CN111222439B - Sea fog identification method based on support vector machine - Google Patents
Sea fog identification method based on support vector machine Download PDFInfo
- Publication number
- CN111222439B CN111222439B CN201911408481.9A CN201911408481A CN111222439B CN 111222439 B CN111222439 B CN 111222439B CN 201911408481 A CN201911408481 A CN 201911408481A CN 111222439 B CN111222439 B CN 111222439B
- Authority
- CN
- China
- Prior art keywords
- data
- std
- satellite channel
- support vector
- vector machine
- 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 40
- 238000012706 support-vector machine Methods 0.000 title claims abstract description 39
- 238000012549 training Methods 0.000 claims abstract description 26
- 238000012360 testing method Methods 0.000 claims abstract description 8
- 238000012886 linear function Methods 0.000 claims description 13
- 238000012545 processing Methods 0.000 claims description 10
- 238000012795 verification Methods 0.000 claims description 8
- 238000013507 mapping Methods 0.000 claims description 5
- 239000013598 vector Substances 0.000 claims description 5
- 238000011425 standardization method Methods 0.000 claims description 3
- 238000003909 pattern recognition Methods 0.000 abstract description 4
- 230000006870 function Effects 0.000 description 9
- 241000208818 Helianthus Species 0.000 description 5
- 235000003222 Helianthus annuus Nutrition 0.000 description 5
- 238000012544 monitoring process Methods 0.000 description 5
- 239000011159 matrix material Substances 0.000 description 3
- 238000007619 statistical method Methods 0.000 description 3
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 206010039203 Road traffic accident Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000009833 condensation Methods 0.000 description 1
- 230000005494 condensation Effects 0.000 description 1
- 239000013078 crystal Substances 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000012567 pattern recognition method Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/13—Satellite images
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01W—METEOROLOGY
- G01W1/00—Meteorology
- G01W1/10—Devices for predicting weather conditions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]
-
- 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
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Environmental & Geological Engineering (AREA)
- General Engineering & Computer Science (AREA)
- Remote Sensing (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Astronomy & Astrophysics (AREA)
- Evolutionary Computation (AREA)
- Multimedia (AREA)
- Artificial Intelligence (AREA)
- Atmospheric Sciences (AREA)
- Biodiversity & Conservation Biology (AREA)
- Ecology (AREA)
- Environmental Sciences (AREA)
- Image Processing (AREA)
Abstract
The invention provides a sea fog identification method based on a support vector machine, which comprises the following steps: step S1: acquiring satellite channel disc projection data and ship observation original data in a period of time; step S2: decoding is carried out, and satellite channel lattice point data x, visibility observation data y and corresponding time ship position information are obtained; step S3: obtaining satellite channel standardized data x std And observation point visibility binarization data y std The method comprises the steps of carrying out a first treatment on the surface of the Step S4: generating a training and testing sample set, and training a support vector machine model; and performing sea fog recognition on the data to be detected through the trained support vector machine model. Compared with the traditional threshold method in the prior art, the support vector machine has the advantages of solving the problems of small sample, nonlinearity and high-dimensional pattern recognition, and improving the accuracy of satellite sea fog recognition.
Description
Technical Field
The invention belongs to the technical field of weather prediction, and particularly relates to a sea fog identification method based on a support vector machine.
Background
Sea fog refers to dangerous weather that occurs in the lower atmosphere above the sea, shore and islands, and the large number of water droplets or ice crystals generated due to water vapor condensation make horizontal visibility less than 1 km. The visibility decrease caused by the sea fog has serious influence on ship navigation, and is an important cause for causing various accidents in offshore and coastal areas. The movement and landing of sea fog have important influence on coastal areas, and are easy to cause flight delay, train delay, even serious consequences such as flight take-off and landing accidents, expressway traffic accidents and the like. Sea fog has serious influence on production and life in offshore and coastal areas, so accurate identification and monitoring of sea fog are necessary.
The conventional sea fog monitoring method is to arrange stations on the land and sea surface for manual or instrument automatic observation, so that a large amount of manpower and material resources are consumed, the density of the observation stations is difficult to meet the monitoring requirement, and few or even no observation stations are arranged on the sea. Compared with site monitoring, the meteorological satellite observation data has the characteristics of wide coverage range and high space-time resolution, so that the meteorological satellite observation data has unique advantages in the aspect of monitoring fog generation and elimination dynamics. Therefore, it is important to study sea fog recognition methods based on satellite observation data.
The traditional sea fog identification method based on satellite data is mainly based on a threshold method, and mainly comprises the steps of determining a channel threshold value or a double-channel difference threshold value of a sea fog area through statistical analysis of satellite channel values, so that sea fog is distinguished from targets such as clouds and sea surfaces, and the purpose of identifying sea fog is achieved.
With the continuous progress of satellite observation technology, the satellite-mounted sensor has more and more channels, and the limitation of the statistical analysis method in searching for a high-dimensional threshold value is gradually reflected, so that the accuracy of sea fog identification is greatly limited. Therefore, how to apply a new method suitable for solving the high-dimensional threshold problem in the sea fog recognition problem is very important for improving the sea fog recognition accuracy.
In recent years, machine learning algorithms such as artificial neural networks and support vector machines have greatly progressed in image classification, and are increasingly used in research of meteorological problems such as remote sensing image recognition. The support vector machine (Support Vector Machine, SVM) is a new pattern recognition method developed on the basis of a statistical learning theory, and the method has the advantages of the statistical method and the traditional neural network on solving the problems of small sample, nonlinearity and high-dimensional pattern recognition, and can be popularized to machine learning problems such as function fitting, classification and the like. The basic idea of the support vector machine is: the input space is transformed into a high-dimensional space in which the optimal classification plane is found by a nonlinear transformation, which is performed by defining an appropriate inner kernel function that satisfies the Mercer condition.
Disclosure of Invention
The sea fog identification method based on the support vector machine is small in structural risk, suitable for learning problems with low sample size requirements, and capable of enabling a final decision function to be determined by only a few support vectors and avoiding dimension disasters.
The invention particularly provides a sea fog identification method based on a support vector machine, which is characterized by comprising the following steps of:
step S1: acquiring satellite channel disk projection data x over a period of time 0 And ship observation raw data y 0 ；
Step S2: for the satellite channel disk projection data x 0 And ship observation raw data y 0 Decoding is carried out, and satellite channel lattice point data x, visibility observation data y and corresponding time ship position information are obtained;
step S3: the satellite channel lattice point data x is subjected to standardization processing to obtain satellite channel standardized data x std The method comprises the steps of carrying out a first treatment on the surface of the Carrying out foggy/non-foggy binarization processing on the visibility data y to obtain the visibility binarization data y of the observation point std ；
Step S4: normalizing the grid point of the satellite channel corresponding to time and position to obtain data x std And the visibility data y std Fusing to generate a training and testing sample set, and training a support vector machine model;
step S5: and performing sea fog recognition on the data to be detected through the trained support vector machine model.
Further, in step S2, the satellite channel disk projection data x is recorded 0 Decoding is carried out by adopting the position of a pixel point in a coordinate system taking a satellite as an origin, the equatorial radius and the polar radius of the earth and calculating the longitude and latitude of the pixel point according to the geometric principleAnd acquiring longitude and latitude information of the pixel point in the circular chart according to the degree coordinates, and interpolating the longitude and latitude information to equidistant theodolite points to obtain satellite channel lattice point data x.
Further, in step S2, raw data y is observed for the vessel 0 Decoding raw data y observed from the vessel 0 And extracting the visibility observation data y and the corresponding ship position information.
Further, in step S3, the method further includes the steps of:
step S31: empirical constant x for extracting maximum value of satellite channel data max And an empirical constant x of minimum min
Step S32: and respectively carrying out standardization processing on the satellite channel data x of 16 channels by adopting a maximum value and minimum value standardization method:
in the formula, x std Is satellite channel standardized data, x is satellite channel lattice point data, x max Is the empirical constant of the maximum value of the grid point data of the satellite channel, x min Is an empirical constant of the minimum of the satellite channel grid point data.
Further, in step S4, the method further includes the following steps:
step S41: for a given time t, selecting the visibility y at the time t std And corresponding longitude and latitude coordinates (lon, lat), n groups in total;
step S42: selecting 16 satellite channel standardized data x at t moment std ；
Step S43: for n groups of the visibility y screened in the step S41 std And corresponding longitude and latitude coordinate (lon, lat) data, respectively reading 16 satellite channel standardized data x at t time std The values of the nearest lattice points of the medium-distance coordinates (lon, lat) form a characteristic quantity f;
step S44: repeating for all times T in the sample setStep S41-step S43, obtaining the visibility y std A corresponding feature set F;
step S45: dividing the feature set F into a training sample and a verification sample, wherein the training sample comprises a foggy or non-foggy label, and the verification sample does not comprise a foggy or non-foggy label;
step S46: and inputting the training sample into a support vector machine model for training, and checking the training accuracy of the support vector machine model through the verification sample.
Further, in step S5, the method further includes the following steps:
step S51: normalizing data x for low-dimensional satellite channels std ∈R N By a kernel function k (x std ,x i ) Mapping to high-dimensional space R H Constructing an optimal hyperplane in the H;
step S52: in a high-dimensional space R H Constructing a linear function in an optimal hyperplane:
step S53: and calculating the linear function through a step function, and judging an output result.
Further, in step S51, the low-dimensional satellite channel standardized data x std ∈R N By a kernel function k (x std ,x i ) Mapping to high-dimensional space R H The method comprises the following steps:
(φ(x std ),φ(x i ))＝k(x std ,x i )
in the formula, phi is a feature map; x is x i Is a support vector, and is obtained by training a support vector machine model. Further, in step S52, the linear function is:
f(x std )＝w·k(x std ,x i )+b
where w is the slope of the linear function and b is the intercept of the linear function.
Further, in step S53, the decision on the output result is:
y std ＝sgn(f(x std ))
in the formula, sgn is a step function; if y std If the value is 1, the fog is determined; y is std And=0, then it is determined that there is no fog.
The beneficial effects of the invention are as follows:
the invention adopts a new generation of stationary meteorological satellite sunflower No. 8 as model input data; compared with the prior art which uses more satellites such as GOES, MTSAT, FY-2, the sunflower satellite No. 8 has a more complete channel and higher spatial resolution (2 km), and can obtain finer identification results. Meanwhile, aiming at the problem of low recognition accuracy of the traditional threshold method in the prior art, the sea fog recognition technology based on the support vector machine is provided. Compared with the traditional threshold method in the prior art, the support vector machine has great advantages in solving the problems of small sample, nonlinearity and high-dimensional pattern recognition, and can improve the accuracy of satellite sea fog recognition.
Drawings
Fig. 1 is a schematic flow chart of a sea fog recognition method based on a support vector machine according to an embodiment of the present invention;
fig. 2 is a schematic view of satellite disk projection coordinates in a sea fog recognition method based on a support vector machine according to an embodiment of the present invention;
fig. 3 is a schematic diagram of a portion of a sample of standardized satellite channel data in a sea fog recognition method based on a support vector machine according to an embodiment of the present invention;
fig. 4 is a schematic diagram of a sea fog recognition result in southeast coastal areas of China in the sea fog recognition method based on the support vector machine provided by the embodiment of the invention.
Detailed Description
The technical scheme of the invention is further specifically described below by means of examples and with reference to fig. 1-4.
As shown in fig. 1, the invention aims to provide a sea fog identification method based on a support vector machine, which comprises the following steps:
step S1: acquiring satellite channel disk projection data x over a period of time 0 And ship observation raw data y 0 ；
Step S2: for satellite channel disk projection data x 0 And ship viewMeasuring raw data y 0 Decoding is carried out, and satellite channel lattice point data x, visibility observation data y and corresponding time ship position information are obtained;
step S3: the satellite channel grid point data x is subjected to standardization processing to obtain satellite channel standardized data x std The method comprises the steps of carrying out a first treatment on the surface of the Carrying out foggy/foggy binarization processing on the visibility data y to obtain visibility binarization data y of the observation point std ；
Step S4: normalizing data x of satellite channel lattice points corresponding to time and position std And visibility data y std Fusing to generate a training and testing sample set, and training a support vector machine model;
step S5: and performing sea fog identification on the test sample set through the trained support vector machine model.
In step S1, satellite channel satellite dish data x is obtained from sunflower satellite No. 8 0 The method comprises the steps of carrying out a first treatment on the surface of the Acquisition of Ship observation raw data y from International maritime weather Integrated database (ICOADS) 0 Ship observation original data y 0 Including coordinates, visibility, and time of observation points.
As shown in fig. 2, in step S2, the data decoding further includes the steps of:
step S21: the 16-channel circular projection data provided by sunflower number 8 stationary satellite was decoded.
Specifically, the longitude and latitude coordinates of the pixel point are calculated according to the geometric principle by the position of the pixel point in a coordinate system taking a satellite as an origin and the equatorial radius and polar radius of the earth, the longitude and latitude information of the pixel point in a circular chart is further obtained, and the longitude and latitude information is interpolated on equidistant theodolite points to obtain satellite channel lattice point data x. The satellite channel grid point data x is a two-dimensional matrix, and the numerical value of each position in the two-dimensional matrix represents the earth surface radiation value data acquired by the satellite of the theodolite point at the corresponding position; since the satellites provide 16 channels of data, 16 values (or 16 two-dimensional matrices) are associated with each satellite channel for each location in the two-dimensional matrix.
Step S22: from the International Meteorological comprehensive database (I)COADS) ship observation raw data y 0 The visibility observation data y and the corresponding ship position (longitude and latitude) information are extracted.
As shown in fig. 3, in step S3, the data normalization process further includes the steps of:
step S31: empirical constant x for extracting maximum value of satellite channel data max And an empirical constant x of minimum min 。
Step S32: and (3) respectively carrying out standardization processing on satellite channel data x of 16 channels by adopting a maximum value and minimum value standardization method:
in the formula, x std Is satellite channel standardized data, x is satellite channel lattice point data, x max Is the empirical constant of the maximum value of the grid point data of the satellite channel, x min Is an empirical constant of the minimum of the satellite channel grid point data.
Step S33: according to the regulation of visibility record in ICOADS data description document, a threshold Th is set, th=93 is taken as the threshold of whether sea fog exists or not, and ship observation visibility data is subjected to binarization processing according to the following formula:
in the formula, y is observation point visibility observation data, y std Is the observation point visibility binarization data.
In step S4, data fusion includes:
step S41: for a given time t, selecting observation point visibility binarized data y at the time t std And corresponding longitude and latitude coordinates (lon, lat), n groups in total; wherein n is the number of ship observation points obtained from an International Meteorological Integrated database (ICOADS), and a group of observation point visibility binarization data y is added when each ship observation data is obtained std And corresponding warpLatitude coordinates (lon, lat);
step S42: selecting 16 satellite channel standardized data x at t moment std ；
Step S43: the visibility binarization data y of the n groups of observation points screened in the step S41 std And corresponding longitude and latitude coordinate (lon, lat) data, respectively reading 16 satellite channel standardized data x of sunflower number 8 satellite at t moment std The values of the nearest lattice points of the medium-distance coordinates (lon, lat) form a characteristic quantity f;
step S44: repeating the steps S41-S43 for all the moments T in the sample set to obtain the visibility binarized data y of the observation point std A corresponding feature set F;
step S45: dividing the feature set F into a training sample and a verification sample, wherein the training sample comprises a foggy or non-foggy label, and the verification sample does not comprise a foggy or non-foggy label;
step S46: and inputting the training samples into a support vector machine model for training, and checking the training accuracy of the support vector machine model by verifying the samples.
In step S5, the sea fog identification based on satellite multichannel data has the problems of few sea fog label samples, high characteristic dimension and the like, and belongs to the problem of small sample and high dimension mode identification; the support vector machine has great advantages in solving the problem of high-dimensional pattern recognition. Therefore, the sea fog identification method selects the support vector machine model for sea fog identification.
The support vector machine model for identifying sea fog further comprises the following steps:
step S51: standardized data x of low-dimensional satellite channel to be detected std ∈R N By a kernel function k (x std ,x i ) Mapping to high-dimensional space R H ：
(φ(x std ),φ(x i ))＝k(x std ,x i )
In the formula, phi is a feature map; x is x i Is a support vector, and is obtained by training a support vector machine model.
Constructing an optimal hyperplane in the H;
step S52: in a high-dimensional space R H Constructing a linear function in an optimal hyperplane:
f(x std )＝w·k(x std ,x i )+b
where w is the slope of the linear function and b is the intercept of the linear function.
Step S53: and judging the output result:
y std ＝sgn(f(x std ))
in the formula, sgn is a step function; if y std If the value is 1, the fog is determined; y is std And=0, then it is determined that there is no fog.
In one embodiment, to verify model robustness and recognition accuracy, the model is verified with a separate measured dataset. The independent test set was determined using a random extraction method and contained 4200 sets of samples, 1200 of which were fog-free and 3000 of which were non-fog samples. Identifying the fog samples in the test results to obtain 1013 identified fog results and 187 identified no fog results; and identifying the haze-free samples to obtain 362 identified haze results and 2638 identified haze-free results.
Test results show that the accuracyFalse alarm rate->
In another embodiment, as shown in fig. 4, the present invention supports the recognition result of the vector machine model on sea fog in the southeast coastal area of China. From the above, the invention can effectively identify the sea fog in the coastal region of the China.
While the invention has been disclosed in terms of preferred embodiments, the embodiments are not intended to limit the invention. Any equivalent changes or modifications can be made without departing from the spirit and scope of the present invention, and are intended to be within the scope of the present invention. The scope of the invention should therefore be determined by the following claims.
Claims (4)
1. The sea fog identification method based on the support vector machine is characterized by comprising the following steps of:
step S1: acquiring satellite channel disk projection data x over a period of time 0 And ship observation raw data y 0 ；
Step S2: for the satellite channel disk projection data x 0 And ship observation raw data y 0 Decoding is carried out, and satellite channel lattice point data x, visibility observation data y and corresponding time ship position information are obtained;
step S3: the satellite channel lattice point data x is subjected to standardization processing to obtain satellite channel standardized data x std The method comprises the steps of carrying out a first treatment on the surface of the Carrying out foggy/non-foggy binarization processing on the visibility observation data y to obtain visibility binarization data y of the observation point std ；
Step S4: normalizing the grid point of the satellite channel corresponding to time and position to obtain data x std Binary data y of visibility with the observation point std Fusing to generate a training and testing sample set, and training a support vector machine model;
step S5: performing sea fog recognition on the data to be detected through the trained support vector machine model;
in step S2, the satellite channel disk projection data x 0 Decoding, namely calculating longitude and latitude coordinates of the pixel point according to a geometric principle by using the position of the pixel point in a coordinate system taking a satellite as an origin and the equatorial radius and polar radius of the earth, further obtaining longitude and latitude information of the pixel point in a circular chart, and interpolating the longitude and latitude information to equidistant theodolite points to obtain satellite channel lattice point data x;
observing the original data y of the ship 0 Decoding raw data y observed from the vessel 0 Extracting the visibility observation data y and the corresponding ship position information;
in step S3, the method further comprises the steps of:
step S31: empirical constant x for extracting maximum value of satellite channel data max And an empirical constant x of minimum min
Step S32: and respectively carrying out standardization processing on the satellite channel data x of 16 channels by adopting a maximum value and minimum value standardization method:
in the formula, x std Is satellite channel standardized data, x is satellite channel lattice point data, x max Is the empirical constant of the maximum value of the grid point data of the satellite channel, x min Is an experience constant of the minimum value of the satellite channel lattice point data;
in step S4, the method further includes the following steps:
step S41: for a given time t, selecting the observation point visibility binarized data y at the time t std And corresponding longitude and latitude coordinates (lon, lat), n groups in total;
step S42: selecting 16 satellite channel standardized data x at t moment std ；
Step S43: the visibility binarization data y of n groups of observation points screened in the step S41 std And corresponding longitude and latitude coordinate (lon, lat) data, respectively reading 16 satellite channel standardized data x at t time std The values of the nearest lattice points of the medium-distance coordinates (lon, lat) form a characteristic quantity f;
step S44: repeating the steps S41-S43 for all the moments T in the sample set to obtain the visibility binarized data y of the observation point std A corresponding feature set F;
step S45: dividing the feature set F into a training sample and a verification sample, wherein the training sample comprises a foggy or non-foggy label, and the verification sample does not comprise a foggy or non-foggy label;
step S46: inputting the training sample into a support vector machine model for training, and checking the training accuracy of the support vector machine model through the verification sample;
in step S5, the method further includes the following steps:
step S51: standardized data x of low-dimensional satellite channel to be detected std ∈R N By a kernel function k (x std ,x i ) Mapping to high-dimensional space R H Constructing an optimal hyperplane in the H;
step S52: in a high-dimensional space R H Constructing a linear function in an optimal hyperplane:
step S53: and calculating the linear function through a step function, and judging an output result.
2. The sea fog recognition method according to claim 1, wherein in step S51, the low-dimensional satellite channel standardized data x std ∈R N By a kernel function k (x std ,x i ) Mapping to high-dimensional space R H The method comprises the following steps:
(φ(x std ),φ(x i ))＝k(x std ,x i )
in the formula, phi is a feature map; x is x i Is a support vector, and is obtained by training a support vector machine model.
3. The sea fog recognition method according to claim 1, wherein in step S52, the linear function is:
f(x std )＝w·k(x std ,x i )+b
where w is the slope of the linear function and b is the intercept of the linear function.
4. The sea fog recognition method according to claim 1, wherein in step S53, the decision on the output result is:
y std ＝sgn(f(x std ))
in the formula, sgn is a step function; if y std If the value is 1, the fog is determined; y is std And=0, then it is determined that there is no fog.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911408481.9A CN111222439B (en) | 2019-12-31 | 2019-12-31 | Sea fog identification method based on support vector machine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911408481.9A CN111222439B (en) | 2019-12-31 | 2019-12-31 | Sea fog identification method based on support vector machine |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111222439A CN111222439A (en) | 2020-06-02 |
| CN111222439B true CN111222439B (en) | 2024-03-22 |
Family
ID=70825872
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911408481.9A Active CN111222439B (en) | 2019-12-31 | 2019-12-31 | Sea fog identification method based on support vector machine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111222439B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103885067A (en) * | 2014-03-24 | 2014-06-25 | 中国海洋大学 | Method for verifying satellite sea fog remote sensing ratio measurement by satellite-borne laser radar |
| CN104217215A (en) * | 2014-08-28 | 2014-12-17 | 哈尔滨工程大学 | Classification and identification method for foggy water surface image and clear water surface image |
| CN110208880A (en) * | 2019-06-05 | 2019-09-06 | 北京邮电大学 | A kind of sea fog detection method based on deep learning and satellite remote sensing technology |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108761484A (en) * | 2018-04-26 | 2018-11-06 | 江苏省气象台 | A kind of sea fog monitoring method based on Multi-sensor satellite remote sensing |
-
2019
- 2019-12-31 CN CN201911408481.9A patent/CN111222439B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103885067A (en) * | 2014-03-24 | 2014-06-25 | 中国海洋大学 | Method for verifying satellite sea fog remote sensing ratio measurement by satellite-borne laser radar |
| CN104217215A (en) * | 2014-08-28 | 2014-12-17 | 哈尔滨工程大学 | Classification and identification method for foggy water surface image and clear water surface image |
| CN110208880A (en) * | 2019-06-05 | 2019-09-06 | 北京邮电大学 | A kind of sea fog detection method based on deep learning and satellite remote sensing technology |
Non-Patent Citations (1)
| Title |
|---|
| 张春桂 ; 林炳青 ; .基于FY-2E卫星数据的福建沿海海雾遥感监测.国土资源遥感.2018,(第01期),全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111222439A (en) | 2020-06-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP2927709B1 (en) | Identifying and tracking convective weather cells | |
| Chen et al. | Ground and top of canopy extraction from photon-counting LiDAR data using local outlier factor with ellipse searching area | |
| JP2019211342A (en) | Weather analyzer, weather analysis method, and program | |
| Ekhtesasi et al. | Determining area affected by dust storms in different wind speeds, using satellite images | |
| Han et al. | Evaluation of summer passive microwave sea ice concentrations in the Chukchi Sea based on KOMPSAT-5 SAR and numerical weather prediction data | |
| CN116467565B (en) | Enteromorpha green tide plaque optimal search area forecasting method | |
| Chen et al. | Rain-contaminated region segmentation of X-band marine radar images with an ensemble of SegNets | |
| CN109948697B (en) | Method for extracting urban built-up area by using multi-source data to assist remote sensing image classification | |
| CN117173590A (en) | Water body abnormality monitoring method based on multisource time sequence remote sensing image | |
| Koo et al. | Semi-automated tracking of iceberg B43 using Sentinel-1 SAR images via Google Earth Engine | |
| CN117171533B (en) | Real-time acquisition and processing method and system for geographical mapping operation data | |
| Xu et al. | Obtaining accurate water level measurements in lakes: analysis of changes in icesat altimetry accuracy with buffer changes | |
| CN111222439B (en) | Sea fog identification method based on support vector machine | |
| Sun et al. | Deep learning for check dam area extraction with optical images and digital elevation model: A case study in the hilly and gully regions of the Loess Plateau, China | |
| Xie et al. | Machine-Learning-Method-Based Inversion of Shallow Bathymetric Maps Using ICESat-2 ATL03 Data | |
| CN106709473B (en) | Voxel-based airborne LIDAR road extraction method | |
| CN115436966A (en) | Batch extraction method for laser radar reference water depth control points | |
| Hill et al. | Extension of the QuikSCAT sea ice extent data set with OSCAT data | |
| Wei et al. | Long-term observation of global nuclear power plants thermal plumes using Landsat images and deep learning | |
| Cao et al. | Ocean Mesoscale Eddies Identification Based on Yolof | |
| CN113281716A (en) | Photon counting laser radar data denoising method | |
| Rajeswari et al. | Classification based land use/land cover change detection through Landsat images | |
| Rana et al. | HYBRID TECHNIQUE FOR DETECTION OF WATER USING SATELLITE IMAGES. | |
| CN113807409B (en) | Coastal zone classification method based on discriminant analysis | |
| Lasko et al. | Automated method for artificial impervious surface area mapping in temperate, tropical, and arid environments using hyperlocal training data with Sentinel-2 imagery |
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 |