CN110851789B - Island reef shallow sea water depth prediction method based on extreme gradient lifting - Google Patents
Island reef shallow sea water depth prediction method based on extreme gradient lifting Download PDFInfo
- Publication number
- CN110851789B CN110851789B CN201910945145.1A CN201910945145A CN110851789B CN 110851789 B CN110851789 B CN 110851789B CN 201910945145 A CN201910945145 A CN 201910945145A CN 110851789 B CN110851789 B CN 110851789B
- Authority
- CN
- China
- Prior art keywords
- water depth
- shallow sea
- extreme gradient
- reflectivity
- island
- 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
- 238000000034 method Methods 0.000 title claims abstract description 44
- 239000013535 sea water Substances 0.000 title claims abstract description 36
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 179
- 238000002310 reflectometry Methods 0.000 claims abstract description 100
- 239000013049 sediment Substances 0.000 claims abstract description 56
- 238000012549 training Methods 0.000 claims abstract description 20
- 238000003066 decision tree Methods 0.000 claims description 31
- 238000012795 verification Methods 0.000 claims description 17
- 230000006870 function Effects 0.000 claims description 9
- 238000004590 computer program Methods 0.000 claims description 7
- 238000012937 correction Methods 0.000 claims description 5
- 238000007781 pre-processing Methods 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 238000005259 measurement Methods 0.000 description 9
- 238000005516 engineering process Methods 0.000 description 4
- 238000005070 sampling Methods 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 241001481665 Protophormia terraenovae Species 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000004088 simulation Methods 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000001035 drying Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000000691 measurement method Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000005303 weighing Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C13/00—Surveying specially adapted to open water, e.g. sea, lake, river or canal
- G01C13/008—Surveying specially adapted to open water, e.g. sea, lake, river or canal measuring depth of open water
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Mathematical Optimization (AREA)
- Theoretical Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Pure & Applied Mathematics (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Algebra (AREA)
- Hydrology & Water Resources (AREA)
- Radar, Positioning & Navigation (AREA)
- Operations Research (AREA)
- Probability & Statistics with Applications (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Remote Sensing (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The application relates to an island shallow sea water depth prediction method, an island shallow sea water depth prediction device, a storage medium and computer equipment based on extreme gradient lifting, wherein the method comprises the following steps: acquiring multispectral reflectivity of the island shallow sea area; acquiring an actual water depth value of a water depth control point of the island shallow sea area; calculating a predicted value of the suspended sediment content of the water depth control point according to the near infrared and infrared band reflectivity in the multispectral reflectivity; training an extreme gradient lifting model; and inputting the multispectral reflectivity of the island reef shallow sea area into the extreme gradient lifting model to perform time-space sequence continuous inversion to obtain a water depth predicted value. The island shallow sea water depth prediction method based on extreme gradient lifting has the advantages of being capable of efficiently and quickly obtaining the island shallow sea water depth, and accurate and reliable in result.
Description
Technical Field
The application relates to the technical field of geographic information, in particular to an island shallow sea water depth prediction method and device based on extreme gradient lifting, a storage medium and computer equipment.
Background
At present, the water depth measurement of the shallow sea area of the island is mainly carried out on-site measurement in the shallow sea area of the island by an unmanned ship carrying a depth measuring instrument, but because the water depth and the limitation of the island are not clear, the water depth of part of the shallow sea area cannot be measured or the unmanned ship can be damaged, the efficiency is low, and the measurement result is not accurate enough.
With the wide application of the satellite remote sensing technology, the improvement of the water depth measurement method is promoted, wherein the water depth measurement by the remote sensing water depth inversion method becomes a topic worthy of research.
Disclosure of Invention
Based on this, an object of the embodiments of the present application is to provide a method, an apparatus, a storage medium, and a computer device for predicting the shallow sea depth of an island reef based on extreme gradient lift, which have the advantages of efficiently and quickly acquiring the shallow sea depth of the island reef, and accurate and reliable results.
In a first aspect, an embodiment of the application provides an island shallow sea water depth prediction method based on extreme gradient lift, which includes the following steps:
acquiring multispectral reflectivity of the island shallow sea area;
acquiring an actual water depth value of a water depth control point of the island shallow sea area;
calculating a predicted value of the suspended sediment content of the water depth control point according to the near infrared and infrared band reflectivity in the multispectral reflectivity;
in the multispectral reflectivity of the water depth control point, the ratio of the natural logarithm of the reflectivity below the water surface in the same wave band to the natural logarithm of the reflectivity in the deep water area is used as the characteristic component of the characteristic vector, and meanwhile, the predicted value of the suspended sediment content is also used as the characteristic component of the characteristic vector; training an extreme gradient lifting model by taking the actual water depth value of the water depth control point as an expected output;
and inputting the multispectral reflectivity of the island reef shallow sea area into the extreme gradient lifting model to perform time-space sequence continuous inversion to obtain a water depth predicted value.
In one embodiment, after the step of inputting the multispectral reflectivity of the island reef shallow sea area into the extreme gradient elevation model for performing time-space sequence continuous inversion to obtain the island reef shallow sea water depth, the method further comprises:
acquiring an actual water depth value of a water depth verification point of the island shallow sea area;
and calculating inversion accuracy according to the actual water depth value of the water depth verification point and the water depth predicted value of the water depth verification point obtained through space-time sequence continuous inversion, and establishing a relation model between the inversion accuracy and the actual water depth value.
In one embodiment, the acquiring multispectral reflectivity of the shallow sea area of the island reef comprises:
acquiring multispectral remote sensing data with medium and high resolution in the island shallow sea area;
performing data preprocessing on the multispectral remote sensing data through atmospheric correction and wavelet transformation;
and converting the multispectral remote sensing data into multispectral reflectivity.
In one embodiment, the actual water depth value and the suspended sediment content of the water depth control point are obtained by actually measuring the water depth control point of the island shallow sea area according to the transit time of a source satellite of the multispectral remote sensing data.
In one embodiment, the step of calculating the predicted value of the suspended sediment content at the water depth control point according to the near-infrared and infrared band reflectivities in the multispectral reflectivity is specifically calculated by using a relational model as follows:
in the formula, RnirAnd RredThe reflectivities of a near infrared band and a red light band are respectively, TSS is suspended sediment concentration with the unit of mg/L, and a, b and c are regression coefficients.
In one embodiment, after the step of calculating the predicted value of the suspended sediment content of the water depth control point according to the near infrared band reflectivity and the infrared band reflectivity in the multispectral reflectivity, the method further comprises the following steps:
acquiring a measured value of suspended sediment content of a water depth control point;
and carrying out precision verification on the predicted value of the suspended sediment content and calibrating the relation model according to the actually measured value of the suspended sediment content.
In one embodiment, in the multispectral reflectivity of the water depth control point, the ratio of the natural logarithm of the reflectivity below the water surface in the same wave band to the natural logarithm of the reflectivity in the deep water area is used as a characteristic component of a characteristic vector, and the predicted value of the content of suspended sediment is also used as the characteristic component of the characteristic vector; the method comprises the following steps of training an extreme gradient lifting model by taking an actual water depth value of a water depth control point as expected output, wherein the steps comprise:
setting data setsThe loss function of the learning unit of the extreme gradient lifting model isNumber of iterationst and a tree k of the decision tree, traversing the iteration times t and the tree k of the decision tree in a set data set, verifying the precision, and selecting the iteration times t with the highest precision and the tree k of the decision tree as the extreme gradient lifting model parameters; wherein x isiIs a feature vector, n is the number of samples in the dataset, yiIn order to actually output the result of the output,outputting a result for the extreme gradient lifting model;
obtaining the simulation predicted value of the t-th time of the extreme gradient lifting model according to the following mode
Wherein,for the output result of the extreme gradient lifting model t-1 times, ft(xi) Outputting a result for the t iteration of the decision tree;
randomly putting back and extracting a plurality of subsets K in the training data set, generating a decision tree for each subset, training each decision tree to obtain residual errors
The training loss function is minimized in the following manner:
wherein gamma is a regular term coefficient;
summing the prediction results of the decision trees to obtain an actual water depth value of the water depth control point:
wherein f iskIs the result of a prediction of a single decision tree,is the set of all decision trees CART.
In a second aspect, the embodiment of the present application further provides an island shallow sea water depth prediction device based on extreme gradient lift, including:
the first data acquisition module is used for acquiring the multispectral reflectivity of the island shallow sea area;
the second data acquisition module is used for acquiring the actual water depth value of the water depth control point of the island reef shallow sea area;
the suspended sediment content prediction module is used for calculating a predicted suspended sediment content value of the water depth control point according to the near-infrared and infrared band reflectivity in the multispectral reflectivity;
the extreme ladder lifting model building module is used for taking the ratio of the natural logarithm of the reflectivity below the water surface of the same wave band to the natural logarithm of the reflectivity of the deep water area in the multispectral reflectivity of the water depth control point as the characteristic component of the characteristic vector, and simultaneously taking the predicted value of the suspended sediment content as the characteristic component of the characteristic vector; training an extreme gradient lifting model by taking the actual water depth value of the water depth control point as an expected output;
and the water depth prediction module is used for inputting the multispectral reflectivity of the island reef shallow sea area into the extreme gradient lifting model to perform time-space sequence continuous inversion so as to obtain a water depth prediction value.
In a third aspect, an embodiment of the present application further provides a computer storage medium, wherein the computer storage medium stores a plurality of instructions adapted to be loaded by a processor and execute the steps of the extreme gradient lift-based island reef shallow sea water depth prediction method according to any one of the above.
In a fourth aspect, an embodiment of the present application further provides a computer device, including:
a processor and a memory;
wherein the memory stores a computer program; the computer program is adapted to be loaded by the processor and to perform the steps of the extreme gradient lift based island shallow sea water depth prediction method as described in any one of the above.
In the technical scheme of the embodiment of the application, in the multispectral reflectivity of the water depth control point, the ratio of the natural logarithm of the reflectivity below the water surface of the same waveband to the natural logarithm of the reflectivity of the deep water area is used as the characteristic component of the characteristic vector, and the predicted value of the suspended sediment content is also used as the characteristic component of the characteristic vector; training an extreme gradient lifting model by taking the actual water depth value of the water depth control point as expected output, applying the multispectral reflectivity of the island reef shallow sea area to the extreme gradient lifting model to invert and predict the water depth, constructing the extreme gradient lifting model only by actually measuring the water depth values of the water depth control points with less quantity, and efficiently and quickly inverting by using the extreme gradient lifting model to obtain the water depth values of different points of the whole island reef shallow sea area; and because the change of the water surface reflectivity caused by the suspended sediment content is considered, the suspended sediment content is also used as the influence factor of the water depth when a model is constructed, so that the predicted value of the water depth is accurate and reliable. The extreme gradient lifting model is the gradient lifting integration of the decision tree, has the beneficial effects of high precision, high robustness and interpretability on the analysis of the multispectral reflectivity of unclean data, and eliminates overfitting to a certain extent compared with general classification and regression tree modeling.
For a better understanding and practice, the present application is described in detail below with reference to the accompanying drawings.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a flowchart illustrating steps of a shallow sea water depth prediction method for an island based on extreme gradient lift according to an embodiment of the present application;
fig. 2 is a flowchart illustrating a step S100 of an extreme gradient lift-based island reef shallow sea water depth prediction method according to an embodiment of the present application;
fig. 3 is a flowchart illustrating steps included after step S300 of the extreme gradient lift-based island shallow sea water depth prediction method according to the embodiment of the present application;
fig. 4 is a flowchart illustrating steps included after step S500 of the island shallow sea water depth prediction method based on extreme gradient lift according to the embodiment of the present application;
fig. 5 is a schematic structural diagram of an extreme gradient lifting-based island reef shallow sea water depth prediction device provided in an embodiment of the present application;
fig. 6 is a schematic structural diagram of a computer device according to an embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more clear, embodiments of the present application will be described in further detail below with reference to the accompanying drawings.
It should be understood that the embodiments described are only a few embodiments of the present application, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the application, as detailed in the appended claims.
In the description of the present application, it is to be understood that the terms "first," "second," "third," and the like are used solely to distinguish one from another and are not necessarily used to describe a particular order or sequence, nor are they to be construed as indicating or implying relative importance. The specific meaning of the above terms in the present application can be understood by those of ordinary skill in the art as appropriate. Further, in the description of the present application, "a plurality" means two or more unless otherwise specified. "and/or" describes the association relationship of the associated objects, meaning that there may be three relationships, e.g., a and/or B, which may mean: a exists alone, A and B exist simultaneously, and B exists alone. The character "/" generally indicates that the former and latter associated objects are in an "or" relationship.
The embodiment of the application discloses an island shallow sea water depth prediction method based on extreme gradient lifting, and please refer to fig. 1, which is a flow chart of the island shallow sea water depth prediction method based on extreme gradient lifting provided by the embodiment of the application. The method comprises the following steps:
step S100: and acquiring the multispectral reflectivity of the shallow sea area of the island.
In one embodiment, referring to fig. 2, the step of obtaining the multispectral reflectivity of the shallow sea area of the island reef comprises:
step S110: acquiring multispectral remote sensing data with medium and high resolution in the island shallow sea area;
step S120: performing data preprocessing on the multispectral remote sensing data through atmospheric correction and wavelet transformation;
step S130: and converting the multispectral remote sensing data into multispectral reflectivity.
The source satellites of the multispectral remote sensing data with medium and high resolution comprise high score-1/2 (GF-1 and GF-2), resource No. three (ZY-3), SPOT-6/7, Landsat OLI and Sentinel-2. The step S130 is to convert DN values of the multispectral bands into multispectral reflectivities. The multispectral reflectivity is related to the water depth, suspended sediment content, salinity conditions and the like of the island shallow sea area.
Step S200: and acquiring the actual water depth value of the water depth control point of the island reef shallow sea area.
The water depth control points are selected in advance and are located in the shallow sea area of the island, and the control points are used for actual measurement of the unmanned ship. In one embodiment, the actual water depth value and the suspended sediment content of the water depth control point are obtained by actually measuring the water depth control point in the island shallow sea area by selecting clear and cloudless time according to the transit time of a source satellite of the multispectral remote sensing data. Specifically, an unmanned ship carrying a single-wave velocity depth finder, a global positioning system and a real-time dynamic carrier phase difference technology is used for carrying out water depth sampling, the sampling is uniformly carried out in a buffer area of 2000 m around an island reef, the sampling interval is 40m, and the precision error is +/-1 cm. Considering that the water body does certain regular fluctuation motion in the vertical direction under the action of induced tidal force, the actual water depth value of the water depth control point of the water depth remote sensing inversion is the instantaneous water depth at the moment of acquiring the multispectral remote sensing data. The depth starting surface of the embodiment of the application is a theoretical depth reference surface and is consistent with a tidal height reference surface in a tidal table, so that the actual water depth value at a certain moment is equal to the actual water depth value plus the tidal height at the moment. And respectively carrying out tide correction on the actually measured water depth value according to the transit time of each scene image to obtain an actual water depth value.
Step S300: and calculating a predicted value of the suspended sediment content of the water depth control point according to the near infrared band reflectivity and the infrared band reflectivity in the multispectral reflectivity.
In an embodiment, based on the higher correlation between the near infrared and red light bands and the concentration of suspended sediment in the water body, the step S300 of calculating the predicted value of the content of suspended sediment at the water depth control point according to the near infrared and infrared band reflectivities in the multispectral reflectivity is specifically calculated by the following relation model:
in the formula, RnirAnd RredThe reflectivities of a near infrared band and a red light band are respectively, TSS is suspended sediment concentration with the unit of mg/L, and a, b and c are regression coefficients.
In this application embodiment, suspended sediment concentration is the proportion of suspended sediment content at the sampling water, calculates the suspended sediment concentration of water depth control point through near-infrared and infrared band reflectivity, also can calculate suspended sediment content predicted value promptly.
In an embodiment, referring to fig. 3, after the step S300 of calculating the predicted value of the suspended sediment content at the water depth control point according to the near-infrared and infrared band reflectivities in the multispectral reflectivity, the method further includes:
s300 a: acquiring a measured value of suspended sediment content of a water depth control point;
s300 b: and carrying out precision verification on the predicted value of the suspended sediment content and calibrating the relation model according to the actually measured value of the suspended sediment content.
When the actual measurement value of the suspended sediment content of the water depth control point is obtained, the on-site water sample collection for measuring the suspended sediment content is carried out according to the marine survey specification (GB/T12763.1-2007) and the marine measurement specification (GB/T12763.4-2007) of the people's republic of China, the on-site water sample is stored in a black bottle in a sealed and light-proof mode, and the black bottle is conveyed to a professional analysis laboratory to carry out suspended sediment content measurement by adopting a GB 11901-89 drying and weighing method.
Step S400: in the multispectral reflectivity of the water depth control point, the ratio of the natural logarithm of the reflectivity below the water surface in the same wave band to the natural logarithm of the reflectivity in the deep water area is used as the characteristic component of the characteristic vector, and meanwhile, the predicted value of the suspended sediment content is also used as the characteristic component of the characteristic vector; and training an extreme gradient lifting model by taking the actual water depth value of the water depth control point as expected output.
The extreme gradient lifting model takes a plurality of decision trees as learning units, a next decision tree is fitted according to a residual error between an output result of a previous decision tree and an actual value, and an island reef shallow sea depth predicted value based on extreme gradient lifting is obtained by summing a plurality of decision tree output results; and the extreme gradient lifting model is an optimization algorithm based on a proper cost function, the ith tree is fitted on the predicted residual error of the (i-1) th tree to correct the error of the next tree, and the final prediction result is obtained by summing the output results of each tree.
In one embodiment, in the multispectral reflectivity of the water depth control point, the ratio of the natural logarithm of the reflectivity below the water surface in the same wave band to the natural logarithm of the reflectivity in the deep water area is used as a characteristic component of a characteristic vector, and the predicted value of the content of suspended sediment is also used as the characteristic component of the characteristic vector; the method comprises the following steps of training an extreme gradient lifting model by taking an actual water depth value of a water depth control point as expected output, wherein the steps comprise:
setting data setsThe loss function of the learning unit of the extreme gradient lifting model isThe iteration times t and the tree k of the decision tree are traversed in a set data set, the precision is verified, and the iteration times t with the highest precision and the tree k of the decision tree are selected as the extreme gradient lifting model parameters; wherein x isiIs a feature vector, n is the number of samples in the dataset, yiIn order to actually output the result of the output,outputting a result for the extreme gradient lifting model;
obtaining the simulation predicted value of the t-th time of the extreme gradient lifting model according to the following mode
Wherein,for the output result of the extreme gradient lifting model t-1 times, ft(xi) Outputting a result for the t iteration of the decision tree;
randomly having a number of payout draws in the training data setAnd a subset K, each subset generating a decision tree, each decision tree being trained to obtain a residual error
The training loss function is minimized in the following manner:
wherein gamma is a regular term coefficient;
summing the prediction results of the decision trees to obtain an actual water depth value of the water depth control point:
wherein f iskIs the result of a prediction of a single decision tree,is the set of all decision trees CART.
The extreme gradient lifting model finally constructed is as follows:
λ1……λnrefer to different wave bands, Rw(λ1) Is the band reflectivity, R, of the first band∞(λ1) Is the deep water zone reflectivity of the first band;is the natural logarithm ln (R)w(λ1) And the natural logarithm ln (R)∞(λ1) The other bands are the same.
Step S500: and inputting the multispectral reflectivity of the island reef shallow sea area into the extreme gradient lifting model to perform time-space sequence continuous inversion to obtain a water depth predicted value. Because the suspended sediment content is related to the reflectivity of the red light wave band and the near infrared wave band, the input of the extreme gradient lifting model is actually the multispectral reflectivity of the island shallow sea area, and the predicted water depth value obtained based on the multispectral reflectivity of the island shallow sea area is the value of the corresponding leaf node reached when the y sample is transmitted in the tree.
In an embodiment, in order to verify the accuracy of the constructed extreme gradient elevation model, please refer to fig. 4, the step S500 of inputting the multispectral reflectivity of the island reef shallow sea area into the extreme gradient elevation model for performing spatiotemporal sequence continuous inversion to obtain the island reef shallow sea water depth further includes:
s601: acquiring an actual water depth value of a water depth verification point of the island shallow sea area;
s602: and calculating inversion accuracy according to the actual water depth value of the water depth verification point and the water depth predicted value of the water depth verification point obtained through space-time sequence continuous inversion, and establishing a relation model between the inversion accuracy and the actual water depth value.
The water depth verification point is similar to the water depth control point, is a plurality of verification points which are pre-selected and are positioned in the island reef shallow sea area and are used for unmanned ship actual measurement, and mainly forms a verification set for verifying and correcting the constructed model. The method for obtaining the actual water depth value of the water depth verification point is the same as the actual water depth value of the water depth control point.
According to the method and the device, the accuracy and the maximum depth of the optical remote sensing water depth inversion under different water color conditions can be determined according to the relation model of the inversion accuracy and the actual water depth value, and the extreme gradient lifting model can be subjected to model correction according to the relation model of the inversion accuracy and the actual water depth value.
According to the technical scheme of the embodiment of the application, an extreme gradient lifting model of shallow sea water depth, multispectral reflectivity and suspended sediment content is constructed by utilizing a classification and regression tree algorithm according to the multispectral reflectivity, the actual water depth value and the predicted suspended sediment content value of a water depth control point, then the multispectral reflectivity of the island reef shallow sea area is applied to the extreme gradient lifting model, the water depth can be inversely predicted, the extreme gradient lifting model can be constructed only by actually measuring the water depth values of a small number of water depth control points, and then the extreme gradient lifting model is used for efficiently and quickly inverting to obtain the water depth values of different points of the whole island reef shallow sea area; and because the change of the water surface reflectivity caused by the suspended sediment content is considered, the suspended sediment content is also used as the influence factor of the water depth when a model is constructed, so that the predicted value of the water depth is accurate and reliable.
In a second aspect, referring to fig. 5, an embodiment of the present application further provides an island shallow sea water depth prediction apparatus based on extreme gradient lift, including:
the first data acquisition module 1 is used for acquiring the multispectral reflectivity of the island shallow sea area;
the second data acquisition module 2 is used for acquiring the actual water depth value of the water depth control point of the island reef shallow sea area;
the suspended sediment content prediction module 3 is used for calculating a suspended sediment content prediction value of a water depth control point according to the near-infrared and infrared band reflectivity in the multispectral reflectivity;
the extreme gradient lifting model building module 4 is used for taking the ratio of the natural logarithm of the reflectivity below the water surface of the same wave band to the natural logarithm of the reflectivity of the deep water area in the multispectral reflectivity of the water depth control point as the characteristic component of the characteristic vector, and simultaneously taking the predicted value of the suspended sediment content as the characteristic component of the characteristic vector; training an extreme gradient lifting model by taking the actual water depth value of the water depth control point as an expected output;
and the water depth prediction module 5 is used for inputting the multispectral reflectivity of the island reef shallow sea area into the extreme gradient lifting model to perform time-space sequence continuous inversion so as to obtain a water depth prediction value.
Based on the same inventive concept, the island reef shallow sea water depth prediction device based on extreme gradient lifting has the same beneficial effects as the island reef shallow sea water depth prediction method based on extreme gradient lifting.
In a third aspect, an embodiment of the present application further provides a computer storage medium, wherein the computer storage medium stores a plurality of instructions adapted to be loaded by a processor and execute the steps of the extreme gradient lift-based island reef shallow sea water depth prediction method according to any one of the above.
In a fourth aspect, referring to fig. 6, an embodiment of the present application further provides a computer device, including:
a processor 10 and a memory 20;
wherein the memory 20 stores a computer program; the computer program is adapted to be loaded by the processor 10 and to perform the steps of the extreme gradient lift based island shallow sea water depth prediction method as described in any one of the above.
In this embodiment, the processor 10 and memory 20 are connected by a bus, and the memory 20 may take the form of a computer program product embodied on one or more storage media having program code embodied therein (including, but not limited to, disk storage, CD-ROM, optical storage, etc.). Computer readable storage media, which include both non-transitory and non-transitory, removable and non-removable media, may implement any method or technology for storage of information. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of the storage medium of the computer include, but are not limited to: phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technologies, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic tape storage or other magnetic storage devices, or any other non-transmission medium, may be used to store information that may be accessed by a computing device.
The processor 10 may be one or more Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs), Digital Signal Processing Devices (DSPDs), Programmable Logic Devices (PLDs), Field Programmable Gate Arrays (FPGAs), controllers, micro-controllers, microprocessors, or other electronic components. In this embodiment, the processor 10 may also be multiple, or the processor 10 may include one or more processing cores.
In the several embodiments provided in the present application, it should be understood that the disclosed system, apparatus and method may be implemented in other manners. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the units is only one logical division, and other divisions may be realized in practice, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application.
The units may or may not be physically separate, and components displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application.
Claims (10)
1. An island reef shallow sea water depth prediction method based on extreme gradient lifting is characterized by comprising the following steps:
acquiring multispectral reflectivity of the island shallow sea area;
acquiring an actual water depth value of a water depth control point of the island shallow sea area;
calculating a suspended sediment content predicted value of a water depth control point according to the near infrared and red light wave band reflectivities in the multispectral reflectivity;
in the multispectral reflectivity of the water depth control point, the ratio of the natural logarithm of the reflectivity below the water surface in the same wave band to the natural logarithm of the reflectivity in the deep water area is used as the characteristic component of the characteristic vector, and meanwhile, the predicted value of the suspended sediment content is also used as the characteristic component of the characteristic vector; training an extreme gradient lifting model by taking the actual water depth value of the water depth control point as an expected output;
and inputting the multispectral reflectivity of the island reef shallow sea area into the extreme gradient lifting model to perform time-space sequence continuous inversion to obtain a water depth predicted value.
2. The extreme gradient lift-based island reef shallow sea water depth prediction method according to claim 1, wherein: after the step of inputting the multispectral reflectivity of the island reef shallow sea area into the extreme gradient lifting model for performing time-space sequence continuous inversion to obtain a water depth predicted value, the method further comprises the following steps:
acquiring an actual water depth value of a water depth verification point of the island shallow sea area;
and calculating inversion accuracy according to the actual water depth value of the water depth verification point and the water depth predicted value of the water depth verification point obtained through space-time sequence continuous inversion, and establishing a relation model between the inversion accuracy and the actual water depth value.
3. The extreme gradient lift-based island reef shallow sea water depth prediction method according to claim 1, wherein: the acquiring of the multispectral reflectivity of the island reef shallow sea area comprises the following steps:
acquiring multispectral remote sensing data with medium and high resolution in the island shallow sea area;
performing data preprocessing on the multispectral remote sensing data through atmospheric correction and wavelet transformation;
and converting the preprocessed multispectral remote sensing data into multispectral reflectivity.
4. The extreme gradient lift-based island reef shallow sea water depth prediction method according to claim 1 or 2, wherein: and the actual water depth value is obtained by actually measuring and tide correcting a water depth control point or a water depth verification point of the island shallow sea area according to the transit time of a source satellite of the multispectral remote sensing data.
5. The extreme gradient lift-based island reef shallow sea water depth prediction method according to claim 1, wherein: the step of calculating the predicted value of the suspended sediment content of the water depth control point according to the near infrared and red light wave band reflectivities in the multispectral reflectivity is specifically calculated by the following relation model:
in the formula, RnirAnd RredThe reflectivities of a near infrared band and a red light band are respectively, TSS is suspended sediment concentration with the unit of mg/L, and a, b and c are regression coefficients.
6. The extreme gradient lift-based island reef shallow sea water depth prediction method according to claim 5, wherein: after the step of calculating the predicted value of the suspended sediment content of the water depth control point according to the near infrared band reflectivity and the infrared band reflectivity in the multispectral reflectivity, the method further comprises the following steps of:
acquiring a measured value of suspended sediment content of a water depth control point;
and carrying out precision verification on the predicted value of the suspended sediment content and calibrating the relation model according to the actually measured value of the suspended sediment content.
7. The extreme gradient lift-based island reef shallow sea water depth prediction method according to claim 6, wherein: in the multispectral reflectivity of the water depth control point, the ratio of the natural logarithm of the reflectivity below the water surface in the same wave band to the natural logarithm of the reflectivity in the deep water area is used as the characteristic component of the characteristic vector, and meanwhile, the predicted value of the suspended sediment content is also used as the characteristic component of the characteristic vector; the method comprises the following steps of training an extreme gradient lifting model by taking an actual water depth value of a water depth control point as expected output, wherein the steps comprise:
setting data setsThe loss function of the learning unit of the extreme gradient lifting model isFor the iteration times t and the number k of the decision tree, traversing the iteration times t and the number k of the decision tree in a set data set, verifying the precision, and selecting the iteration times t with the highest precision and the number k of the decision tree as the extreme gradient lifting model parameters; wherein x isiIs a feature vector, n is the number of samples in the dataset, yiIn order to actually output the result of the output,outputting a final output result of the extreme gradient lifting model;
obtaining the output result of the t-th iteration of the extreme gradient lifting model according to the following mode
Wherein,for the initial output of the extreme gradient boost model, is the output result of the t-1 iteration of the extreme gradient lifting model,outputting a result of the t iteration of the extreme gradient lifting model; f. oft(xi) Outputting a result for the t iteration of the decision tree;
randomly putting back and extracting a plurality of subsets K in the training data set, generating a decision tree for each subset, training each decision tree to obtain residual errors
The training loss function is minimized in the following manner:
wherein gamma is a regular term coefficient; l (-) is a logarithmic loss function; obj(t)The output result of the minimized training loss function after the t iteration is obtained;
summing the prediction results of each decision tree to obtain the actual water depth value of the water depth control point:
8. An island reef shallow sea water depth prediction device based on extreme gradient lifting is characterized by comprising:
the first data acquisition module is used for acquiring the multispectral reflectivity of the island shallow sea area;
the second data acquisition module is used for acquiring the actual water depth value of the water depth control point of the island reef shallow sea area;
the suspended sediment content prediction module is used for calculating a suspended sediment content prediction value of a water depth control point according to the near infrared and red light wave band reflectivity in the multispectral reflectivity;
the extreme gradient lifting model building module is used for taking the ratio of the natural logarithm of the reflectivity below the water surface of the same wave band to the natural logarithm of the reflectivity of the deep water area in the multispectral reflectivity of the water depth control point as the characteristic component of the characteristic vector, and simultaneously taking the predicted value of the suspended sediment content as the characteristic component of the characteristic vector; training an extreme gradient lifting model by taking the actual water depth value of the water depth control point as an expected output;
and the water depth prediction module is used for inputting the multispectral reflectivity of the island reef shallow sea area into the extreme gradient lifting model to perform time-space sequence continuous inversion so as to obtain a water depth prediction value.
9. A computer storage medium storing a plurality of instructions adapted to be loaded by a processor and to perform the steps of the extreme gradient lift based island shallow sea water depth prediction method according to any one of claims 1 to 7.
10. A computer device, comprising:
a processor and a memory;
wherein the memory stores a computer program; the computer program is adapted to be loaded by the processor and to perform the steps of the extreme gradient lift based island shallow sea water depth prediction method according to any one of claims 1 to 7.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910945145.1A CN110851789B (en) | 2019-09-30 | 2019-09-30 | Island reef shallow sea water depth prediction method based on extreme gradient lifting |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910945145.1A CN110851789B (en) | 2019-09-30 | 2019-09-30 | Island reef shallow sea water depth prediction method based on extreme gradient lifting |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110851789A CN110851789A (en) | 2020-02-28 |
CN110851789B true CN110851789B (en) | 2020-12-08 |
Family
ID=69597376
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910945145.1A Active CN110851789B (en) | 2019-09-30 | 2019-09-30 | Island reef shallow sea water depth prediction method based on extreme gradient lifting |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110851789B (en) |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107894381A (en) * | 2017-12-29 | 2018-04-10 | 广州和时通电子科技有限公司 | Husky apparatus and method are surveyed in a kind of southern river |
CN110082300A (en) * | 2019-05-20 | 2019-08-02 | 中国水利水电科学研究院 | Canopy of winter wheat moisture content monitoring model method for building up based on spectral parameter |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105865424B (en) * | 2016-04-13 | 2019-09-17 | 中测新图(北京)遥感技术有限责任公司 | A kind of multispectral remote sensing inversion method and device based on nonlinear model |
CN108663330B (en) * | 2018-04-19 | 2019-08-16 | 中国自然资源航空物探遥感中心 | A kind of vegetation-covered area soil copper inversion method based on blade measured spectra |
CN109283144A (en) * | 2018-10-26 | 2019-01-29 | 浙江省水利河口研究院 | The strong long remote sensing calculation method for lasting variation of tidal height muddiness river mouth suspension bed sediment |
CN109409441A (en) * | 2018-11-16 | 2019-03-01 | 福州大学 | Based on the coastal waters chlorophyll-a concentration remote sensing inversion method for improving random forest |
-
2019
- 2019-09-30 CN CN201910945145.1A patent/CN110851789B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107894381A (en) * | 2017-12-29 | 2018-04-10 | 广州和时通电子科技有限公司 | Husky apparatus and method are surveyed in a kind of southern river |
CN110082300A (en) * | 2019-05-20 | 2019-08-02 | 中国水利水电科学研究院 | Canopy of winter wheat moisture content monitoring model method for building up based on spectral parameter |
Also Published As
Publication number | Publication date |
---|---|
CN110851789A (en) | 2020-02-28 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110823190B (en) | Island reef shallow sea water depth prediction method based on random forest | |
CN105865424B (en) | A kind of multispectral remote sensing inversion method and device based on nonlinear model | |
CN110849334B (en) | Island reef shallow sea water depth prediction method based on classification and regression tree | |
CN112180369B (en) | Depth learning-based sea surface wind speed inversion method for one-dimensional synthetic aperture radiometer | |
CN113592786B (en) | Deep learning-based ocean mesoscale vortex detection method | |
CN113408742A (en) | High-precision sea surface temperature inversion method based on machine learning | |
CN110163287A (en) | A kind of mesoscale eddy detection method and device | |
CN112666328A (en) | Chlorophyll a remote sensing inversion method and device based on genetic neural network model | |
CN106940219A (en) | A kind of spectral response acquisition methods of broadband satellite remote sensor in orbit | |
CN116562414A (en) | Land water reserve change prediction method based on multisource hydrographic meteorological data | |
CN113920255B (en) | High-efficient mapping system based on point cloud data | |
CN116977311A (en) | Flood disaster area detection method, system, electronic equipment and storage medium | |
Mateus et al. | Assessment of two techniques to merge ground-based and TRMM rainfall measurements: a case study about Brazilian Amazon Rainforest | |
Vanden Boomen et al. | Application of deep learning for imaging‐based stream gaging | |
CN113237554B (en) | Method and device for generating surface temperature image under cloud and terminal equipment | |
CN116879192B (en) | Water bloom prediction method, device, equipment and medium based on satellite remote sensing data | |
CN112288744B (en) | SAR image change detection method based on integer reasoning quantification CNN | |
Chen et al. | Improving surface current estimation from Geostationary Ocean Color Imager using tidal ellipse and angular limitation | |
CN110851789B (en) | Island reef shallow sea water depth prediction method based on extreme gradient lifting | |
CN110852475B (en) | Extreme gradient lifting algorithm-based vegetation index prediction method, system and equipment | |
CN110852415B (en) | Vegetation index prediction method, system and equipment based on neural network algorithm | |
Platt et al. | Primary production: sensitivity to surface irradiance and implications for archiving data | |
CN116822716A (en) | Typhoon prediction method, system, equipment and medium based on space-time attention | |
Ivanda et al. | An application of 1D convolution and deep learning to remote sensing modelling of Secchi depth in the northern Adriatic Sea | |
CN114494501B (en) | Method and device for reconstructing chlorophyll a of water body |
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 | ||
CP01 | Change in the name or title of a patent holder | ||
CP01 | Change in the name or title of a patent holder |
Address after: 510075 one of the compound No. 100, Xianlie Middle Road, Yuexiu District, Guangzhou City, Guangdong Province Patentee after: Guangzhou Institute of geography, Guangdong Academy of Sciences Address before: 510075 one of the compound No. 100, Xianlie Middle Road, Yuexiu District, Guangzhou City, Guangdong Province Patentee before: GUANGZHOU INSTITUTE OF GEOGRAPHY |