CN112686363A - Natural disaster assessment method based on three-dimensional neural network - Google Patents
Natural disaster assessment method based on three-dimensional neural network Download PDFInfo
- Publication number
- CN112686363A CN112686363A CN202011593053.0A CN202011593053A CN112686363A CN 112686363 A CN112686363 A CN 112686363A CN 202011593053 A CN202011593053 A CN 202011593053A CN 112686363 A CN112686363 A CN 112686363A
- Authority
- CN
- China
- Prior art keywords
- neural network
- dimensional neural
- network model
- natural disaster
- data set
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 20
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 14
- 238000003062 neural network model Methods 0.000 claims abstract description 32
- 230000006378 damage Effects 0.000 claims abstract description 19
- 238000006424 Flood reaction Methods 0.000 claims description 4
- 206010037844 rash Diseases 0.000 claims description 4
- 238000012549 training Methods 0.000 claims description 4
- 238000011156 evaluation Methods 0.000 description 7
- 230000004927 fusion Effects 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 241000282414 Homo sapiens Species 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000004083 survival effect Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a natural disaster assessment method based on a three-dimensional neural network, which relates to the technical field of natural disaster assessment and comprises the following steps: acquiring source picture data in advance and constructing a data set; constructing a three-dimensional neural network model based on the acquired data set and using the three-dimensional neural network model as a prediction model; and taking the target picture as an input based on the three-dimensional neural network model, and acquiring a grade output result of the damage of the object as a prediction result. According to the invention, the damage grade can be automatically evaluated by collecting pictures before and after the building and crops encounter natural disasters, so that the rescue timeliness is improved, the labor cost is reduced, and the income of related departments is indirectly improved.
Description
Technical Field
The invention relates to the technical field of natural disaster assessment, in particular to a natural disaster assessment method based on a three-dimensional neural network.
Background
At present, natural disasters such as earthquakes, volcanic eruption, floods and the like are common natural disasters, and can cause serious damage to buildings, crops and the like.
Natural disasters bring a lot of harm to production and survival of human beings all the time, once the natural disasters are hidden, irreparable loss is easily caused, so that the disaster early warning has epoch-crossing significance to human life; at present, image processing technology is generally fused for early warning research of natural disasters, image fusion becomes an important and useful new technology in the field of image understanding and computer vision, and multi-source remote sensing image data fusion also becomes a research hotspot in the field of remote sensing, and the aim is to intelligently synthesize image data from multiple information sources to generate more accurate and reliable description and judgment compared with single sensor data, so that a fused image is more consistent with visual characteristics of people and machines, and further image understanding and analysis such as target detection and identification are facilitated. Therefore, the damage level of things after the natural disaster occurs cannot be rapidly evaluated, and the timeliness of rescue is reduced.
The retrieval patent CN108182678A discloses a natural disaster monitoring and early warning system with accurate early warning, which realizes accurate evaluation of the fusion effect of remote sensing images, the subjective evaluation value has the advantages of simplicity and intuition, obvious image information can be evaluated quickly and conveniently, the objective evaluation value can avoid the subjective defects of personnel, the images are evaluated objectively, and the comprehensive evaluation value combines the advantages of the subjective evaluation and the objective evaluation, thereby being beneficial to realizing accurate evaluation of the fusion effect, accurately determining whether natural disasters occur in a monitoring scene and carrying out early warning processing; however, the early warning system uses an image fusion technology to perform early warning on disaster situations which are already or are about to occur, for example, the flood level is close to a threshold value, and early warning is started.
An effective solution to the problems in the related art has not been proposed yet.
Disclosure of Invention
Aiming at the problems in the related art, the invention provides a natural disaster assessment method based on a three-dimensional neural network, and the damage grade can be automatically assessed by collecting pictures before and after a building and crops encounter a natural disaster after the natural disaster occurs, so that the rescue timeliness is improved, the labor cost is reduced, the income of related departments is indirectly improved, and the technical problems in the prior related art are solved.
The technical scheme of the invention is realized as follows:
a natural disaster assessment method based on a three-dimensional neural network comprises the following steps:
step S1, acquiring source picture data in advance and constructing a data set;
step S2, constructing a three-dimensional neural network model based on the acquired data set and using the three-dimensional neural network model as a prediction model;
and step S3, taking the target picture as input based on the three-dimensional neural network model, and acquiring a grade output result of the damage of the object as a prediction result.
Further, the step of obtaining the source picture data in advance and constructing the data set includes the following steps:
the method comprises the steps of collecting pictures of objects before and after natural disasters occur in advance and marking damage levels, wherein the pictures of each object before and after the natural disasters occur are a recording unit;
and repeatedly acquiring pictures of objects before and after the natural disaster, and constructing a data set.
Further, the natural disasters include earthquakes, volcanic eruptions, and floods, and the things include buildings and crops.
Further, the step of generating the three-dimensional neural network model further includes the steps of:
and training the three-dimensional neural network model by using the acquired source picture data set.
The invention has the beneficial effects that:
the natural disaster assessment method based on the three-dimensional neural network is realized by acquiring source picture data and constructing a data set in advance, constructing a three-dimensional neural network model based on the acquired data set and using the three-dimensional neural network model as a prediction model, taking a target picture as input based on the three-dimensional neural network model, acquiring a grade output result of damage of an object and using the grade output result as a prediction result, so that the damage grade can be automatically assessed by acquiring pictures before and after the natural disaster occurs to buildings and crops, the timeliness of rescue is improved, the labor cost is reduced, and the income of related departments is indirectly improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without creative efforts.
Fig. 1 is a schematic flowchart of a natural disaster assessment method based on a three-dimensional neural network according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments that can be derived by one of ordinary skill in the art from the embodiments given herein are intended to be within the scope of the present invention.
According to the embodiment of the invention, a natural disaster assessment method based on a three-dimensional neural network is provided.
As shown in fig. 1, a natural disaster assessment method based on a three-dimensional neural network according to an embodiment of the present invention includes the following steps:
step S1, acquiring source picture data in advance and constructing a data set;
step S2, constructing a three-dimensional neural network model based on the acquired data set and using the three-dimensional neural network model as a prediction model;
and step S3, taking the target picture as input based on the three-dimensional neural network model, and acquiring a grade output result of the damage of the object as a prediction result.
The method for pre-acquiring the source picture data and constructing the data set comprises the following steps:
the method comprises the steps of collecting pictures of objects before and after natural disasters occur in advance and marking damage levels, wherein the pictures of each object before and after the natural disasters occur are a recording unit;
and repeatedly acquiring pictures of objects before and after the natural disaster, and constructing a data set.
Wherein the natural disasters include earthquakes, volcanic eruptions, and floods, and the things include buildings and crops.
Wherein, the three-dimensional neural network model further comprises the following steps:
and training the three-dimensional neural network model by using the acquired source picture data set.
By means of the technical scheme, the method and the device have the advantages that source picture data are obtained in advance and a data set is built, the three-dimensional neural network model is built on the basis of the obtained data set and serves as a prediction model, the target picture serves as input on the basis of the three-dimensional neural network model, the grade output result of the object damaged is obtained and serves as the prediction result, the purpose that after a natural disaster occurs, the damage grade can be automatically evaluated by collecting pictures of buildings and crops before and after the natural disaster occurs is achieved, the rescue timeliness is improved, the labor cost is reduced, and the income of relevant departments is indirectly improved.
Specifically, the three-dimensional neural network model is as follows:
note: in the table, the size of batch _ size and class _ num are classified
In addition, the neural network model is trained according to the constructed data set until the loss of the neural network model is stabilized at a smaller value, which indicates that the neural network model has converged. And obtaining the trained neural network model. When a certain natural disaster occurs again, pictures before and after the natural disaster occurs to a certain object (building and crops) are collected and input into the trained neural network model, so that the damage level of the object can be predicted.
In addition, no related technology for predicting damage levels of things (buildings and crops) after natural disasters occur by utilizing a computer exists at present, and after the scheme is provided, the damage levels can be rapidly predicted only by collecting pictures before and after the natural disasters occur, so that the effectiveness is improved. The invention can automatically predict the bad grade for the computer, and reduces the labor cost. After the model provided by the invention is trained, the model can be used anytime and anywhere, one-time training is realized, and the model is convenient to use everywhere. After the method is implemented, the labor cost is reduced, and the income of related departments is indirectly improved.
In summary, according to the technical scheme of the invention, by acquiring the source picture data and constructing the data set in advance, constructing the three-dimensional neural network model based on the acquired data set and using the three-dimensional neural network model as the prediction model, and acquiring the grade output result of the object damaged based on the target picture as the input of the three-dimensional neural network model and using the grade output result as the prediction result, the purpose that the damage grade can be automatically evaluated by acquiring pictures before and after the natural disaster occurs to the building and the crops is achieved, so that the timeliness of rescue is improved, the labor cost is reduced, and the income of relevant departments is indirectly improved.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (4)
1. A natural disaster assessment method based on a three-dimensional neural network is characterized by comprising the following steps:
acquiring source picture data in advance and constructing a data set;
constructing a three-dimensional neural network model based on the acquired data set and using the three-dimensional neural network model as a prediction model;
and taking the target picture as an input based on the three-dimensional neural network model, and acquiring a grade output result of the damage of the object as a prediction result.
2. The natural disaster assessment method based on the three-dimensional neural network as claimed in claim 1, wherein the step of pre-acquiring the source picture data and constructing the data set comprises the following steps:
the method comprises the steps of collecting pictures of objects before and after natural disasters occur in advance and marking damage levels, wherein the pictures of each object before and after the natural disasters occur are a recording unit;
and repeatedly acquiring pictures of objects before and after the natural disaster, and constructing a data set.
3. The three-dimensional neural network-based natural disaster assessment method according to claim 2, wherein said natural disasters include earthquakes, volcanic eruptions and floods, and said things include buildings and crops.
4. The natural disaster assessment method based on three-dimensional neural network as claimed in claim 1, wherein the step of three-dimensional neural network model further comprises the steps of:
and training the three-dimensional neural network model by using the acquired source picture data set.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011593053.0A CN112686363A (en) | 2020-12-29 | 2020-12-29 | Natural disaster assessment method based on three-dimensional neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011593053.0A CN112686363A (en) | 2020-12-29 | 2020-12-29 | Natural disaster assessment method based on three-dimensional neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112686363A true CN112686363A (en) | 2021-04-20 |
Family
ID=75454964
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011593053.0A Pending CN112686363A (en) | 2020-12-29 | 2020-12-29 | Natural disaster assessment method based on three-dimensional neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112686363A (en) |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112001565A (en) * | 2020-09-08 | 2020-11-27 | 清华大学合肥公共安全研究院 | Earthquake disaster loss prediction and evaluation method and system based on Softmax regression model |
-
2020
- 2020-12-29 CN CN202011593053.0A patent/CN112686363A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112001565A (en) * | 2020-09-08 | 2020-11-27 | 清华大学合肥公共安全研究院 | Earthquake disaster loss prediction and evaluation method and system based on Softmax regression model |
Non-Patent Citations (1)
| Title |
|---|
| S.L.ULLO等: "LANDSLIDE GEOHAZARD ASSESSMENT WITH CONVOLUTIONAL NEURAL NETWORKS USING SENTINEL-2 IMAGERY DATA", 《IGARSS 2019 - 2019 IEEE INTERNATIONAL GEOSCIENCE AND REMOTE SENSING SYMPOSIUM》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111462451B (en) | Straw burning detection alarm system based on video information | |
| CN110659397B (en) | Behavior detection method and device, electronic equipment and storage medium | |
| CN101383000B (en) | Information processing apparatus, and information processing method | |
| CN111626188B (en) | Indoor uncontrollable open fire monitoring method and system | |
| CN109743547A (en) | A kind of artificial intelligence security monitoring management system | |
| KR102149832B1 (en) | Automated Violence Detecting System based on Deep Learning | |
| JP4999794B2 (en) | Still region detection method and apparatus, program and recording medium | |
| US20200349360A1 (en) | Population density determination from multi-camera sourced imagery | |
| CN109117771B (en) | System and method for detecting violence events in image based on anchor nodes | |
| CN109873979A (en) | Camera-based static image difference comparison method and device | |
| KR102511287B1 (en) | Image-based pose estimation and action detection method and appratus | |
| CN114120171A (en) | Fire smoke detection method, device and equipment based on video frame and storage medium | |
| CN110740298A (en) | Distributed classroom discipline behavior detection system, method and medium | |
| CN113139530B (en) | Method and device for detecting sleep post behavior and electronic equipment thereof | |
| JP2021007055A (en) | Discriminator learning device, discriminator learning method, and computer program | |
| CN112686363A (en) | Natural disaster assessment method based on three-dimensional neural network | |
| KR102521725B1 (en) | Fire detection system based on artificial intelligence, fire detection device and method thereof | |
| CN214151721U (en) | Recognition all-in-one machine system based on face algorithm | |
| CN112926542B (en) | Sex detection method and device, electronic equipment and storage medium | |
| CN115116136A (en) | Abnormal behavior detection method, device and medium | |
| CN206452471U (en) | Visual identifying system | |
| CN112241671A (en) | Personnel identity identification method, device and system | |
| CN117994808A (en) | Method, device, server, medium and program product for detecting smoking behavior | |
| Nawfal et al. | Performance Evaluation Between Tiny Yolov3 and MobileNet SSDv1 for Object Detection | |
| WO2023281897A1 (en) | Video surveillance system and video surveillance method |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210420 |