Urban inland inundation prevention and control task generation method based on machine vision

Abstract

The invention discloses an urban inland inundation prevention and control task generation method based on machine vision, and relates to the technical field of big data mining, and the method comprises the steps: evaluating the rainfall index of each region of a city according to a meteorological factor and a topographic structure; initializing the drainage capacity of each region of the city; collecting real-time image data of a drainage system in each region in the city to identify a drainage port state of each region, and evaluating a drainage blockage index of each region in the city; correcting the drainage capacity of each region of the initialized city by using the drainage blockage index of each region of the city to obtain the real-time drainage capacity of each region of the city; based on the real-time drainage capacity of each region of the city and the rainfall index of each region of the city, analyzing a waterlogging risk index of each region of the city; generating an urban inland inundation prevention and control task based on the inland inundation risk level corresponding to the inland inundation risk index of each region of the city; the method has the advantages that the management efficiency of an urban drainage system is improved, and the urban inland inundation prevention and treatment efficiency is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of big data mining, and in particular to a method for generating urban waterlogging prevention and control tasks based on machine vision. Background Art

[0002] Urban waterlogging prevention and control refers to reducing or eliminating urban waterlogging through a series of engineering, management and technical means, ensuring the normal operation of urban drainage systems, and preventing the impact of floods caused by heavy rains or extreme weather on urban life, economic activities, public safety and other aspects.

[0003] The existing urban surface is mainly composed of asphalt or cement roads. When it rains, rainwater cannot penetrate into the ground quickly. In addition, urban garbage drifts with the accumulated water, causing blockage in urban drainage outlets, making areas with low terrain prone to floods. The mainstream prevention and control method still relies on manual mobile inspections, but the efficiency is low. Summary of the invention

[0004] In order to solve the above technical problems, a method for generating urban waterlogging prevention and control tasks based on machine vision is provided. This technical solution solves the problem that the existing urban surfaces mentioned above are mainly asphalt or cement roads. When such roads are faced with rainfall, rainwater cannot quickly penetrate into the ground, and because urban garbage drifts with the accumulated water, the city drainage outlets are blocked, making areas with low terrain prone to floods and waterlogging. The mainstream prevention and control method still relies on manual flow inspections, but the efficiency is low.

[0005] In order to achieve the above purpose, the technical solution adopted by the present invention is: A method for generating urban waterlogging prevention tasks based on machine vision, comprising: Obtain meteorological data and urban topography, and evaluate precipitation indicators in various areas of the city based on meteorological factors and topographic structure; Initialize the drainage capacity of each area of ​​the city based on the structural parameters of the urban drainage system; Based on edge sensors, real-time image data of drainage systems in various areas of the city are collected to identify the status of drainage outlets in each area and evaluate the drainage blockage index in various areas of the city; The drainage blockage index of each area of ​​the city is used to correct the initial drainage capacity of each area of ​​the city to obtain the real-time drainage capacity of each area of ​​the city; Analyze the waterlogging risk indicators of various areas in the city based on the real-time drainage capacity and precipitation indicators of various areas in the city; Generate urban waterlogging prevention and control tasks based on the waterlogging risk levels corresponding to the waterlogging risk indicators of various areas in the city; The real-time drainage capacity of each area in the city is specifically: D'i =D i ×G i ; Where D' i is the real-time drainage capacity of the ith area in the city, D i is the drainage capacity index of the i-th area in the city, G i is the drainage blockage index of the ith area in the city.

[0006] Preferably, obtaining meteorological data and urban topography, and evaluating precipitation indicators in various areas of the city based on meteorological factors and topographic structure specifically include: Based on the city meteorological station, obtain the city meteorological data, mark the meteorological factors in the meteorological data of each area of ​​the city, and analyze the rainfall in each area of ​​the city; the city meteorological data includes: temperature, humidity, cloud volume, and air pressure; Based on the public data of the Urban Surveying and Mapping Bureau, obtain the urban topography data, and evaluate the environmental water accumulation coefficient of each area of ​​the city according to the topographic factors of each area of ​​the city; the urban topography data includes: terrain elevation, terrain slope, and terrain type; Substitute the rainfall in each area of ​​the city into the urban internal precipitation state prediction model, take the environmental water accumulation coefficient of each area of ​​the city as the influencing variable, build the urban precipitation multivariate regression model, and predict the precipitation index of each area of ​​the city;

[0007] The rainfall in each area of ​​the analyzed city is specifically: i =α0+α1A 11 +α2A 12 +…+α j A ij +ε; Where, X i is the rainfall in the ith area of ​​the city, A ij is the jth meteorological factor of the i-th area of ​​the city, α0 is the intercept term, α1, α2, α j are linear regression coefficients, and ε is the error term; The environmental water accumulation coefficient of each area in the city is specifically: i =α0+α1B 11 +α2B 12 +…+α k B ik +ε; Where Y i is the environmental water accumulation coefficient of the i-th area in the city, B ik is the kth terrain factor of the i-th area of ​​the city, α0 is the intercept term, α1, α2, α k are linear regression coefficients, and ε is the error term; The precipitation indexes for each area of ​​the predicted city are specifically: ; Where P i is the precipitation index of the ith region of the city, β0 is a constant term, and n is the total number of regions in the city.

[0008] Preferably, based on the structural parameters of the urban drainage system, initializing the drainage capacity of each area of ​​the city specifically includes: Get the number and diameter of drainage pipes in the drainage system of each area of ​​the city; Determine the correction parameters of drainage pipes in various areas of the city based on the historical maintenance tasks of the drainage system in various areas of the city; Get the number and width of drainage roads for drainage coefficients in each area of ​​the city; Determine the parameters for the correction of drainage roads in various areas of the city based on the historical maintenance tasks of the drainage system in various areas of the city; Based on the number of drainage pipes, drainage pipe diameter, drainage pipe correction parameters and the number of drainage roads, drainage width and drainage road correction parameters of the drainage system in each area of ​​the city, a regional drainage capacity assessment model is constructed to initialize the drainage capacity indicators of each area of ​​the city;

[0009] The regional drainage capacity assessment model is specifically as follows: ; Where D i is the drainage capacity index of the i-th area in the city, S i is the number of drainage pipes in the i-th area of ​​the city, SD i is the diameter of the drainage pipe in the ith area of ​​the city, Φ1 is the correction parameter of the drainage pipes in each area of ​​the city, V i is the number of drainage roads in the ith area of ​​the city, VD i The width of the drainage road in the i-th area of ​​the city, Φ2 is the drainage road correction parameter in each area of ​​the city, and γ1, γ2, γ3, γ4, and γ5 are all regression coefficients.

[0010] Preferably, based on edge sensors, real-time image data of drainage systems in various areas of the city are collected to identify the status of drainage outlets in various areas, and the drainage blockage index of various areas of the city is evaluated, specifically including: Based on the real-time image data of the drainage system in various areas of the city, data is cropped at the frame interval of video sampling to obtain a real-time image frame sample data set of the drainage system in various areas of the city; The real-time image frame sample datasets of the drainage system in various areas of the city are preprocessed, and the MaskR-CNN convolutional neural network is used to extract and mark the real-time image frame samples of each drain outlet blockage, thus obtaining the positive and negative sample datasets of the real-time image frames of the drainage system in various areas of the city. According to the positive and negative sample data sets of real-time image frames of the drainage system in various areas of the city, the real-time image frame samples of each blocked drain outlet are used as positive samples, and the real-time image frame samples of each unblocked drain outlet are used as negative samples to train the SVM support vector machine, create a classification model for the blocked state of the drain outlet, and generate the real-time state indicators of the drainage system in various areas of the city; According to the real-time status indicators of the drainage system in each area of ​​the city, the drainage blockage weight is allocated according to the blockage level of the real-time status indicators of the drainage system, and the drainage blockage index of each area of ​​the city is calculated;

[0011] The drainage blockage index of each area of ​​the city is specifically: ; In the formula, G i is the drainage blockage index of the i-th area in the city, δ i is the drainage blockage weight of the i-th area in the city, Z i It is the real-time status indicator of the drainage system in the i-th area of ​​the city.

[0012] Preferably, based on the real-time drainage capacity of each area of ​​the city and the precipitation index of each area of ​​the city, the waterlogging risk index of each area of ​​the city is analyzed, which specifically includes: Based on the total rainfall per unit time in each area of ​​the city, the precipitation index of each area of ​​the city is determined, and the precipitation index time series data of each area of ​​the city is established; Based on the precipitation index time series data of each area of ​​the city and the real-time drainage capacity of each area of ​​the city, the drainage trend of each area of ​​the city is analyzed, and the waterlogging risk index of each area of ​​the city is calculated; The specific waterlogging risk indicators for each area of ​​the city are: ; Where FRI i is the waterlogging risk index of the ith area in the city, P i (t) is the precipitation index of the t-th unit time in the i-th area of ​​the city, D' i (t) is the real-time drainage capacity of the ith area of ​​the city at t unit time, v is the adjustment coefficient, and R(t) is the ratio of the drainage capacity to the precipitation load at t unit time.

[0013] Compared with the prior art, the present invention has the following beneficial effects: The present invention proposes a machine vision-based urban waterlogging prevention task generation scheme, which is based on meteorological data, urban topography information and drainage system parameters, by evaluating the precipitation index of each area of ​​the city and initializing the drainage capacity of each area of ​​the city, using edge sensors to collect the drainage outlet status in real time and evaluate the drainage blockage index of each area of ​​the city, correcting and initializing the drainage capacity of each area of ​​the city, and fitting and analyzing the real-time drainage capacity of each area of ​​the city and the precipitation index of each area of ​​the city, calculating the waterlogging risk level of each area, and generating prevention tasks accordingly. The beneficial effects are: improving the management efficiency of the urban drainage system and reducing the efficiency of urban waterlogging prevention and control. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 Generate a flow chart for a machine vision-based urban waterlogging prevention task method; Figure 2 A flow chart of the method for initializing the drainage capacity of each area in the city; Figure 3 A flow chart of the method to assess precipitation indicators in various areas of the city; Figure 4 A flow chart of the method to assess the drainage blockage index in various areas of the city; Figure 5 Flowchart of the method for analyzing waterlogging risk indicators in various areas of the city. DETAILED DESCRIPTION

[0015] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art may think of other obvious variations.

[0016] Reference Figure 1 As shown, a method for generating urban waterlogging prevention tasks based on machine vision includes: Obtain meteorological data and urban topography, and evaluate precipitation indicators in various areas of the city based on meteorological factors and topographic structure; Initialize the drainage capacity of each area of ​​the city based on the structural parameters of the urban drainage system; Based on edge sensors, real-time image data of drainage systems in various areas of the city are collected to identify the status of drainage outlets in each area and evaluate the drainage blockage index in various areas of the city; The drainage blockage index of each area of ​​the city is used to correct the initial drainage capacity of each area of ​​the city to obtain the real-time drainage capacity of each area of ​​the city; Analyze the waterlogging risk indicators of various areas in the city based on the real-time drainage capacity and precipitation indicators of various areas in the city; Generate urban waterlogging prevention and control tasks based on the waterlogging risk levels corresponding to the waterlogging risk indicators of various areas in the city; The real-time drainage capacity of each area in the city is specifically: D' i =D i ×G i ; Where D' i is the real-time drainage capacity of the ith area in the city, D i is the drainage capacity index of the i-th area in the city, G i is the drainage blockage index of the ith area in the city.

[0017] This solution is based on meteorological data, urban topography information and drainage system parameters. By evaluating the precipitation index of each area of ​​the city and initializing the drainage capacity of each area of ​​the city, edge sensors are used to collect the status of the drainage outlets in real time and evaluate the drainage blockage index of each area of ​​the city. The drainage capacity of each area of ​​the city is corrected and initialized, and the real-time drainage capacity of each area of ​​the city and the precipitation index of each area of ​​the city are fitted and analyzed. The waterlogging risk level of each area is calculated and prevention and control tasks are generated accordingly. The beneficial effects are: improving the management efficiency of the urban drainage system and reducing the efficiency of urban waterlogging prevention and control.

[0018] Reference Figure 2 As shown in the figure, meteorological data and urban topography are obtained, and precipitation indicators of various areas of the city are evaluated based on meteorological factors and topographic structure, including: Based on the city meteorological station, obtain the city meteorological data, mark the meteorological factors in the meteorological data of each area of ​​the city, and analyze the rainfall in each area of ​​the city; the city meteorological data includes: temperature, humidity, cloud volume, and air pressure; Based on the public data of the Urban Surveying and Mapping Bureau, obtain the urban topography data, and evaluate the environmental water accumulation coefficient of each area of ​​the city according to the topographic factors of each area of ​​the city; the urban topography data includes: terrain elevation, terrain slope, and terrain type; Substitute the rainfall in each area of ​​the city into the urban internal precipitation state prediction model, take the environmental water accumulation coefficient of each area of ​​the city as the influencing variable, build the urban precipitation multivariate regression model, and predict the precipitation index of each area of ​​the city; The rainfall in each area of ​​the analyzed city is specifically: i =α0+α1A 11 +α2A 12 +…+α j A ij +ε; Where, X i is the rainfall in the ith area of ​​the city, A ij is the jth meteorological factor of the i-th area of ​​the city, α0 is the intercept term, α1, α2, α j are linear regression coefficients, and ε is the error term; The environmental water accumulation coefficient of each area in the city is specifically: i =α0+α1B 11 +α2B 12 +…+α k B ik +ε; Where Y i is the environmental water accumulation coefficient of the i-th area in the city, B ik is the kth terrain factor of the i-th area of ​​the city, α0 is the intercept term, α1, α2, α k are linear regression coefficients, and ε is the error term; The precipitation indexes for each area of ​​the predicted city are specifically: ; Where P i is the precipitation index of the ith region of the city, β0 is a constant term, and n is the total number of regions in the city.

[0019] It is understandable that whether each area of ​​the city will be flooded after rainfall is greatly affected by topographic factors. Therefore, by predicting the rainfall in each area of ​​the city and the environmental water accumulation index of each area, and taking the topography of the city as the influencing variable of the water flow direction of rainfall, it is possible to accurately evaluate and predict the rainfall in each area of ​​the city and its potential water accumulation. Reference Figure 3 As shown in the figure, based on the construction parameters of the urban drainage system, the initialization of the drainage capacity of each area in the city specifically includes: Get the number and diameter of drainage pipes in the drainage system of each area of ​​the city; Determine the correction parameters of drainage pipes in various areas of the city based on the historical maintenance tasks of the drainage system in various areas of the city; Get the number and width of drainage roads for drainage coefficients in each area of ​​the city; Determine the parameters for the correction of drainage roads in various areas of the city based on the historical maintenance tasks of the drainage system in various areas of the city; Based on the number of drainage pipes, drainage pipe diameter, drainage pipe correction parameters and the number of drainage roads, drainage width and drainage road correction parameters of the drainage system in each area of ​​the city, a regional drainage capacity assessment model is constructed to initialize the drainage capacity indicators of each area of ​​the city; The regional drainage capacity assessment model is specifically as follows: ; Where D i is the drainage capacity index of the i-th area in the city, S i is the number of drainage pipes in the i-th area of ​​the city, SD iis the diameter of the drainage pipe in the ith area of ​​the city, Φ1 is the correction parameter of the drainage pipes in each area of ​​the city, V i is the number of drainage roads in the ith area of ​​the city, VD i The width of the drainage road in the i-th area of ​​the city, Φ2 is the drainage road correction parameter in each area of ​​the city, and γ1, γ2, γ3, γ4, and γ5 are all regression coefficients.

[0020] This scheme constructs a regional drainage capacity assessment model by comprehensively considering structural parameters such as the number of drainage pipes in various areas of the city, drainage pipe diameters, the number and width of drainage roads, as well as correction parameters determined based on historical maintenance tasks. It accurately evaluates and initializes the drainage capacity indicators of various areas of the city, and provides core parameter objects for subsequent judgment of whether each area has experienced waterlogging.

[0021] Reference Figure 4 As shown in the figure, based on edge sensors, real-time image data of drainage systems in various areas of the city are collected to identify the status of drainage outlets in each area, and the drainage blockage index of each area in the city is evaluated, including: Based on the real-time image data of the drainage system in various areas of the city, data is cropped at the frame interval of video sampling to obtain a real-time image frame sample data set of the drainage system in various areas of the city; The real-time image frame sample datasets of the drainage system in various areas of the city are preprocessed, and the MaskR-CNN convolutional neural network is used to extract and mark the real-time image frame samples of each drain outlet blockage, thus obtaining the positive and negative sample datasets of the real-time image frames of the drainage system in various areas of the city. According to the positive and negative sample data sets of real-time image frames of the drainage system in various areas of the city, the real-time image frame samples of each blocked drain outlet are used as positive samples, and the real-time image frame samples of each unblocked drain outlet are used as negative samples to train the SVM support vector machine, create a classification model for the blocked state of the drain outlet, and generate the real-time state indicators of the drainage system in various areas of the city; According to the real-time status indicators of the drainage system in each area of ​​the city, the drainage blockage weight is allocated according to the blockage level of the real-time status indicators of the drainage system, and the drainage blockage index of each area of ​​the city is calculated; The drainage blockage index of each area of ​​the city is specifically: ; In the formula, G i is the drainage blockage index of the i-th area in the city, δ i is the drainage blockage weight of the i-th area in the city, Z i It is the real-time status indicator of the drainage system in the i-th area of ​​the city.

[0022] This solution collects real-time image data of drainage systems in various areas of the city based on edge sensors, and uses video frame interval cropping to generate drainage system image data sets. The Mask R-CNN convolutional neural network is used to preprocess and annotate image data, extract samples of whether the drain outlet is blocked or not, and form a positive and negative sample data set. Secondly, the support vector machine (SVM) is used to train the drainage outlet blockage status classification model, evaluate the real-time status of the drainage system in each area, and calculate and generate the drainage blockage index in combination with the drainage blockage level. The beneficial effect is: identifying potential blockage problems in advance and improving the emergency response capability of the drainage system.

[0023] Reference Figure 5 As shown in the figure, based on the real-time drainage capacity of each area of ​​the city and the precipitation index of each area of ​​the city, the waterlogging risk indicators of each area of ​​the city are analyzed, including: Based on the total rainfall per unit time in each area of ​​the city, the precipitation index of each area of ​​the city is determined, and the precipitation index time series data of each area of ​​the city is established; Based on the precipitation index time series data of each area of ​​the city and the real-time drainage capacity of each area of ​​the city, the drainage trend of each area of ​​the city is analyzed, and the waterlogging risk index of each area of ​​the city is calculated; The specific waterlogging risk indicators for each area of ​​the city are: ; Where FRI i is the waterlogging risk index of the ith area in the city, P i (t) is the precipitation index of the t-th unit time in the i-th area of ​​the city, D' i (t) is the real-time drainage capacity of the ith area of ​​the city at t unit time, v is the adjustment coefficient, and R(t) is the ratio of the drainage capacity to the precipitation load at t unit time.

[0024] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. The present invention may be subject to various changes and improvements without departing from the spirit and scope of the present invention. These changes and improvements fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.

Claims

1. A method for generating urban waterlogging prevention tasks based on machine vision, characterized in that: include: Obtain meteorological data and urban topography, and evaluate precipitation indicators in various areas of the city based on meteorological factors and topographic structure; Initialize the drainage capacity of each area in the city based on the structural parameters of the urban drainage system; Based on edge sensors, real-time image data of drainage systems in various areas of the city are collected to identify the status of drainage outlets in each area and evaluate the drainage blockage index in various areas of the city; The drainage blockage index of each area of ​​the city is used to correct the initial drainage capacity of each area of ​​the city to obtain the real-time drainage capacity of each area of ​​the city; Analyze the waterlogging risk indicators of various areas in the city based on the real-time drainage capacity of each area in the city and the precipitation indicators of each area in the city; Generate urban waterlogging prevention and control tasks based on the waterlogging risk levels corresponding to the waterlogging risk indicators of various areas in the city; The real-time drainage capacity of each area in the city is specifically: D' i =D i ×G i ; Where D' i is the real-time drainage capacity of the ith area in the city, D i is the drainage capacity index of the i-th area in the city, G i is the drainage blockage index of the i-th area in the city.

2. The method for generating urban waterlogging prevention tasks based on machine vision according to claim 1 is characterized in that: Obtain meteorological data and urban topography, and evaluate precipitation indicators in various areas of the city based on meteorological factors and topographic structure, including: Based on the city meteorological station, obtain the city meteorological data, mark the meteorological factors in the meteorological data of each area of ​​the city, and analyze the rainfall in each area of ​​the city; the city meteorological data includes: temperature, humidity, cloud volume, and air pressure; Based on the public data of the Urban Surveying and Mapping Bureau, obtain the urban topography data, and evaluate the environmental water accumulation coefficient of each area of ​​the city according to the topographic factors of each area of ​​the city; the urban topography data includes: terrain elevation, terrain slope, and terrain type; The rainfall in various areas of the city is substituted into the urban internal precipitation status prediction model. The environmental water accumulation coefficient of each area of ​​the city is taken as the influencing variable to construct an urban precipitation multivariate regression model to predict the precipitation indicators of various areas of the city.

3. The method for generating urban waterlogging prevention tasks based on machine vision according to claim 2 is characterized in that: The specific rainfall in each area of ​​the analyzed city is: i =α0+α1A 11 +α2A 12 +…+α j A ij +ε; Where, X i is the rainfall in the ith area of ​​the city, A ij is the jth meteorological factor of the i-th area of ​​the city, α0 is the intercept term, α1, α2, α j are linear regression coefficients, and ε is the error term; The environmental water accumulation coefficient of each area in the city is specifically: i =α0+α1B 11 +α2B 12 +…+α k B ik +ε; Where Y i is the environmental water accumulation coefficient of the ith area in the city, B ik is the kth terrain factor of the i-th area of ​​the city, α0 is the intercept term, α1, α2, α k are linear regression coefficients, and ε is the error term; The precipitation indexes for each area of ​​the predicted city are specifically: ; Where P i is the precipitation index of the ith region of the city, β0 is a constant term, and n is the total number of regions in the city.

4. The method for generating urban waterlogging prevention tasks based on machine vision according to claim 3 is characterized in that: Based on the structural parameters of the urban drainage system, the initialization of the drainage capacity of each area of ​​the city includes: Get the number and diameter of drainage pipes in the drainage system of each area of ​​the city; Determine the correction parameters of drainage pipes in various areas of the city based on the historical maintenance tasks of the drainage system in various areas of the city; Get the number and width of drainage roads for drainage coefficients in each area of ​​the city; Determine the parameters for the correction of drainage roads in various areas of the city based on the historical maintenance tasks of the drainage system in various areas of the city; Based on the number of drainage pipes, drainage pipe diameters, drainage pipe correction parameters and the number of drainage roads, drainage widths and drainage road correction parameters of the drainage systems in various areas of the city, a regional drainage capacity assessment model is constructed to initialize the drainage capacity indicators of various areas in the city.

5. The method for generating urban waterlogging prevention tasks based on machine vision according to claim 4 is characterized in that: The regional drainage capacity assessment model is specifically as follows: ； Where D i is the drainage capacity index of the i-th area in the city, S i is the number of drainage pipes in the i-th area of ​​the city, SD i is the diameter of the drainage pipe in the ith area of ​​the city, Φ1 is the correction parameter of the drainage pipes in each area of ​​the city, V i is the number of drainage roads in the ith area of ​​the city, VD i The width of the drainage road in the i-th area of ​​the city, Φ2 is the drainage road correction parameter in each area of ​​the city, and γ1, γ2, γ3, γ4, and γ5 are all regression coefficients.

6. The method for generating urban waterlogging prevention tasks based on machine vision according to claim 5 is characterized in that: Based on edge sensors, real-time image data of drainage systems in various areas of the city are collected to identify the status of drainage outlets in each area and evaluate the drainage blockage index in various areas of the city. Specifically, the following are included: Based on the real-time image data of the drainage system in various areas of the city, data is cropped at the frame interval of video sampling to obtain a real-time image frame sample data set of the drainage system in various areas of the city; The real-time image frame sample datasets of the drainage system in various areas of the city are preprocessed, and the Mask R-CNN convolutional neural network is used to extract and mark the real-time image frame samples of each drain outlet blockage, thus obtaining the positive and negative sample datasets of the real-time image frames of the drainage system in various areas of the city. According to the positive and negative sample data sets of real-time image frames of the drainage system in various areas of the city, the real-time image frame samples of each blocked drain outlet are used as positive samples, and the real-time image frame samples of each unblocked drain outlet are used as negative samples to train the SVM support vector machine, create a classification model for the blocked state of the drain outlet, and generate the real-time state indicators of the drainage system in various areas of the city; According to the real-time status indicators of the drainage system in various areas of the city, drainage blockage weights are allocated according to the blockage levels of the real-time status indicators of the drainage system, and the drainage blockage index of various areas of the city is calculated.

7. The method for generating urban waterlogging prevention tasks based on machine vision according to claim 6, characterized in that: The drainage blockage index of each area of ​​the city is specifically as follows: ; In the formula, G i is the drainage blockage index of the i-th area in the city, δ i is the drainage blockage weight of the i-th area in the city, Z i It is the real-time status indicator of the drainage system in the i-th area of ​​the city.

8. The method for generating urban waterlogging prevention tasks based on machine vision according to claim 7, characterized in that: Based on the real-time drainage capacity of each area of ​​the city and the precipitation indicators of each area of ​​the city, the waterlogging risk indicators of each area of ​​the city are analyzed, including: Based on the total rainfall per unit time in each area of ​​the city, the precipitation index of each area of ​​the city is determined, and the precipitation index time series data of each area of ​​the city is established; Based on the precipitation index time series data of each area of ​​the city and the real-time drainage capacity of each area of ​​the city, the drainage trend of each area of ​​the city is analyzed, and the waterlogging risk index of each area of ​​the city is calculated; The specific waterlogging risk indicators for each area of ​​the city are: ; Where FRI i is the waterlogging risk index of the ith area in the city, P i (t) is the precipitation index of the t-th unit time in the i-th area of ​​the city, D' i (t) is the real-time drainage capacity of the ith area of ​​the city at t unit time, v is the adjustment coefficient, and R(t) is the ratio of the drainage capacity to the precipitation load at t unit time.

Images