CN-122002003-A - Urban disaster prevention and reduction intelligent early warning method based on machine vision

CN122002003ACN 122002003 ACN122002003 ACN 122002003ACN-122002003-A

Abstract

The invention provides an intelligent early warning method for urban disaster prevention and reduction based on machine vision, and relates to the field of urban disaster prevention early warning systems. The intelligent early warning method for urban disaster prevention and reduction based on machine vision is controlled by a main control center, wherein the main control center comprises a vision monitoring module, a dispatching center module, a data center module and an early warning center module, the main control center is used for a central processing center for the whole urban disaster prevention and reduction, the vision monitoring center is used for controlling monitoring camera shooting required by the whole urban disaster prevention and reduction and processing images, and the data center is used for intelligent analysis and data recording of data acquired by vision monitoring. The invention presents the monitoring result in visual images and videos, so that the manager and emergency rescue personnel can more intuitively know the situation of the disaster scene, including the position, the range, the severity and the like of the disaster, and more comprehensive and more accurate disaster monitoring and early warning are realized.

Inventors

Wu ningbo
Han Fahai
YANG XUESONG
Mao Jianxiu
CHEN YONGZHU
LUO CHENGYUAN
LUO XU
LIU YU
ZHU YANBING
LUO HUAN
CHANG JUANJUAN
Jin Qingfu
Luo Yuncan
Huang Xiuyin

Assignees

贵州联建土木工程质量检测监控中心有限公司

Dates

Publication Date: 20260508
Application Date: 20241231

Claims (8)

1. The intelligent early warning system for urban disaster prevention and reduction based on machine vision is characterized in that the intelligent early warning system for urban disaster prevention and reduction based on machine vision is controlled by a general control center, the general control center comprises a vision monitoring module, a dispatching center module, a data center module and an early warning center module, the general control center is used for a central processing center for whole urban disaster prevention and reduction, the vision monitoring center is used for controlling monitoring camera shooting and processing images required by whole urban disaster prevention and reduction, the data center is used for intelligently analyzing and recording data acquired by vision monitoring, the dispatching center is used for fast early warning and processing urban disaster reduction, and the early warning center is used for summarizing and executing data analyzed by the data center.
2. The machine vision-based urban disaster prevention and reduction intelligent early warning system is characterized in that the vision monitoring center comprises a data acquisition module and an image preprocessing module, the data acquisition module comprises camera deployment, and the image preprocessing module comprises an image graying function, an image noise reduction function and an image geometry correction function.
3. The machine vision-based urban disaster prevention and reduction intelligent early warning system is characterized in that a data analysis module of the data center module comprises a feature extraction function and a model optimization function, the feature extraction function comprises a target detection function, a behavior analysis function and a change detection function, the target detection function detects and identifies specific targets in images by utilizing a neural network, the behavior analysis function carries out movement analysis prediction on moving objects in image adaptation, judges whether abnormal behaviors exist or not, the change detection function compares images at different times, detects change areas in a scene, and timely discovers newly-appearing disaster hidden dangers.
4. The intelligent early warning system for urban disaster prevention and reduction based on machine vision according to claim 1, wherein the model optimization function of the data center module comprises a data set construction function, a deep learning function and a model optimization function, the data set construction function collects a large amount of image data comprising various disaster scenes and normal scenes, the deep learning function selects a YOLO deep learning model to conduct intelligent training of the data model, and the model optimization function adopts a model compression quantization technology to optimize the trained model.
5. The machine vision-based intelligent early warning system for urban disaster prevention and reduction, as set forth in claim 1, wherein the data recording function of the data center module comprises a local data storage function and a blockchain cloud storage function, and the blockchain cloud storage function uses a plurality of servers to store synchronous backup storage in different places.
6. The machine vision-based urban disaster prevention and reduction intelligent early warning system is characterized in that the early warning center comprises an early warning analysis function and an early warning resource scheduling function, the early warning analysis function carries out scheme generation on disasters through data analysis results generated by the data center, the scheme generation comprises continuous state monitoring, disaster risk assessment and disaster early warning decision scheme generation on the disasters, the early warning resource scheduling function fuses visual disasters with sensor monitoring data of other departments through cross-department cooperation, information interconnection and intercommunication among the departments are achieved, and the disaster prevention and reduction early warning accuracy is improved through multidimensional data information.
7. The machine vision-based urban disaster prevention and reduction intelligent early warning system is characterized in that the dispatching center issues and executes a disaster early warning decision scheme specific implementation scheme generated by the early warning center, and the dispatching center shares data information and the decision scheme to an emergency management center, a fire control center and an emergency treatment center in real time, so that information interconnection and intercommunication among departments are realized, and support is provided for collaborative emergency response.
8. The intelligent early warning method for urban disaster prevention and reduction based on machine vision is characterized by being realized by an intelligent early warning system for urban disaster prevention and reduction based on machine vision, and comprises the following steps: S1, data acquisition, namely installing machine vision equipment such as a high-definition camera, an infrared thermal imager and the like in a critical area and a disaster-prone place of a city, ensuring the omnibearing coverage of a key area, and carrying a camera on an unmanned aerial vehicle to carry out regular or irregular inspection; S2, image data processing, namely performing image denoising, geometric correction and image graying on the image to greatly compress the image, so as to realize the effect of high-speed transmission of mass image data; S3, target detection and identification, namely performing quick feature identification on the processed image by using a convolutional neural network deep learning algorithm, determining the category of a target object which is likely to have disasters, accurately determining the position and the range of the target in the image by target positioning, and marking the disaster target; S4, after the early warning center receives the data analysis, carrying out continuous state detection and risk assessment on the abnormal data, and according to the results of disaster analysis and assessment, sending out early warning signals in time when the disaster risk reaches a preset threshold value, and generating disaster prevention and reduction decisions; And S5, carrying out emergency response by the dispatching center, sharing disaster information acquired by the machine vision system to related units such as a disaster prevention command center, a fire department, a traffic management department and the like of the city in real time, realizing information interconnection and intercommunication among the departments, and providing support for collaborative emergency response.

Description

Urban disaster prevention and reduction intelligent early warning method based on machine vision Technical Field The invention relates to the field of urban disaster prevention early warning systems, in particular to an intelligent urban disaster prevention and reduction early warning method based on machine vision. Background Along with the acceleration of the urban process, urban population is highly dense, buildings and infrastructure are increasingly complex, the city faces various disaster threats, once disasters occur, huge casualties and property losses are often caused, serious challenges are formed for the safety and stability of the city, the traditional urban disaster prevention early warning system mainly depends on modes of manual inspection, sensor monitoring and the like, the problems of limited monitoring range, low instantaneity, low accuracy, false alarm omission and the like exist, the traditional early warning system is insufficient in information transmission and visualization, disaster information cannot be timely and intuitively transmitted to related personnel and departments, and the efficiency and effect of emergency response are affected. Therefore, the invention provides an intelligent early warning method for urban disaster prevention and reduction based on machine vision, which effectively solves the trouble and the difficult problem by a visual monitoring mode. Disclosure of Invention Aiming at the defects of the prior art, the invention provides an intelligent early warning method for urban disaster prevention and reduction based on machine vision, which can monitor under the condition of not directly contacting a target object through a camera device, avoid the interference and damage to a monitored object, monitor roads, bridges, tunnels and the like in real time on the premise of not influencing normal traffic and urban operation, timely find potential safety hazards, quickly adjust the monitoring range according to the number of monitoring cameras and unmanned aerial vehicles, realize the coverage monitoring of a large area by installing the cameras in each key area and high points of the city, and realize the omnibearing dead-angle-free key monitoring of key areas in some cities by combining with unmanned aerial vehicle mobile monitoring equipment, thereby improving the accuracy and timeliness of visual monitoring early warning. Technical proposal The intelligent early warning system for urban disaster prevention and reduction based on machine vision is controlled by a total control center, wherein the total control center comprises a vision monitoring module, a dispatching center module, a data center module and an early warning center module, the total control center is used for a central processing center for whole urban disaster prevention and reduction, the vision monitoring center is used for controlling monitoring camera shooting and processing images required by whole urban disaster prevention and reduction, the data center is used for intelligent analysis and data recording of data acquired by vision monitoring, the dispatching center is used for fast early warning and processing of urban disaster reduction, and the early warning center is used for summarizing and executing data analyzed by the data center; Through the technical scheme, the high-definition cameras are installed in the key areas of the city, such as places with large flow of people and easy occurrence of disasters, such as high-rise buildings, bridges, subway stations and railway stations, video image data are continuously collected for 24 hours, a basis is provided for subsequent analysis, and the unmanned aerial vehicle carries the high-definition cameras to carry out regular or irregular aerial inspection, especially in the areas where natural disasters possibly occur, such as mountain areas, near rivers and the like, so that more comprehensive image information is timely obtained. Preferably, the visual monitoring center comprises a data acquisition module and an image preprocessing module, wherein the data acquisition module comprises camera deployment, and the image preprocessing module comprises an image graying function, an image noise reduction function and an image geometric correction function; Through the technical scheme, the acquired color image can be converted into the gray image by graying, the data volume is reduced, the subsequent processing speed is accelerated, meanwhile, the contour and texture information of the image are highlighted, the image is noise-reduced, the noise in the image is removed by using a median filtering algorithm, the definition and quality of the image are improved, the interference of the noise on the subsequent analysis is avoided, the image deformation caused by factors such as the shooting angle and the position of a camera is corrected by adopting a Gaussian algorithm in the geometric correction of the image, and the accuracy and the consistency of