CN108304809A - The damaged appraisal procedure of near real-time based on aerial images after shake - Google Patents
The damaged appraisal procedure of near real-time based on aerial images after shake Download PDFInfo
- Publication number
- CN108304809A CN108304809A CN201810119671.8A CN201810119671A CN108304809A CN 108304809 A CN108304809 A CN 108304809A CN 201810119671 A CN201810119671 A CN 201810119671A CN 108304809 A CN108304809 A CN 108304809A
- Authority
- CN
- China
- Prior art keywords
- building
- collapse
- distribution map
- simulant
- earthquake
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/13—Satellite images
Abstract
The invention discloses a kind of damaged appraisal procedures of the near real-time based on aerial images after shake, the practical building collapse distribution map that this method passes through the identification earthquake region from aerial images after shake, and a large amount of simulant building is obtained by earthquake disaster simulation and is collapsed distribution map, then it selects and collapses distribution map with practical building collapse distribution map optimal simulant building the most matched from a large amount of simulant building distribution map that collapses by similarity mode method, collapsed the damaged assessment of distribution map progress according to optimal simulant building.This method utilizes the important information that earthquake region scene aerial photograph provides, and improves the precision of damaged assessment, has high computational efficiency, can provide the assessment result of near real-time in 48 hours after shake.
Description
Technical field
The present invention relates to technical field of civil engineering more particularly to a kind of near real-time based on aerial images after shake is damaged comments
Estimate method.
Background technology
China is earthquake-prone countries, and between statistical data shows 1900-2016, earthquake is damaged to the economy that China is brought
Lose the ranking whole world second.Economic loss caused by quickly and accurately prediction earthquake building is destroyed, to formulate the rational disaster relief and
Reconstruction model has substantial worth.
Earthquake loss statistical method includes mainly field investigation or the loss of selective examination statistics, is damaged according to loss forecasting model prediction
Lose, lost according to remote sensing or data prediction of taking photo by plane etc..Field investigation is relatively the most accurate, but time-consuming longer, often needs several weeks
Even as long as some months, and it is easy to be influenced by human factor.Loss forecasting model or aerial photograph prediction loss are time-consuming short, have
It can adapt to the current demand that Fast Evaluation is damaged after shaking.But loss forecasting model needs rational input earthquake；And it takes photo by plane
Photo is difficult to building interior and destroys situation, causes to underestimate loss.Therefore, existing method is still not enough to take into account damaged assessment
The requirement of efficiency and precision.
Invention content
The purpose of the present invention is intended to solve above-mentioned one of technical problem at least to a certain extent.
For this purpose, first purpose of the present invention is the damaged assessment side of the near real-time based on aerial images after shake proposed
Method obtains largely by identifying the practical building collapse distribution map in earthquake region from aerial images after shake, and by earthquake disaster simulation
Simulant building collapse distribution map, then by similarity mode method from a large amount of simulant building collapse distribution map select with it is practical
Building collapse distribution map optimal simulant building the most matched collapses distribution map, is carried out according to the optimal simulant building distribution map that collapses
Damaged assessment.This method utilizes the important information that earthquake region scene aerial photograph provides, and improves the precision of damaged assessment, has pole
High computational efficiency can provide the assessment result of near real-time after shake in 48 hours.
To achieve the goals above, the near real-time based on aerial images after shake of first aspect present invention embodiment is damaged comments
Estimate method, including：
The practical building collapse state for identifying each building in aerial images after shaking obtains the practical building collapse point in earthquake region
Butut, wherein practical building collapse distribution map includes the practical building collapse state of each building in earthquake region；
Operating mode is analyzed according at least one Nonlinear Time course, earthquake region building earthquake disaster simulation is executed to buildings model, obtain
Each simulant building in earthquake region collapses distribution map, wherein a kind of Nonlinear Time course analysis operating mode corresponds to a simulant building
Collapse distribution map, simulant building collapse distribution map include each building in earthquake region the detailed collapse state of simulation；
Each simulant building distribution map that collapses is matched with practical building collapse distribution map, obtains each simulant building
Collapse the similarity score of distribution map；
Optimal simulant building is selected to collapse distribution map according to the collapse similarity score of distribution map of each simulant building；
Each building is carried out according to the optimal simulant building detailed collapse state of simulation of each building in distribution map that collapses
Damaged assessment.
Method as described above, it is described to be destroyed in detail according to the simulation of each building in distribution map of collapsing of optimal simulant building
State carries out the damaged assessment of each building, including：
To each building：
According to formula Lh=S × Dh× P calculates house failure loss Lh, wherein S is house architectural area, and P is house weight
Set unit price, DhFor the corresponding house failure loss ratio of the optimal detailed collapse state of simulation；
According to formula Ld=γ1×γ2×(ξ×S)×Dd× (η × P) calculates finishing failure loss Ld, wherein γ1To examine
Consider the correction factor of each regional economy situation difference, γ2To consider that the correction factor of different building occupancies, ξ are medium-to-high grade fill
Estate building coverage of repairing the house accounts for the ratio in total house, and η is the ratio of house decoration expense and main house body cost, DdFor optimal simulation
The corresponding finishing failure loss ratio of detailed collapse state；
According to house failure loss LhWith finishing failure loss LdCalculate the house Seismic Economic Losses each built.
Method as described above, it is described that each simulant building collapses the progress of distribution map and practical building collapse distribution map
Match, obtains each simulant building and collapse the similarity score of distribution map, including：
Collapse distribution map to i-th of simulant building：
According to formulaJ-th of corresponding i-th of simulant building of building is calculated to collapse the similar of distribution map
Spend score Saij；
According to formulaI-th of simulant building is calculated to collapse the similarity score S of distribution mapAi；
Wherein, yijCollapse the detailed collapse state of simulation of distribution map for j-th of corresponding i-th of simulant building of building, yjFor
The practical building collapse state of j-th of building, i are positive integer, and m is positive integer, and m is the building sum in earthquake region, and j is just whole
Number, j values in 1 to m.
Method as described above, it is described that each simulant building collapses the progress of distribution map and practical building collapse distribution map
Match, obtains each simulant building and collapse the similarity score of distribution map, including：
Collapse distribution map to i-th of simulant building：
According to formulaJ-th of corresponding i-th of simulant building of building is calculated to collapse the similar of distribution map
Spend score Sbij；
According to formulaI-th of simulant building is calculated to collapse the similarity score S of distribution mapBi；
Wherein, the weight of j-th of buildingpjThe collapse probability built for j-th；
Wherein, yijCollapse the detailed collapse state of simulation of distribution map for j-th of corresponding i-th of simulant building of building, yjFor
The practical building collapse state of j-th of building, i are positive integer, and m is positive integer, and m is the building sum in earthquake region, and j is just whole
Number, j values in 1 to m.
Method as described above further includes：
According to formulaCalculate the collapse probability of j-th of building
pj；
Wherein, x is collapse probability factor vector, and θ is the ginseng trained based on the logical division algorithm in machine learning
The codomain of number vector θ, logical function h (z) are (0,1).
Method as described above further includes：
Determine building collapse distributed model P (y | x；θ)=h (θTx)y[1-h(θTx)]1-y；
Determine that the maximum likelihood estimator model of parameter vector θ is according to building collapse distributed model
From sample data concentrate choose training set, using training set train maximum likelihood estimator model, obtain parameter to
Measure the maximum likelihood estimator of θ, wherein the collapse probability factor x of j-th of buildingjAnd practical building collapse state yjAs sample
J-th of sample data of data set；
The corresponding parameter vector θ of maximum likelihood estimator is determined as based on the logical division algorithm training in machine learning
Obtained parameter vector θ.
Method as described above, the optimal simulation of similarity score selection of the distribution map that collapsed according to each simulant building
Building collapse distribution map, including：
The collapse similarity score of distribution map of each simulant building is ranked up, is selected and is met the simulation of preset condition and build
The distribution map that collapses is built as optimal simulant building to collapse distribution map.
Method as described above, it is described that shake is executed to buildings model according at least one Nonlinear Time course analysis operating mode
Area builds earthquake disaster simulation, and each simulant building for obtaining earthquake region collapses distribution map, including：
Determine the buildings model of each building in earthquake region, wherein to the median low structure in earthquake region, buildings model is more
Particle shears series model, and to the skyscraper in earthquake region, buildings model is more particle parallel connection shear-bow Coordination Models；
Operating mode is analyzed to each Nonlinear Time course：Using the earthquake motion after amplitude modulation corresponding with each building to phase
Corresponding buildings model carries out earthquake disaster simulation, obtains the detailed collapse state of simulation of each building；
The corresponding mould of each Nonlinear Time course analysis operating mode is exported according to the detailed collapse state of the simulation of each building
Quasi- building collapse distribution map.
Method as described above further includes：
Determine at least one ground motion predictive equation；
At least one seismic motion record is obtained from earthquake motion database；
At least one acquired seismic motion record is carried out using identified at least one ground motion predictive equation
Amplitude modulation analyzes operating mode to build at least one Nonlinear Time course.
Method as described above, aerial images carry out image procossing after described pair of shake, obtain the practical building collapse in earthquake region
Distribution map, including：
The visually practical building collapse state of each building after interpretation method identification shake in aerial images, according to earthquake region
Building distribution map and the practical building collapse state of each building that recognizes, obtain the practical building collapse distribution in earthquake region
Figure.
The additional aspect of the present invention and advantage will be set forth in part in the description, and will partly become from the following description
Obviously, or practice through the invention is recognized.
Description of the drawings
Above-mentioned and/or additional aspect and advantage of the invention will become from the following description of the accompanying drawings of embodiments
Obviously and it is readily appreciated that, wherein
Fig. 1 is the flow signal of the damaged appraisal procedure of the near real-time based on aerial images after shake of one embodiment of the invention
Figure；
Fig. 2 is the schematic diagram of distribution of illustratively collapsing；
Fig. 3 is the practical building collapse distribution map of Ludian earthquake；
Fig. 4 is that the optimal simulant building of Ludian earthquake collapses distribution map；
Fig. 5 is the relational graph of similarity score and economic loss；
Fig. 6 is the block diagram of economic loss.
Specific implementation mode
The embodiment of the present invention is described below in detail, examples of the embodiments are shown in the accompanying drawings, wherein from beginning to end
Same or similar label indicates same or similar element or element with the same or similar functions.Below with reference to attached
The embodiment of figure description is exemplary, it is intended to for explaining the present invention, and is not considered as limiting the invention.
Below with reference to the accompanying drawings the damaged appraisal procedure of the near real-time based on aerial images after shake of the embodiment of the present invention is described.
Fig. 1 is the flow signal of the damaged appraisal procedure of the near real-time based on aerial images after shake of one embodiment of the invention
Figure.As shown in Figure 1, the damaged appraisal procedure of the near real-time provided in this embodiment based on aerial images after shake, includes the following steps：
S101, the practical building collapse state for identifying each building in aerial images after shaking, obtain the practical building in earthquake region
Collapse distribution map, wherein practical building collapse distribution map includes the practical building collapse state of each building in earthquake region.
Specifically, earthquake region can take photo by plane aerial images after acquisition shake using unmanned plane, then to shadow of taking photo by plane after shake
As the image preprocessings such as being enhanced, being restored, compressed, to obtain aerial images after high-visible shake.In the present embodiment, real
Border building collapse state includes two states, the first collapses for building, builds and do not collapse, based on what is recognized
The practical building collapse distribution map in the practical building collapse state output earthquake region of each building.
Specifically, visual interpretation method refers to that people use abundant specialty background knowledge, is observed by the naked eye, by synthesis
Analysis, reasoning from logic, verification detection with detecting target in image comprehensive information extraction and parse.The present embodiment passes through special
Industry personnel's is rigid in checking up, and the practical building collapse state of each building is more precisely identified from aerial images after shake, is connect
The building distribution map in conjunction with earthquake region, determines the practical building collapse distribution map in earthquake region.
S102, operating mode is analyzed according at least one Nonlinear Time course to buildings model execution earthquake region building earthquake mould
Quasi-, each simulant building for obtaining earthquake region collapses distribution map, wherein a kind of Nonlinear Time course analysis operating mode corresponds to a mould
Quasi- building collapse distribution map, simulant building collapse distribution map include each building in earthquake region the detailed collapse state of simulation.
Specifically, the simulant building for simulating to obtain earthquake region by Nonlinear Time course collapses distribution map, each building
It simulates detailed collapse state and is embodied in analog result i.e. simulant building and collapse in distribution map.In the present embodiment, simulation is broken in detail
Bad state may include intact, slight, medium, serious, collapse.
Before executing earthquake region building earthquake disaster simulation, the Nonlinear Time course analysis operating mode for first building magnanimity is needed.
It is described below and how to build Nonlinear Time course analysis operating mode.
In one possible implementation, the specific implementation mode of " structure Nonlinear Time course analyzes operating mode " is：
Determine at least one ground motion predictive equation；At least one seismic motion record is obtained from earthquake motion database；Using really
Fixed at least one ground motion predictive equation carries out amplitude modulation at least one acquired seismic motion record, with structure at least one
A Nonlinear Time course analyzes operating mode.
For example, p kinds (for example, p is the positive integer more than 30) Different Ground motion prediction equation is selected first
(GMPE, Ground Motion Prediction Equation), further according to focal shock parameter, for each GMPE from earthquake motion
Q items (q is the positive integer more than 20) seismic motion record is selected in database, can construct pq Nonlinear Time course point in this way
Analyse operating mode.
In the present embodiment, the specific implementation of step S102 is：
S21, the buildings model for determining each building in earthquake region.
Specifically, when building buildings model, the median low structure in earthquake region is simulated using more particles shearing series model,
The skyscraper in earthquake region is simulated using more particle parallel connection shear-bow Coordination Models.
S22, operating mode is analyzed to each Nonlinear Time course：Utilize the earthquake motion after amplitude modulation corresponding with each building
As input, earthquake disaster simulation is carried out to corresponding buildings model, obtains the detailed collapse state of simulation of each building.
Specifically, multiple Nonlinear Time course analysis operating modes are constructed in advance, specifically, according to ground motion prediction side
The Earthquake Intensity that journey calculates each building position builds every, according to calculating every selected seismic motion record
Obtained Earthquake Intensity carries out amplitude modulation, and operating mode is analyzed to build multiple Nonlinear Time courses.
For example, the sum of Nonlinear Time course analysis operating mode is 60, and the building sum in earthquake region is 50.To the 1st
Nonlinear Time course analyzes operating mode, and the result of earthquake disaster simulation is：Building 1：Intact, building 2：Slightly, 3 are built：Medium, building
4：Seriously, 5 are built：Collapse, build 6：Collapse, build 7：Collapse, build 8：Seriously ... build 49：Seriously, 50 are built：
It collapses.And so on, it obtains the 2nd Nonlinear Time course and analyzes the shake that operating mode analyzes operating mode to the 60th Nonlinear Time course
The result of evil simulation.
S23, each Nonlinear Time course analysis operating mode correspondence is exported according to the detailed collapse state of simulation of each building
Simulant building collapse distribution map.
For example, the 1st Nonlinear Time course analyzes operating mode, and the simulation for having respectively obtained building 1 to building 50 is detailed
The detailed collapse state of simulation of building 1 to building 50 is embodied to simulant building and is collapsed distribution map by thin collapse state.Class successively
It pushes away, obtains the simulant building that the 2nd Nonlinear Time course analysis operating mode analyzes operating mode to the 60th Nonlinear Time course and fall
Collapse distribution map.
In the present embodiment, earthquake disaster simulation is carried out using the Nonlinear Time course of magnanimity analysis operating mode, obtains multiple moulds
Quasi- building collapse distribution map is distributed similarity analysis for building collapse and provides basis.
S103, distribution map that each simulant building collapses are matched with practical building collapse distribution map, obtain each mould
The similarity score of quasi- building collapse distribution map.
In the present embodiment, it is a kind of two-value event that whether building, which collapses, therefore two-value similitude can be used to weigh simulation
The similitude of building collapse distribution map and practical building collapse distribution map.In order to find in collapsing distribution map from one group of simulant building
With practical building collapse distribution map distribution map the most matched, this implementation passes through the similarity of calculating simulation building collapse distribution map
Score is found and practical building collapse distribution map distribution map the most matched according to the height of similarity score.
Specifically, the similarity score that point-by-point matching method carrys out calculating simulation building collapse distribution map, this method may be used
It is simple and effective.In point-by-point matching method, whether the detailed collapse state of simulation of more every building falls with practical building one by one
State of collapsing is identical, and 1 point is counted if identical, otherwise counts 0 point.
If stochastic variable y indicates that building collapse situation, y obey Bernoulli Jacob's distribution, i.e. y~B (1, p).Y=1 expressions are collapsed,
Y=0 expressions are not collapsed.
Assuming that building sum in earthquake region is m, m is positive integer, and the collapse sum of distribution map of acquired simulant building is I, I
For positive integer.
J-th is built, yijShape is destroyed in detail for the collapse simulation of distribution map of j-th of corresponding i-th of simulant building of building
State, yjThe practical building collapse state built for j-th, wherein i is positive integer, and the value in 1 to I, j is positive integer, and j is 1
To value in m.
Then to point-by-point matching method, the specific implementation of step S103 is：
S31, collapse to i-th of simulant building distribution map：
According to formula：
J-th of corresponding i-th of simulant building of building is calculated to collapse the similarity score S of distribution mapaij。
Specifically, collapse distribution map to any simulant building, judge the detailed collapse state of simulation each built whether with
Practical building collapse state is identical, if identical, the detailed collapse state of simulation of the building is similar to practical building collapse state
Degree is scored at 1 point of meter, otherwise counts 0 point.
S32, according to formula：
I-th of simulant building is calculated to collapse the similarity score S of distribution mapAi。
Specifically, collapse distribution map to any simulant building, the simulation that each building can be obtained according to step S31 is detailed
The similarity score of collapse state and practical building collapse state builds the detailed collapse state of the simulation of each building with practical
The similarity score of collapsed state is first summed to be averaged again, is obtained the simulant building and is collapsed distribution map similarity score.
Specifically, the present embodiment can also be using the point-by-point matching method of weighting come the similar of calculating simulation building collapse distribution map
Spend score.
The collapse probability of difference building is different, depends on architectural character and situation of building.Position in order to introduce building is sat
Influence of other important informations such as mark, structure type for result defines a weight related with the collapse probability of building, will
It is each of final that the similarity score of the detailed collapse state of simulation and practical building collapse state each built, which is multiplied by weight,
The similarity score of the simulation detailed collapse state and practical building collapse state of building.
If stochastic variable y indicates that building collapse state, y obey Bernoulli Jacob's distribution, i.e. y~B (1, p).Y=1 expressions are collapsed,
Y=0 expressions are not collapsed.
Assuming that building sum in earthquake region is m, m is positive integer, and the collapse sum of distribution map of acquired simulant building is I, I
For positive integer.
J-th is built, yijShape is destroyed in detail for the collapse simulation of distribution map of j-th of corresponding i-th of simulant building of building
State, yjThe practical building collapse state built for j-th, wherein i is positive integer, and the value in 1 to I, j is positive integer, and j is 1
To value in m.
Then to weighting point-by-point matching method, the specific implementation of step S103 is：
S33, collapse to i-th of simulant building distribution map：
According to formula：
J-th of corresponding i-th of simulant building of building is calculated to collapse the similarity score S of distribution mapbij。
Specifically, collapse distribution map to any simulant building, judge the detailed collapse state of simulation each built whether with
Practical building collapse state is identical, if identical, the detailed collapse state of simulation of the building is similar to practical building collapse state
Degree is scored at 1 point of meter, otherwise counts 0 point.
S34, according to formula：
I-th of simulant building is calculated to collapse the similarity score S of distribution mapBi。
Wherein, the weight of j-th of buildingpjThe collapse probability built for j-th.
Fig. 2 is the schematic diagram of distribution of illustratively collapsing.In fig. 2, including 12 buildings.Wherein, Fig. 2 (a) is practical
Collapse distribution, Fig. 2 (b) is that simulation is collapsed distribution 1, and Fig. 2 (c) is that simulation is collapsed distribution 2, and Fig. 2 (d) is that simulation is collapsed distribution 3.
When calculating similarity score using point-by-point matching method, simulation, which is collapsed, is distributed 1 and differing distribution maximum of actually collapsing,
Corresponding relatively low (the S of similarity scoreA1=7/12 point), simulation, which is collapsed, is distributed 2 corresponding similarity score (SA2=10/12 point)
Higher, simulation, which is collapsed, is distributed 3 corresponding similarity score (SA3=10/12 point) it is higher.
When using point-by-point matching method calculating similarity score is weighted, simulation, which is collapsed, is distributed 1 corresponding similarity score SB1
=0.596 point, simulation, which is collapsed, is distributed 2 corresponding similarity score SB2=0.857 point, simulation, which is collapsed, is distributed 3 corresponding similarities
Score SB3=0.882 point.It can be by the position coordinates of building, knot it can be found that weighting the relatively point-by-point matching method of point-by-point matching method
Influence of other important informations such as structure type for result takes into account.
In the present embodiment, the collapse probability of each building can be obtained based on the method for machine learning.
First, determine that the determinant of building collapse probability is collapse probability factor vector x.For example, however, it is determined that build
Build coordinate, structure type, build the age, the factors such as the number of plies be building collapse determinant, then collapse probability factor vector x by
Building structure, structure type, building age, number of plies structure, but be not limited to illustrate.
It is then determined building collapse probability.Specifically：
The collapse probability p of j-th of buildingjFor：
Then, the not collapse probability 1-p of j-th of buildingjFor：
1-pj=P (y=0 | x=xj；θ)=1-h (θTxj)(6)；
Wherein, x is collapse probability factor vector, xjIndicate the collapse probability factor vector of j-th of building, θ is based on patrolling
The codomain for collecting parameter vector θ, logical function h (z) that classification algorithm training obtains is (0,1).
It is a known quantity by above formula it is found that x collapse probability factor vectors are pre-determined determinant, as long as
Parameter vector θ is determined, the collapse probability and not collapse probability of each building are would know that according to formula (5) and formula (6).
Be described below how the parameter vector θ trained based on the logical division algorithm in machine learning.
In one possible implementation, " parameter trained based on the logical division algorithm in machine learning to
Amount θ " specific implementation be：
S201, building collapse distributed model is determined：
P(y|x；θ)=h (θTx)y[1-h(θTx)]1-y(7)。
Specifically, by the observation to formula (5) and formula (6), merged, obtain building collapse distributed model.
S202, determine that the maximum likelihood estimator model of parameter vector θ is according to building collapse distributed model：
Specifically, according to being not difficult after determining building collapse distributed model to the regulation of likelihood function in the prior art
Obtain the maximum likelihood estimator model of parameter vector θ.
S203, it is concentrated from sample data and chooses training set, trained maximum likelihood estimator model using training set, joined
The maximum likelihood estimator of number vector θ.
Specifically, the building sum in earthquake region is m, and the practical building of each building has been recognized from aerial images after shake
Collapsed state, and the collapse probability factor of each building is determined.For example, being built to j-th, collapse probability factor is xj,
Its practical building collapse state is yj.According to known information, structure total sample number is the sample data set of m, and j-th is built
The collapse probability factor x builtjAnd practical building collapse state yjJ-th of sample data as sample data set.
In the present embodiment, it can choose to concentrate from sample data and choose a part of sample as training set, training is maximum
Likelihood estimator model obtains the maximum likelihood estimator of parameter vector θ
Further, in order to avoid over-fitting, the non-negative regularization parameter λ of addition in formula (8) forms formula
(9)：
In the present embodiment, it can choose to concentrate from sample data and choose a part of sample as cross validation collection, determine
The value of regularization parameter λ.A part of sample is chosen as the survey of cross validation collection it is, of course, also possible to choose and concentrated from sample data
Examination collection, test machine learn precision.
S204, the corresponding parameter vector θ of maximum likelihood estimator is determined as calculating based on the logical division in machine learning
The parameter vector θ that method is trained.
Specifically, it after obtaining parameter vector θ, is substituted into formula (5) and would know that each building with formula (6)
Collapse probability and not collapse probability.
In the present embodiment, it is calculated using training to obtain parameter vector θ based on the logical division algorithm in machine learning
The collapse probability of each building and not collapse probability, result of calculation are more reliable.
S104, the distribution map that collapsed according to each simulant building similarity score select optimal simulant building to collapse distribution
Figure.
Specifically, similarity score and the predetermined threshold value of distribution map of each simulant building can collapsing carries out size ratio
Compared with, the simulant building distribution map that collapses that similarity score is more than to predetermined threshold value collapses distribution map as optimal simulant building,
It can be to being ranked up to the collapse similarity score of distribution map of each simulant building, by the highest simulant building of similarity score
The distribution map that collapses collapses distribution map as optimal simulant building, and but it is not limited to this.
In one possible implementation, the specific implementation of step S104 is：Collapse point to each simulant building
The similarity score of Butut is ranked up, and is selected and is met the simulant building of preset condition and collapse distribution map as optimal simulant building
Collapse distribution map.In the present embodiment, preset condition can be with sets itself, the phase of distribution map for example, each simulant building collapses
Like degree score carry out descending sort, using the simulant building of the similarity score for the predetermined number for coming front collapse distribution map as
Optimal simulant building collapses distribution map (i.e. optimal simulant building collapse distribution map have multiple)；For example, similarity score is highest
The simulant building distribution map that collapses collapses distribution map as optimal simulant building；If similarity score highest is there are multiple, more
A highest simulant building of similarity score distribution map that collapses collapses distribution map as optimal simulant building.
The detailed collapse state of the simulation of each building carries out each build in S105, the distribution map that collapsed according to optimal simulant building
The damaged assessment built.
Specifically, house Seismic Economic Losses are house failure loss LhWith finishing failure loss LdThe sum of, but be not limited to
This, the computational methods of house Seismic Economic Losses are formulated according to practical situation.
In the present embodiment, to each building, according to formula：
Lh=S × Dh×P(10)
Calculate house failure loss Lh。
In above formula, S is house architectural area, unit m2；P is house resetting unit price, and unit is member/m2, DhIt is optimal
Simulate the corresponding house failure loss ratio of detailed collapse state.
Table 1 is that each collapse state lower room is destroyed and finishing failure loss compares median.In the present embodiment, DhAccording to table
1 value.
1 each collapse state lower room of table is destroyed and finishing failure loss compares median
Table 2 is building replacement cost median.In the present embodiment, P is according to 2 value of table.
Table 2 builds replacement cost median, unit：Member/m2
In the present embodiment, to each building, according to formula：
Ld=γ1×γ2×(ξ×S)×Dd×(η×P)(11)
Calculate finishing failure loss Ld。
In above formula, γ1To consider the correction factor of each regional economy situation difference, γ2To consider different building occupancies
Correction factor, it is the ratio that medium-to-high grade decorating house construction area accounts for total house that 1.0, ξ is taken if not considering, η is house dress
Repair the ratio of expense and main house body cost, DdFor the corresponding finishing failure loss ratio of the optimal detailed collapse state of simulation.At this
In embodiment, DdAccording to 1 value of table.
It should be pointed out that from table 1 it follows that building economy costing bio disturbance requires the detailed collapse state of building to make
For input, the distribution situation of collapsing of building is only identified from the aerial photograph of disaster area, is not sufficient to for calculating economic loss.But
The distribution situation of collapsing of building can be used for identifying optimal analog result from a large amount of analog results, include in optimal analog result
The detailed collapse state of building, so as to be used to calculate economic loss.
The damaged appraisal procedure of near real-time provided in this embodiment based on aerial images after shake, passes through the aerial images after shake
The practical building collapse distribution map in middle identification earthquake region, and a large amount of simulant building is obtained by earthquake disaster simulation and is collapsed distribution map,
Then it is selected from a large amount of simulant building distribution map that collapses with practical building collapse distribution map the most by similarity mode method
The optimal simulant building matched collapses distribution map, and damaged assessment is carried out according to the optimal simulant building distribution map that collapses.This method utilizes
The important information that earthquake region scene aerial photograph provides, improves the precision of damaged assessment, has high computational efficiency, Ke Yi
The assessment result of near real-time is provided after shake in 48 hours.
Below by taking Ludian earthquake as an example, the damaged appraisal procedure of near real-time based on aerial images after shake is carried out further
Explanation.
Fig. 3 is the practical building collapse distribution map of Ludian earthquake；Fig. 4 is that the optimal simulant building of Ludian earthquake collapses distribution
Figure；Fig. 5 is the relational graph of similarity score and economic loss；Fig. 6 is the block diagram of economic loss.
By taking Ludian earthquake Longtoushan town in 2014 is built as an example, the focus of Ludian earthquake is would know that by inquiring pertinent literature
Information：Earthquake magnitude is M6.5, depth of focus 12km, and earthquake centre is located at 27.189 ° of N, 103.409 ° of E, earthquake to 9km outside Longtoushan town
Bring more serious destruction.Second day after the earthquake, China Seismology Bureau just obtained a large amount of disaster areas using unmanned plane and takes photo by plane
Photo；The detailed destruction information of building is then to go to scene expansion seimic disaster census after obtained from nearly one month as expert.
The damaged appraisal procedure of near real-time introduced below based on aerial images after shake is applied to the process of Ludian earthquake.
First, the practical building collapse distribution map (i.e. Fig. 3) of Ludian earthquake is obtained.Specifically, by China Seismology Bureau
The live aerial photograph that a large amount of disaster area aerial photographs of publication and major news media provide is identified, and can obtain Shandong rapidly
The practical building collapse distribution map (i.e. Fig. 3) in pasture Longtoushan town.
Secondly, (12) linear fading function such as formula is defined, it is shown in Fig. 3 by taking coordinate system shown in Fig. 3 as an example
In coordinate system, x' axis positive directions are east, and y' axis positive directions are northern (N), to the plane coordinates (y', x') of Center for Architecture point,
The coordinate maximum value of y' axis directions is y'max, it is x' in the coordinate maximum value of x' axis directionsmax, y' axis directions coordinate most
Small value is y'min, it is x' in the coordinate minimum value of x' axis directionsmin。
PGAmaxIt is a coefficient, codomain is taken as { 0.2g, 0.4g, 0.6g, 0.8g, 1.0g, 1.2g }, therefore, defines
30 kinds of different ground motion predictive equations.28 Near-field ground motion records are selected from earthquake motion database, to construct 840 kinds
Different Nonlinear Time courses analyze operating mode.
Median low structure is simulated using more particles shearing series model, more particle parallel connection shear-bow Coordination Model simulations are high
Layer building carries out the Nonlinear Time course mould for building earthquake in region using 840 kinds of different Nonlinear Time course analysis operating modes
It is quasi-, obtain the detailed collapse state (intact, slight, medium, serious, collapse) of every building.
Then, using the distribution Similarity Match Method that collapses, each simulant building can be calculated and collapse distribution map
Score and optimal simulant building collapse distribution map i.e. Fig. 4.
Finally, the relational graph of similarity score and economic loss, such as Fig. 5 are drawn.And Fig. 5 is counted, form such as Fig. 6
Economic loss block diagram, the V in Fig. 6optAccording to optimal simulant building collapse each building in distribution map simulation it is broken in detail
Bad state computation obtains economic loss.
In Figure 5, in addition to the loss V of simulationj, it is also added into actual loss Vactual, Seismic Vulnerability Matrixes loss Vyin, real
Earthquake operating mode in border loses Vrecorded。
It should be pointed out that the loss V of simulationjBe according to simulant building collapse each building in distribution map simulation it is detailed
Economic loss is calculated in collapse state.
It is pointed out that actual loss VactualScene expansion seimic disaster census is gone to be obtained after nearly one month for expert
The economic loss arrived.
It should be pointed out that Strong Earthquakes In China, which moves platform net, obtains the practically vibrations near the place, Strong Earthquakes In China is utilized
The practically vibrations of dynamic platform net record carry out Nonlinear Time course analysis, obtain actual seismic operating mode loss Vrecorded。
It should be pointed out that use traditional Seismic Vulnerability Matrixes method to the building in this secondary earthquake Longtoushan town lose into
Row assessment, economic loss are denoted as Seismic Vulnerability Matrixes loss Vyin。
As can be seen from the results：
(1) the optimal simulant building that the point-by-point matching method of weighting proposed by the present invention obtains collapses distribution map (i.e. Fig. 4) and reality
Border building collapse distribution map (i.e. Fig. 3) is still more similar.
(2) make the corresponding loss V of the optimal analog result being obtained by the present inventionoptWith actual loss VactualIt connects
Closely, it is much better than the loss V that Seismic Vulnerability Matrixes method providesyin, as shown in Figure 6.Its key reason be it is proposed that method fill
Divide and practical this important information of building collapse distribution situation is utilized.
(3) the loss forecasting result V obtained using the dynamic record of actual measurement Ludian earthquakerecordedWith actual loss VactualAlso it kisses
It closes good.It should be noted that the seismic motion record of macroseism station actual measurement is likely difficult to embody what each building was subject to completely
Actual seismic acts on, therefore is also impossible to completely the same with practical earthquake.
It should be noted that above-mentioned 840 kinds of Nonlinear Time courses analysis operating mode is parallel on a multi-core computer
Operation (CPU：Intel E5-2695v4@2.10Hz, 36 cores；Memory：64GB), only about 4min is taken, distribution similarity of collapsing
Be even more can be completed in several seconds.It is just expected in 24 hours obtain a large amount of calamities with remote sensing or the progress of unmanned air vehicle technique, after shake
The satellite or aerial photograph in area, to which identification obtains the building collapse situation in disaster area rapidly.
In conclusion by above example, it can summarize and learn：Method provided by the invention is a kind of earthquake of near real-time
Economic loss evaluation method can be provided after shake in a couple of days by the important information for making full use of disaster area aerial photograph to provide
The close damaged estimation with the expert investigation result of time-consuming several weeks.
In the description of this specification, reference term " one embodiment ", " some embodiments ", " example ", " specifically show
The description of example " or " some examples " etc. means specific features, structure, material or spy described in conjunction with this embodiment or example
Point is included at least one embodiment or example of the invention.In the present specification, schematic expression of the above terms are not
It must be directed to identical embodiment or example.Moreover, particular features, structures, materials, or characteristics described can be in office
It can be combined in any suitable manner in one or more embodiments or example.In addition, without conflicting with each other, the skill of this field
Art personnel can tie the feature of different embodiments or examples described in this specification and different embodiments or examples
It closes and combines.
In addition, term " first ", " second " are used for description purposes only, it is not understood to indicate or imply relative importance
Or implicitly indicate the quantity of indicated technical characteristic.Define " first " as a result, the feature of " second " can be expressed or
Implicitly include at least one this feature.In the description of the present invention, the meaning of " plurality " is at least two, such as two, three
It is a etc., unless otherwise specifically defined.
Any process described otherwise above or method description are construed as in flow chart or herein, and expression includes
It is one or more for realizing custom logic function or process the step of executable instruction code module, segment or portion
Point, and the range of the preferred embodiment of the present invention includes other realization, wherein can not press shown or discuss suitable
Sequence, include according to involved function by it is basic simultaneously in the way of or in the opposite order, to execute function, this should be of the invention
Embodiment person of ordinary skill in the field understood.
Expression or logic and/or step described otherwise above herein in flow charts, for example, being considered use
In the order list for the executable instruction for realizing logic function, may be embodied in any computer-readable medium, for
Instruction execution system, device or equipment (system of such as computer based system including processor or other can be held from instruction
The instruction fetch of row system, device or equipment and the system executed instruction) it uses, or combine these instruction execution systems, device or set
It is standby and use.For the purpose of this specification, " computer-readable medium " can any can be included, store, communicating, propagating or passing
Defeated program is for instruction execution system, device or equipment or the dress used in conjunction with these instruction execution systems, device or equipment
It sets.The more specific example (non-exhaustive list) of computer-readable medium includes following：Electricity with one or more wiring
Interconnecting piece (electronic device), portable computer diskette box (magnetic device), random access memory (RAM), read-only memory
(ROM), erasable edit read-only storage (EPROM or flash memory), fiber device and portable optic disk is read-only deposits
Reservoir (CDROM).In addition, computer-readable medium can even is that can on it the paper of print routine or other suitable be situated between
Matter, because can be for example by carrying out optical scanner to paper or other media, then into edlin, interpretation or when necessary with other
Suitable method is handled electronically to obtain program, is then stored in computer storage.
It should be appreciated that each section of the present invention can be realized with hardware, software, firmware or combination thereof.Above-mentioned
In embodiment, software that multiple steps or method can in memory and by suitable instruction execution system be executed with storage
Or firmware is realized.Such as, if realized in another embodiment with hardware, following skill well known in the art can be used
Any one of art or their combination are realized：With for data-signal realize logic function logic gates from
Logic circuit is dissipated, the application-specific integrated circuit with suitable combinational logic gate circuit, programmable gate array (PGA), scene can compile
Journey gate array (FPGA) etc..
Those skilled in the art are appreciated that realize all or part of step that above-described embodiment method carries
Suddenly it is the program that relevant hardware can be instructed to complete by program can be stored in a kind of computer readable storage medium,
The program includes the steps that one or a combination set of embodiment of the method when being executed.
In addition, each functional unit in each embodiment of the present invention can be integrated in a processing module, it can also
That each unit physically exists alone, can also two or more units be integrated in a module.Above-mentioned integrated mould
The form that hardware had both may be used in block is realized, can also be realized in the form of software function module.If integrated module with
The form of software function module realizes and when sold or used as an independent product, can also be stored in one it is computer-readable
It takes in storage medium.
Storage medium mentioned above can be read-only memory, disk or CD etc..Although having been shown and retouching above
The embodiment of the present invention is stated, it is to be understood that above-described embodiment is exemplary, and should not be understood as the limit to the present invention
System, those skilled in the art can be changed above-described embodiment, change, replace and become within the scope of the invention
Type.
Claims (10)
1. a kind of damaged appraisal procedure of near real-time based on aerial images after shake, which is characterized in that including：
The practical building collapse state for identifying each building in aerial images after shaking obtains the practical building collapse distribution in earthquake region
Figure, wherein practical building collapse distribution map includes the practical building collapse state of each building in earthquake region；
Operating mode is analyzed according at least one Nonlinear Time course, earthquake region building earthquake disaster simulation is executed to buildings model, obtain earthquake region
Each simulant building collapse distribution map, wherein a kind of Nonlinear Time course analysis operating mode corresponds to a simulant building and collapses
Distribution map, simulant building collapse distribution map include each building in earthquake region the detailed collapse state of simulation；
Each simulant building distribution map that collapses is matched with practical building collapse distribution map, each simulant building is obtained and collapses
The similarity score of distribution map；
Optimal simulant building is selected to collapse distribution map according to the collapse similarity score of distribution map of each simulant building；
The damaged of each building is carried out according to the optimal simulant building detailed collapse state of simulation of each building in distribution map that collapses
Assessment.
2. the method as described in claim 1, which is characterized in that described to be collapsed each in distribution map build according to optimal simulant building
The detailed collapse state of simulation built carries out the damaged assessment of each building, including：
To each building：
According to formula Lh=S × Dh× P calculates house failure loss Lh, wherein S is house architectural area, and P is that house resetting is single
Valence, DhFor the corresponding house failure loss ratio of the optimal detailed collapse state of simulation；
According to formula Ld=γ1×γ2×(ξ×S)×Dd× (η × P) calculates finishing failure loss Ld, wherein γ1It is each to consider
The correction factor of a regional economy situation difference, γ2To consider that the correction factor of different building occupancies, ξ are medium-to-high grade finishing room
Estate building coverage accounts for the ratio in total house, and η is the ratio of house decoration expense and main house body cost, DdIt is detailed for optimal simulation
The corresponding finishing failure loss ratio of collapse state；
According to house failure loss LhWith finishing failure loss LdCalculate the house Seismic Economic Losses each built.
3. the method as described in claim 1, which is characterized in that described to build each simulant building distribution map that collapses with practical
The distribution map that collapses is matched, and is obtained each simulant building and is collapsed the similarity score of distribution map, including：
Collapse distribution map to i-th of simulant building：
According to formulaThe collapse similarity of distribution map of j-th of corresponding i-th of simulant building of building is calculated to obtain
Divide Saij；
According to formulaI-th of simulant building is calculated to collapse the similarity score S of distribution mapAi；
Wherein, yijCollapse the detailed collapse state of simulation of distribution map for j-th of corresponding i-th of simulant building of building, yjIt is j-th
The practical building collapse state of building, i are positive integer, and m is positive integer, and m is the building sum in earthquake region, and j is positive integer, and j is 1
To value in m.
4. the method as described in claim 1, which is characterized in that described to build each simulant building distribution map that collapses with practical
The distribution map that collapses is matched, and is obtained each simulant building and is collapsed the similarity score of distribution map, including：
Collapse distribution map to i-th of simulant building：
According to formulaThe collapse similarity of distribution map of j-th of corresponding i-th of simulant building of building is calculated to obtain
Divide Sbij；
According to formulaI-th of simulant building is calculated to collapse the similarity score S of distribution mapBi；
Wherein, the weight of j-th of buildingpjThe collapse probability built for j-th；
Wherein, yijCollapse the detailed collapse state of simulation of distribution map for j-th of corresponding i-th of simulant building of building, yjIt is j-th
The practical building collapse state of building, i are positive integer, and m is positive integer, and m is the building sum in earthquake region, and j is positive integer, and j is 1
To value in m.
5. method as claimed in claim 4, which is characterized in that further include：
According to formulaCalculate the collapse probability p of j-th of buildingj；
Wherein, x is collapse probability factor vector, θ be the parameter trained based on the logical division algorithm in machine learning to
θ is measured, the codomain of logical function h (z) is (0,1).
6. method as claimed in claim 4, which is characterized in that further include：
Determine building collapse distributed model P (y | x；θ)=h (θTx)y[1-h(θTx)]1-y；
Determine that the maximum likelihood estimator model of parameter vector θ is according to building collapse distributed model
It is concentrated from sample data and chooses training set, trained maximum likelihood estimator model using training set, obtain parameter vector θ's
Maximum likelihood estimator, wherein the collapse probability factor x of j-th of buildingjAnd practical building collapse state yjAs sample data
J-th of sample data of collection；
The corresponding parameter vector θ of maximum likelihood estimator is determined as training to obtain based on the logical division algorithm in machine learning
Parameter vector θ.
7. the method as described in claim 1, which is characterized in that the similarity of the distribution map that collapsed according to each simulant building
The optimal simulant building of component selections collapses distribution map, including：
The collapse similarity score of distribution map of each simulant building is ranked up, is selected and is met the simulant building of preset condition and fall
The distribution map that collapses collapses distribution map as optimal simulant building.
8. the method as described in claim 1, which is characterized in that described to analyze operating mode according at least one Nonlinear Time course
Earthquake region is executed to buildings model and builds earthquake disaster simulation, each simulant building for obtaining earthquake region collapses distribution map, including：
Determine the buildings model of each building in earthquake region, wherein to the median low structure in earthquake region, buildings model is more particles
Series model is sheared, to the skyscraper in earthquake region, buildings model is more particle parallel connection shear-bow Coordination Models；
Operating mode is analyzed to each Nonlinear Time course：Using the earthquake motion after amplitude modulation corresponding with each building as defeated
Enter, earthquake disaster simulation is carried out to corresponding buildings model, obtains the detailed collapse state of simulation of each building.
The corresponding simulation of each Nonlinear Time course analysis operating mode is exported according to the detailed collapse state of the simulation of each building to build
Build the distribution map that collapses.
9. the method as described in claim 1, which is characterized in that further include：
Determine at least one ground motion predictive equation；
At least one seismic motion record is obtained from earthquake motion database；
Amplitude modulation is carried out at least one acquired seismic motion record using identified at least one ground motion predictive equation,
Operating mode is analyzed to build at least one Nonlinear Time course.
10. the method as described in claim 1, which is characterized in that aerial images carry out image procossing after described pair of shake, obtain shake
The practical building collapse distribution map in area, including：
The visually practical building collapse state of each building after interpretation method identification shake in aerial images, according to building for earthquake region
The practical building collapse state for building distribution map and each building recognized, obtains the practical building collapse distribution map in earthquake region.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810119671.8A CN108304809B (en) | 2018-02-06 | 2018-02-06 | Near real-time earthquake damage assessment method based on post-earthquake aerial image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810119671.8A CN108304809B (en) | 2018-02-06 | 2018-02-06 | Near real-time earthquake damage assessment method based on post-earthquake aerial image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108304809A true CN108304809A (en) | 2018-07-20 |
| CN108304809B CN108304809B (en) | 2020-03-27 |
Family
ID=62864450
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810119671.8A Active CN108304809B (en) | 2018-02-06 | 2018-02-06 | Near real-time earthquake damage assessment method based on post-earthquake aerial image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108304809B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111985355A (en) * | 2020-08-01 | 2020-11-24 | 桂林理工大学 | Remote sensing building earthquake damage assessment method and system based on deep learning and cloud computing |
| CN114862687A (en) * | 2021-02-04 | 2022-08-05 | 四川大学 | Self-adaptive compressed image restoration method driven by depth deblocking operator |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101979961A (en) * | 2010-05-18 | 2011-02-23 | 中国地震局地球物理研究所 | Disaster condition acquisition system |
| CN201910818U (en) * | 2011-01-05 | 2011-07-27 | 贾群林 | Seismic disaster relief simulation system |
| CN103116810A (en) * | 2013-01-28 | 2013-05-22 | 北京师范大学 | Method for evaluating population loss in earthquake |
| CN103678937A (en) * | 2013-12-29 | 2014-03-26 | 中国地震局工程力学研究所 | Method for evaluating overall earthquake damage level of reinforced concrete frame structure based on equivalent single-degree-of-freedom system |
| CN103761740A (en) * | 2014-01-23 | 2014-04-30 | 武汉大学 | Construction damage assessment method based on single post-earthquake POLSAR image |
| CN104239885A (en) * | 2014-09-05 | 2014-12-24 | 北京航天控制仪器研究所 | Earthquake disaster damage degree evaluation method based on unmanned aerial vehicle aerial photos |
| CN104346538A (en) * | 2014-11-26 | 2015-02-11 | 中国测绘科学研究院 | Earthquake hazard evaluation method based on control of three disaster factors |
| CN105809266A (en) * | 2014-12-31 | 2016-07-27 | 郑静晨 | Method for preliminary evaluation and prediction on earthquake casualty |
| CN106251032A (en) * | 2016-06-28 | 2016-12-21 | 中国人民解放军后勤工程学院 | A kind of Earthquake-landslide disaster Disaster rapid evaluation method |
| CN106651039A (en) * | 2016-12-27 | 2017-05-10 | 中国地震局工程力学研究所 | City building group seismic hazard matrix dynamic prediction method |
| US20170169534A1 (en) * | 2015-12-09 | 2017-06-15 | One Concern, Inc. | Damage data propagation in predictor of structural damage |
| CN106875096A (en) * | 2017-01-17 | 2017-06-20 | 山西省地震局 | A kind of earthquake house loss appraisal system |
-
2018
- 2018-02-06 CN CN201810119671.8A patent/CN108304809B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101979961A (en) * | 2010-05-18 | 2011-02-23 | 中国地震局地球物理研究所 | Disaster condition acquisition system |
| CN201910818U (en) * | 2011-01-05 | 2011-07-27 | 贾群林 | Seismic disaster relief simulation system |
| CN103116810A (en) * | 2013-01-28 | 2013-05-22 | 北京师范大学 | Method for evaluating population loss in earthquake |
| CN103678937A (en) * | 2013-12-29 | 2014-03-26 | 中国地震局工程力学研究所 | Method for evaluating overall earthquake damage level of reinforced concrete frame structure based on equivalent single-degree-of-freedom system |
| CN103761740A (en) * | 2014-01-23 | 2014-04-30 | 武汉大学 | Construction damage assessment method based on single post-earthquake POLSAR image |
| CN104239885A (en) * | 2014-09-05 | 2014-12-24 | 北京航天控制仪器研究所 | Earthquake disaster damage degree evaluation method based on unmanned aerial vehicle aerial photos |
| CN104346538A (en) * | 2014-11-26 | 2015-02-11 | 中国测绘科学研究院 | Earthquake hazard evaluation method based on control of three disaster factors |
| CN105809266A (en) * | 2014-12-31 | 2016-07-27 | 郑静晨 | Method for preliminary evaluation and prediction on earthquake casualty |
| US20170169534A1 (en) * | 2015-12-09 | 2017-06-15 | One Concern, Inc. | Damage data propagation in predictor of structural damage |
| CN106251032A (en) * | 2016-06-28 | 2016-12-21 | 中国人民解放军后勤工程学院 | A kind of Earthquake-landslide disaster Disaster rapid evaluation method |
| CN106651039A (en) * | 2016-12-27 | 2017-05-10 | 中国地震局工程力学研究所 | City building group seismic hazard matrix dynamic prediction method |
| CN106875096A (en) * | 2017-01-17 | 2017-06-20 | 山西省地震局 | A kind of earthquake house loss appraisal system |
Non-Patent Citations (1)
| Title |
|---|
| 甘甜等: "面向对象的高分辨率遥感影像建筑物震害信息提取", 《测绘工程》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111985355A (en) * | 2020-08-01 | 2020-11-24 | 桂林理工大学 | Remote sensing building earthquake damage assessment method and system based on deep learning and cloud computing |
| CN111985355B (en) * | 2020-08-01 | 2022-09-27 | 桂林理工大学 | Remote sensing building earthquake damage assessment method and system based on deep learning and cloud computing |
| CN114862687A (en) * | 2021-02-04 | 2022-08-05 | 四川大学 | Self-adaptive compressed image restoration method driven by depth deblocking operator |
| CN114862687B (en) * | 2021-02-04 | 2023-05-09 | 四川大学 | Self-adaptive compressed image restoration method driven by depth deblocking operator |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108304809B (en) | 2020-03-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108537215A (en) | A kind of flame detecting method based on image object detection | |
| CN109446925A (en) | A kind of electric device maintenance algorithm based on convolutional neural networks | |
| Aghamohammadi et al. | Seismic human loss estimation for an earthquake disaster using neural network | |
| CN108763763A (en) | A kind of bridge structure strain-responsive abnormity early warning method | |
| CN113496104B (en) | Precipitation prediction correction method and system based on deep learning | |
| US11709979B1 (en) | Bridge damage identification method considering uncertainty | |
| CN110378232B (en) | Improved test room examinee position rapid detection method of SSD dual-network | |
| CN106991666A (en) | A kind of disease geo-radar image recognition methods suitable for many size pictorial informations | |
| CN108596883A (en) | It is a kind of that method for diagnosing faults is slid based on the Aerial Images stockbridge damper of deep learning and distance restraint | |
| CN108492298A (en) | Based on the multispectral image change detecting method for generating confrontation network | |
| CN110379036A (en) | Intelligent substation patrol recognition methods, system, device and storage medium | |
| CN108304809A (en) | The damaged appraisal procedure of near real-time based on aerial images after shake | |
| CN112685812A (en) | Dynamic supervision method, device, equipment and storage medium | |
| CN116884192A (en) | Power production operation risk early warning method, system and equipment | |
| CN114863352B (en) | Personnel group behavior monitoring method based on video analysis | |
| CN114331009B (en) | Method for compiling immovable cultural relic high wind disaster risk graph | |
| Vervoort | Application of neural networks to rate of spread estimation in shrublands | |
| CN113361968B (en) | Power grid infrastructure worker safety risk assessment method based on artificial intelligence and big data | |
| CN109523509A (en) | Detection method, device and the electronic equipment of wheat heading stage | |
| CN111860258A (en) | Examination room global event detection method and system based on three-dimensional convolutional neural network | |
| CN113283817B (en) | Disaster assessment method and system | |
| CN116091272B (en) | Campus abnormal activity monitoring method, device, equipment and medium | |
| Yu et al. | Sports event level measurement indicator system using multisensor information fusion | |
| LIN et al. | Application of Neural Network Entropy Algorithm and Convolution Neural Network for Structural Health Monitoring | |
| Ma | An Improved Convolutional Neural Network with Reservoir Computing for Boiler Four Tube |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |