning method, community space elements with high safety risks are extracted, correlation analysis is carried out on the community space elements and post-disaster statistical data, and therefore an intelligent model capable of predicting the community earthquake disaster safety is trained; the method effectively solves the problem of insufficient prediction means of the urban community earthquake disaster safety risk, can more objectively and massively predict the urban community disaster safety, and can more continuously guide the urban earthquake prevention and disaster reduction resource allocation and the related space management and updating work.
Exemplary apparatus
Based on the above embodiment, the present invention further provides a terminal, including: the system comprises a processor, a memory, an interface, a display screen and a communication module which are connected through a system bus; wherein the processor is configured to provide computing and control capabilities; the memory includes a storage medium and an internal memory; the storage medium stores an operating system and a computer program; the internal memory provides an environment for the operation of the operating system and computer programs in the storage medium; the interface is used for connecting external equipment, such as mobile terminals, computers and other equipment; the display screen is used for displaying corresponding information; the communication module is used for communicating with a cloud server or a mobile terminal.
The computer program, when executed by the processor, is operative to implement a community seismic disaster safety prediction method based on transfer learning.
It will be appreciated by those skilled in the art that the functional block diagram shown in fig. 2 is merely a block diagram of some of the structures associated with the present inventive arrangements and is not limiting of the terminal to which the present inventive arrangements may be applied, and that a particular terminal may include more or less components than those shown, or may combine some of the components, or have a different arrangement of components.
In one embodiment, a terminal is provided, including: the community earthquake disaster safety prediction system comprises a processor and a memory, wherein the memory stores a community earthquake disaster safety prediction program based on transfer learning, and the community earthquake disaster safety prediction program based on transfer learning is used for realizing the operation of the community earthquake disaster safety prediction method based on transfer learning.
In one embodiment, a computer readable storage medium is provided, wherein the computer readable storage medium stores a migration learning-based community seismic disaster safety prediction program for implementing the operations of the migration learning-based community seismic disaster safety prediction method as described above when executed by the processor.
Those skilled in the art will appreciate that implementing all or part of the above-described methods may be accomplished by way of a computer program comprising instructions for the relevant hardware, the computer program being stored on a non-volatile storage medium, the computer program when executed comprising the steps of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in embodiments provided herein may include non-volatile and/or volatile memory.
In summary, the invention provides a community earthquake disaster safety prediction method, a terminal and a storage medium based on migration learning, wherein the method comprises the following steps: acquiring a disaster front light index and a disaster rear light index of each space grid at night in a research area, and calculating a disaster front-disaster rear difference value of the light indexes of each space grid at night; acquiring remote sensing images of a research area, extracting characteristic elements of the remote sensing images in corresponding space grids according to pre-disaster-post-disaster differences of night light indexes of each space grid, and sorting according to the characteristic elements of the remote sensing images to obtain post-disaster statistical data; performing dimension reduction processing on the remote sensing image characteristic elements, and performing ridge regression analysis on the proportion of each remote sensing image characteristic element after dimension reduction and the post-disaster statistical data to obtain a seismic disaster safety prediction model; the invention provides a novel urban community space earthquake disaster safety prediction method based on remote sensing images and transfer learning, which can guide urban earthquake prevention and disaster reduction resource allocation and related space management and update work.
It is to be understood that the invention is not limited in its application to the examples described above, but is capable of modification and variation in light of the above teachings by those skilled in the art, and that all such modifications and variations are intended to be included within the scope of the appended claims.
Claims (10)
1. The community earthquake disaster safety prediction method based on transfer learning is characterized by comprising the following steps of:
acquiring a disaster front light index and a disaster rear light index of each space grid at night in a research area, and calculating a disaster front-disaster rear difference value of the light indexes of each space grid at night;
acquiring remote sensing images of a research area, extracting characteristic elements of the remote sensing images in corresponding space grids according to pre-disaster-post-disaster differences of night light indexes of each space grid, and sorting according to the characteristic elements of the remote sensing images to obtain post-disaster statistical data;
and performing dimension reduction processing on the remote sensing image characteristic elements, and performing ridge regression analysis on the proportion of each remote sensing image characteristic element after dimension reduction and the post-disaster statistical data to obtain a seismic disaster safety prediction model.
2. The method for predicting the safety of earthquake disasters in communities based on transfer learning according to claim 1, wherein the steps of obtaining the pre-disaster light index and the post-disaster light index of each space grid at night in the research area, and calculating the pre-disaster-post-disaster difference value of the light index of each space grid at night comprise the following steps:
acquiring global night light data of the previous year of the earthquake and the year of the earthquake in the research area;
determining a clipping region according to administrative division data of the research region, clipping the acquired night light data according to the preset space grid size, and generating night light data of the previous year and the earthquake year of each space grid earthquake;
and calculating the difference value of night light data of the previous year of earthquake and the year of earthquake of each space grid to obtain the disaster pre-disaster post difference value of night light indexes of each space grid.
3. The community earthquake disaster safety prediction method based on transfer learning according to claim 2, wherein the calculating the difference between the night light data of the previous year and the seismic year of each space grid to obtain the pre-disaster-post-disaster difference of the night light index of each space grid further comprises:
and dividing the degree of influence of the disasters of each space grid into pre-disaster-post-disaster differences of 2-10, 11-30 and more than 30 according to the pre-disaster-post-disaster differences of the night light indexes of each space grid.
4. The community earthquake disaster safety prediction method based on transfer learning as set forth in claim 3, wherein the extracting the feature elements of the remote sensing image in the corresponding space grid according to the pre-disaster-post-disaster difference values of the night light indexes of the space grids comprises:
identifying space grids with the difference exceeding 2 according to the pre-disaster-post-disaster difference value of the night light indexes of each space grid, and extracting remote sensing images in the corresponding space grids;
classifying and labeling the corresponding categories of the space grid remote sensing images according to the degree of influence of disasters on each space grid;
and identifying the remote sensing image characteristic elements corresponding to each space grid remote sensing image category by adopting a convolutional neural network.
5. The community earthquake disaster safety prediction method based on transfer learning of claim 4, wherein the identifying the remote sensing image feature elements corresponding to each spatial grid remote sensing image category by using the convolutional neural network comprises:
the proportion of each type of characteristic elements in the remote sensing image is identified by adopting a convolutional neural network, and the disaster influence degree of each space grid is predicted according to the proportion of each type of characteristic elements;
and extracting the characteristic elements of the remote sensing image according to the class of the degree of the influence of the disaster on the space grid.
6. The community earthquake disaster safety prediction method based on transfer learning according to claim 1, wherein the obtaining post-disaster statistical data by sorting according to the remote sensing image feature elements comprises:
and extracting the post-disaster statistical data according to the characteristic elements of the remote sensing images, and sorting according to administrative division to obtain the post-disaster statistical data of each area.
7. The community earthquake disaster safety prediction method based on transfer learning of claim 1, wherein the performing dimension reduction processing on the remote sensing image feature elements, and performing ridge regression analysis on the proportion of each remote sensing image feature element after dimension reduction and the post-disaster statistical data to obtain an earthquake disaster safety prediction model comprises:
performing dimension reduction processing on the remote sensing image characteristic elements to obtain dimension reduced remote sensing image characteristic elements;
and calculating the proportion of the feature elements of the remote sensing image after the dimension reduction, and carrying out ridge regression analysis on the feature elements and the post-disaster statistical data to obtain the earthquake disaster safety prediction model.
8. The community earthquake disaster safety prediction method based on transfer learning of claim 7, wherein the performing dimension reduction processing on the remote sensing image feature elements comprises:
and performing dimension reduction processing on the remote sensing image characteristic elements by adopting a principal component analysis method, and calculating the average value of the dimension reduced remote sensing image characteristic elements.
9. A terminal, comprising: the system comprises a processor and a memory, wherein the memory stores a community earthquake disaster safety prediction program based on transfer learning, and the community earthquake disaster safety prediction program based on transfer learning is used for realizing the operation of the community earthquake disaster safety prediction method based on transfer learning according to any one of claims 1-8 when being executed by the processor.
10. A computer-readable storage medium storing a migration learning-based community seismic disaster safety prediction program that, when executed by a processor, is operable to implement the migration learning-based community seismic disaster safety prediction method of any one of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311455130.XA CN117474164A (en) | 2023-11-02 | 2023-11-02 | Community earthquake disaster safety prediction method and terminal based on transfer learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311455130.XA CN117474164A (en) | 2023-11-02 | 2023-11-02 | Community earthquake disaster safety prediction method and terminal based on transfer learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117474164A true CN117474164A (en) | 2024-01-30 |
Family
ID=89625206
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311455130.XA Pending CN117474164A (en) | 2023-11-02 | 2023-11-02 | Community earthquake disaster safety prediction method and terminal based on transfer learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117474164A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107085662A (en) * | 2017-05-12 | 2017-08-22 | 首都师范大学 | A kind of method that earthquake intensity is extracted based on nighttime light data |
| CN109241846A (en) * | 2018-08-06 | 2019-01-18 | 广州市城市规划勘测设计研究院 | Change in time and space estimating and measuring method, device and the storage medium of remote sensing image |
| CN109523125A (en) * | 2018-10-15 | 2019-03-26 | 广州地理研究所 | A kind of poor Measurement Method based on DMSP/OLS nighttime light data |
| CN113435268A (en) * | 2021-06-09 | 2021-09-24 | 武汉理工大学 | Earthquake disaster area remote sensing image interpretation method based on graph transformation knowledge embedding algorithm |
| CN114821149A (en) * | 2022-03-08 | 2022-07-29 | 天津科技大学 | Hyperspectral remote sensing image identification method based on deep forest transfer learning |
| CN115830456A (en) * | 2022-12-28 | 2023-03-21 | 云南省地质环境监测院(云南省环境地质研究院) | High-resolution multispectral earthquake secondary landslide disaster remote sensing monitoring method and system |
-
2023
- 2023-11-02 CN CN202311455130.XA patent/CN117474164A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107085662A (en) * | 2017-05-12 | 2017-08-22 | 首都师范大学 | A kind of method that earthquake intensity is extracted based on nighttime light data |
| CN109241846A (en) * | 2018-08-06 | 2019-01-18 | 广州市城市规划勘测设计研究院 | Change in time and space estimating and measuring method, device and the storage medium of remote sensing image |
| CN109523125A (en) * | 2018-10-15 | 2019-03-26 | 广州地理研究所 | A kind of poor Measurement Method based on DMSP/OLS nighttime light data |
| CN113435268A (en) * | 2021-06-09 | 2021-09-24 | 武汉理工大学 | Earthquake disaster area remote sensing image interpretation method based on graph transformation knowledge embedding algorithm |
| CN114821149A (en) * | 2022-03-08 | 2022-07-29 | 天津科技大学 | Hyperspectral remote sensing image identification method based on deep forest transfer learning |
| CN115830456A (en) * | 2022-12-28 | 2023-03-21 | 云南省地质环境监测院(云南省环境地质研究院) | High-resolution multispectral earthquake secondary landslide disaster remote sensing monitoring method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110992169B (en) | Risk assessment method, risk assessment device, server and storage medium | |
| CN110070391B (en) | Data processing method and device, computer readable medium and electronic equipment | |
| CN109345130B (en) | Method and device for commercial site selection, computer equipment and storage medium | |
| CN112328909B (en) | Information recommendation method and device, computer equipment and medium | |
| US10984518B2 (en) | Methods and systems for assessing the quality of geospatial data | |
| US20170301028A1 (en) | Processing system to generate attribute analysis scores for electronic records | |
| CN113570867B (en) | Urban traffic state prediction method, device, equipment and readable storage medium | |
| CN112184304A (en) | Method, system, server and storage medium for assisting decision | |
| CN114503139A (en) | Systems, methods, and platforms for performing multi-level catastrophic risk exposure analysis on a portfolio | |
| CN112613569A (en) | Image recognition method, and training method and device of image classification model | |
| US10489427B2 (en) | Document classification system, document classification method, and document classification program | |
| CN107798615A (en) | Declaration form renews charge difficulty Forecasting Methodology and device | |
| CN111815576B (en) | Method, device, equipment and storage medium for detecting corrosion condition of metal part | |
| CN115660143A (en) | Power failure duration prediction method, system, equipment and medium under typhoon disaster | |
| CN117037482A (en) | Real-time road traffic running state sensing method, system, equipment and storage medium | |
| JP2023105078A (en) | Information processing apparatus, information processing method, and information processing program | |
| CN117474164A (en) | Community earthquake disaster safety prediction method and terminal based on transfer learning | |
| CN114445711B (en) | Image detection method, image detection device, electronic equipment and storage medium | |
| CN113298438B (en) | Regional risk level assessment method and device, computer equipment and storage medium | |
| Karmshahi et al. | Application of an integrated CA-Markov model in simulating spatiotemporal changes in forest cover: a case study of Malekshahi county forests, Ilam province | |
| CN112199434B (en) | Data processing method, device, electronic equipment and storage medium | |
| CN116403403B (en) | Traffic early warning method, system, equipment and medium based on big data analysis | |
| CN110751567A (en) | Vehicle information processing method, device, computer equipment and storage medium | |
| CN117975107A (en) | Urban space evolution prediction method based on interest points and remote sensing images | |
| CN116777515A (en) | Space-time characteristic-based addressing method, device, equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |