CN117876362A - Deep learning-based natural disaster damage assessment method and device - Google Patents
Deep learning-based natural disaster damage assessment method and device Download PDFInfo
- Publication number
- CN117876362A CN117876362A CN202410268719.7A CN202410268719A CN117876362A CN 117876362 A CN117876362 A CN 117876362A CN 202410268719 A CN202410268719 A CN 202410268719A CN 117876362 A CN117876362 A CN 117876362A
- Authority
- CN
- China
- Prior art keywords
- target
- facility
- disaster
- stricken
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 40
- 238000013135 deep learning Methods 0.000 title claims abstract description 25
- 238000012545 processing Methods 0.000 claims abstract description 40
- 238000013210 evaluation model Methods 0.000 claims abstract description 16
- 238000011156 evaluation Methods 0.000 claims description 26
- 238000011084 recovery Methods 0.000 claims description 18
- 238000012549 training Methods 0.000 claims description 10
- 238000004590 computer program Methods 0.000 claims description 8
- 238000013527 convolutional neural network Methods 0.000 claims description 6
- 238000012216 screening Methods 0.000 claims description 3
- 238000005516 engineering process Methods 0.000 abstract description 9
- 230000008569 process Effects 0.000 description 10
- 230000004913 activation Effects 0.000 description 6
- 238000004364 calculation method Methods 0.000 description 6
- 230000003287 optical effect Effects 0.000 description 6
- 238000004422 calculation algorithm Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 238000011176 pooling Methods 0.000 description 5
- 230000004927 fusion Effects 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 238000013136 deep learning model Methods 0.000 description 3
- 238000001514 detection method Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 230000008439 repair process Effects 0.000 description 3
- 238000000605 extraction Methods 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 230000002093 peripheral effect Effects 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 239000013598 vector Substances 0.000 description 2
- 238000006424 Flood reaction Methods 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000009432 framing Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Landscapes
- Image Analysis (AREA)
Abstract
The application provides a natural disaster damage assessment method and device based on deep learning. The method comprises the following steps: acquiring a target current image and a target reconstruction image of a target disaster-stricken facility set; processing the current image of the target by using a type recognition model to obtain the type of the target facility of each target disaster-stricken facility; processing the target current image and the target reconstruction image by using a scale evaluation model to obtain a target current scale and a target reconstruction scale of each target disaster-stricken facility; determining the target damage proportion of each target disaster-stricken facility according to the current target scale and the target reconstruction scale; and determining a target loss value of each target disaster-stricken facility according to the target damage proportion, the target facility type, the target reconstruction scale and the target area value. By introducing the deep learning technology, a large amount of data can be automatically processed, the loss of the building structure and the infrastructure caused by the natural disasters can be rapidly and accurately estimated, and the efficiency and consistency of natural disaster loss estimation are improved.
Description
Technical Field
The application relates to the technical field of image processing, in particular to a natural disaster loss assessment method and device based on deep learning.
Background
Natural disasters, such as earthquakes, floods, typhoons, etc., often result in significant economic losses to building structures and infrastructure. Timely and accurate assessment of these losses is critical for post-disaster reconstruction and insurance claims.
Currently, natural disaster damage assessment relies mainly on manual processes, i.e. judging the damage level of a facility by professional assessors based on experience and observation to estimate the damage.
Although manual evaluation can obtain evaluation results with rich details, the problems of lower evaluation efficiency and poor result consistency exist.
Disclosure of Invention
In view of the above-mentioned problems, the present application has been made to provide a deep learning-based natural disaster damage assessment method and apparatus for overcoming the problems or at least partially solving the problems, including:
a natural disaster damage assessment method based on deep learning comprises the following steps:
acquiring a target current image and a target reconstruction image of a target disaster-stricken facility set, and a target area value of an area to which the target disaster-stricken facility set belongs;
Processing the target current image by using a pre-constructed type recognition model to obtain a target facility type of each target disaster-stricken facility;
processing the target current image and the target reconstructed image respectively by using a pre-constructed scale evaluation model to obtain a target current scale and a target reconstructed scale of each target disaster-stricken facility;
determining a target damage proportion of each target disaster-stricken facility according to the current target scale and the target reconstruction scale;
and determining a target loss value of each target disaster-stricken facility according to the target damage proportion, the target facility type, the target reconstruction scale and the target area value.
Preferably, the method further comprises:
acquiring a target current image of a target disaster-stricken facility set;
and generating a target reconstruction image of the target disaster facility set according to the target current image.
Preferably, the step of generating the target reconstructed image of the target disaster facility set according to the target current image includes:
processing the target current image by using a pre-constructed type recognition model to obtain a target facility type of each target disaster-stricken facility;
Generating a target reconstruction unit image of each target disaster-stricken facility according to the target current unit image of the target disaster-stricken facility in the target current image and the target current unit images of other target disaster-stricken facilities with the same type as the target facility of the target disaster-stricken facility;
and generating a target reconstruction image of the target disaster facility set according to all the target reconstruction unit images.
Preferably, the step of generating a target reconstructed unit image of each target disaster-stricken facility according to the target current unit image of the target disaster-stricken facility in the target current image and the target current unit images of other target disaster-stricken facilities with the same type as the target facility of the target disaster-stricken facility includes:
processing the target current image by using a pre-constructed scale evaluation model to obtain a target current scale of each target disaster-stricken facility;
screening out a preset number of target reference facilities with the maximum current standard from all target disaster-stricken facilities with the same target facility type;
when the target disaster-stricken facility belongs to the target reference facility, generating a target reconstruction unit image of the target disaster-stricken facility according to a target current unit image of the target disaster-stricken facility in the target current image;
When the target disaster-stricken facility does not belong to the target reference facility, generating a target reconstruction unit image of the target disaster-stricken facility according to a target current unit image of the target disaster-stricken facility in the target current image and a target current unit image of the target reference facility, which is the same as the target facility of the target disaster-stricken facility in type.
Preferably, the step of determining a target loss value of each target disaster recovery facility according to the target damage proportion, the target facility type, the target reconstruction scale and the target area value includes:
determining a target facility value of each target disaster-stricken facility according to the target facility type, the target reconstruction scale and the target area value;
and determining a target loss value of each target disaster-stricken facility according to the target damage proportion and the target facility value.
Preferably, the method further comprises:
acquiring a sample image of a sample disaster-stricken facility set and a sample type of each sample disaster-stricken facility;
training the initial type recognition model by using the sample image and the sample type to obtain a type recognition model.
Preferably, the initial type recognition model is a convolutional neural network.
A deep learning-based natural disaster damage assessment device, comprising:
the target information acquisition module is used for acquiring a target current image and a target reconstruction image of a target disaster-stricken facility set and a target area value of an area to which the target disaster-stricken facility set belongs;
the target type recognition module is used for processing the target current image by using a pre-constructed type recognition model to obtain a target facility type of each target disaster-stricken facility;
the target scale evaluation module is used for respectively processing the target current image and the target reconstruction image by using a pre-constructed scale evaluation model to obtain a target current scale and a target reconstruction scale of each target disaster-stricken facility;
the target damage evaluation module is used for determining the target damage proportion of each target disaster-stricken facility according to the current target scale and the target reconstruction scale;
and the target loss evaluation module is used for determining a target loss value of each target disaster-stricken facility according to the target damage proportion, the target facility type, the target reconstruction scale and the target area value.
A computer device comprising a processor, a memory and a computer program stored on the memory and capable of running on the processor, which computer program, when executed by the processor, implements the assessment method as claimed in any one of the preceding claims.
A computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the assessment method according to any of the preceding claims.
The application has the following advantages:
in the embodiment of the application, compared with the problems of lower evaluation efficiency and poorer result consistency of the existing evaluation method, the application provides a solution for introducing a deep learning technology to automatically evaluate natural disaster loss, which is specifically as follows: acquiring a target current image and a target reconstruction image of a target disaster-stricken facility set, and a target area value of an area to which the target disaster-stricken facility set belongs; processing the target current image by using a pre-constructed type recognition model to obtain a target facility type of each target disaster-stricken facility; processing the target current image and the target reconstructed image respectively by using a pre-constructed scale evaluation model to obtain a target current scale and a target reconstructed scale of each target disaster-stricken facility; determining a target damage proportion of each target disaster-stricken facility according to the current target scale and the target reconstruction scale; and determining a target loss value of each target disaster-stricken facility according to the target damage proportion, the target facility type, the target reconstruction scale and the target area value. By introducing the deep learning technology, a large amount of data can be automatically processed, the loss of the building structure and the infrastructure caused by the natural disasters can be rapidly and accurately estimated, and the efficiency and consistency of natural disaster loss estimation are improved.
Drawings
In order to more clearly illustrate the technical solutions of the present application, the drawings that are needed in the description of the present application will be briefly introduced below, it being obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of steps of a natural disaster damage assessment method based on deep learning according to an embodiment of the present application;
FIG. 2 is a flow chart of steps of a method for estimating natural disaster damage based on deep learning according to another embodiment of the present application;
FIG. 3 is a block diagram of a natural disaster damage assessment device based on deep learning according to an embodiment of the present application;
fig. 4 is a block diagram of a natural disaster damage assessment device based on deep learning according to another embodiment of the present application;
fig. 5 is a schematic structural diagram of a computer device according to an embodiment of the present application.
Reference numerals in the drawings of the specification are as follows:
12. a computer device; 14. an external device; 16. a processing unit; 18. a bus; 20. a network adapter; 22. an I/O interface; 24. a display; 28. a memory; 30. a random access memory; 32. a cache memory; 34. a storage system; 40. program/utility; 42. program modules.
Detailed Description
In order to make the objects, features and advantages of the present application more comprehensible, the present application is described in further detail below with reference to the accompanying drawings and detailed description. It will be apparent that the embodiments described are some, but not all, of the embodiments of the present application. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are within the scope of the present disclosure.
Referring to fig. 1, a natural disaster damage assessment method based on deep learning according to an embodiment of the present application is shown, including:
s110, acquiring a target current image and a target reconstruction image of a target disaster-stricken facility set, and a target area value of an area to which the target disaster-stricken facility set belongs;
s120, processing the target current image by using a pre-constructed type recognition model to obtain a target facility type of each target disaster-stricken facility;
s130, respectively processing the target current image and the target reconstructed image by using a pre-constructed scale evaluation model to obtain a target current scale and a target reconstructed scale of each target disaster-stricken facility;
S140, determining the target damage proportion of each target disaster-stricken facility according to the current target scale and the target reconstruction scale;
and S150, determining a target loss value of each target disaster-stricken facility according to the target damage proportion, the target facility type, the target reconstruction scale and the target area value.
In the embodiment of the application, compared with the problems of lower evaluation efficiency and poorer result consistency of the existing evaluation method, the application provides a solution for introducing a deep learning technology to automatically evaluate natural disaster loss, which is specifically as follows: acquiring a target current image and a target reconstruction image of a target disaster-stricken facility set, and a target area value of an area to which the target disaster-stricken facility set belongs; processing the target current image by using a pre-constructed type recognition model to obtain a target facility type of each target disaster-stricken facility; processing the target current image and the target reconstructed image respectively by using a pre-constructed scale evaluation model to obtain a target current scale and a target reconstructed scale of each target disaster-stricken facility; determining a target damage proportion of each target disaster-stricken facility according to the current target scale and the target reconstruction scale; and determining a target loss value of each target disaster-stricken facility according to the target damage proportion, the target facility type, the target reconstruction scale and the target area value. By introducing the deep learning technology, a large amount of data can be automatically processed, the loss of the building structure and the infrastructure caused by the natural disasters can be rapidly and accurately estimated, and the efficiency and consistency of natural disaster loss estimation are improved.
Next, a natural disaster damage evaluation method based on deep learning in the present exemplary embodiment will be further described.
As described in the step S110, a target current image and a target reconstructed image of a target disaster recovery facility set are acquired, and a target area value of an area to which the target disaster recovery facility set belongs is obtained.
The target current image refers to a real-time or near real-time image taken of a target set of disaster-stricken facilities (e.g., buildings, bridges, roads, etc.) after the occurrence of a natural disaster. These images may be taken from multiple angles (e.g., aerial and sideways) to fully capture the extent of damage to the targeted collection of disaster-stricken facilities.
The target reconstructed image is an image before the damage of the target disaster facility set which is restored based on the target current image. These images may be generated by a deep learning algorithm in combination with historical data and similar scenes.
The target regional value refers to the economic or social value of the region to which the target disaster-stricken facility set belongs. The value of the different areas affects the outcome of the loss assessment, as the same damage may lead to greater losses in areas of higher economic value.
And (3) processing the target current image by using a pre-constructed type recognition model to obtain a target facility type of each target disaster-stricken facility as shown in the step S120.
The current images of the targets are input into a type recognition model, which processes the images and recognizes the type of each target disaster-affected facility (e.g., residential, commercial building, bridge, road, etc.). The type recognition model is a pre-trained deep learning model for recognizing the type of object and facility in the image.
Particularly, when the current image of the target is obtained by shooting at multiple angles, that is, the number of the input images is greater than one, the model integrates the features of the images at different angles through a feature fusion algorithm (such as feature superposition, feature fusion network and attention mechanism) to obtain comprehensive features, so that the most important part in the image is focused, and a classification decision is made according to the comprehensive features to identify the type of the target disaster facility.
And as shown in the step S130, the target current image and the target reconstructed image are respectively processed by using a pre-constructed scale evaluation model, so as to obtain a target current scale and a target reconstructed scale of each target disaster facility.
Inputting the current images of the targets into a scale evaluation model, processing the images by the model, and calculating the current scale (such as occupied area and height) of each target disaster facility based on reference information (such as the scale of the known object and shooting parameters, etc.); the target reconstructed images are input into a scale assessment model, which processes the images and calculates a reconstruction scale (e.g., floor area, height, etc.) for each target disaster recovery facility based on reference information (e.g., scale of known objects, shooting parameters, etc.). The scale assessment model contains a series of visual algorithms for extracting objects in the image and calculating the facility scale.
Particularly, when the current image of the target is obtained through multi-angle shooting, that is, the number of the input images is greater than one, the model can identify the same characteristic points in different images through a characteristic matching algorithm, the shooting position and direction of each image are determined through a structure from a motion technology, 3D point clouds of the target disaster-stricken facilities are generated through triangulation based on the matched characteristic points and shooting parameters, and then the facility scale is extracted.
As described in step S140, a target damage proportion of each target disaster-stricken facility is determined according to the target current scale and the target reconstruction scale.
And calculating the damage proportion of the target according to a preset calculation rule based on the current scale of the target and the reconstruction scale of the target. As an example, the calculation formula of the target damage ratio is as follows:
；（1）
wherein,for the target damage proportion, +.>For the current footprint of the target disaster recovery facility,reconstructing the floor space for the target disaster-stricken facility, < > for>For the current height of the target disaster facility, < +.>The reconstruction height of the target disaster-stricken facility.
As described in the step S150, a target loss value of each target disaster recovery facility is determined according to the target damage proportion, the target facility type, the target reconstruction scale and the target area value.
Calculating a target facility value according to a preset calculation rule based on the target facility type, the target reconstruction scale and the target area value, and calculating a target loss value based on the target damage proportion and the target facility value. As an example, the calculation formula of the target loss value is as follows:
；（2）
wherein, loss is the target Loss value,value coefficient for the type of target facility, +.>Reconstructing the floor space for the target disaster-stricken facility, < > for>Reconstruction height for target disaster facility, < +.>For the value of the target area, +.>Is the target damage proportion.
Referring to fig. 2, in an embodiment of the present application, the evaluation method further includes:
s010, acquiring a target current image of a target disaster-stricken facility set;
s020, generating a target reconstruction image of the target disaster facility set according to the target current image.
In an embodiment of the present application, the step of generating the target reconstructed image of the target disaster facility set according to the target current image includes:
processing the target current image by using a pre-constructed type recognition model to obtain a target facility type of each target disaster-stricken facility;
generating a target reconstruction unit image of each target disaster-stricken facility according to the target current unit image of the target disaster-stricken facility in the target current image and the target current unit images of other target disaster-stricken facilities with the same type as the target facility of the target disaster-stricken facility;
And generating a target reconstruction image of the target disaster facility set according to all the target reconstruction unit images.
In an embodiment of the present application, the step of generating a target reconstructed unit image of each target disaster recovery facility according to the target current unit image of the target disaster recovery facility in the target current image and the target current unit images of other target disaster recovery facilities having the same type as the target facility of the target disaster recovery facility includes:
processing the target current image by using a pre-constructed scale evaluation model to obtain a target current scale of each target disaster-stricken facility;
screening out a preset number of target reference facilities with the maximum current standard from all target disaster-stricken facilities with the same target facility type;
when the target disaster-stricken facility belongs to the target reference facility, generating a target reconstruction unit image of the target disaster-stricken facility according to a target current unit image of the target disaster-stricken facility in the target current image.
Considering that the damage range of the target reference facility is small, the image of the target reference facility can be directly reconstructed through a deep learning technology. Specifically, a pre-built type recognition model is used for processing a target current unit image to obtain a target facility type, a pre-built image reconstruction model is used for processing the target current unit image and the target facility type to obtain the target reconstruction unit image, and the image reconstruction model is a pre-trained deep learning model and is used for building a complete image according to the type of an object and an input image.
When the target disaster-stricken facility does not belong to the target reference facility, generating a target reconstruction unit image of the target disaster-stricken facility according to a target current unit image of the target disaster-stricken facility in the target current image and a target current unit image of the target reference facility, which is the same as the target facility of the target disaster-stricken facility in type.
For a cluster of facilities within a defined range, considering that the same type of facilities are more similar in appearance, an image of a non-target reference facility with a larger damaged range can be reconstructed in an auxiliary manner through an image of a target reference facility with a smaller damaged range. Specifically, the target current unit image (first image) of the target disaster facility and the target current unit image (second image) of the target reference facility are aligned through feature point matching or image registration technology, good areas corresponding to damaged areas in the first image are identified in the second image through similarity of image features, and the good areas identified in the second image are fused into the first image through image fusion technology to obtain a fused image. The fused image is then reconstructed by a deep learning technique. Specifically, a pre-built type recognition model is used for processing a target current unit image to obtain a target facility type, a pre-built image reconstruction model is used for processing the fusion image and the target facility type to obtain the target reconstruction unit image, and the image reconstruction model is a pre-trained deep learning model and is used for building a complete image according to the type of an object and an input image.
In an embodiment of the present application, the step of determining a target loss value of each of the target disaster recovery facilities according to the target damage proportion, the target facility type, the target reconstruction scale and the target area value includes:
determining a target facility value of each target disaster-stricken facility according to the target facility type, the target reconstruction scale and the target area value;
and determining a target loss value of each target disaster-stricken facility according to the target damage proportion and the target facility value.
In an embodiment of the present application, the evaluation method further includes:
acquiring a sample image of a sample disaster-stricken facility set and a sample type of each sample disaster-stricken facility;
training the initial type recognition model by using the sample image and the sample type to obtain a type recognition model.
In one embodiment of the present application, the type-recognition model is a Convolutional Neural Network (CNN) that includes an input layer, a convolutional layer, a pooling layer, a fully-connected layer, and an output layer. The input layer is used for receiving a sample image, and the size of the image is adjusted according to actual conditions; the convolution layers are used to extract features of the image using convolution kernels, and may include multiple convolution layers, each followed by a batch normalization and activation function; the pooling layer is used for reducing the space dimension of the feature map, and can adopt maximum pooling or average pooling; the full connection layer is used for converting the feature images extracted by the convolution layer and the pooling layer into one-dimensional feature vectors and classifying the feature vectors; the output layer is a fully connected layer with the number of nodes equal to the number of categories, and the output is converted into a probability distribution using a softmax activation function.
The training method of the type recognition model comprises the steps of preprocessing (such as scaling, normalization and the like) an input sample image, selecting a proper optimization algorithm to update the weight of the model, using cross entropy loss as a loss function, inputting training data into the model in batches, training each batch containing a certain number of samples, and repeating batch training until the performance of the model reaches a satisfactory level or reaches a preset iteration number.
In one embodiment of the present application, the scale assessment model includes an input layer, an object detection layer, a scale extraction layer, a monocular ranging layer, and a scale calculation layer. The input layer is used for receiving the sample image and the scale and shooting parameters (including camera calibration parameters) of the known object in the sample image; the object detection layer is used for performing object detection by using a Convolutional Neural Network (CNN), identifying all objects in the image and framing out; the scale extraction layer is used for extracting pixel scale information from the boundary box of the detected object; the monocular distance measuring layer is used for calculating the distance from the camera to the object to be processed based on the scale of the known object and the camera calibration parameters and the extracted pixel scale; the scale calculation layer is used for combining the scale of the known object and the distance from the camera to the object to be processed, calculating the pixel scale of the object to be processed in the image, and further converting the pixel scale to obtain the actual scale.
In an embodiment of the present application, the evaluation method further includes:
acquiring a sample initial image of a sample disaster-stricken facility set and a sample reconstruction image of each sample disaster-stricken facility;
training an initial image reconstruction model by using the sample initial image and the sample reconstruction image to obtain an image reconstruction model.
In an embodiment of the present application, the image reconstruction model is a model based on generating a countermeasure network (GAN), and is composed of a Generator (Generator) and a Discriminator (Discriminator).
The generator is for receiving the corrupted initial image as input and generating a restored reconstructed image as output, and includes an input layer, a convolution layer, an activation layer, an upsampling/transpose convolution layer, and an output layer. Wherein the input layer is used for receiving images; the convolution layer is used for extracting features of the image, and the features are used for understanding the difference between the damaged part and the undamaged part of the image by learning the mode in the image; the activation layer is used for introducing nonlinearity and helping the network learn complex image restoration tasks; the up-sampling/transpose convolution layer is used for amplifying the feature map and gradually constructing higher-resolution image output; the output layer is a convolution layer and is used for converting the output of the network into an image data format to generate a repaired image. The working principle is as follows: the generator attempts to generate a near-true undamaged image by learning the mapping relationship between the damaged image and its corresponding undamaged image. Through the iterative training process, the generator gradually improves its output to more accurately repair the damaged portion in the image.
The discriminator is used for distinguishing the repair image generated by the generator from the real undamaged image, and comprises an input layer, a convolution layer, an activation layer, a full connection layer and an output layer. The input layer is used for receiving two inputs, namely an image generated by the generator and a real undamaged image; the convolution layers are used to extract image features using multiple convolution layers that help the discriminator learn to distinguish differences between the generated image and the real image; the activation layer is used for introducing nonlinearity and enhancing the discrimination capability of the model; the full-connection layer is used for summarizing the characteristics to form a final discrimination result; the output layer outputs a probability value representing the probability that the input image is a true image. The working principle is as follows: the discriminator trains itself by judging whether the input image is a true image or an image produced by the generator. During training, the discriminator is continuously improving its discrimination capability, and the generator attempts to fool the discriminator to generate a more realistic image. This countermeasure process facilitates the learning of the generator, ultimately producing a higher quality repair image.
For the device embodiments, since they are substantially similar to the method embodiments, the description is relatively simple, and reference is made to the description of the method embodiments for relevant points.
Referring to fig. 3, a natural disaster damage assessment device based on deep learning according to an embodiment of the present application is shown, including:
the target information obtaining module 310 is configured to obtain a target current image and a target reconstructed image of a target disaster-stricken facility set, and a target area value of an area to which the target disaster-stricken facility set belongs;
the target type recognition module 320 is configured to process the target current image by using a pre-constructed type recognition model, so as to obtain a target facility type of each target disaster-stricken facility;
a target scale evaluation module 330, configured to process the target current image and the target reconstructed image by using a pre-constructed scale evaluation model, so as to obtain a target current scale and a target reconstructed scale of each target disaster-stricken facility;
a target damage assessment module 340, configured to determine a target damage proportion of each target disaster-stricken facility according to the current target scale and the target reconstruction scale;
a target loss evaluation module 350, configured to determine a target loss value of each target disaster recovery facility according to the target damage proportion, the target facility type, the target reconstruction scale, and the target area value.
Referring to fig. 4, in an embodiment of the present application, the evaluation device further includes:
a target image obtaining module 210, configured to obtain a target current image of a target disaster facility set;
the target image generating module 220 is configured to generate a target reconstructed image of the target disaster facility set according to the target current image.
Referring to FIG. 5, there is shown a computer device of the present application, the computer device 12 being embodied in the form of a general purpose computing device; the computer device 12 comprises: one or more processors or processing units 16, a memory 28, and a bus 18 that connects the various system components, including the memory 28 and the processing unit 16.
Bus 18 may be one or more of several types of bus structures including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor, or a local bus using any of a variety of bus architectures. By way of example, and not limitation, such architectures include Industry Standard Architecture (ISA) bus, micro channel architecture (MAC) bus, enhanced ISA bus, video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnect (PCI) bus.
Computer device 12 typically includes a variety of computer system readable media. Such media can be any available media that is accessible by computer device 12 and includes both volatile and nonvolatile media, removable and non-removable media.
Memory 28 may include computer system readable media in the form of volatile memory, such as random access memory 30 and/or cache memory 32. The computer device 12 may further include other removable/non-removable, volatile/nonvolatile computer storage media. By way of example only, storage system 34 may be used to read from or write to non-removable, nonvolatile magnetic media (commonly referred to as a "hard disk drive"). Although not shown in fig. 5, a magnetic disk drive for reading from and writing to a removable non-volatile magnetic disk (e.g., a "floppy disk"), and an optical disk drive for reading from or writing to a removable non-volatile optical disk such as a CD-ROM, DVD-ROM, or other optical media may be provided. In such cases, each drive may be coupled to bus 18 through one or more data medium interfaces. The memory may include at least one program product having a set (e.g., at least one) of program modules 42, the program modules 42 being configured to carry out the functions of the various embodiments of the application.
A program/utility 40 having a set (at least one) of program modules 42 may be stored, for example, in a memory, such program modules 42 including an operating system, one or more application programs, other program modules 42, and program data, each or some combination of which may include an implementation of a network environment. Program modules 42 generally perform the functions and/or methods in the embodiments described herein.
The computer device 12 may also communicate with one or more external devices 14 (e.g., keyboard, pointing device, display 24, camera, etc.), one or more devices that enable an operator to interact with the computer device 12, and/or any devices (e.g., network card, modem, etc.) that enable the computer device 12 to communicate with one or more other computing devices. Such communication may occur through the I/O interface 22. Moreover, computer device 12 may also communicate with one or more networks such as a Local Area Network (LAN), a Wide Area Network (WAN), and/or a public network such as the Internet, through network adapter 20. As shown in fig. 5, the network adapter 20 communicates with other modules of the computer device 12 via the bus 18. It should be appreciated that although not shown in fig. 5, other hardware and/or software modules may be used in connection with computer device 12, including, but not limited to: microcode, device drivers, redundant processing units 16, external disk drive arrays, RAID systems, tape drives, data backup storage systems 34, and the like.
The processing unit 16 executes various functional applications and data processing by running programs stored in the memory 28, for example, implementing the evaluation method provided in any of the embodiments of the present application.
That is, the processing unit 16 may implement: acquiring a target current image and a target reconstruction image of a target disaster-stricken facility set, and a target area value of an area to which the target disaster-stricken facility set belongs; processing the target current image by using a pre-constructed type recognition model to obtain a target facility type of each target disaster-stricken facility; processing the target current image and the target reconstructed image respectively by using a pre-constructed scale evaluation model to obtain a target current scale and a target reconstructed scale of each target disaster-stricken facility; determining a target damage proportion of each target disaster-stricken facility according to the current target scale and the target reconstruction scale; and determining a target loss value of each target disaster-stricken facility according to the target damage proportion, the target facility type, the target reconstruction scale and the target area value.
In an embodiment of the present application, there is also provided a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the evaluation method provided by any of the embodiments of the present application.
That is, the program, when executed by the processor, may implement: acquiring a target current image and a target reconstruction image of a target disaster-stricken facility set, and a target area value of an area to which the target disaster-stricken facility set belongs; processing the target current image by using a pre-constructed type recognition model to obtain a target facility type of each target disaster-stricken facility; processing the target current image and the target reconstructed image respectively by using a pre-constructed scale evaluation model to obtain a target current scale and a target reconstructed scale of each target disaster-stricken facility; determining a target damage proportion of each target disaster-stricken facility according to the current target scale and the target reconstruction scale; and determining a target loss value of each target disaster-stricken facility according to the target damage proportion, the target facility type, the target reconstruction scale and the target area value.
Any combination of one or more computer readable media may be employed. The computer readable medium may be a computer readable signal medium or a computer readable storage medium. The computer readable storage medium may be an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable storage medium would include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
The computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, either in baseband or as part of a carrier wave. Such a propagated data signal may take any of a variety of forms, including electro-magnetic, optical, or any suitable combination of the preceding. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.
Computer program code for carrying out operations of the present application may be written in one or more programming languages, including an object oriented programming language such as Java, smalltalk, C ++ and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the operator's computer, partly on the operator's computer, as a stand-alone software package, partly on the operator's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the operator computer through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computer (e.g., connected through the internet using an internet service provider). In this specification, each embodiment is described in a progressive manner, and each embodiment is mainly described by differences from other embodiments, and identical and similar parts between the embodiments are all enough to be referred to each other.
While preferred embodiments of the present embodiments have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. It is therefore intended that the following claims be interpreted as including the preferred embodiments and all such alterations and modifications as fall within the scope of the embodiments of the present application.
Finally, it is further noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or terminal that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or terminal. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article or terminal device comprising the element.
The method and the device for evaluating natural disaster damage based on deep learning provided by the application are described in detail, and specific examples are applied to illustrate the principle and the implementation of the application, and the description of the above examples is only used for helping to understand the method and the core idea of the application; meanwhile, as those skilled in the art will have modifications in the specific embodiments and application scope in accordance with the ideas of the present application, the present description should not be construed as limiting the present application in view of the above.
Claims (10)
1. The natural disaster damage assessment method based on deep learning is characterized by comprising the following steps of:
acquiring a target current image and a target reconstruction image of a target disaster-stricken facility set, and a target area value of an area to which the target disaster-stricken facility set belongs;
processing the target current image by using a pre-constructed type recognition model to obtain a target facility type of each target disaster-stricken facility;
processing the target current image and the target reconstructed image respectively by using a pre-constructed scale evaluation model to obtain a target current scale and a target reconstructed scale of each target disaster-stricken facility;
Determining a target damage proportion of each target disaster-stricken facility according to the current target scale and the target reconstruction scale;
and determining a target loss value of each target disaster-stricken facility according to the target damage proportion, the target facility type, the target reconstruction scale and the target area value.
2. The evaluation method according to claim 1, characterized by further comprising:
acquiring a target current image of a target disaster-stricken facility set;
and generating a target reconstruction image of the target disaster facility set according to the target current image.
3. The method of assessing a condition of claim 2, wherein the step of generating a target reconstructed image of the set of target disaster recovery facilities from the target current image includes:
processing the target current image by using a pre-constructed type recognition model to obtain a target facility type of each target disaster-stricken facility;
generating a target reconstruction unit image of each target disaster-stricken facility according to the target current unit image of the target disaster-stricken facility in the target current image and the target current unit images of other target disaster-stricken facilities with the same type as the target facility of the target disaster-stricken facility;
And generating a target reconstruction image of the target disaster facility set according to all the target reconstruction unit images.
4. The evaluation method according to claim 3, wherein the step of generating a target reconstructed unit image for each target disaster recovery facility from a target current unit image of the target disaster recovery facility in the target current image and target current unit images of other target disaster recovery facilities of the same type as the target facilities of the target disaster recovery facility, comprises:
processing the target current image by using a pre-constructed scale evaluation model to obtain a target current scale of each target disaster-stricken facility;
screening out a preset number of target reference facilities with the maximum current standard from all target disaster-stricken facilities with the same target facility type;
when the target disaster-stricken facility belongs to the target reference facility, generating a target reconstruction unit image of the target disaster-stricken facility according to a target current unit image of the target disaster-stricken facility in the target current image;
when the target disaster-stricken facility does not belong to the target reference facility, generating a target reconstruction unit image of the target disaster-stricken facility according to a target current unit image of the target disaster-stricken facility in the target current image and a target current unit image of the target reference facility, which is the same as the target facility of the target disaster-stricken facility in type.
5. The method of evaluating according to claim 1, wherein the step of determining a target loss value for each of the target disaster recovery facilities in accordance with the target damage proportion, the target facility type, the target reconstruction scale, and the target area value comprises:
determining a target facility value of each target disaster-stricken facility according to the target facility type, the target reconstruction scale and the target area value;
and determining a target loss value of each target disaster-stricken facility according to the target damage proportion and the target facility value.
6. The evaluation method according to claim 1, characterized by further comprising:
acquiring a sample image of a sample disaster-stricken facility set and a sample type of each sample disaster-stricken facility;
training the initial type recognition model by using the sample image and the sample type to obtain a type recognition model.
7. The method of evaluation of claim 6, wherein the initial type recognition model is a convolutional neural network.
8. A natural disaster damage assessment device based on deep learning, characterized by comprising:
the target information acquisition module is used for acquiring a target current image and a target reconstruction image of a target disaster-stricken facility set and a target area value of an area to which the target disaster-stricken facility set belongs;
The target type recognition module is used for processing the target current image by using a pre-constructed type recognition model to obtain a target facility type of each target disaster-stricken facility;
the target scale evaluation module is used for respectively processing the target current image and the target reconstruction image by using a pre-constructed scale evaluation model to obtain a target current scale and a target reconstruction scale of each target disaster-stricken facility;
the target damage evaluation module is used for determining the target damage proportion of each target disaster-stricken facility according to the current target scale and the target reconstruction scale;
and the target loss evaluation module is used for determining a target loss value of each target disaster-stricken facility according to the target damage proportion, the target facility type, the target reconstruction scale and the target area value.
9. A computer device comprising a processor, a memory and a computer program stored on the memory and capable of running on the processor, which computer program, when executed by the processor, implements the method according to any of claims 1-7.
10. A computer readable storage medium, characterized in that the computer readable storage medium has stored thereon a computer program which, when executed by a processor, implements the method according to any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410268719.7A CN117876362B (en) | 2024-03-11 | 2024-03-11 | Deep learning-based natural disaster damage assessment method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410268719.7A CN117876362B (en) | 2024-03-11 | 2024-03-11 | Deep learning-based natural disaster damage assessment method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117876362A true CN117876362A (en) | 2024-04-12 |
| CN117876362B CN117876362B (en) | 2024-05-28 |
Family
ID=90595152
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410268719.7A Active CN117876362B (en) | 2024-03-11 | 2024-03-11 | Deep learning-based natural disaster damage assessment method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117876362B (en) |
Citations (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103455708A (en) * | 2013-07-24 | 2013-12-18 | 安徽省电力科学研究院 | Power transmission line disaster monitoring and risk assessment platform based on satellite and weather information |
| CN103927389A (en) * | 2014-04-30 | 2014-07-16 | 北京中有联科技有限公司 | Method for establishing flood disaster geographical analysis and evaluation dynamic model |
| CN109635441A (en) * | 2018-12-13 | 2019-04-16 | 青岛理工大学 | A kind of groups of building earthquake disaster simulation visualization system and method based on BIM |
| CN110222935A (en) * | 2019-05-08 | 2019-09-10 | 深圳中大环保科技创新工程中心有限公司 | Natural calamity loss appraisal method and apparatus |
| CN110321774A (en) * | 2019-04-04 | 2019-10-11 | 平安科技（深圳）有限公司 | Crops evaluation methods for disaster condition, device, equipment and computer readable storage medium |
| CN110969183A (en) * | 2019-09-20 | 2020-04-07 | 北京方位捷讯科技有限公司 | Method and system for determining damage degree of target object according to image data |
| CN113344327A (en) * | 2021-05-06 | 2021-09-03 | 河南省科达水利勘测设计有限公司 | Flood control influence evaluation system |
| CN113657732A (en) * | 2021-07-30 | 2021-11-16 | 哈尔滨工业大学 | Disaster degree index calculation method and equipment based on analytic hierarchy process |
| CN114091897A (en) * | 2021-11-19 | 2022-02-25 | 云南电网有限责任公司电力科学研究院 | Neural network-based building damage assessment method after natural disaster |
| CN114936730A (en) * | 2022-03-16 | 2022-08-23 | 兰州大方电子有限责任公司 | Disaster damage assessment system for disaster-affected area |
| CN114971872A (en) * | 2022-06-10 | 2022-08-30 | 中国银行股份有限公司 | Disaster-suffering loan assessment method, system, electronic device and storage medium |
| CN115311169A (en) * | 2022-08-29 | 2022-11-08 | 上海大学 | Damaged image repairing method and device, electronic equipment and storage medium |
| CN115420688A (en) * | 2022-09-13 | 2022-12-02 | 安徽联智创新软件有限公司 | Agricultural disaster information remote sensing extraction loss evaluation method based on Internet of things |
| CN115700645A (en) * | 2022-09-15 | 2023-02-07 | 北京师范大学 | Method for evaluating indirect damage of disaster |
| KR102530929B1 (en) * | 2022-08-10 | 2023-05-10 | 주식회사 언더라이터 | Method and apparatus for calculating house fire insurance using machine learning and software |
| CN116205799A (en) * | 2022-12-21 | 2023-06-02 | 深圳大学 | Image restoration method, device and computer readable storage medium |
| CN116228521A (en) * | 2022-12-20 | 2023-06-06 | 中国船舶集团有限公司系统工程研究院 | Heterologous image registration and fusion method |
| CN116645616A (en) * | 2023-04-20 | 2023-08-25 | 武汉大学 | Method for evaluating damage degree of building in natural disasters by using remote sensing images |
-
2024
- 2024-03-11 CN CN202410268719.7A patent/CN117876362B/en active Active
Patent Citations (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103455708A (en) * | 2013-07-24 | 2013-12-18 | 安徽省电力科学研究院 | Power transmission line disaster monitoring and risk assessment platform based on satellite and weather information |
| CN103927389A (en) * | 2014-04-30 | 2014-07-16 | 北京中有联科技有限公司 | Method for establishing flood disaster geographical analysis and evaluation dynamic model |
| CN109635441A (en) * | 2018-12-13 | 2019-04-16 | 青岛理工大学 | A kind of groups of building earthquake disaster simulation visualization system and method based on BIM |
| CN110321774A (en) * | 2019-04-04 | 2019-10-11 | 平安科技（深圳）有限公司 | Crops evaluation methods for disaster condition, device, equipment and computer readable storage medium |
| CN110222935A (en) * | 2019-05-08 | 2019-09-10 | 深圳中大环保科技创新工程中心有限公司 | Natural calamity loss appraisal method and apparatus |
| CN110969183A (en) * | 2019-09-20 | 2020-04-07 | 北京方位捷讯科技有限公司 | Method and system for determining damage degree of target object according to image data |
| CN113344327A (en) * | 2021-05-06 | 2021-09-03 | 河南省科达水利勘测设计有限公司 | Flood control influence evaluation system |
| CN113657732A (en) * | 2021-07-30 | 2021-11-16 | 哈尔滨工业大学 | Disaster degree index calculation method and equipment based on analytic hierarchy process |
| CN114091897A (en) * | 2021-11-19 | 2022-02-25 | 云南电网有限责任公司电力科学研究院 | Neural network-based building damage assessment method after natural disaster |
| CN114936730A (en) * | 2022-03-16 | 2022-08-23 | 兰州大方电子有限责任公司 | Disaster damage assessment system for disaster-affected area |
| CN114971872A (en) * | 2022-06-10 | 2022-08-30 | 中国银行股份有限公司 | Disaster-suffering loan assessment method, system, electronic device and storage medium |
| KR102530929B1 (en) * | 2022-08-10 | 2023-05-10 | 주식회사 언더라이터 | Method and apparatus for calculating house fire insurance using machine learning and software |
| CN115311169A (en) * | 2022-08-29 | 2022-11-08 | 上海大学 | Damaged image repairing method and device, electronic equipment and storage medium |
| CN115420688A (en) * | 2022-09-13 | 2022-12-02 | 安徽联智创新软件有限公司 | Agricultural disaster information remote sensing extraction loss evaluation method based on Internet of things |
| CN115700645A (en) * | 2022-09-15 | 2023-02-07 | 北京师范大学 | Method for evaluating indirect damage of disaster |
| CN116228521A (en) * | 2022-12-20 | 2023-06-06 | 中国船舶集团有限公司系统工程研究院 | Heterologous image registration and fusion method |
| CN116205799A (en) * | 2022-12-21 | 2023-06-02 | 深圳大学 | Image restoration method, device and computer readable storage medium |
| CN116645616A (en) * | 2023-04-20 | 2023-08-25 | 武汉大学 | Method for evaluating damage degree of building in natural disasters by using remote sensing images |
Non-Patent Citations (2)
| Title |
|---|
| 洪惠群;林金发;: "灾情图中建筑物识别算法的研究", 绵阳师范学院学报, no. 08, 15 August 2018 (2018-08-15) * |
| 陈子翰;窦爱霞;王晓青;吴玮莹;王书民;: "基于高分辨率影像的九寨沟7.0级地震道路震害评估", 中国地震, no. 04, 15 December 2017 (2017-12-15) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117876362B (en) | 2024-05-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111401201B (en) | Aerial image multi-scale target detection method based on spatial pyramid attention drive | |
| CN107784288B (en) | Iterative positioning type face detection method based on deep neural network | |
| CN111523414A (en) | Face recognition method and device, computer equipment and storage medium | |
| CN112861575A (en) | Pedestrian structuring method, device, equipment and storage medium | |
| CN112419202B (en) | Automatic wild animal image recognition system based on big data and deep learning | |
| CN116579616B (en) | Risk identification method based on deep learning | |
| CN112801227B (en) | Typhoon identification model generation method, device, equipment and storage medium | |
| CN115457492A (en) | Target detection method and device, computer equipment and storage medium | |
| CN114519819B (en) | Remote sensing image target detection method based on global context awareness | |
| CN115937546A (en) | Image matching method, three-dimensional image reconstruction method, image matching device, three-dimensional image reconstruction device, electronic apparatus, and medium | |
| CN116189139A (en) | Traffic sign detection method based on Transformer | |
| CN114694130A (en) | Method and device for detecting telegraph poles and pole numbers along railway based on deep learning | |
| CN114202473A (en) | Image restoration method and device based on multi-scale features and attention mechanism | |
| CN113393385A (en) | Unsupervised rain removal method, system, device and medium based on multi-scale fusion | |
| CN117876362B (en) | Deep learning-based natural disaster damage assessment method and device | |
| CN116052097A (en) | Map element detection method and device, electronic equipment and storage medium | |
| CN116415020A (en) | Image retrieval method, device, electronic equipment and storage medium | |
| CN114332715A (en) | Method, device and equipment for identifying snow through automatic meteorological observation and storage medium | |
| CN114155524A (en) | Single-stage 3D point cloud target detection method and device, computer equipment and medium | |
| CN113191364A (en) | Vehicle appearance component identification method, device, electronic device and medium | |
| CN117173385B (en) | Detection method, device, medium and equipment of transformer substation | |
| CN116805337B (en) | Crowd positioning method based on trans-scale visual transformation network | |
| CN113657214B (en) | Building damage assessment method based on Mask RCNN | |
| CN112396648B (en) | Target identification method and system capable of positioning mass center of target object | |
| CN118154856A (en) | Target detection method, device and related equipment |
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 |