CN113686788A - Conventional water quality monitoring system and method based on remote sensing wave band combination - Google Patents
Conventional water quality monitoring system and method based on remote sensing wave band combination Download PDFInfo
- Publication number
- CN113686788A CN113686788A CN202111097212.2A CN202111097212A CN113686788A CN 113686788 A CN113686788 A CN 113686788A CN 202111097212 A CN202111097212 A CN 202111097212A CN 113686788 A CN113686788 A CN 113686788A
- Authority
- CN
- China
- Prior art keywords
- remote sensing
- layer
- data
- water quality
- wave band
- 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.)
- Granted
Links
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 title claims abstract description 71
- 238000012544 monitoring process Methods 0.000 title claims abstract description 59
- 238000000034 method Methods 0.000 title claims abstract description 46
- 238000012417 linear regression Methods 0.000 claims abstract description 51
- 238000012545 processing Methods 0.000 claims abstract description 18
- 238000007781 pre-processing Methods 0.000 claims abstract description 13
- 238000006243 chemical reaction Methods 0.000 claims abstract description 11
- 238000013079 data visualisation Methods 0.000 claims abstract description 9
- IJGRMHOSHXDMSA-UHFFFAOYSA-N Atomic nitrogen Chemical compound N#N IJGRMHOSHXDMSA-UHFFFAOYSA-N 0.000 claims description 18
- 238000004422 calculation algorithm Methods 0.000 claims description 18
- 230000002159 abnormal effect Effects 0.000 claims description 13
- 230000008569 process Effects 0.000 claims description 13
- XKMRRTOUMJRJIA-UHFFFAOYSA-N ammonia nh3 Chemical compound N.N XKMRRTOUMJRJIA-UHFFFAOYSA-N 0.000 claims description 12
- 238000012549 training Methods 0.000 claims description 10
- OAICVXFJPJFONN-UHFFFAOYSA-N Phosphorus Chemical compound [P] OAICVXFJPJFONN-UHFFFAOYSA-N 0.000 claims description 9
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 claims description 9
- 229910052757 nitrogen Inorganic materials 0.000 claims description 9
- 229910052760 oxygen Inorganic materials 0.000 claims description 9
- 239000001301 oxygen Substances 0.000 claims description 9
- 229910052698 phosphorus Inorganic materials 0.000 claims description 9
- 239000011574 phosphorus Substances 0.000 claims description 9
- 238000001514 detection method Methods 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 6
- 230000001133 acceleration Effects 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 4
- 230000004927 fusion Effects 0.000 claims description 4
- 238000003702 image correction Methods 0.000 claims description 4
- 230000009466 transformation Effects 0.000 claims description 4
- 238000002372 labelling Methods 0.000 claims description 3
- 208000009934 Huntington disease-like 1 Diseases 0.000 claims description 2
- 101100339395 Hydra viridissima Hln-1 gene Proteins 0.000 claims description 2
- 101100339396 Hydra viridissima Hln-2 gene Proteins 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 6
- 238000007689 inspection Methods 0.000 description 6
- 238000012795 verification Methods 0.000 description 5
- 230000005540 biological transmission Effects 0.000 description 4
- 230000003595 spectral effect Effects 0.000 description 4
- 230000003993 interaction Effects 0.000 description 3
- 238000012552 review Methods 0.000 description 3
- 102000006463 Talin Human genes 0.000 description 2
- 108010083809 Talin Proteins 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000001228 spectrum Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 238000010200 validation analysis Methods 0.000 description 2
- 238000010220 Pearson correlation analysis Methods 0.000 description 1
- RTAQQCXQSZGOHL-UHFFFAOYSA-N Titanium Chemical compound [Ti] RTAQQCXQSZGOHL-UHFFFAOYSA-N 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013502 data validation Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
- G01N21/35—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light
- G01N21/3577—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light for analysing liquids, e.g. polluted water
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N2021/1765—Method using an image detector and processing of image signal
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N2021/1793—Remote sensing
-
- 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
- Y02A20/00—Water conservation; Efficient water supply; Efficient water use
- Y02A20/152—Water filtration
Abstract
The invention provides a conventional water quality monitoring system and method based on remote sensing waveband combination, wherein the system comprises: a hardware layer, a logic layer, a network layer and an application layer; the hardware layer is connected with the logic layer, the logic layer is connected with the network layer, and the network layer is connected with the application layer; the hardware layer comprises a plurality of satellites, and remote sensing waveband data is obtained through the satellites; remote sensing preprocessing is carried out on the remote sensing waveband data through a logic layer, and cloud picture reexamination processing is carried out on the filtered remote sensing waveband data through a parallel thread pool to obtain remote sensing image data; the network layer calculates the remote sensing image data based on a Bayesian linear regression model to obtain ground data; the application layer performs real-time or timed data visualization conversion on the ground data and the remote sensing image data, acquires water quality monitoring index data and stores the data. The invention can ensure the universality, accuracy, reliability and stability of the system.
Description
Technical Field
The invention relates to the technical field of water quality monitoring, in particular to a conventional water quality monitoring system and method based on remote sensing waveband combination.
Background
With the continuous development of remote sensing technology, the remote sensing technology is more and more applied to the field of water quality monitoring. The remote sensing technology is used as a regional monitoring means, and can overcome the defects of the conventional water quality monitoring method. The remote sensing data is used for calculating the water quality parameters, and the process of calculation is carried out according to a preset inversion model.
However, in the prior art, the conventional water quality data is inverted mainly by using single-spectrum, multi-spectrum and other satellite remote sensing data, and the problems exist: (1) the inversion model is simple to select, most models consider linear fitting methods, one part uses an exponential fitting method, and the other part selects a basic model of a machine learning algorithm. The simple model is used for water quality fitting (even if a plurality of simple models are parallel), so that the application range is narrow, the universality is poor, finally, only one or two types of water quality inversion can be realized, and the situation of special water quality (such as abnormal water quality) cannot be adapted; (2) the remote sensing wave band used by the inversion model is single, most of the prior art selects a certain wave band with strongest correlation or a simple combination of one wave band and two wave bands, for example, a sentinel 2 satellite has 12 wave bands to select, the utilization rate of model data is low, if the error of the selected single wave band or one wave band and two wave bands is large, for example, the selected wave band does not well reflect the condition that the acquisition section has errors (other wave bands which are not selected can often reflect), when the step of cloud picture re-inspection is not carried out, the wave band data is used for carrying out conventional water quality fitting, the model fitting can be greatly interfered, and an error item caused by the condition cannot be corrected, only the whole group of data can be removed, so that the accuracy of the model is reduced, and the fitting goodness is poor; (3) the method is characterized in that a cloud picture reexamination step is lacked in the wave band data input into the inversion model, so that the model is often fitted or trained by using wrong remote sensing data, and the fitting or training direction of the whole model is influenced by points on one or more wrong sections. And only the wave band data is analyzed, whether in a table form or a database form, the result which is not subjected to visual verification by using a satellite map is unreliable, and the result output by the model is greatly different from the actual result by adopting the fitting of the wave band data and has no reliability.
Disclosure of Invention
Based on this, it is necessary to provide a conventional water quality monitoring system and method based on remote sensing band combination to solve the above technical problems.
A conventional water quality monitoring system based on remote sensing wave band combination comprises: a hardware layer, a logic layer, a network layer and an application layer; the hardware layer is connected with the logic layer, the logic layer is connected with the network layer, and the network layer is connected with the application layer; the hardware layer comprises a plurality of satellites, and remote sensing waveband data is obtained through the satellites and sent to the logic layer; the logic layer carries out remote sensing preprocessing on the remote sensing waveband data, and carries out cloud picture reexamination processing on the filtered remote sensing waveband data through a parallel thread pool to obtain remote sensing image data and send the remote sensing image data to the network layer; the network layer calculates the remote sensing image data based on a Bayesian linear regression model, obtains ground data, and sends the ground data and the remote sensing image data to the application layer; and the application layer performs real-time or timed data visualization conversion on the ground data and the remote sensing image data, acquires water quality monitoring index data and stores the data.
In one embodiment, the method further comprises the following steps: and the hardware layer, the logic layer, the network layer and the application layer are all connected with the auxiliary layer, and the auxiliary layer is used for providing data support and acceleration service.
In one embodiment, the plurality of satellites are high-rank satellites, resource-series satellites, sentinel-series satellites, and other camping satellites with detection bands located in visible light bands or infrared bands.
In one embodiment, the remote sensing preprocessing comprises: image scaling, image enhancement, image correction, orthorectification, image mosaicing, data fusion, image transformation, information extraction, and content classification.
In one embodiment, the logic layer is provided with a parallel thread pool, the parallel thread pool can be distributed to a plurality of processes for cloud image re-inspection processing, and the processes comprise three threads which are respectively a first thread, a second thread and a third thread; the thread conducts cloud picture reexamination from the first section of the remote sensing image data; sensing whether the first section is a cloud point or not through a first thread, if so, performing cloud point marking on the first section, sensing whether the first section is a shore point or not through a second thread, and if so, performing shore point marking on the first section; sensing whether the first section is an abnormal point or not through a third thread, and if so, carrying out abnormal point marking on the first section; meanwhile, manual labeling is assisted, and the network layer ignores labeled cloud points, shore points and abnormal points during calculation.
In one embodiment, the network Layer is formed by a bayesian linear regression model, the bayesian linear regression model includes an input Layer, an output Layer and Hidden layers, the size of the input Layer is the number of remote sensing bands, the size of the output Layer is the number of water quality monitoring indexes, the number of the Hidden layers is calculated by using a Hidden Layer number algorithm, and the formula is as follows:
HLN2=[log2IN+5];(2)
HLN1 and HLN2 are calculation results of a root equation method and a logarithm method respectively, HLN is the size of a hidden layer of final output, IN is the size of an input layer, ON is the size of an output layer, and an operator "[ x ]" indicates that x is rounded.
In one embodiment, the water quality monitoring indicators include dissolved oxygen, conductivity, turbidity, permanganate, ammonia nitrogen, total phosphorus and total nitrogen; when the auxiliary layer is not adopted, the model and the wave band combination of the dissolved oxygen, the conductivity, the turbidity, the ammonia nitrogen and the total phosphorus are Bayes linear regression + full wave band, the model and the wave band combination of the permanganate are Bayes linear regression + inter-wave band values, and the model and the wave band combination of the total nitrogen are the result of feeding back the Bayes linear regression + full wave band + index ammonia nitrogen.
In one embodiment, when the auxiliary layer is used, the model of conductivity and band combination becomes bayesian linear regression + band centered at 650 nm/band centered at 700nm + (band centered at 750 nm-band centered at 800 nm) + other bands; a Bayesian linear regression device with a three-layer network structure is adopted to take wave band combination as input, take conventional water quality monitoring indexes as output, take all the input as a training set, and obtain ground data through fitting.
In one embodiment, the application layer comprises a real-time DV converter and a timing DV converter, and the remote sensing image data and the ground data are converted and stored through the real-time DV converter or the timing DV converter; the real-time DV converter comprises a delayer, a remote sensing image format reader, a ground data format reader, a prediction reading algorithm/merging algorithm, a format merger, a universal converter, high-performance supporting equipment or an integrated environment, a universal converter and a display; the timing DV converter comprises a decision device, a remote sensing image format reader, a ground data format reader, a preset reading algorithm/merging algorithm, a format merger, a universal converter and display equipment.
A conventional water quality monitoring method based on remote sensing wave band combination comprises the following steps: acquiring remote sensing waveband data through a plurality of satellites of a hardware layer, and sending the remote sensing waveband data to the logic layer; performing remote sensing preprocessing on the remote sensing waveband data according to the logic layer, performing cloud picture reexamination processing on the filtered remote sensing waveband data through a parallel thread pool, acquiring remote sensing image data, and sending the remote sensing image data to a network layer; in the network layer, calculating the remote sensing image data based on a Bayesian linear regression model to obtain ground data, and sending the ground data and the remote sensing image data to the application layer; and performing real-time or timed data visualization conversion on the ground data and the remote sensing image data according to the application layer, acquiring and storing water quality monitoring index data.
Compared with the prior art, the invention has the advantages and beneficial effects that:
1. the invention provides and realizes a rapid and feasible conventional water quality index monitoring model based on remote sensing satellite waveband combination and an integral system operating by depending on the model. Data processing is carried out through a Bayesian linear regression model, and cloud picture reexamination processing is realized through a parallel pool, so that the universality, accuracy, reliability and stability of the system can be ensured.
2. The invention has the advantages of no damage, high speed, strong applicability and reusability, and can be widely used in the aspects of environmental protection construction, city planning, emergency monitoring, agricultural supervision, industrial layout planning and the like.
3. The logic layer of the invention adopts the cloud picture reinspection technology, so that the network layer can obtain high-precision input data, thereby reducing the loss caused by the deviation of the training direction, and the output water quality monitoring index data has higher reliability and stability.
4. The invention can carry out accurate parameter adjustment on different water quality indexes through the network layer, has wide application range and strong universality, and can realize inversion of various types of water quality; meanwhile, a wave band combination method is adopted, the auxiliary layer is combined, remote sensing wave band data are fully utilized, the input layer of the Bayesian linear regression model is expanded, and the accuracy and the fitting goodness of the model are improved.
Drawings
FIG. 1 is a schematic structural diagram of a conventional water quality monitoring system based on remote sensing band combination in one embodiment;
FIG. 2 is a schematic diagram of a schematic structure of a conventional water quality monitoring system based on remote sensing band combination in one embodiment;
FIG. 3 is a schematic diagram of a re-inspection of the cloud chart of FIG. 2;
FIG. 4 is a schematic diagram of a real-time DV converter in one embodiment;
FIG. 5 is a block diagram of a timing DV converter in one embodiment;
FIG. 6 is a schematic flow chart of a conventional water quality monitoring method based on remote sensing band combination in one embodiment.
In the figure, a hardware layer 10, a logic layer 20, a network layer 30, an application layer 40, a real-time DV converter 41, a timing DV converter 42, and an auxiliary layer 50.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings by way of specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
In one embodiment, as shown in fig. 1 and 2, there is provided a conventional water quality monitoring system based on remote sensing band combination, including: a hardware layer 10, a logic layer 20, a network layer 30, and an application layer 40; the hardware layer 10 is connected with the logic layer 20, the logic layer 20 is connected with the network layer 30, and the network layer 30 is connected with the application layer 40; the hardware layer 10 comprises a plurality of satellites, and remote sensing waveband data is obtained through the satellites and sent to the logic layer 20; the logic layer 20 performs remote sensing preprocessing on the remote sensing band data, performs cloud image re-inspection processing on the filtered remote sensing band data through the parallel thread pool, acquires remote sensing image data, and sends the remote sensing image data to the network layer 30; the network layer 30 calculates the remote sensing image data based on the bayesian linear regression model, obtains ground data, and sends the ground data and the remote sensing data to the application layer 40; the application layer 40 performs real-time or timed data visualization conversion on the ground data and the remote sensing image data to acquire and store water quality monitoring index data.
In this embodiment, the hardware layer 10 is connected to the logic layer 20, the logic layer 20 is connected to the network layer 30, the network layer 30 is connected to the application layer 40, the hardware layer 10 includes a plurality of satellites, and the remote sensing band data is obtained through the plurality of satellites and sent to the logic layer 20; the logic layer 20 performs remote sensing preprocessing on the remote sensing band data, performs cloud image re-inspection processing on the filtered remote sensing band data through the parallel thread pool, acquires remote sensing image data, and sends the remote sensing image data to the network layer 30; the network layer 30 calculates the remote sensing image data based on the bayesian linear regression model, obtains ground data, and sends the ground data and the remote sensing data to the application layer 40; the application layer 40 performs real-time or timed data visualization conversion on the ground data and the remote sensing image data to acquire and store water quality monitoring index data, performs data processing through a Bayesian linear regression model, and simultaneously adopts a parallel pool to realize cloud image reexamination, thereby ensuring the universality, accuracy, reliability and stability of the system.
Wherein, still include: the auxiliary layer 50, the hardware layer 10, the logic layer 20, the network layer 30 and the application layer 40 are all connected with the auxiliary layer 50, and the auxiliary layer 50 is used for providing data support and acceleration services.
Specifically, the auxiliary layer 50 can provide data support or acceleration services, etc. for the hardware layer 10, the logic layer 20, the network layer 30, and the application layer 40 using various public, common resources, and the deletion of the auxiliary layer 50 only affects the system performance without disabling the system.
In particular, with respect to the hardware layer 10 and the logic layer 20, the auxiliary layer 50 can additionally provide a wired or wireless network transmission protocol (TCP/IP) for enhanced data interaction, so that the system can obtain a separable characteristic and a long-distance transmission characteristic without depending on a limitation of a data transmission line.
For the logic layer 20, the auxiliary layer 50 can additionally provide a verification method, pearson correlation verification, for further reducing errors caused by the model input remote sensing image data.
After the pearson correlation verification is carried out on the remote sensing image data, the correlation of the second waveband of the sentinel with the conventional water quality is shown in the following table 1 according to the strong and weak sequence:
TABLE 1 correlation between each water quality monitoring index and corresponding band
Index (I) | Correlation>0.8 | Correlation>0.7 | Correlation>0.6 | Correlation>0.5 | Correlation>0.4 |
O2 | Band_3 | Band_4 | Band_2 | Band_5 | Band_1 |
ELE | Band_4 | Band_5 | Band_6 | Band_7 | Band_1 |
NTU | Band_5 | Band_4 | Band_3 | Band_2 | Band_6 |
Mn | Band_5 | Band_6 | Band_4 | Band_7 | Band_8 |
N | Band_9 | Band_11 | Band_8a | Band_1 | / |
TP | Band_5 | Band_4 | Band_6 | Band_7 | Band_3 |
TN | Band_4 | Band_5 | Band_8 | Band_7 | Band_6 |
For the network layer 30, the auxiliary layer 50 can provide more remote sensing data and ground data as a bayesian training set for further improving the accuracy of the model.
For the application layer 40, the auxiliary layer 50 can provide ground station data to stabilize the source of the model input, so that the system obtains stable input, thereby ensuring the stability of the output of the system and improving the automation degree of the system.
The plurality of satellites are high-resolution serial satellites, resource series satellites, sentinel series satellites and other camping satellites with detection bands located in visible light bands or infrared bands.
Specifically, the plurality of satellites are: the existing high-resolution series satellites (including a first high-resolution satellite, a second high-resolution satellite, a third high-resolution satellite, a fourth high-resolution satellite and the satellites which are put into operation in the future), the resource series satellites (the first resource, the second resource, the third resource and the satellites which are put into operation in the future), the sentry series satellites (the second sentry, the third sentry, the fourth sentry and the satellites which are put into operation in the future) and the existing high-resolution series satellites which are not in the list but have detection bands located in visible light bands or infrared bands.
In addition to satellites, other detection instruments using spectral loading techniques are also included: unmanned (Da Jiang DJI, Chinese remote, Yihang, and future, operational), ground spectrometer (shared, SPECTRO, and future, operational), and off the list above, spectral instrumentation that can use spectral loading techniques to obtain spectral data in the visible to infrared bands.
Wherein, the remote sensing preliminary treatment includes: image scaling, image enhancement, image correction, orthorectification, image mosaicing, data fusion, image transformation, information extraction, and content classification.
Specifically, after the remote sensing band data is acquired from the hardware layer 10, remote sensing preprocessing needs to be performed on the remote sensing band data, and the related technologies mainly include image scaling, image enhancement, image correction, orthorectification, image mosaic, data fusion, image transformation, information extraction, and content classification.
The logic layer is provided with a parallel thread pool, and the parallel thread pool can be distributed to a plurality of processes for perception processing; the process comprises three threads, namely a first thread, a second thread and a third thread; the thread starts to sense from a first section of the remote sensing image data; whether the first section is a point on the cloud or not is sensed through the first thread, and if the first section is the point on the cloud, point on the cloud is marked on the first section; sensing whether the first section is a shore point or not through a second thread, and if so, carrying out shore point marking on the first section; sensing whether the first section is an abnormal point or not through a third thread, and if so, carrying out abnormal point marking on the first section; meanwhile, manual labeling is assisted, and the network layer ignores the labeled cloud points, the labeled shore points and the labeled outliers during calculation.
Specifically, as shown in fig. 3, the step performed synchronously in the logic layer 20 is cloud image review, after the remote sensing image data entering the logic layer 20 is captured, a parallel thread pool is started, and the parallel thread pool can be allocated to a plurality of processes for cloud image review processing, where 1 process is taken as an example, in 3 threads of 1 process, starting from a first section of an image, the cloud image review processing process in the parallel pool is:
(1) the computer senses whether the first cross section is a point on the cloud or not from the remote sensing image data through a first thread, and if the first cross section is the point on the cloud, cloud point marking is carried out on the first cross section; meanwhile, the engineering personnel perceive whether the first section is a cloud point or not from the remote sensing image data, and if so, the engineering personnel inform the computer to enable the computer to correspondingly mark the first section in the next judgment;
(2) the computer senses whether the first cross section is a shore point from the remote sensing image data through a second thread, and if the first cross section is the shore point, the shore point marking is carried out on the first cross section; meanwhile, the engineering personnel perceive whether the first section is an onshore point or not from the remote sensing image data, and if so, the engineering personnel inform the computer to enable the computer to correspondingly mark the first section in the next judgment;
(3) the computer senses whether the first section is an abnormal point or not from the remote sensing image data through a third thread, for example, the first section is an on-board point or other aquatic object points, and if the first section is the abnormal point, abnormal point marking is carried out on the first section; and meanwhile, the engineering personnel can perceive whether the first section is an abnormal point or not from the remote sensing image data, and if so, the engineering personnel can inform the computer to correspondingly label the first section in the next judgment.
Through the processing of the cloud picture reexamination, the accuracy of remote sensing image data can be improved, and the accuracy and the reliability of model fitting and training are improved.
Specifically, the logic layer 20 makes full use of the superiority of human-computer interaction, and after the cloud image re-inspection technology is performed, the network layer 30 can obtain high-precision input data, so that loss caused by deviation of a training direction is reduced or even avoided, and the output result of the system has higher reliability and stability.
Wherein, the network Layer 30 comprises Bayesian linear regression model, and Bayesian linear regression model is including input Layer, output Layer and Hidden Layer, and the input Layer size is the number of remote sensing wave band, and the output Layer size is the number of water quality monitoring index, and the number of Hidden Layer uses the number of Hidden Layer Numbers algorithm to calculate, and the formula is as follows:
HLN2=[log2IN+5];(2)
wherein, HLN1And HLN2Computing results of a root method and a logarithm method respectively, HLN is the size of a hidden layer of final output, IN is the size of an input layer, ON is the size of an output layer, and an operator' [ x ]]"means to round off x.
Wherein the water quality monitoring indexes comprise dissolved oxygen, conductivity, turbidity, permanganate, ammonia nitrogen, total phosphorus and total nitrogen; when the auxiliary layer 50 is not adopted, the model and the wave band combination of the dissolved oxygen, the conductivity, the turbidity, the ammonia nitrogen and the total phosphorus are Bayes linear regression + full wave band, the model and the wave band combination of the permanganate are Bayes linear regression + inter-wave band values, and the model and the wave band combination of the total nitrogen are the result of feedback Bayes linear regression + full wave band + index ammonia nitrogen.
Specifically, the water quality monitoring indexes of the embodiment include seven types, namely dissolved oxygen, conductivity, turbidity, permanganate, ammonia nitrogen, total phosphorus and total nitrogen. When the auxiliary layer 50 is adopted or not adopted, the wave band combinations corresponding to the water quality monitoring indexes have differences.
When the auxiliary layer 50 is adopted, the combination of the conductivity model and the wave band becomes Bayesian linear regression + wave band with 650nm center/wave band with 700nm center + (wave band with 750nm center-wave band with 800nm center) + other wave bands; a Bayesian linear regression device with a three-layer network structure is adopted to take wave band combination as input, take conventional water quality monitoring indexes as output, take all the input as a training set, and obtain ground data through fitting.
Specifically, the network Layer 30 is formed by a machine-learned bayesian linear regression model, the bayesian linear regression model includes an input Layer, an output Layer and a Hidden Layer, the size of the input Layer is the number of remote sensing bands obtained according to the remote sensing band data, the size of the output Layer is the number of water quality monitoring indexes, and the number of the Hidden layers is calculated and obtained according to a Hidden Layer number algorithm.
Specifically, when the auxiliary layer 50 is not used, the band combinations of the input layers are as shown in table 2 below:
TABLE 2 combination of model and band corresponding to each water quality monitoring index
Index (I) | Model and wave band combination | Goodness of fit |
O2 | Bayesian linear regression + full band | R2=0.88307 |
ELE | Bayesian linear regression + full band | R2=0.9526 |
NTU | Bayesian linear regression + full band | R2=0.7381 |
Mn | Bayesian linear regression + inter-band values | R2=0.86055 |
N | Bayesian linear regression + full band | R2=0.86288 |
TP | Bayesian linear regression + full band | R2=0.7424 |
TN | Feedback Bayes linear regression + full-wave band + index N result | R2=0.8913 |
When the auxiliary layer 50 is used, the band combination of the index ELE (conductivity) will be changed to the following test result obtained from pearson correlation: bayesian linear regression + band centered at 650 nm/band centered at 700nm + (band centered at 750 nm-band centered at 800 nm) + other bands.
Specifically, a Bayesian regression device with a three-layer network structure is adopted to take wave band combination as input, a ground conventional numerical detection index as output, and the matching test of the model completely depends on ground data, so that the input is totally taken as a training set; the fitted ground data is then presented to the application layer 40 along with the input remote sensing image data.
Specifically, the network layer 30 uses a bayesian linear regression model based on a probability density function, and performs accurate parameter adjustment for different water quality monitoring indexes; the size of a hidden layer of the Bayesian linear regression network model is scientific and reasonable, and the generated remote sensing band-based inversion model theoretically covers common rivers, lakes, reservoirs, ditches, pits and ponds, paddy fields, even black and odorous water areas, dead water and wetlands. Therefore, the model has wide application range and strong universality and can realize various types of water quality inversion.
In addition, the network layer 30 uses a band combination method, and combines with the pearson correlation analysis method of the auxiliary layer 50, so that the remote sensing data is fully utilized, the input layer of the bayesian linear regression model is enlarged, the uncertainty is reduced, and the accuracy and the goodness of fit of the model are increased.
The application layer 40 comprises a real-time DV converter 41 and a timing DV converter 42, and the remote sensing image data and the ground data are visually converted through the real-time DV converter 41 or the timing DV converter 42 to obtain and store water quality monitoring index data; the real-time DV converter 41 includes a delay device, a remote-sensing image format reader, a ground data format reader, a predictive reading algorithm/merging algorithm, a format merger, a universal converter, a high-performance support device or an integrated environment, a universal converter, and a display; the timing DV converter 42 includes a decision device, a remote sensing image format reader, a ground data format reader, a preset reading algorithm/merging algorithm, a format merger, a universal converter, and a display device.
Specifically, the application layer 40 receives a set of remote sensing image data and a set of ground data, and stores the two sets of data in a format specified by the system after the two sets of data are converted by the real-time DV converter 41 or the timing DV converter 42. The format can be converted into a network general format (for example, jpg, png, svg, etc.) through the real-time DV converter 41 or the timing DV converter 42 and displayed, and the display method can be either the network popular visualization (see the figure) software or various display devices on sale.
Referring to fig. 4 and fig. 5, there are schematic structural diagrams of the real-time DV converter 41 and the timing DV converter 42, respectively, wherein the real-time DV converter 41 needs to be supported by a high-performance integrated environment or a high-performance device, for example, a Titan FPGA: Pango PGT180H & Intel Additional I350T2, and the system delay τ is greater than or equal to 3.5 ms.
In one embodiment, as shown in fig. 6, a conventional water quality monitoring method based on remote sensing band combination is provided, which includes the following steps:
step S601, obtaining remote sensing wave band data through a plurality of satellites of a hardware layer, and sending the remote sensing wave band data to a logic layer.
And step S602, performing remote sensing preprocessing on the remote sensing waveband data according to the logic layer, performing cloud picture reexamination processing on the filtered remote sensing waveband data through the parallel thread pool, acquiring remote sensing image data, and sending the remote sensing image data to the network layer.
And step S603, calculating the remote sensing image data based on the Bayesian linear regression model in the network layer, acquiring ground data, and sending the ground data and the remote sensing image data to the application layer.
And step S604, performing real-time or timed data visualization conversion on the ground data and the remote sensing image data according to the application layer, acquiring water quality monitoring index data, and displaying the data.
In this embodiment, remote sensing band data is obtained through a plurality of satellites in a hardware layer, remote sensing preprocessing is performed on the remote sensing band data according to a logic layer, a cloud image reexamination processing is performed on the filtered remote sensing band data through a parallel pool, remote sensing image data is obtained, the remote sensing image data is calculated through a Bayesian linear regression model on a network layer, ground data is obtained, and finally real-time or timed data visualization conversion is performed on the ground data and the remote sensing image data through an application layer, so that water quality monitoring index data is obtained and displayed, and therefore universality, accuracy, reliability and stability of the method can be guaranteed.
In one embodiment, the present invention undergoes data validation and field validation: and selecting the remote sensing data of the sentinel II at 28 days, 7 months and 2021, verifying the data of the model, and comparing the predicted ground data (seven types including dissolved oxygen, conductivity, turbidity, permanganate, ammonia nitrogen, total phosphorus and total nitrogen) results with the actual results to find that the distribution trend of the system output result is the same as that of the actual ground data.
The same data is selected, and as shown in table 3, a Bayesian linear regression model is subjected to grouping comparison verification, in order to verify the Bayesian linear regression model of the method, a comparison model GRNN generalized neural network is introduced, the GRNN generalized neural network is a radial basis network, prediction is carried out according to nonparametric estimation and the maximum probability principle, input and output are extended to a high-dimensional space, the linear relation of the input and output is searched, and the model is an efficient and high-precision curve fitting model.
TABLE 3 original ground data of each water quality monitoring index after grouping
The same input layer size is used and both models are verified with ground data.
Specifically, the validation results showed that, in 7 groups, 49 indexes in total:
(1) the bayesian linear regression model had 39 sets of excellent fitting results, 6 sets of better fitting results, 3 sets of poor fitting results, and 1 set of poor fitting results.
(2) The GRNN high-dimensional projection regression model has 22 sets of excellent fit results, 12 sets of better fit results, 4 sets of poor fit results, and 10 sets of poor fit results.
Obviously, the bayesian linear regression model provided by the invention has more excellent performance than the GRNN high-dimensional projection regression model, and is not compared with a linear model, an exponential model and the like which have simpler structures than the GRNN model.
In one embodiment, taking a set of randomly sampled remote sensing section data (hereinafter referred to as experimental data) as an example, a conventional water quality monitoring model and a conventional water quality monitoring system based on remote sensing waveband combination are adopted to monitor conventional water quality monitoring indexes (dissolved oxygen, conductivity, turbidity, permanganate, ammonia nitrogen, total phosphorus and total nitrogen) of the water quality monitoring model, and an entity water sample is not required to be extracted, and the engineering requirement error is less than or equal to 20%. The specific implementation steps are as follows:
(1) in the hardware layer 10, the system downloads a sentinel second remote sensing image (10m) containing section data by using a wired transmission protocol, and should download data of 12 wave bands in total, as shown in table 4.
Table 4 downloaded section data (part)
(2) In the logic layer 20, the following two partial steps are performed:
the remote sensing image data is subjected to remote sensing preprocessing, so that a good imaging effect is presented.
And secondly, performing cloud picture reexamination on the remote sensing image data.
In the substep II, man-machine interaction operation is executed, the remote sensing image data is processed by the engineering personnel and the system according to the method shown in the figure 3, and when the auxiliary layer 50 is adopted, the system can also screen out the first few strong correlation wave bands corresponding to the water quality monitoring index in the remote sensing image data of the sentinel II by using a Pearson correlation method.
(3) In the network layer 30, the bayesian linear regression model is trained using a previously introduced bayesian training set, and the model is input by the layer size: 11-12, output layer size: 7, obtaining the size of the Hidden Layer by a Hidden Layer number algorithm: 9.
(4) in the application layer 40, the result output from the network layer 30 is subjected to real-time DV conversion, and the result is obtained when τ is 10 ms. When format conversion is performed, to verify the model, the data is converted into an x-xlsx format and stored in a common mechanical hard disk, and a stored file is opened by Microsoft Excel software, which is shown as follows:
TABLE 5 Experimental data and System output results Table (part)
Note: the number of decimal places of experimental data and output results is not just 2, and is intended here for convenience of presentation.
Finally, error detection is carried out on all data, and average value statistics is carried out on errors, so that the obtained average value of the errors is 2.32% and is far less than 20%, and therefore, the engineering requirements are met.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by a computer program, which can be stored in a computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like.
It will be apparent to those skilled in the art that the modules or steps of the invention described above may be implemented in a general purpose computing device, they may be centralized on a single computing device or distributed across a network of computing devices, and optionally they may be implemented in program code executable by a computing device, such that they may be stored on a computer storage medium (ROM/RAM, magnetic disks, optical disks) and executed by a computing device, and in some cases, the steps shown or described may be performed in an order different than that described herein, or they may be separately fabricated into individual integrated circuit modules, or multiple ones of them may be fabricated into a single integrated circuit module. Thus, the present invention is not limited to any specific combination of hardware and software.
The foregoing is a more detailed description of the present invention that is presented in conjunction with specific embodiments, and the practice of the invention is not to be considered limited to those descriptions. For those skilled in the art to which the invention pertains, several simple deductions or substitutions can be made without departing from the spirit of the invention, and all shall be considered as belonging to the protection scope of the invention.
Claims (10)
1. A conventional water quality monitoring system based on remote sensing wave band combination is characterized by comprising: a hardware layer, a logic layer, a network layer and an application layer;
the hardware layer is connected with the logic layer, the logic layer is connected with the network layer, and the network layer is connected with the application layer;
the hardware layer comprises a plurality of satellites, and remote sensing waveband data is obtained through the satellites and sent to the logic layer;
the logic layer carries out remote sensing preprocessing on the remote sensing waveband data, and carries out cloud picture reexamination processing on the filtered remote sensing waveband data through a parallel thread pool to obtain remote sensing image data and send the remote sensing image data to the network layer;
the network layer calculates the remote sensing image data based on a Bayesian linear regression model, obtains ground data, and sends the ground data and the remote sensing image data to the application layer;
and the application layer performs real-time or timed data visualization conversion on the ground data and the remote sensing image data, acquires water quality monitoring index data and stores the data.
2. The conventional water quality monitoring system based on remote sensing wave band combination according to claim 1, characterized by further comprising: and the hardware layer, the logic layer, the network layer and the application layer are all connected with the auxiliary layer, and the auxiliary layer is used for providing data support and acceleration service.
3. The system of claim 1, wherein the hardware layer comprises a high-resolution satellite, a resource satellite, a sentry satellite and other camping satellites with detection bands located in visible or infrared bands.
4. The conventional water quality monitoring system based on remote sensing waveband combination according to claim 1, wherein the remote sensing preprocessing comprises: image scaling, image enhancement, image correction, orthorectification, image mosaicing, data fusion, image transformation, information extraction, and content classification.
5. The conventional water quality monitoring system based on remote sensing waveband combination according to claim 1, wherein a parallel thread pool is arranged on the logic layer, the parallel thread pool can be distributed to a plurality of processes for cloud picture reinspection processing, and the processes comprise three threads which are respectively a first thread, a second thread and a third thread; the thread conducts cloud picture reexamination from the first section of the remote sensing image data;
sensing whether the first section is a point on the cloud or not through a first thread, if so, carrying out point on the cloud on the first section,
sensing whether the first section is a shore point or not through a second thread, and if so, carrying out shore point marking on the first section;
sensing whether the first section is an abnormal point or not through a third thread, and if so, carrying out abnormal point marking on the first section;
meanwhile, manual labeling is assisted, and the network layer ignores labeled cloud points, shore points and abnormal points during calculation.
6. The conventional water quality monitoring system based on remote sensing wave band combination as claimed in claim 1, wherein the network Layer is composed of a Bayesian linear regression model, the Bayesian linear regression model comprises an input Layer, an output Layer and Hidden layers, the input Layer size is the number of remote sensing wave bands, the output Layer size is the number of water quality monitoring indexes, the number of Hidden layers is calculated by using a Hidden Layer number algorithm, and the formula is as follows:
HLN2=[log2IN+5]; (2)
HLN1 and HLN2 are calculation results of a root equation method and a logarithm method respectively, HLN is the size of a hidden layer of final output, IN is the size of an input layer, ON is the size of an output layer, and an operator "[ x ]" indicates that x is rounded.
7. The conventional water quality monitoring system based on remote sensing wave band combination as claimed in claim 2, wherein the water quality monitoring index comprises dissolved oxygen, conductivity, turbidity, permanganate, ammonia nitrogen, total phosphorus and total nitrogen; when the auxiliary layer is not adopted, the model and the wave band combination of the dissolved oxygen, the conductivity, the turbidity, the ammonia nitrogen and the total phosphorus are Bayes linear regression + full wave band, the model and the wave band combination of the permanganate are Bayes linear regression + inter-wave band values, and the model and the wave band combination of the total nitrogen are the result of feeding back the Bayes linear regression + full wave band + index ammonia nitrogen.
8. The conventional water quality monitoring system based on remote sensing wave band combination as claimed in claim 7, wherein when the auxiliary layer is adopted, the model of the conductivity and the wave band combination become Bayesian linear regression + wave band with 650nm center/wave band with 700nm center + (wave band with 750nm center-wave band with 800nm center) + other wave bands; a Bayesian linear regression device with a three-layer network structure is adopted to take wave band combination as input, take conventional water quality monitoring indexes as output, take all the input as a training set, and obtain ground data through fitting.
9. The conventional water quality monitoring system based on remote sensing waveband combination according to claim 1, wherein the application layer comprises a real-time DV converter and a timing DV converter, and the remote sensing image data and the ground data are converted and stored through the real-time DV converter or the timing DV converter;
the real-time DV converter comprises a delayer, a remote sensing image format reader, a ground data format reader, a prediction reading algorithm/merging algorithm, a format merger, a universal converter, high-performance supporting equipment or an integrated environment, a universal converter and a display;
the timing DV converter comprises a decision device, a remote sensing image format reader, a ground data format reader, a preset reading algorithm/merging algorithm, a format merger, a universal converter and display equipment.
10. A conventional water quality monitoring method based on remote sensing wave band combination is characterized by comprising the following steps:
acquiring remote sensing waveband data through a plurality of satellites of a hardware layer, and sending the remote sensing waveband data to the logic layer;
performing remote sensing preprocessing on the remote sensing waveband data according to the logic layer, performing cloud picture reexamination processing on the filtered remote sensing waveband data through a parallel thread pool, acquiring remote sensing image data, and sending the remote sensing image data to a network layer;
in the network layer, calculating the remote sensing image data based on a Bayesian linear regression model to obtain ground data, and sending the ground data and the remote sensing image data to the application layer;
and performing real-time or timed data visualization conversion on the ground data and the remote sensing image data according to the application layer, acquiring and storing water quality monitoring index data.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111097212.2A CN113686788B (en) | 2021-09-18 | 2021-09-18 | Conventional water quality monitoring system and method based on remote sensing wave band combination |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111097212.2A CN113686788B (en) | 2021-09-18 | 2021-09-18 | Conventional water quality monitoring system and method based on remote sensing wave band combination |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113686788A true CN113686788A (en) | 2021-11-23 |
CN113686788B CN113686788B (en) | 2022-08-09 |
Family
ID=78586861
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111097212.2A Active CN113686788B (en) | 2021-09-18 | 2021-09-18 | Conventional water quality monitoring system and method based on remote sensing wave band combination |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113686788B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116910920A (en) * | 2023-09-12 | 2023-10-20 | 陕西万禾数字科技有限公司 | Aeroengine comprehensive health management system and method based on augmented reality technology |
Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106023091A (en) * | 2016-04-22 | 2016-10-12 | 西安电子科技大学 | Image real-time defogging method based on graphics processor |
CN106525762A (en) * | 2016-11-07 | 2017-03-22 | 航天恒星科技有限公司 | Water quality monitoring method and water quality monitoring device based on adaptive model |
US20180007073A1 (en) * | 2016-06-29 | 2018-01-04 | Paypal, Inc. | Network Operation Application Monitoring |
WO2018076885A1 (en) * | 2016-10-31 | 2018-05-03 | 华讯方舟科技有限公司 | Security check method and system based on microwave imaging of human bodies |
CN109297968A (en) * | 2018-11-21 | 2019-02-01 | 河南工业职业技术学院 | A kind of method of generation face domain water quality monitoring result |
CN110347499A (en) * | 2019-06-13 | 2019-10-18 | 武汉大学 | A kind of remote sensing image tile generates and the method for deployed in real time |
CN111007021A (en) * | 2019-12-31 | 2020-04-14 | 北京理工大学重庆创新中心 | Hyperspectral water quality parameter inversion system and method based on one-dimensional convolution neural network |
CN111307727A (en) * | 2020-03-13 | 2020-06-19 | 生态环境部卫星环境应用中心 | Water body water color abnormity identification method and device based on time sequence remote sensing image |
CN112051222A (en) * | 2020-08-30 | 2020-12-08 | 山东锋士信息技术有限公司 | River and lake water quality monitoring method based on high-resolution satellite image |
US10872417B1 (en) * | 2019-07-04 | 2020-12-22 | FlyPard Analytics GmbH | Automatic delineation agricultural field management zones using remote sensing and field data |
CN112816421A (en) * | 2021-01-25 | 2021-05-18 | 中国科学院南京地理与湖泊研究所 | Land-based remote sensing monitoring method for nutritive salt and chemical oxygen demand of water body |
-
2021
- 2021-09-18 CN CN202111097212.2A patent/CN113686788B/en active Active
Patent Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106023091A (en) * | 2016-04-22 | 2016-10-12 | 西安电子科技大学 | Image real-time defogging method based on graphics processor |
US20180007073A1 (en) * | 2016-06-29 | 2018-01-04 | Paypal, Inc. | Network Operation Application Monitoring |
WO2018076885A1 (en) * | 2016-10-31 | 2018-05-03 | 华讯方舟科技有限公司 | Security check method and system based on microwave imaging of human bodies |
CN106525762A (en) * | 2016-11-07 | 2017-03-22 | 航天恒星科技有限公司 | Water quality monitoring method and water quality monitoring device based on adaptive model |
CN109297968A (en) * | 2018-11-21 | 2019-02-01 | 河南工业职业技术学院 | A kind of method of generation face domain water quality monitoring result |
CN110347499A (en) * | 2019-06-13 | 2019-10-18 | 武汉大学 | A kind of remote sensing image tile generates and the method for deployed in real time |
US10872417B1 (en) * | 2019-07-04 | 2020-12-22 | FlyPard Analytics GmbH | Automatic delineation agricultural field management zones using remote sensing and field data |
CN111007021A (en) * | 2019-12-31 | 2020-04-14 | 北京理工大学重庆创新中心 | Hyperspectral water quality parameter inversion system and method based on one-dimensional convolution neural network |
CN111307727A (en) * | 2020-03-13 | 2020-06-19 | 生态环境部卫星环境应用中心 | Water body water color abnormity identification method and device based on time sequence remote sensing image |
CN112051222A (en) * | 2020-08-30 | 2020-12-08 | 山东锋士信息技术有限公司 | River and lake water quality monitoring method based on high-resolution satellite image |
CN112816421A (en) * | 2021-01-25 | 2021-05-18 | 中国科学院南京地理与湖泊研究所 | Land-based remote sensing monitoring method for nutritive salt and chemical oxygen demand of water body |
Non-Patent Citations (3)
Title |
---|
PIERRE COURBIN ·IRINA LUPU ·JOËL GOOSSENS: "Scheduling of hard real-time multi-phase multi-thread (MPMT) periodic tasks", 《REAL-TIME SYST》 * |
YISHAN ZHANG .ET AL: ""Retrieval of Water Quality Parameters from Hyperspectral Images Using Hybrid Bayesian Probabilistic Neural Network"", 《REMOTE SENSING》 * |
季富华: ""农作物类型遥感识别算法及国产高分卫星"", 《中 国 农 业 资 源 与 区 划》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116910920A (en) * | 2023-09-12 | 2023-10-20 | 陕西万禾数字科技有限公司 | Aeroengine comprehensive health management system and method based on augmented reality technology |
CN116910920B (en) * | 2023-09-12 | 2023-12-05 | 陕西万禾数字科技有限公司 | Aeroengine comprehensive health management system and method based on augmented reality technology |
Also Published As
Publication number | Publication date |
---|---|
CN113686788B (en) | 2022-08-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102662949A (en) | Method and system for retrieving specified object based on multi-feature fusion | |
CN111242127A (en) | Vehicle detection method with granularity level multi-scale characteristics based on asymmetric convolution | |
CN110992366B (en) | Image semantic segmentation method, device and storage medium | |
CN116363440B (en) | Deep learning-based identification and detection method and system for colored microplastic in soil | |
CN114998744B (en) | Agricultural machinery track field dividing method and device based on motion and vision dual-feature fusion | |
CN112258254A (en) | Internet advertisement risk monitoring method and system based on big data architecture | |
CN112419268A (en) | Method, device, equipment and medium for detecting image defects of power transmission line | |
CN113686788B (en) | Conventional water quality monitoring system and method based on remote sensing wave band combination | |
Lauko et al. | Local color and morphological image feature based vegetation identification and its application to human environment street view vegetation mapping, or how green is our county? | |
CN114612902A (en) | Image semantic segmentation method, device, equipment, storage medium and program product | |
CN116012653A (en) | Method and system for classifying hyperspectral images of attention residual unit neural network | |
CN114998703A (en) | Remote sensing image change detection method based on high-resolution convolutional neural network | |
CN111651411A (en) | Complex terrain remote sensing soil moisture product downscaling method | |
Han et al. | Statistical analysis of infrared thermogram for CNN-based electrical equipment identification methods | |
CN116563649B (en) | Tensor mapping network-based hyperspectral image lightweight classification method and device | |
Akhyar et al. | A beneficial dual transformation approach for deep learning networks used in steel surface defect detection | |
CN116403071B (en) | Method and device for detecting few-sample concrete defects based on feature reconstruction | |
CN107170004A (en) | To the image matching method of matching matrix in a kind of unmanned vehicle monocular vision positioning | |
CN116681962A (en) | Power equipment thermal image detection method and system based on improved YOLOv5 | |
CN114463764A (en) | Table line detection method and device, computer equipment and storage medium | |
Dai et al. | MDC-Net: A multi-directional constrained and prior assisted neural network for wood and leaf separation from terrestrial laser scanning | |
Havens et al. | A new formulation of the coVAT algorithm for visual assessment of clustering tendency in rectangular data | |
CN113705346B (en) | Remote sensing image ground object classification generalization improving method and device | |
CN110674820A (en) | License plate calibration and identification method and system based on convolutional neural network and electronic equipment | |
CN117351015B (en) | Tamper detection method and system based on edge supervision and multi-domain cross correlation |
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 |