CN112766146B - Multi-source data-based dynamic reservoir monitoring system and method - Google Patents
Multi-source data-based dynamic reservoir monitoring system and method Download PDFInfo
- Publication number
- CN112766146B CN112766146B CN202110058297.7A CN202110058297A CN112766146B CN 112766146 B CN112766146 B CN 112766146B CN 202110058297 A CN202110058297 A CN 202110058297A CN 112766146 B CN112766146 B CN 112766146B
- Authority
- CN
- China
- Prior art keywords
- reservoir
- layer
- remote sensing
- information
- dynamic
- 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
- 238000012544 monitoring process Methods 0.000 title claims abstract description 77
- 238000000034 method Methods 0.000 title claims abstract description 70
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 32
- 238000010276 construction Methods 0.000 claims abstract description 26
- 238000001514 detection method Methods 0.000 claims abstract description 26
- 238000013135 deep learning Methods 0.000 claims abstract description 23
- 238000011176 pooling Methods 0.000 claims description 70
- 238000013136 deep learning model Methods 0.000 claims description 24
- 230000009193 crawling Effects 0.000 claims description 14
- 238000012549 training Methods 0.000 claims description 13
- 230000011218 segmentation Effects 0.000 claims description 11
- 238000000605 extraction Methods 0.000 abstract description 13
- 238000005516 engineering process Methods 0.000 abstract description 8
- 238000011161 development Methods 0.000 description 9
- 230000018109 developmental process Effects 0.000 description 9
- 230000008859 change Effects 0.000 description 6
- 230000003203 everyday effect Effects 0.000 description 5
- 238000012423 maintenance Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 5
- 238000012545 processing Methods 0.000 description 4
- 210000000746 body region Anatomy 0.000 description 3
- 238000009835 boiling Methods 0.000 description 3
- 230000007547 defect Effects 0.000 description 3
- 238000003973 irrigation Methods 0.000 description 3
- 230000002262 irrigation Effects 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000013527 convolutional neural network Methods 0.000 description 2
- 230000002354 daily effect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 239000003344 environmental pollutant Substances 0.000 description 2
- 231100000719 pollutant Toxicity 0.000 description 2
- 238000010248 power generation Methods 0.000 description 2
- 230000003595 spectral effect Effects 0.000 description 2
- 238000003860 storage Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000009395 breeding Methods 0.000 description 1
- 230000001488 breeding effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000009472 formulation Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000000746 purification Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 239000002689 soil Substances 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
- 238000003911 water pollution Methods 0.000 description 1
- 238000001363 water suppression through gradient tailored excitation Methods 0.000 description 1
Images
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
- G06V20/13—Satellite images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/35—Categorising the entire scene, e.g. birthday party or wedding scene
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A10/00—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE at coastal zones; at river basins
- Y02A10/40—Controlling or monitoring, e.g. of flood or hurricane; Forecasting, e.g. risk assessment or mapping
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Economics (AREA)
- Molecular Biology (AREA)
- Software Systems (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Biomedical Technology (AREA)
- Remote Sensing (AREA)
- Astronomy & Astrophysics (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Image Analysis (AREA)
Abstract
The invention provides a multi-source data-based dynamic reservoir monitoring system and method, belongs to the dynamic reservoir monitoring technology, and comprises a data acquisition module, a multi-source remote sensing data sample module, a deep learning module, a dynamic reservoir text updating module, a dynamic reservoir monitoring module and a cloud platform. The invention provides a method for constructing a nationwide full-factor reservoir information database; carrying out reservoir target detection by using an object-oriented classification method in combination with a deep learning technology and giving reservoir attribute information; and updating the reservoir information database according to two methods of remote sensing monitoring and ubiquitous network monitoring. The dynamic active monitoring method for the nationwide reservoir can actively discover, track and lock reservoir construction and management hotspots and key targets, realize automatic extraction of deep reservoir information, and provide great convenience for people to use water conservancy information.
Description
Technical Field
The invention belongs to the technical field of dynamic reservoir monitoring, and particularly relates to a dynamic reservoir monitoring system and method based on multi-source data.
Background
China is rich in water resources, particularly in southern areas, flood disasters are easy to happen in the first flood season, and in the past, the mode of acquiring water resource information of relevant areas is to perform field measurement immediately, is limited by range and cannot acquire the information in time. In view of the fact that water safety and water crisis become prominent factors restricting the development of the society and the economy of China, and reservoir engineering has great significance for downstream flood control and irrigation, the method has great significance for mastering information such as reservoir types, areas, capacities, distribution characteristics, hazard development rules, dynamic change trends and influences on the surrounding environment and the like, and plays a great role in protecting life safety, property safety, economic construction and the like of people.
In the prior art, although the water regime information early warning can be effectively carried out, the application range is limited to a certain reservoir and the surrounding environment thereof, the used reservoir information is less, the large-area and multi-type data can not be used in a combined manner, and the defects of data use lag and information ineffective utilization exist. In view of this, how to provide a dynamic reservoir monitoring method with strong timeliness and wide application is a technical problem to be solved by those skilled in the art.
Disclosure of Invention
Aiming at the defects in the prior art, the dynamic reservoir monitoring system and method based on multi-source data provided by the invention realize dynamic and information management of information such as reservoir type, area, capacity, distribution characteristics, hazard development rule, dynamic change trend and influence condition of the reservoir on the surrounding environment.
In order to achieve the above purpose, the invention adopts the technical scheme that:
the scheme provides a dynamic reservoir monitoring system based on multi-source data, which comprises a data acquisition module, a multi-source remote sensing data sample module, a deep learning module, a dynamic reservoir text updating module, a dynamic reservoir monitoring module and a cloud platform;
the data acquisition module is used for acquiring constructed and under-construction reservoir information and constructing a full-factor reservoir database according to the reservoir information;
the multi-source remote sensing data sample module is used for carrying out target detection on the reservoir by using an object-oriented method according to the remote sensing image and constructing a reservoir-based multi-source remote sensing data sample library according to the detected target reservoir;
the deep learning module is used for constructing a deep learning model by using a deep learning method, training the deep learning model by using a reservoir multi-source remote sensing data sample library to obtain a reservoir identification model, carrying out reservoir detection on a remote sensing image by using the reservoir identification model, and endowing attribute information of a full-element reservoir database on the detected reservoir;
the reservoir dynamic text updating module is used for crawling dynamic information of the reservoir according to the reservoir detection result to obtain updated reservoir information;
the reservoir dynamic monitoring module is used for dynamically monitoring the updated reservoir information by combining the remote sensing image;
and the cloud platform is used for finishing reservoir dynamic monitoring based on multi-source data according to the reservoir identification model, the reservoir dynamic information and the updated reservoir dynamic information.
Based on the system, the invention also provides a reservoir dynamic monitoring method based on multi-source data, which comprises the following steps:
s1, acquiring constructed and under-construction reservoir information, and constructing a full-element reservoir database according to the reservoir information;
s2, carrying out target detection on the reservoir by using an object-oriented method according to the remote sensing image, and constructing a reservoir-based multi-source remote sensing data sample library according to the detected target reservoir;
s3, constructing a deep learning model by using a deep learning method;
s4, training a deep learning model by using a reservoir multi-source remote sensing data sample library to obtain a reservoir identification model;
s5, reservoir detection is carried out on the remote sensing image by using the reservoir identification model, and attribute information of the full-element reservoir database is given to the detected reservoir;
s6, crawling dynamic reservoir information according to the reservoir detection result to obtain updated reservoir information, and combining the remote sensing image to dynamically monitor the updated reservoir information;
and S7, uploading the reservoir identification model, the reservoir dynamic information and the updated reservoir dynamic information to a cloud platform, and completing reservoir dynamic monitoring based on multi-source data.
Further, the step S2 includes the steps of:
s201, identifying a reservoir region in the remote sensing image by using an object-oriented method according to the remote sensing image;
s202, cutting the reservoir area into a data set with the size of 224 x 224, and constructing a reservoir multi-source remote sensing data sample library.
Still further, the remote sensing image in step S201 includes a remote sensing image generated by a multi-scale segmentation method and a remote sensing image extracted by fuzzy logic classification.
Still further, the deep learning model includes an input layer, a first convolution layer connected to the input layer, a first pooling layer connected to the first convolution layer, a second convolution layer connected to the first pooling layer, a second pooling layer connected to the second convolution layer, a third convolution layer connected to the second pooling layer, a third pooling layer connected to the third convolution layer, a fourth convolution layer connected to the third pooling layer, a fourth pooling layer connected to the fourth convolution layer, a fifth convolution layer connected to the fourth pooling layer, a fifth pooling layer connected to the fifth convolution layer, a first fully-connected layer connected to the fifth pooling layer, a second fully-connected layer connected to the first fully-connected layer, a third fully-connected layer connected to the second fully-connected layer, a sixth convolution layer connected to the input layer, a sixth pooling layer connected to the sixth pooling layer, a seventh convolution layer connected to the seventh convolution layer, a seventh convolution layer connected to the fourth fully-connected to the fifth pooling layer, and a fifth pooling layer connected to the fifth convolution layer; the fifth fully-connected layer is connected with the third fully-connected layer.
Still further, the convolution kernel sizes of the first convolution layer, the second convolution layer, the third convolution layer, the fourth convolution layer, the fifth convolution layer, the sixth convolution layer and the seventh convolution layer are all 3 × 3, the step lengths are all 1, and the effective filling size is all 1;
the sizes of convolution kernels of the first pooling layer, the second pooling layer, the third pooling layer, the fourth pooling layer, the fifth pooling layer, the sixth pooling layer and the seventh pooling layer are all 2 x 2.
Still further, the number of convolution kernels of the first convolution layer and the sixth convolution layer is 64, the number of convolution kernels of the second convolution layer and the seventh convolution layer is 128, the number of convolution kernels of the third convolution layer is 256, and the number of convolution kernels of the fourth convolution layer and the fifth convolution layer is 512.
Still further, the cloud platform in step S7 includes two working modes: reservoir remote sensing-network cooperative monitoring and network-remote sensing cooperative monitoring.
The invention has the beneficial effects that:
(1) The invention provides a nationwide full-factor reservoir information database; carrying out reservoir target detection by using an object-oriented classification method in combination with a deep learning technology and giving reservoir attribute information; and updating the reservoir information database according to two methods of remote sensing monitoring and ubiquitous network monitoring. The dynamic active monitoring method for the nationwide reservoir can actively discover, track and lock reservoir construction and management hotspots and key targets, realize automatic extraction of deep reservoir information, and provide great convenience for people to use water conservancy information.
(2) The invention integrates remote sensing monitoring and ubiquitous network monitoring methods, creates a new way for nationwide reservoir dynamic monitoring, forms two working service modes of reservoir remote sensing-ubiquitous network cooperative monitoring and ubiquitous network-remote sensing cooperative monitoring, constructs a dynamic monitoring system based on reservoir engineering, and utilizes the obtained information to realize dynamic and informatization management of information such as reservoir type, area, capacity, distribution characteristics, damage development rule, dynamic change trend, and influence on the surrounding environment.
(3) The invention provides a new reservoir sample library, which is constructed by using various remote sensing image data such as satellite remote sensing images, unmanned aerial vehicle images and the like, compared with a fixed sample data set in the past, the sample data set construction method is wider, and the identification rate is more accurate.
(4) The convolutional neural network is used for training a large amount of reservoir sample data set training models, but the method does not use the traditional classic deep learning model, but improves the deep learning model so as to improve the operation efficiency of the improved model.
(5) The invention provides a rapid reservoir model identification cloud platform, a large amount of remote sensing data are uploaded to the cloud platform, reservoir information identification is automatically completed on the cloud platform, local memory data do not need to be occupied, the use place is not limited, the operation cost and the storage space are saved, and the application range is wider.
Drawings
FIG. 1 is a schematic diagram of the system of the present invention,
FIG. 2 is a flow chart of the method of the present invention.
FIG. 3 is a schematic structural diagram of the deep learning model of the present invention.
Detailed Description
The following description of the embodiments of the present invention is provided to facilitate the understanding of the present invention by those skilled in the art, but it should be understood that the present invention is not limited to the scope of the embodiments, and it will be apparent to those skilled in the art that various changes may be made without departing from the spirit and scope of the invention as defined and defined in the appended claims, and all matters produced by the invention using the inventive concept are protected.
Example 1
In the past, although the method can effectively perform water regime information early warning, the application range is limited to a certain reservoir and the surrounding environment thereof, the used reservoir information is less, large-area and multi-type data cannot be combined for use, and the defects that the data use is lagged and the information cannot be effectively utilized exist. In view of the above, how to provide a dynamic reservoir monitoring method with strong timeliness and wide application is a technical problem to be solved by technical personnel in the field, and the invention provides a method for constructing a dynamic reservoir engineering monitoring system by using ubiquitous data and remote sensing data technology, and utilizing all obtained information to realize dynamic and information management of information such as reservoir type, area, capacity, distribution characteristics, damage development rules, dynamic change trend, and influence on ambient environment. As shown in fig. 1, the system comprises a data acquisition module, a multi-source remote sensing data sample module, a deep learning module, a reservoir dynamic text updating module, a reservoir dynamic monitoring module and a cloud platform; the data acquisition module is used for acquiring constructed and under-construction reservoir information and constructing a full-factor reservoir database according to the reservoir information; the multi-source remote sensing data sample module is used for carrying out target detection on the reservoir by using an object-oriented method according to the remote sensing image and constructing a reservoir-based multi-source remote sensing data sample library according to the detected target reservoir; the deep learning module is used for constructing a deep learning model by using a deep learning method, training the deep learning model by using a reservoir multi-source remote sensing data sample library to obtain a reservoir identification model, carrying out reservoir detection on a remote sensing image by using the reservoir identification model, and endowing attribute information of a full-element reservoir database to the detected reservoir; the reservoir dynamic text updating module is used for crawling dynamic reservoir information according to the reservoir detection result to obtain updated reservoir information; the reservoir dynamic monitoring module is used for dynamically monitoring the updated reservoir information in combination with the remote sensing image; and the cloud platform is used for finishing reservoir dynamic monitoring based on multi-source data according to the reservoir identification model, the reservoir dynamic information and the updated reservoir dynamic information.
In the embodiment, the national full-factor reservoir database is constructed by combining the established and in-progress information of the national reservoir; automatically constructing a reservoir sample by using an object-oriented classification method, then detecting a reservoir target by using a deep learning method, and endowing each reservoir with attribute information according to a national reservoir database; a remote sensing monitoring and ubiquitous network monitoring method is integrated, a new way for nationwide reservoir dynamic monitoring is created, and two working service modes of reservoir remote sensing-ubiquitous network cooperative monitoring and ubiquitous network-remote sensing cooperative monitoring are formed. The dynamic active monitoring method for the nationwide reservoir can actively find, track and lock reservoir construction and management hot spots and key targets, realize automatic extraction of deep reservoir information, and provide great convenience for people to use water conservancy information.
Example 2
As shown in fig. 2, the invention provides a dynamic reservoir monitoring method based on multi-source data, which is implemented as follows:
s1, acquiring constructed and under-constructed reservoir information, and constructing a full-factor reservoir database according to the reservoir information.
In this embodiment, the data is acquired through web crawler microblog data, news reports, historical data, yearbook, government announcements, flood and drought disaster bulletins, chinese and english literature bases and the like, and the information of the hydraulic engineering built and under construction nationwide is collected.
In this embodiment, a nationwide reservoir full-factor and long-term data set is constructed according to names, geographical locations, engineering construction conditions, construction purposes, project subjects, engineering investments, operating conditions, major events and the like, based on the collected nationwide constructed and under-construction hydraulic engineering information.
S2, carrying out target detection on the reservoir by using an object-oriented method according to the remote sensing image, and constructing a reservoir-based multi-source remote sensing data sample library according to the detected target reservoir, wherein the implementation method comprises the following steps:
s201, identifying a reservoir region in the remote sensing image by using an object-oriented method according to the remote sensing image;
s202, cutting the reservoir area into a data set with the size of 224 x 224, and constructing a reservoir multi-source remote sensing data sample library.
In this embodiment, the invention uses the remote sensing data such as the satellite remote sensing image, the space remote sensing image, the unmanned aerial vehicle remote sensing image and the like which are updated every day, and uses an object-oriented classification method in view of the spectral characteristics of the water body to automatically identify the water body region in the image, cut the water body region into a data set with the size of 224 x 224, construct a water body sample data set, automatically extract the water body region, and avoid manually sketching the water body sample.
In this embodiment, the object-oriented remote sensing image classification has two independent modules, object generation (image segmentation) and information extraction (image classification):
(1) The object is generated by using a multi-scale segmentation method, the multi-scale segmentation of the image starts from any pixel, and the object is formed by adopting a region combination method from bottom to top. The reservoir area extracted by the invention is obvious, can be obviously distinguished from surrounding ground objects, and the optimal segmentation scale is input, so that the segmentation step can be automatically completed. The optimal segmentation scale is determined by integrating spectral information, image textures and graph structures according to the obvious characteristics of the water body ground objects to fine tune and correct the optimal segmentation scale, and finally, the segmentation scale capable of completing the complete extraction of the reservoir region is the optimal segmentation scale according to the input segmentation threshold.
(2) The information extraction is based on the idea of fuzzy logic classification, a discrimination rule system of characteristic attributes is established, the probability that each object belongs to a certain class is calculated, and the purposes of classification identification and information extraction are achieved. In the information extraction step, other categories except the water body area can be divided into a non-water body according to the division result, the image is finally divided into the water body and the non-water body, the vector boundary of the water body area is derived, and the water body sample delineation of the 224 x 224 size data set is automatically completed.
And S3, constructing a deep learning model by using a deep learning method.
As shown in fig. 3, the deep learning model includes an input layer, a first convolution layer connected to the input layer, a first pooling layer connected to the first convolution layer, a second convolution layer connected to the first pooling layer, a second pooling layer connected to the second convolution layer, a third convolution layer connected to the second pooling layer, a third pooling layer connected to the third convolution layer, a fourth convolution layer connected to the third pooling layer, a fourth pooling layer connected to the fourth convolution layer, a fifth convolution layer connected to the fourth pooling layer, a fifth pooling layer connected to the fifth convolution layer, a first full-connection layer connected to the fifth pooling layer, a second full-connection layer connected to the first full-connection layer, a third full-connection layer connected to the second full-connection layer, a sixth convolution layer connected to the input layer, a sixth pooling layer connected to the sixth pooling layer, a seventh convolution layer connected to the sixth convolution layer, a seventh convolution layer connected to the seventh full-connection layer, a seventh convolution layer connected to the fourth full-connection layer, and a fifth full-connection layer connected to the fifth convolution layer; the fifth fully-connected layer is connected with the third fully-connected layer. The convolution kernels of the first convolution layer, the second convolution layer, the third convolution layer, the fourth convolution layer, the fifth convolution layer, the sixth convolution layer and the seventh convolution layer are all 3 x 3 in size, the step length is 3, the step length is 1, and the effective filling size is 1; the sizes of convolution kernels of the first pooling layer, the second pooling layer, the third pooling layer, the fourth pooling layer, the fifth pooling layer, the sixth pooling layer and the seventh pooling layer are all 2 × 2. The number of convolution kernels of the first convolution layer and the sixth convolution layer is 64, the number of convolution kernels of the second convolution layer and the seventh convolution layer is 128, the number of convolution kernels of the third convolution layer is 256, and the number of convolution kernels of the fourth convolution layer and the fifth convolution layer is 512.
And S4, training a deep learning model by utilizing a reservoir multi-source remote sensing data sample library to obtain a reservoir identification model.
In this embodiment, the training process is as follows: the method comprises the steps that multi-source remote sensing data samples serve as input samples, the reservoir range sketched in the samples serves as training labels, after the deep learning model is determined, a certain number of samples and labels corresponding to the samples are taken at each time and input into the deep learning model, the deep learning model can automatically learn error values between the samples and the labels, the error values of the deep learning model are smaller and smaller along with continuous increase of training times, finally, the trained model is high in precision, and the training process is finished. The correctly trained model can correctly output the water body range in the image of any image.
And S5, performing reservoir detection on the remote sensing image by using the reservoir identification model, and endowing the attribute information of the full-element reservoir database to the detected reservoir.
In the embodiment, the deep learning technology is utilized to rapidly position the reservoir target for the data sources such as the space remote sensing image, the aviation remote sensing image and the low-altitude remote sensing image, and the detected target reservoir is endowed with the constructed full-element reservoir database information.
In this embodiment, reservoir target detection is performed through deep learning, and the implementation steps are as follows:
(1) Constructing a binary image by using the cut 224 x 224 original image data set and the water body vector boundary extracted in the step 3, and inputting the binary image into an improved convolutional neural network;
(2) The improved deep learning model of the invention is as follows:
constructing a deep learning network comprising 5 convolutional layers and 3 full-connection layers, and extracting deep information of the image; extracting shallow layer information of an image by using a model comprising 2 convolutional layers and 1 full-link layer; and connecting the features extracted by the deep network with the features extracted by the shallow network through a full connection layer, and finally outputting an identification result. The deep information extracted by the improved method of the invention is combined with the shallow information, the accuracy of model identification can be improved, and the improved model frame is shown in figure 3:
deep learning convolutional layer process: the size of the initial convolution kernel is 3 × 3 × 3, the size of the stride is 1, the size of the effective padding is 1, and the pooling layer pooling adopts a maximum pooling function max pooling of 2 × 2.
Structure of first to fifth 5 convolutional layers and first to third 3 fully-connected layers:
a1, performing convolution treatment by using 64 convolution kernels at a time, and performing pooling layer posing treatment at a time;
a2, performing convolution treatment by using 128 convolution kernels at a time, and performing pooling layer boiling treatment at a time;
a3, carrying out convolution treatment by using 256 convolution kernels once, and carrying out pooling layer posing treatment once;
a4, performing convolution treatment by using 512 convolution kernels once, and performing pooling layer posing treatment once;
a5, performing convolution treatment by using 512 convolution kernels once, and performing pooling layer posing treatment once;
a6, three full-link layers Fc _ layer were used.
Structures of sixth to seventh convolution layers and fourth full-link layer:
b1, performing convolution treatment by using 64 convolution kernels once, and performing pooling layer boiling treatment once;
b2, performing convolution treatment by using 128 convolution kernels at a time, and performing pooling layer boiling treatment at a time;
and B3, processing by using a fifth full connection layer Fc _ layer.
(3) Training with the constructed 224 x 224 dataset using the deep learning network constructed herein;
(4) Storing the trained reservoir identification model;
(5) And carrying out nationwide reservoir target detection on multi-source data such as aerospace remote sensing images, aviation remote sensing images, low-altitude remote sensing images and the like nationwide by using the trained reservoir identification model.
And S6, crawling dynamic information of the reservoir according to the detection result of the reservoir, and combining the remote sensing image to dynamically monitor the updated reservoir information.
In this embodiment, the dynamic text of the ubiquitous network data reservoir is updated: the information on the internet is rich in variety, and crawlers are regularly carried out on microblog data, news reports, historical data, yearbook, government announcements, flood and drought disaster bulletins, chinese and English literature bases and other data, and according to keywords such as 'reservoir', 'dam', 'water conservancy facility', 'water gate', 'water inlet', 'time', 'place', and the like, reservoir maintenance and construction information occurring at a certain time in a certain area is obtained, and support is provided for updating dynamic reservoir information every day. The method for acquiring the reservoir maintenance and construction information comprises the following steps:
(1) When a crawler system crawls pages, the pages are sequenced according to time sequence by using a topic web crawler algorithm based on keywords, and the results are crawled and stored, and at the moment, the latest time point of data release of the crawled pages in the previous round is recorded;
(2) Comparing the crawling time recorded with the last time in the next round of crawling, if the crawling time is earlier than the last time, the crawling time is a time point which is later than the last time, the crawling is performed if the crawling time is indicated to be the crawled webpage, and if the crawling time is later than the last time point, the crawling is performed;
(3) When the data is stored, simple processing is carried out on the content captured by the crawler, such as title extraction, content extraction, time extraction and the like, and duplicate removal processing is also carried out during storage, so that more processing resources are not wasted.
In this embodiment, according to the reservoir and the related information updated on the internet every day, an acquisition mode of network government affair business information related to reservoir management covering the four levels of the central, provincial, prefecture and county levels and an acquisition mode of information related to reservoir utilization, maintenance, modification and development activities of a coverage mechanism and netizens are established, dynamic information of all elements of the constructed national reservoir is updated in time, and a dynamic reservoir monitoring system based on ubiquitous network data is formed.
In this embodiment, the remote sensing reservoir dynamic monitoring is updated: according to the remote sensing data such as satellite remote sensing images, unmanned aerial vehicle remote sensing images and ground actual observation data acquired every day, nationwide reservoir dynamic detection is realized and information such as reservoir types, areas, capacities, distribution characteristics, hazard development rules, dynamic change trends and influences on the surrounding environment are updated according to the deep learning reservoir target detection method.
S7, uploading the reservoir identification model, the reservoir dynamic information and the updated reservoir dynamic information to a cloud platform, and completing reservoir dynamic monitoring based on multi-source data, wherein the cloud platform comprises two working modes: reservoir remote sensing-network cooperative monitoring and network-remote sensing cooperative monitoring.
In the embodiment, the built deep learning model and the daily updated remote sensing data are uploaded to the built cloud platform, the updated reservoir information is automatically acquired in the network every day through a crawler technology, the acquired remote sensing images are uploaded to the cloud platform, and the monitoring and updating of the data are automatically completed in the cloud platform. The method integrates remote sensing monitoring and ubiquitous network monitoring, creates a new way for nationwide reservoir dynamic monitoring, establishes a nationwide reservoir dynamic active monitoring system, actively discovers, tracks and locks reservoir construction and management hotspots and key targets, and realizes automatic extraction of deep reservoir information. According to the remote sensing image and the primary and secondary scores of the ubiquitous network, two working modes of reservoir remote sensing-ubiquitous network cooperative monitoring and ubiquitous network-remote sensing cooperative monitoring can be divided.
In this embodiment, the use of the reservoir update data:
(1) When rainstorm comes temporarily, critical rainfall and reservoir accumulated rainfall process lines can be determined by the aid of the simulation analysis module according to pre-rain information and post-rain information stored in the reservoir database, and then the real-time early warning module judges and gives early warning according to the critical rainfall and the reservoir accumulated rainfall process lines, so that accurate, timely and effective early warning data can be provided for rainstorm early warning, and the problems of untimely early warning and early warning transition are avoided. The timely monitoring of the water level information is beneficial to peak clipping and peak shifting, dam protection and flood discharge, and casualties and huge property loss are avoided. Most small reservoirs lack monitoring facilities and past observation data, so that many reservoirs cannot compile scheduling plans, and can only be temporarily scheduled according to needs or experience, and serious potential safety hazards exist. The road is easy to collapse in flood season, people cannot be guaranteed to pass when the reservoir is in a dangerous condition, and the rescue opportunity is easy to be delayed if necessary survey and report is lacked, so that the information construction problem of reservoir safety is particularly important.
(2) With the continuous development of the technology, reservoir facilities are inconsistent, the construction time is different, and the quality is uneven, so that the dam safety monitoring is carried out on the reservoir by using the monitoring means, the personnel investment can be reduced, the deformation quality information of the reservoir can be timely obtained, and the early warning can be timely carried out once the deformation information reaches the threshold value, so that the maintenance or reconstruction can be conveniently carried out by related personnel, and the existence of potential safety hazards can be avoided.
(3) Because most reservoirs are wide in distribution areas and are located in remote mountain areas, the traffic and communication facilities of the reservoirs fall behind, and maintainers cannot reach target areas in time to carry out daily maintenance detection on the reservoirs.
(4) The reservoir plays a role in irrigation, power generation, breeding, travel, ecology and the like which is difficult to replace. The method has the advantages that the water and soil loss type, the strength and the distribution characteristics, the damage and the influence condition, the occurrence and development rule and the dynamic change trend of the target area are known, and the method has important significance for the comprehensive treatment of the reservoir, the macroscopic decision of ecological environment construction and the scientific, reasonable and systematic formulation of various measures. The method can be used for timely mastering the pollution condition of the reservoir, and at the initial stage of pollution condition discovery, the method can be used for formulating solution measures according to the types of pollutants, timely adsorbing various pollutants in the water, achieving the expected purification purpose, eliminating water pollution and having important significance in the aspects of irrigation, power generation, cultivation, tourism, ecology and the like.
The invention provides a nationwide full-factor reservoir information database; carrying out reservoir target detection by using an object-oriented classification method in combination with a deep learning technology and giving reservoir attribute information; and updating the reservoir information database according to two methods of remote sensing monitoring and ubiquitous network monitoring. The dynamic active monitoring method for the nationwide reservoir can actively discover, track and lock reservoir construction and management hotspots and key targets, realize automatic extraction of deep reservoir information, and provide great convenience for people to use water conservancy information.
Claims (8)
1. A reservoir dynamic monitoring system based on multi-source data is characterized by comprising a data acquisition module, a multi-source remote sensing data sample module, a deep learning module, a reservoir dynamic text updating module, a reservoir dynamic monitoring module and a cloud platform;
the data acquisition module is used for acquiring constructed and under-construction reservoir information and constructing a full-factor reservoir database according to the reservoir information;
the multi-source remote sensing data sample module is used for carrying out target detection on the reservoir by using an object-oriented method according to the remote sensing image and constructing a reservoir-based multi-source remote sensing data sample library according to the detected target reservoir;
the deep learning module is used for constructing a deep learning model by using a deep learning method, training the deep learning model by using a reservoir multi-source remote sensing data sample library to obtain a reservoir identification model, carrying out reservoir detection on a remote sensing image by using the reservoir identification model, and endowing attribute information of a full-element reservoir database to the detected reservoir;
the reservoir dynamic text updating module is used for crawling dynamic information of the reservoir according to the reservoir detection result to obtain updated reservoir information;
the reservoir dynamic monitoring module is used for dynamically monitoring the updated reservoir information by combining the remote sensing image;
the cloud platform is used for completing reservoir dynamic monitoring based on multi-source data according to the reservoir identification model, the reservoir dynamic information and the updated reservoir dynamic information, wherein,
and according to the collected water conservancy project information built and under construction in the whole country, constructing a full-factor reservoir database according to the name, the geographical position, the engineering construction condition, the construction purpose, the project subject, the engineering investment, the operation condition and the major event.
2. A reservoir dynamic monitoring method based on multi-source data is characterized by comprising the following steps:
s1, acquiring constructed and under-constructed reservoir information, and constructing a full-factor reservoir database according to the reservoir information;
s2, carrying out target detection on the reservoir by using an object-oriented method according to the remote sensing image, and constructing a reservoir-based multi-source remote sensing data sample library according to the detected target reservoir;
s3, constructing a deep learning model by using a deep learning method;
s4, training a deep learning model by utilizing a reservoir multi-source remote sensing data sample library to obtain a reservoir identification model;
s5, performing reservoir detection on the remote sensing image by using the reservoir identification model, and endowing attribute information of the full-element reservoir database on the detected reservoir;
s6, crawling dynamic reservoir information according to the reservoir detection result to obtain updated reservoir information, and combining the remote sensing image to dynamically monitor the updated reservoir information;
s7, uploading the reservoir identification model, the reservoir dynamic information and the updated reservoir dynamic information to a cloud platform to complete reservoir dynamic monitoring based on multi-source data, wherein,
and according to the collected water conservancy project information built and under construction in the whole country, constructing a full-factor reservoir database according to the name, the geographical position, the engineering construction condition, the construction purpose, the project subject, the engineering investment, the operation condition and the major event.
3. The dynamic reservoir monitoring method based on multi-source data according to claim 2, wherein the step S2 comprises the following steps:
s201, identifying a reservoir region in the remote sensing image by using an object-oriented method according to the remote sensing image;
s202, cutting the reservoir area into a data set with the size of 224 x 224, and constructing a reservoir multi-source remote sensing data sample library.
4. The method for dynamically monitoring the reservoir based on the multi-source data according to claim 3, wherein the remote sensing image in the step S201 comprises a remote sensing image generated by a multi-scale segmentation method and a remote sensing image extracted by fuzzy logic classification.
5. The multi-source data-based reservoir dynamic monitoring method of claim 4, wherein the deep learning model comprises an input layer, a first convolutional layer connected to the input layer, a first pooling layer connected to the first convolutional layer, a second convolutional layer connected to the first pooling layer, a second pooling layer connected to the second convolutional layer, a third convolutional layer connected to the second pooling layer, a third pooling layer connected to the third convolutional layer, a fourth convolutional layer connected to the third pooling layer, a fourth pooling layer connected to the fourth convolutional layer, a fifth convolutional layer connected to the fourth pooling layer, a fifth pooling layer connected to the fifth convolutional layer, a first fully-connected layer connected to the fifth pooling layer, a second fully-connected layer connected to the first fully-connected layer, a third fully-connected layer connected to the second fully-connected layer, a sixth convolutional layer connected to the input layer, a sixth convolutional layer connected to the sixth convolutional layer, a sixth convolutional layer connected to the sixth pooling layer, a seventh fully-connected to the sixth pooling layer, a fourth fully-connected to the fifth fully-connected layer, a seventh fully-connected to the fifth convolutional layer, and a seventh fully-connected to the seventh convolutional layer; the fifth fully-connected layer is connected with the third fully-connected layer.
6. The multi-source data-based dynamic reservoir monitoring method according to claim 5, wherein the convolution kernel sizes of the first convolution layer, the second convolution layer, the third convolution layer, the fourth convolution layer, the fifth convolution layer, the sixth convolution layer and the seventh convolution layer are all 3 x 3, the step lengths are all 1, and the effective filling size is all 1;
the sizes of convolution kernels of the first pooling layer, the second pooling layer, the third pooling layer, the fourth pooling layer, the fifth pooling layer, the sixth pooling layer and the seventh pooling layer are all 2 x 2.
7. The multi-source data-based dynamic reservoir monitoring method of claim 6, wherein the number of convolution kernels of the first convolution layer and the sixth convolution layer is 64, the number of convolution kernels of the second convolution layer and the seventh convolution layer is 128, the number of convolution kernels of the third convolution layer is 256, and the number of convolution kernels of the fourth convolution layer and the fifth convolution layer is 512.
8. The dynamic reservoir monitoring method based on multi-source data according to claim 2, wherein the cloud platform in the step S7 comprises two working business modes of reservoir remote sensing-network cooperative monitoring and network-remote sensing cooperative monitoring.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110058297.7A CN112766146B (en) | 2021-01-14 | 2021-01-14 | Multi-source data-based dynamic reservoir monitoring system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110058297.7A CN112766146B (en) | 2021-01-14 | 2021-01-14 | Multi-source data-based dynamic reservoir monitoring system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112766146A CN112766146A (en) | 2021-05-07 |
| CN112766146B true CN112766146B (en) | 2022-11-29 |
Family
ID=75702150
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110058297.7A Active CN112766146B (en) | 2021-01-14 | 2021-01-14 | Multi-source data-based dynamic reservoir monitoring system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112766146B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109740485A (en) * | 2018-12-27 | 2019-05-10 | 中国水利水电科学研究院 | Reservoir or dyke recognition methods based on spectrum analysis and depth convolutional neural networks |
| CN111199195A (en) * | 2019-12-26 | 2020-05-26 | 中科禾信遥感科技（苏州）有限公司 | Pond state full-automatic monitoring method and device based on remote sensing image |
| CN111489370A (en) * | 2020-03-29 | 2020-08-04 | 复旦大学 | Remote sensing image segmentation method based on deep learning |
| CN111666909A (en) * | 2020-06-11 | 2020-09-15 | 中科山水(北京)科技信息有限公司 | Suspected contaminated site space identification method based on object-oriented and deep learning |
| WO2020206914A1 (en) * | 2019-04-11 | 2020-10-15 | 平安科技（深圳）有限公司 | Visualization method and device for multi-source earth observation image processing, and storage medium |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110009213B (en) * | 2019-03-28 | 2021-10-22 | 交通运输部水运科学研究所 | Method for tracking, monitoring and evaluating ecological influence of channel engineering based on long-time sequence satellite remote sensing data |
-
2021
- 2021-01-14 CN CN202110058297.7A patent/CN112766146B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109740485A (en) * | 2018-12-27 | 2019-05-10 | 中国水利水电科学研究院 | Reservoir or dyke recognition methods based on spectrum analysis and depth convolutional neural networks |
| WO2020206914A1 (en) * | 2019-04-11 | 2020-10-15 | 平安科技（深圳）有限公司 | Visualization method and device for multi-source earth observation image processing, and storage medium |
| CN111199195A (en) * | 2019-12-26 | 2020-05-26 | 中科禾信遥感科技（苏州）有限公司 | Pond state full-automatic monitoring method and device based on remote sensing image |
| CN111489370A (en) * | 2020-03-29 | 2020-08-04 | 复旦大学 | Remote sensing image segmentation method based on deep learning |
| CN111666909A (en) * | 2020-06-11 | 2020-09-15 | 中科山水(北京)科技信息有限公司 | Suspected contaminated site space identification method based on object-oriented and deep learning |
Non-Patent Citations (3)
| Title |
|---|
| Convolutional Neural Network based Deep Feature Extraction in Remote Sensing Images;P. Dharani Devi et al;《2020 International Conference on Smart Electronics and Communication (ICOSEC)》;20201007;441-444页 * |
| 基于BP神经网络算法的密云水库水质参数反演研究;马丰魁等;《生态环境学报》;20200331;第29卷(第3期);569-579页 * |
| 黄河水沙关系协调度与骨干水库的调节作用;张金良等;《水利学报》;20200723;1-10页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112766146A (en) | 2021-05-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108710625B (en) | Automatic thematic knowledge mining system and method | |
| CN106651211B (en) | Flood disaster risk assessment method for different scale areas | |
| CN110163472B (en) | Large-range extreme drought emergency monitoring and influence evaluation method and system | |
| CN106021508A (en) | Sudden event emergency information mining method based on social media | |
| Hooshangi et al. | Evaluation of potential sites in Iran to localize solar farms using a GIS-based Fermatean Fuzzy TOPSIS | |
| CN111968019A (en) | Landslide hazard susceptibility prediction method based on CNN and RNN integration | |
| CN115983522B (en) | Rural habitat quality assessment and prediction method | |
| Abdelmoty et al. | Deduction and deductive databases for geographic data handling | |
| CN112750138A (en) | Method for identifying space distribution of silt dam of yellow river basin | |
| Balti et al. | Big data based architecture for drought forecasting using LSTM, ARIMA, and Prophet: Case study of the Jiangsu Province, China | |
| Tayal et al. | To Predict the Fire Outbreak in Australia using Historical Database | |
| CN113780175B (en) | Remote sensing identification method for typhoon and storm landslide in high vegetation coverage area | |
| Mohan et al. | A brief review of recent developments in the integration of deep learning with GIS | |
| Beshir et al. | Trend analysis, past dynamics and future prediction of land use and land cover change in upper Wabe-Shebele river basin | |
| CN112766146B (en) | Multi-source data-based dynamic reservoir monitoring system and method | |
| Hu et al. | Urban landscape information atlas and model system based on remote sensing images | |
| CN117151499A (en) | Monitoring and evaluating method and system for homeland space planning | |
| Zhou | Application of artificial intelligence in geography | |
| CN115169718A (en) | Cellular automaton-based regional landslide risk dynamic prediction method and device | |
| Chao | Machine learning-based intelligent weather modification forecast in smart city potential area | |
| Shi et al. | Application of artificial neural network approach and remotely sensed imagery for regional eco-environmental quality evaluation | |
| Xiao et al. | Remote Sensing Image-Based Comprehensive Monitoring Detection Platform for Coastal Tidal Mudflat Ecological Development | |
| CN112231301A (en) | Yellow river water sand change data warehouse | |
| Liu et al. | Design and application of rural road lifecycle management information system | |
| Getachew et al. | Modeling Land-Use/Land-Cover Change, Using Multi-Layer Perceptron and Markov Chain Analysis: A Study on Bahir Dar City, Ethiopia |
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 |