CN118072165B - River network intensive urban black and odorous water body risk division method and system - Google Patents
River network intensive urban black and odorous water body risk division method and system Download PDFInfo
- Publication number
- CN118072165B CN118072165B CN202410186015.5A CN202410186015A CN118072165B CN 118072165 B CN118072165 B CN 118072165B CN 202410186015 A CN202410186015 A CN 202410186015A CN 118072165 B CN118072165 B CN 118072165B
- Authority
- CN
- China
- Prior art keywords
- water body
- data
- remote sensing
- body data
- black
- 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.)
- Active
Links
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 title claims abstract description 419
- 238000000034 method Methods 0.000 title claims abstract description 34
- 238000003066 decision tree Methods 0.000 claims abstract description 79
- 238000012360 testing method Methods 0.000 claims description 63
- 238000012549 training Methods 0.000 claims description 44
- 238000013139 quantization Methods 0.000 claims description 20
- 238000000605 extraction Methods 0.000 claims description 15
- 238000010801 machine learning Methods 0.000 claims description 14
- 238000007781 pre-processing Methods 0.000 claims description 7
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 claims description 4
- PDNNQADNLPRFPG-UHFFFAOYSA-N N.[O] Chemical compound N.[O] PDNNQADNLPRFPG-UHFFFAOYSA-N 0.000 claims description 4
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 claims description 4
- 229910052799 carbon Inorganic materials 0.000 claims description 4
- 229910052760 oxygen Inorganic materials 0.000 claims description 4
- 239000001301 oxygen Substances 0.000 claims description 4
- 238000012545 processing Methods 0.000 abstract description 5
- 239000010865 sewage Substances 0.000 description 12
- 238000001228 spectrum Methods 0.000 description 6
- 238000002310 reflectometry Methods 0.000 description 4
- 239000000463 material Substances 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 238000005070 sampling Methods 0.000 description 3
- 238000013145 classification model Methods 0.000 description 2
- 238000012937 correction Methods 0.000 description 2
- 230000006866 deterioration Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000011002 quantification Methods 0.000 description 2
- 230000003595 spectral effect Effects 0.000 description 2
- 239000000126 substance Substances 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000007499 fusion processing Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000033116 oxidation-reduction process Effects 0.000 description 1
- 238000010183 spectrum analysis Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/182—Network patterns, e.g. roads or rivers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/762—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using clustering, e.g. of similar faces in social networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Software Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computing Systems (AREA)
- Medical Informatics (AREA)
- Evolutionary Computation (AREA)
- Multimedia (AREA)
- Artificial Intelligence (AREA)
- Human Resources & Organizations (AREA)
- Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- General Health & Medical Sciences (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- Quality & Reliability (AREA)
- General Business, Economics & Management (AREA)
- Operations Research (AREA)
- Marketing (AREA)
- Game Theory and Decision Science (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Educational Administration (AREA)
- Tourism & Hospitality (AREA)
- Development Economics (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention relates to the technical field of data processing, in particular to a river network intensive urban black and odorous water body risk division method and system. Comprising the following steps: step S1: acquiring a plurality of first remote sensing images and a plurality of first water body data of a target water body; step S2: establishing a numerical relation between each characteristic value in the second characteristic set and the first water body data; step S3: acquiring a plurality of second remote sensing images and acquiring second water body data; step S4: the model creation unit creates a plurality of decision trees; step S5: inputting the third water body data into a plurality of decision trees, and outputting a risk classification result; step S6: and obtaining the risk grade of the river network intensive urban black and odorous water body. The invention solves the problem of inaccurate black and odorous water body risk classification and river network dense urban black and odorous water body classification, and improves the accuracy of black and odorous water body risk classification and river network dense urban black and odorous water body classification.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to a river network intensive urban black and odorous water body risk division method and system.
Background
River network intensive urban water system is developed, population is dense, industry types are many and distribution is dense, because the current black and odorous water body investigation method is not uniform, black and odorous water body base number is not clear, black and odorous water body supervision and management are not in place, black and odorous return problems frequently occur and other problems, the river network intensive urban black and odorous water body needs to be subjected to risk grade division, and the black and odorous return problems are reduced according to the urban black and odorous water body risk grade in time treatment, in the prior art, for example: chinese patent CN110987825B discloses a classification method of urban black and odorous water based on spectrum matching, which is used for classifying the measured water remote sensing reflectivity into 6 grades according to the chromaticity judged in the wild; taking the average value of the measured water remote sensing reflectivity of each grade as the standard spectrum of each grade; the spectrum difference wave band range between the water bodies of different grades is obtained through standard spectrum analysis of 6 grades; and (3) constructing a classification method of urban black and odorous water based on spectrum matching to distinguish the water of each grade one by one. The urban water bodies with different color grades are classified into water body grades with different black and odorous degrees, and the overall grading precision is 77.03%. Also for example: european patent WO2023134626A1 discloses a black and odorous water body extraction method based on a CART classification model, which relates to the field of environmental monitoring and comprises the following steps: selecting a research area, and designing a plurality of sampling points in the range of the research area; monitoring the related chemical indexes of the water body at each sampling point respectively, extracting remote sensing reflectivity data of the water body, and judging the type of the water body according to the dividing standard of the related chemical indexes of the urban black and odorous water body; comparing and analyzing the remote sensing reflectivity data extracted from each sampling point, and selecting spectral change characteristics of the black and odorous water body and the general water body; constructing each node of a decision tree according to the spectrum change characteristics and the base index, constructing a decision tree classification model, obtaining classification results of black and odorous water bodies and general water bodies, and calculating classification accuracy; and analyzing and obtaining the space-time distribution change of the black and odorous water body in the research area according to the classification result. The invention can be used for solving the problems of insufficient objectivity and less feature selection of the current black and odorous water body threshold setting. Both patents solve the problem of classifying black and odorous water bodies, but classify according to the spectral characteristics of the water bodies, and compared with classifying black and odorous water bodies by measuring the water body data of the water bodies, the problems of inaccurate classification exist, and the photographed pictures can influence the distribution of the spectrum due to weather reasons, so that the error is larger.
Disclosure of Invention
In order to better solve the problems, the invention provides a river network dense urban black and odorous water body risk division method, which comprises the following steps:
step S1: reading a plurality of first remote sensing images of a target water body from a storage unit, preprocessing the first remote sensing images, and simultaneously reading a plurality of first water body data corresponding to the plurality of first remote sensing images of the target water body;
Step S2: extracting a first feature set of the first remote sensing image through an extraction unit, clustering the first feature set to obtain a second feature set, and establishing a numerical relation between each feature value in the second feature set and the first water body data through machine learning;
Step S3: acquiring a plurality of second remote sensing images through the image pickup unit, extracting each characteristic value in a second characteristic set of the plurality of second remote sensing images, and acquiring second water body data corresponding to the plurality of second remote sensing images based on each characteristic value and the numerical relation;
step S4: the model creation unit creates a plurality of decision trees based on the water body data, trains and tests the decision trees by taking a plurality of first water body data and a plurality of second water body data as training data and test data, and acquires an optimal decision tree set;
Step S5: periodically acquiring a third remote sensing image of the target water body through the unmanned aerial vehicle camera unit, repeating the method of the step S3, acquiring third water body data corresponding to the third remote sensing image, inputting the third water body data into a plurality of decision trees, and outputting a risk classification result;
wherein the water body data includes the first water body data, the second water body data, and the third water body data.
As a preferable embodiment of the present invention, the characteristic values include: the first remote sensing image and the second remote sensing image are three-dimensional depth images and multispectral high-resolution images, and the data comprise dissolved oxygen content, transparency, ammonia oxygen content and total organic carbon concentration of the water body.
As a preferred technical solution of the present invention, the step S5 further includes a step S6:
Obtaining a risk classification result of each black and odorous water body of the river network dense urban based on the method from the step S1 to the step S5, setting corresponding weights according to an average value of all the black and odorous water body risk classifications, an urban basic condition quantization value, a pollution control force quantization value, a supervision management condition quantization value and a pollution load quantization value, obtaining a river network dense urban black and odorous water body risk score, and obtaining a river network dense urban black and odorous water body risk grade according to the river network dense urban black and odorous water body risk score, wherein the river network dense urban black and odorous water body risk score is as follows:
Mu 1、μ2、μ3、μ4 and mu 5 are respectively the average value of the black and odorous water body risk classification, the urban basic condition quantized value, the pollution control force quantized value, the supervision and management condition quantized value and the pollution load quantized value, t i is the i-th black and odorous water body risk classification result, x is the urban basic condition quantized value, y is the pollution control force quantized value, z is the supervision and management condition quantized value, and p is the pollution load quantized value.
As a preferred embodiment of the present invention, the step S2 includes:
Step S21: the method comprises the steps of obtaining a first characteristic data set of a first remote sensing image through an extraction unit, comparing any two first remote sensing images in a plurality of first remote sensing images to obtain a first comparison result, and comparing two first water body data corresponding to any two first remote sensing images to obtain a second comparison result;
Step S22: repeating the step S21 to the combination of any two first remote sensing images, obtaining a first comparison result data set and a second comparison result data set, and obtaining the correlation between each characteristic data of each characteristic data set and the first water body data by analyzing the first comparison result data set and the second comparison result data set;
Step S23: clustering the first characteristic data set through a K-means algorithm based on the correlation between each characteristic data and the first water body data to obtain a second characteristic data set;
Step S24: and acquiring a numerical relation between each characteristic data in the second characteristic data set and each data in the first water body data by a learning unit through each characteristic value in the second characteristic data set and corresponding data in the first water body data.
As a preferred embodiment of the present invention, the step S4 includes:
Step S41: extracting water body characteristics of the water body data through the extraction unit, and grouping the water body data according to the numerical range of the water body characteristics, wherein each numerical range group of each water body characteristic comprises the first water body data and the second water body data;
Step S42: extracting n pieces of water body data from each water body data group to serve as a training data set, and taking the remaining m pieces of water body data of each water body data group to serve as a test data set, wherein the training data and the test data both comprise the first water body data and the second water body data, and the ratio of n to m is 7:3, a step of;
step S43: randomly extracting k1 groups of water body features from all the water body features, wherein each group comprises k2 water body features, constructing a plurality of decision trees based on the k1 groups of water body features, and training the decision trees by using the training data and the pre-judgment risk classification results obtained by the training data based on the risk judgment standards;
step S44: and testing a plurality of decision trees through the test data, and adjusting the decision trees and the numerical relation according to the test result.
As a preferred embodiment of the present invention, in the step S42, a ratio of the data amounts of the first water body data and the second water body data in the training data set and the test data set is equal to a ratio of a total amount of the first water body data and a total amount of the second water body data.
As a preferred embodiment of the present invention, the step S44 includes: when a plurality of decision trees are tested through the test data set, in the test data set with the same test result, searching the first water body data and the second water body data with the highest similarity, acquiring the first remote sensing image and the second remote sensing image corresponding to the first water body data and the second water body data, respectively comparing the first remote sensing image with the second characteristic value set data of the second remote sensing image, acquiring a first comparison result, and repeating the step S2 when the first comparison result is larger than or equal to a first threshold value, and re-acquiring the numerical relation between each characteristic value in the second characteristic set and the first water body data.
As a preferable technical scheme of the invention, when a plurality of decision trees are established, when nodes are set, the weight of the water body characteristics in the risk classification of the target water body is set according to the weight of the water body characteristics, wherein the greater the weight of the water body characteristics is, the more the position of the nodes related to the water body characteristics in the corresponding decision trees is.
The invention also provides a river network intensive urban black and odorous water body risk division system, which is used for realizing the river network intensive urban black and odorous water body risk division method, and comprises the following steps:
The storage unit is used for storing and reading a plurality of first remote sensing images of the target water body, preprocessing the plurality of first remote sensing images and storing and reading a plurality of first water body data corresponding to the plurality of first remote sensing images of the target water body;
an extracting unit, configured to extract a plurality of first feature sets of the first remote sensing images;
the analysis unit is used for acquiring a second feature set through clustering the first feature set, and establishing a numerical relation between each feature value in the second feature set and the corresponding first water body data;
the unmanned aerial vehicle camera unit is used for acquiring a plurality of second remote sensing images and extracting feature values in a second feature set of the second remote sensing images;
the extraction unit is further used for obtaining second water body data corresponding to a plurality of second remote sensing images based on the feature values and the numerical relation;
The model creation unit is used for creating a plurality of decision trees based on the water body data, training and testing the decision trees by taking a plurality of first water body data and a plurality of second water body data as training data and testing data, and acquiring an optimal decision tree set;
the unmanned aerial vehicle camera unit is further used for periodically acquiring a third remote sensing image of the target water body, repeating the method of the step S3, acquiring third water body data corresponding to the third remote sensing image, inputting the third water body data into a plurality of decision trees, and outputting risk classification results;
wherein the water body data includes the first water body data, the second water body data, and the third water body data.
Compared with the prior art, the invention has the following beneficial effects:
According to the invention, the first remote sensing images of a plurality of water bodies are obtained through the storage unit, the first remote sensing images are preprocessed, the first remote sensing images and the first water body data are historical data for measuring the target water body, and a foundation is laid for further analyzing the correlation between the characteristics of the first remote sensing images and the water quality data. Because the risk division model needs a large amount of sample data during training, the water body data serving as the sample data needs to be collected in the field and measured through a measuring instrument, a large amount of manpower and material resources are consumed, and the risk division precision is low when the sample data is insufficient, therefore, the first feature set of the first remote sensing image is extracted through the extraction unit, the correlation between the first feature sets of the plurality of first remote sensing images and the water body data is analyzed, the second feature set is obtained through clustering, the numerical relation between the second feature set and the water body data is further established through a machine learning algorithm, and further the water body data corresponding to the water body can be obtained through the second remote sensing image and the numerical relation obtained through the shooting unit, so that more water body sample data are obtained, the manpower cost is saved, enough training data and test data are provided for risk division, and the precision of risk division is improved. By extracting the characteristics of the water body data, creating a plurality of decision trees through a model creation unit according to the characteristics of the water body data and risk classification standards, grouping the first water body data and the second water body data, extracting training data and test data from the grouping, training and testing the decision trees, adjusting the numerical relation according to the obtained test result, improving the accuracy of the water body data, optimizing the decision trees according to the test result, and obtaining an optimal decision tree set, the current accurate risk classification of the black and odorous water body can be obtained through inputting the third water body data into the decision trees according to the output result of the decision trees, and further obtaining the risk level of the black and odorous water body of a river network intensive city, and relevant departments can take corresponding measures according to the risk level to avoid further deterioration of the black and odorous water body.
Drawings
FIG. 1 is a flow chart of a river network dense urban black and odorous water body risk division method;
FIG. 2 is a block diagram of a river network dense urban black and odorous water body risk division system according to the invention.
Detailed Description
The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
The invention provides a river network dense urban black and odorous water body risk division method, which is shown in figure 1 and comprises the following steps:
step S1: reading a plurality of first remote sensing images of a target water body from a storage unit, preprocessing the first remote sensing images, and simultaneously reading a plurality of first water body data corresponding to the plurality of first remote sensing images of the target water body;
Specifically, a first remote sensing image of a plurality of water bodies is obtained through a storage unit, and is preprocessed, wherein the first remote sensing image comprises a three-dimensional image and a multispectral high-resolution image, the first remote sensing image and the first water body data are historical data for measuring the target water body, and the preprocessing of the first remote sensing image comprises: the noise reduction processing of the three-dimensional image, the orthographic correction processing, the atmospheric correction processing and the image fusion processing of the multispectral high-resolution image are performed, meanwhile, the water quality data of the water corresponding to the first remote sensing image, namely the first water data, are also obtained, and a foundation is laid for further analyzing the correlation between the characteristics of the first remote sensing image and the water quality data.
Step S2: extracting a first feature set of the first remote sensing image through an extraction unit, acquiring a second feature set related to the first water body data through analyzing the first feature set, and establishing a numerical relation between each feature value in the second feature set and the corresponding first water body data through machine learning;
Specifically, since a large amount of sample data is required when the risk division model is trained, and the water body data serving as the sample data needs to be collected in the field and measured by the measuring instrument, a large amount of manpower and material resources are consumed, and when the sample data is insufficient, the risk division precision is not high, therefore, the first feature set of the first remote sensing image is extracted by the extracting unit, and the features include: the method comprises the steps of analyzing the correlation between a first feature set of a plurality of first remote sensing images and water data, acquiring a second feature set through clustering, further establishing a numerical relation between the second feature set and the water data through a machine learning algorithm, and further acquiring the water data corresponding to the water through a second remote sensing image acquired by a shooting unit and the numerical relation, so that more water sample data are acquired, labor cost is saved, enough training data and test data are provided for risk classification, and the precision of risk classification is improved.
Step S3: acquiring a plurality of second remote sensing images through an unmanned aerial vehicle camera unit, extracting each characteristic value in a second characteristic set of the plurality of second remote sensing images, and acquiring second water body data corresponding to the plurality of second remote sensing images based on each characteristic value and the numerical relation;
Specifically, a plurality of second remote sensing images are obtained through the unmanned aerial vehicle camera unit, each characteristic value of the second characteristic set is extracted, second water body data corresponding to the plurality of second remote sensing images are obtained based on the numerical relation between each characteristic value of the second characteristic set and the first water body data, and because the cost of shooting the second remote sensing images through the unmanned aerial vehicle camera unit is low, more sample data are obtained under the condition that the water body is not required to be detected in the field, and the labor cost is saved.
Step S4: the model creation unit creates a plurality of decision trees based on the characteristics of the water body data, trains and tests the decision trees by taking a plurality of first water body data and a plurality of second water body data as training data and test data, and acquires an optimal decision tree set;
Specifically, a plurality of decision trees are created through a model creation unit by extracting the characteristics of the water body data and by using the characteristics of the water body data and risk classification standards, the first water body data and the second water body data are grouped, training data and test data are extracted from the groups to train and test the decision trees, the numerical relation is adjusted according to the acquired test results, and meanwhile the decision trees are optimized according to the test results to acquire an optimal decision tree set, so that more accurate black and odorous water risk classification can be acquired by inputting third water body data.
Step S5: periodically acquiring a third remote sensing image of the target water body through the unmanned aerial vehicle camera unit, repeating the method of the step S3, acquiring third water body data corresponding to the third remote sensing image, inputting the third water body data into the plurality of decision trees, and outputting the target water body risk classification result;
wherein the water body data includes the first water body data, the second water body data, and the third water body data.
Specifically, a third remote sensing image of the black and odorous water body is obtained in real time, each characteristic value of a second characteristic set of the third remote sensing image is extracted, third water body data is obtained based on the numerical relation between each characteristic value of the second characteristic set and each data of the third water body data, the third water body data are input into the decision trees, and the current accurate risk level of the black and odorous water body is obtained according to the output results of the decision trees, wherein the black and odorous water body is divided into three levels including mild pollution, moderate pollution and severe pollution, and the higher the level is, the more serious the pollution is, the relevant departments can take corresponding measures according to the risk level of the target water body to avoid further deterioration of the black and odorous water body.
Further, the first remote sensing image and the second remote sensing image are three-dimensional depth images and multispectral high-resolution images, and the data comprise dissolved oxygen content, transparency, ammonia oxygen content and total organic carbon concentration of the water body.
Further, the step S5 further includes a step S6:
Obtaining a risk classification result of each black and odorous water body of the river network dense urban based on the method from the step S1 to the step S5, setting corresponding weights according to an average value of all the black and odorous water body risk classification results, an urban basic condition quantization value, a pollution control force quantization value, a supervision management condition quantization value and a pollution load quantization value, obtaining a river network dense urban black and odorous water body risk score, and obtaining a river network dense urban black and odorous water body risk grade according to the river network dense urban black and odorous water body risk score, wherein the river network dense urban black and odorous water body risk score is as follows:
Wherein mu 1、μ2、μ3、μ4 and mu 5 are respectively the average value of the black and odorous water body risk classification results, the urban basic condition quantization value, the pollution control force quantization value, the supervision and management condition quantization value and the pollution load quantization value weight, t i is the ith black and odorous water body risk classification result, x is the urban basic condition quantization value, y is the pollution control force quantization value, z is the supervision and management condition quantization value, and p is the pollution load quantization value.
Specifically, the urban basic condition quantification value is determined by weighting black and odor risk classification of a water system of a city, a built area of the city, the number of sewage outlets and rainfall, wherein the weight of the black and odor risk classification of the water system is highest, the number of the sewage outlets is secondary, and the built area weight of the city is minimum; the pollution load quantification value is obtained by weighting the domestic sewage source discharge amount, the agricultural sewage source discharge amount and the industrial sewage source discharge amount, wherein the domestic sewage source discharge amount has the largest weight, and the agricultural sewage source discharge amount and the industrial sewage source discharge amount have the same weight; the quantitative value of the pollution control force is obtained by weighting the sewage collection amount, the pipe network coverage length, the sewage treatment efficiency and the operation and maintenance period of the sewage treatment facility, wherein the weights of the four items are the same, and the quantitative value of the supervision and management condition is weighted by the spot check period, the water quality detection period and the rain sewage pipe network management inspection period, and the weights of the three items are the same. the value of t i is a positive integer from 1 to 3, wherein the larger the value of t i is, the higher the risk classification of the black and odorous water body is, the value of x, y, z, p is a positive integer from 1 to 10, the sum of mu 1、μ2、μ3、μ4 and mu 5 is 1, when the value of r is more than or equal to 7.5, the river network dense city is a black and odorous water body high risk city, when the value of r is less than or equal to 7.5 and more, the river network dense city is a black and odorous water body medium risk city, and when the value of r is less than 5, the river network dense city is a black and odorous water body low risk city.
Further, the step S2 includes:
Step S21: the method comprises the steps of obtaining a first characteristic data set of a first remote sensing image through an extraction unit, comparing each characteristic value in the first characteristic data sets of any two first remote sensing images in a plurality of first remote sensing images to obtain a first comparison result, and comparing each data in two first water body data corresponding to any two first remote sensing images to obtain a second comparison result;
Specifically, the extracting unit extracts a first feature data set of the first remote sensing image, where the first remote sensing image includes: a three-dimensional depth image and a multispectral high-resolution image, wherein the first feature data set comprises: the method comprises the steps of comparing each characteristic value in first characteristic data sets of any two first remote sensing images, obtaining a first comparison result, comparing each data of first water body data corresponding to any two first remote sensing images, and obtaining a second comparison result, wherein the first comparison result corresponds to the second comparison result, namely, characteristic value differences corresponding to water body data differences of the same water body at different times are obtained, and a foundation is laid for further analyzing the corresponding relation between each data in the first water body data and each characteristic value in a first characteristic data set under the condition that the first comparison result and the second comparison result sample are more than one.
Step S22: repeating the step S21 until all first comparison result data sets and second comparison result data sets are obtained, and obtaining the correlation relationship between each characteristic data of each characteristic data set and the first water body data by analyzing the first comparison result data sets and the second comparison result data sets;
Specifically, all combinations of any two first remote sensing images are obtained through repeating the step S21, all first comparison results obtained by comparing all values of the second characteristic data sets corresponding to all the any two first remote sensing images with second comparison results obtained by comparing all data corresponding to the first water body data are obtained, so that more first comparison results and more second comparison results are obtained under the condition that the number of the first remote sensing images is certain, and therefore the correlation relation between each data of more accurate water body data and each characteristic value in the first characteristic data sets is obtained through analyzing all the first comparison results and the second comparison results.
Step S23: clustering the first characteristic data set through a K-means algorithm based on the correlation between each characteristic value and the first water body data to obtain a second characteristic data set;
Specifically, since the same data in the first water body data may correspond to the values of the plurality of features in the first feature data set, the first feature data set may be clustered based on each data of the first water body and the plurality of feature values in the corresponding first feature data set, and a clustered second feature data set may be obtained.
Step S24: and acquiring a numerical relation between each characteristic data in the second characteristic data set and each data in the first water body data by a machine learning algorithm from each characteristic value in the second characteristic data set and corresponding data in the first water body data.
Specifically, a machine learning model is built through a machine learning unit, the corresponding relation between each characteristic value in the second characteristic data set and corresponding data in the first water body data is input into the machine learning model, the machine learning model is trained through machine learning, data in the water body data is calculated through the machine learning model and each characteristic value in the second characteristic data set of the remote sensing image, the machine model and the machine learning algorithm are also called as the numerical relation between each characteristic value in the second characteristic data set and each data in the water body data, the water body data can be obtained through the remote sensing image through the numerical relation, and more sample data are provided for black and odorous water body risk classification.
Further, the step S4 includes:
Step S41: taking the category of the water body data as a water body characteristic, and carrying out balanced grouping on the water body data according to the numerical range of each water body characteristic, wherein each grouping comprises the first water body data and the second water body data;
Specifically, the water body characteristics include: the dissolved oxygen content, the oxidation-reduction potential, the ammonia oxygen and the total organic carbon concentration of the water body are grouped according to the numerical range of the water body characteristics, so that the water body data are ensured to have larger numerical values or smaller numerical values in each group, the sample data in each group are distributed more uniformly, and each group comprises the first water body data and the second water body data, so that the sample data in each group are balanced and diversified.
Step S42: extracting n pieces of water body data from each water body data group to serve as a training data set, and taking the remaining m pieces of water body data of each water body data group to serve as a test data set, wherein the training data and the test data both comprise the first water body data and the second water body data, and the ratio of n to m is 7:3, a step of;
Specifically, n pieces of water body data are extracted from the group to serve as training data sets, the rest of water body data are served as test data, the first water body training data and the test data comprise first water body data and second water body data, the ratio of the training data to the test data is 7:3, and the first water body data and the second water body data are different in sources, the precision of the first water body data is larger than that of the second water body data, so that the distribution balance of sample data is ensured, meanwhile, the richness of the sample data is ensured, and the more the training data, the more accurate the result is output through a training decision tree, so that the training data is more than the test data.
Step S43: randomly extracting k1 groups of water body features from all the water body features, wherein each group comprises k2 water body features, constructing a plurality of decision trees based on the k1 groups of water body features, and training the decision trees by using the training data and the pre-judgment risk classification results obtained by the training data based on the risk judgment standards;
Specifically, K1 sets of water features are randomly extracted from the water features, nodes of each decision tree are created based on the water features and risk judgment standards, meanwhile, a plurality of decision trees are trained through training data and pre-judgment risk classification results corresponding to the training data, and iteration optimization is carried out according to output results of the plurality of decision trees to obtain optimal decision trees, so that accuracy of output risk classification is guaranteed, wherein the risk judgment standards can be set by a user by themselves and are not particularly limited.
Step S44: and testing a plurality of decision trees through the test data, and adjusting the decision trees and the numerical relation according to the test result.
Specifically, the iterative adjustment of the decision trees according to the test result to obtain the optimal decision trees is not repeated in the prior art, in the test data set with the same test result, the first water body data and the second water body data with the highest similarity are searched, the second characteristic value set data corresponding to the first water body data and the second water body data are obtained for comparison, the first comparison result is obtained, the difference between the first water body data and the second water body data is smaller, when the first comparison result is greater than or equal to the first threshold value, the first comparison result may not be accurate enough, and therefore, the step S2 is repeated, and the numerical relation between each characteristic value in the second characteristic set and the first water body data is obtained again, otherwise, the numerical relation is accurate, and readjustment is not needed.
Further, in the step S42, a ratio of the data amounts of the first water body data and the second water body data in the training data set and the test data set is equal to a ratio of a total amount of the first water body data and a total amount of the second water body data.
Further, the step S44 includes: when a plurality of decision trees are tested through the test data set, in the test data set with the same test result, searching the first water body data and the second water body data with the highest similarity, acquiring the first remote sensing image and the second remote sensing image corresponding to the first water body data and the second water body data, respectively comparing the first remote sensing image with the second characteristic value set data of the second remote sensing image, acquiring a first comparison result, and repeating the step S2 when the first comparison result is larger than or equal to a first threshold value, and re-acquiring the numerical relation between each characteristic value in the second characteristic set and the first water body data.
Specifically, when the decision trees are tested by the test data, when the test results corresponding to the first water body data and the second water body data are the same, the difference between the first water body data and the second water body data is also small, and because the first water body data and the second water body data are obtained by the corresponding second characteristic data sets extracted by the remote sensing images through the numerical relationship, the difference between the second characteristic data sets corresponding to the first water body data and the second water body data is also small, and when the first comparison result obtained by comparing the second data sets of the first water body data and the second water body data is large, the first water body data is obtained through measurement and is accurate, and therefore, the first water body data and the second water body data are possibly not accurate enough, and therefore, the accuracy of the numerical relationship can be improved by re-obtaining the numerical relationship through repeating the step S2, so that more accurate water body data can be provided through the remote sensing images.
Further, when a plurality of decision trees are established, when nodes are set, the weight of the nodes is set according to the water body characteristics during risk classification, wherein the greater the water body characteristics weight is, the more front the nodes related to the water body characteristics are in the position corresponding to the decision trees.
Specifically, when the decision tree is established, the selection of the nodes is related to the performance of the whole decision tree, and the positions of the nodes are set according to the weights of the water features when the black and odorous risks of the water are classified, wherein the water features with the larger weights are positioned in front of the decision tree, the trend of the classification result of the black and odorous water can be obtained as soon as possible, the robustness of the decision tree is improved, and therefore the accuracy of the output result is improved.
The invention also provides a river network intensive urban black and odorous water body risk division system, which is used for realizing the river network intensive urban black and odorous water body risk division method, as shown in fig. 2, and comprises the following steps:
The storage unit is used for storing and reading a plurality of first remote sensing images of the target water body, preprocessing the first remote sensing images and storing and reading a plurality of first water body data corresponding to the plurality of first remote sensing images of the target water body;
The extraction unit is used for extracting a first feature set of the first remote sensing image;
the analysis unit is used for analyzing the first feature set to obtain a second feature set related to the first water body data, and establishing a numerical relation between each feature value in the second feature set and the corresponding first water body data;
the unmanned aerial vehicle camera unit is used for acquiring a plurality of second remote sensing images and extracting feature values in a second feature set of the second remote sensing images;
the extraction unit is further used for obtaining second water body data corresponding to a plurality of second remote sensing images based on the feature values and the numerical relation;
The model creation unit is used for creating a plurality of decision trees based on the water body data, training and testing the decision trees by taking a plurality of first water body data and a plurality of second water body data as training data and testing data, and acquiring an optimal decision tree set;
the unmanned aerial vehicle camera unit is further used for periodically acquiring a third remote sensing image of the target water body, repeating the method of the step S3, acquiring third water body data corresponding to the third remote sensing image, inputting the third water body data into the plurality of decision trees, and outputting a risk classification result;
Wherein the water body data includes the first water body data and the second water body data.
In summary, the first remote sensing images of a plurality of water bodies are obtained through the storage unit, the first remote sensing images are preprocessed, the first remote sensing images and the first water body data are historical data for measuring the target water body, and a foundation is laid for further analyzing the correlation between the characteristics of the first remote sensing images and the water quality data. Because the risk division model needs a large amount of sample data during training, the water body data serving as the sample data needs to be collected in the field and measured through a measuring instrument, a large amount of manpower and material resources are consumed, and the risk division precision is low when the sample data is insufficient, therefore, the first feature set of the first remote sensing image is extracted through the extraction unit, the correlation between the first feature sets of the plurality of first remote sensing images and the water body data is analyzed, the second feature set is obtained through clustering, the numerical relation between the second feature set and the water body data is further established through a machine learning algorithm, and further the water body data corresponding to the water body can be obtained through the second remote sensing image and the numerical relation obtained through the shooting unit, so that more water body sample data are obtained, the manpower cost is saved, enough training data and test data are provided for black and odorous water body risk division, and the black odorous water body risk classification precision is improved. By extracting the characteristics of the water body data, creating a plurality of decision trees through a model creation unit according to the characteristics of the water body data and black and odorous water body risk classification standards, grouping the first water body data and the second water body data, extracting training data and test data from the grouping to train and test the plurality of decision trees, adjusting the numerical relation according to the obtained test result, improving the accuracy of the water body data, optimizing the plurality of decision trees according to the test result, and obtaining an optimal decision tree set, the current accurate risk level of the black and odorous water body can be obtained through inputting the third water body data into the plurality of decision trees according to the output result of the plurality of decision trees, and further obtaining the river network dense urban black and odorous water body risk level.
The technical features of the foregoing embodiments may be arbitrarily combined, and for brevity, all of the possible combinations of the technical features of the foregoing embodiments are not described, however, they should be considered as the scope of the disclosure as long as there is no contradiction between the combinations of the technical features.
The foregoing examples illustrate only a few embodiments of the invention and are described in detail herein without thereby limiting the scope of the invention. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the invention, which are all within the scope of the invention. Accordingly, the scope of protection of the present invention is to be determined by the appended claims.
The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, and alternatives falling within the spirit and principles of the invention.
Claims (7)
1. A river network intensive urban black and odorous water body risk division method is characterized by comprising the following steps:
Step S1: reading a plurality of first remote sensing images of a target water body from a storage unit, preprocessing the first remote sensing images, and simultaneously reading a plurality of first water body data corresponding to the plurality of first remote sensing images of the target water body from the storage unit;
Step S2: extracting a first feature set of the first remote sensing image through an extraction unit, clustering the first feature set to obtain a second feature set, and establishing a numerical relation between each feature value in the second feature set and the first water body data through machine learning;
Step S3: acquiring a plurality of second remote sensing images through an unmanned aerial vehicle camera unit, extracting each characteristic value in a second characteristic set of the plurality of second remote sensing images, and acquiring second water body data corresponding to the plurality of second remote sensing images based on each characteristic value and the numerical relation;
Step S4: the model creation unit creates a plurality of decision trees based on the water body data, trains and tests the decision trees by taking a plurality of first water body data and a plurality of second water body data as training data and test data, and acquires an optimal decision tree set;
Step S5: periodically acquiring a third remote sensing image of the target water body through the unmanned aerial vehicle camera unit, repeating the method of the step S3, acquiring third water body data corresponding to the third remote sensing image, inputting the third water body data into a plurality of decision trees, and outputting a risk classification result of the target water body;
the water body data comprises the first water body data, the second water body data and the third water body data, the first remote sensing image and the second remote sensing image are three-dimensional depth images and multispectral high-resolution images, and the data comprise dissolved oxygen content, transparency, ammonia oxygen content and total organic carbon concentration of the water body;
The step S2 includes:
Step S21: the method comprises the steps of obtaining a first characteristic data set of a first remote sensing image through an extraction unit, comparing any two first remote sensing images in a plurality of first remote sensing images to obtain a first comparison result, and comparing two first water body data corresponding to any two first remote sensing images to obtain a second comparison result;
Step S22: repeating the step S21 to the combination of any two first remote sensing images, obtaining a first comparison result data set and a second comparison result data set, and obtaining the correlation between each characteristic data of each characteristic data set and the first water body data by analyzing the first comparison result data set and the second comparison result data set;
Step S23: clustering the first characteristic data set through a K-means algorithm based on the correlation relation between each characteristic data and the first water body data to obtain a second characteristic data set;
Step S24: and acquiring a numerical relation between each characteristic data in the second characteristic data set and each data in the first water body data by a learning unit through each characteristic value in the second characteristic data set and corresponding data in the first water body data.
2. The river network dense urban black and odorous water body risk classification method according to claim 1, wherein the step S5 further comprises a step S6:
Obtaining a risk classification result of each black and odorous water body of the river network dense urban based on the method from the step S1 to the step S5, setting corresponding weights according to an average value of all the black and odorous water body risk classifications, an urban basic condition quantization value, a pollution control force quantization value, a supervision management condition quantization value and a pollution load quantization value, obtaining a river network dense urban black and odorous water body risk score, and obtaining a river network dense urban black and odorous water body risk grade according to the river network dense urban black and odorous water body risk score, wherein the river network dense urban black and odorous water body risk score is as follows:
；
wherein, AndRespectively the average value of the black and odorous water body risk classification, the urban basic condition quantized value, the pollution control force quantized value, the supervision and management condition quantized value and the pollution load quantized value,Is the firstThe risk classification result of the black and odorous water body,Quantifying the value of the urban basic condition,Quantifying the pollution control force,Quantifying the value of the supervision and management status,The values are quantified for pollution load.
3. The river network dense urban black and odorous water body risk classification method according to claim 1, wherein the step S4 comprises:
Step S41: extracting water body characteristics of the water body data through the extraction unit, and grouping the water body data according to the numerical range of the water body characteristics, wherein each numerical range group of each water body characteristic comprises the first water body data and the second water body data;
Step S42: extracting n pieces of water body data from each water body data group to serve as a training data set, and taking the remaining m pieces of water body data of each water body data group to serve as a test data set, wherein the training data and the test data both comprise the first water body data and the second water body data, and the ratio of n to m is 7:3, a step of;
Step S43: randomly extracting k1 groups of water body features from all the water body features, wherein each group comprises k2 water body features, constructing a plurality of decision trees based on the k1 groups of water body features, and training the decision trees by using the training data and the pre-judgment risk classification results obtained by the training data based on the risk judgment standards;
step S44: and testing a plurality of decision trees through the test data, and adjusting the decision trees and the numerical relation according to the test result.
4. A river network dense urban black and odorous water body risk classification method according to claim 3, wherein in step S42, the ratio of the data amounts of the first water body data and the second water body data in the training data set and the test data set is equal to the ratio of the total amount of the first water body data and the total amount of the second water body data.
5. The river network dense urban black and odorous water body risk classification method according to claim 3, wherein the step S44 comprises: when a plurality of decision trees are tested through the test data set, in the test data set with the same test result, searching the first water body data and the second water body data with the highest similarity, acquiring the first remote sensing image and the second remote sensing image corresponding to the first water body data and the second water body data, respectively comparing the first remote sensing image with the second characteristic value set data of the second remote sensing image, acquiring a first comparison result, and repeating the step S2 when the first comparison result is larger than or equal to a first threshold value, and re-acquiring the numerical relation between each characteristic value in the second characteristic set and the first water body data.
6. A river network dense urban black and odorous water body risk dividing method according to claim 3, wherein when a plurality of decision trees are established, when nodes are set, the weight of the water body characteristics in the risk dividing process of the target water body is set according to the weight of the water body characteristics, and the greater the weight of the water body characteristics, the more front the position of the nodes related to the water body characteristics in the corresponding decision trees.
7. A river network dense urban black and odorous water body risk classification system, characterized in that the system is used for realizing the river network dense urban black and odorous water body risk classification method according to any one of claims 1-6, and the system comprises:
The storage unit is used for storing and reading a plurality of first remote sensing images of the target water body, preprocessing the plurality of first remote sensing images and storing and reading a plurality of first water body data corresponding to the plurality of first remote sensing images of the target water body;
an extracting unit, configured to extract a plurality of first feature sets of the first remote sensing images;
the analysis unit is used for acquiring a second feature set through clustering the first feature set, and establishing a numerical relation between each feature value in the second feature set and the corresponding first water body data;
the unmanned aerial vehicle camera unit is used for acquiring a plurality of second remote sensing images and extracting feature values in a second feature set of the second remote sensing images;
the extraction unit is further used for obtaining second water body data corresponding to a plurality of second remote sensing images based on the feature values and the numerical relation;
The model creation unit is used for creating a plurality of decision trees based on the water body data, training and testing the decision trees by taking a plurality of first water body data and a plurality of second water body data as training data and testing data, and acquiring an optimal decision tree set;
The unmanned aerial vehicle camera unit is further used for periodically acquiring a third remote sensing image of the target water body, repeating the method of the step S3, acquiring third water body data corresponding to the third remote sensing image, inputting the third water body data into a plurality of decision trees, and outputting a risk classification result of the target water body;
wherein the water body data includes the first water body data, the second water body data, and the third water body data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410186015.5A CN118072165B (en) | 2024-02-20 | 2024-02-20 | River network intensive urban black and odorous water body risk division method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410186015.5A CN118072165B (en) | 2024-02-20 | 2024-02-20 | River network intensive urban black and odorous water body risk division method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118072165A CN118072165A (en) | 2024-05-24 |
| CN118072165B true CN118072165B (en) | 2024-08-16 |
Family
ID=91105318
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410186015.5A Active CN118072165B (en) | 2024-02-20 | 2024-02-20 | River network intensive urban black and odorous water body risk division method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118072165B (en) |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113887493A (en) * | 2021-10-21 | 2022-01-04 | 河北工业大学 | Black and odorous water body remote sensing image identification method based on ID3 algorithm |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2018081404A (en) * | 2016-11-15 | 2018-05-24 | パナソニック インテレクチュアル プロパティ コーポレーション オブ アメリカＰａｎａｓｏｎｉｃ Ｉｎｔｅｌｌｅｃｔｕａｌ Ｐｒｏｐｅｒｔｙ Ｃｏｒｐｏｒａｔｉｏｎ ｏｆ Ａｍｅｒｉｃａ | Discrimination method, discrimination device, discriminator generation method and discriminator generation device |
| US20180260719A1 (en) * | 2017-03-10 | 2018-09-13 | Microsoft Technology Licensing, Llc | Cascaded random decision trees using clusters |
| CN108256419B (en) * | 2017-12-05 | 2018-11-23 | 交通运输部规划研究院 | A method of port and pier image is extracted using multispectral interpretation |
| CN110987955B (en) * | 2019-12-05 | 2022-09-20 | 南京师范大学 | Urban black and odorous water body grading method based on decision tree |
| CN114387529A (en) * | 2022-01-11 | 2022-04-22 | 北华航天工业学院 | Black and odorous water body extraction method based on CART classification model |
| WO2023159070A1 (en) * | 2022-02-16 | 2023-08-24 | Klimanovus Llc | Systems and methods for automatic environmental planning and decision support using artificial intelligence and data fusion techniques on distributed sensor network data |
| CN117111092A (en) * | 2023-07-19 | 2023-11-24 | 苏州梦伯乐信息科技有限公司 | High-spatial-resolution remote sensing water quality detection method based on machine learning |
-
2024
- 2024-02-20 CN CN202410186015.5A patent/CN118072165B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113887493A (en) * | 2021-10-21 | 2022-01-04 | 河北工业大学 | Black and odorous water body remote sensing image identification method based on ID3 algorithm |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118072165A (en) | 2024-05-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112052755B (en) | Semantic convolution hyperspectral image classification method based on multipath attention mechanism | |
| CN111665575B (en) | Medium-and-long-term rainfall grading coupling forecasting method and system based on statistical power | |
| CN101299237B (en) | High spectroscopic data supervision classifying method based on information quantity dimensionality sequence | |
| CN111126511B (en) | Vegetation index fusion-based LAI quantitative model establishment method | |
| CN107895136B (en) | Coal mine area identification method and system | |
| CN116340723B (en) | Rural water pollution quick tracing method and system based on big data | |
| CN109410184A (en) | Live streaming pornographic image detection method based on dense confrontation network semi-supervised learning | |
| CN113936214B (en) | Karst wetland vegetation community classification method based on fusion of aerospace remote sensing images | |
| CN114387529A (en) | Black and odorous water body extraction method based on CART classification model | |
| CN112580864A (en) | Village and town domestic garbage yield prediction system combining with multivariate data application value improvement | |
| CN106919645A (en) | The sight spot meteorological element Intelligent fine Forecasting Methodology at the big scenic spot of complex landform | |
| CN115598162B (en) | Soil heavy metal content detection method based on stacked model | |
| CN114638465A (en) | Acquisition terminal network signal quality evaluation method based on secondary scoring | |
| CN118072165B (en) | River network intensive urban black and odorous water body risk division method and system | |
| CN118014949A (en) | X-ray image quality evaluation system, method and training method | |
| CN117726978A (en) | Water body detection method and device based on wastewater discharged from chemical region | |
| CN108663334A (en) | The method for finding soil nutrient spectral signature wavelength based on multiple Classifiers Combination | |
| CN116933206A (en) | Method and system for predicting heavy metal pollution space risk of polluted site soil | |
| CN117010274A (en) | Intelligent early warning method for harmful elements in underground water based on integrated incremental learning | |
| CN114998719A (en) | Forest fire prediction method based on deep learning and multi-source remote sensing data | |
| CN103984960A (en) | Remote sensing image sample intelligent collecting method | |
| CN111898579A (en) | Extreme gradient lifting-based unbiased semi-supervised classification model for high-resolution remote sensing images | |
| CN118470440B (en) | Tumor early recognition system based on deep learning and hyperspectral image | |
| CN114648074B (en) | Loess cave-trapping object-oriented extraction method integrating topographic features | |
| CN116011745B (en) | Ecological restoration scheme optimization method and system based on cloud service |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |