CN107728121A - A kind of Local Good-fit test method based on variable window - Google Patents
A kind of Local Good-fit test method based on variable window Download PDFInfo
- Publication number
- CN107728121A CN107728121A CN201710762845.8A CN201710762845A CN107728121A CN 107728121 A CN107728121 A CN 107728121A CN 201710762845 A CN201710762845 A CN 201710762845A CN 107728121 A CN107728121 A CN 107728121A
- Authority
- CN
- China
- Prior art keywords
- sea clutter
- local
- amplitude
- fitting
- goodness
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/40—Means for monitoring or calibrating
- G01S7/4052—Means for monitoring or calibrating by simulation of echoes
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention discloses a kind of Local Good-fit test method based on variable window.The test of fitness of fot is the rational judgment to sea clutter Amplitude Distributed Model, overcoming mean square error examines (MSD) and amendment side to examine (MMSD) method to observe the problem of section is more single to amplitude, section is observed by choosing one or more radars local amplitude interested, calculate the weighted results of the selected local section test of fitness of fot, it this approach enhance the comprehensive of statistical model test of fitness of fot result, the design of judgement and sea-surface target CFAR detection (CFAR) method for sea clutter amplitude sequence statistical model provides more structurally sound foundation.
Description
Technical Field
The invention relates to a local goodness-of-fit testing method for sea clutter amplitude statistical distribution based on a variable window, and belongs to the technical field of radars.
Background
The sea target detection is an important task of the radar, however, the sea clutter is an important factor influencing the radar target detection, and in order to effectively realize the target detection under the background of the sea clutter, a statistical distribution model of the sea clutter must be analyzed. For actually measured sea clutter data, it is inaccurate to select the best-fitting sea clutter amplitude model from various distribution models by visual observation only, so a statistic for quantitatively describing the fitting degree of the sea clutter amplitude model and the actually measured data, namely, a goodness-of-fit test, is needed. The traditional classical goodness-of-fit inspection method mainly comprises the following steps: MSD method and MMSD method. However, the MSD method only focuses on global statistical fitting of sea clutter amplitude statistics, and neglects the fitting effect of amplitude local intervals; the MMSD method only focuses on the fitting effect of the sea clutter large-amplitude sequence, namely the tailing fitting effect. The two typical goodness-of-fit inspection methods only consider the global or tail fitting result of the sea clutter amplitude singly, and neglect the fitting condition in the sea clutter local amplitude interval under different sea condition conditions. When the radar works actually, complex sea clutter echo signals under different sea conditions can be met, and the model fitting result for analyzing the sea clutter from the whole world or the tail part only is not comprehensive and is not accurate.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: the method for testing the local goodness of fit based on the variable window overcomes the defects of the prior art, can test the fitting condition in the local amplitude interval of the sea clutter under different sea conditions, enhances the reliability and comprehensiveness of the goodness of fit test, and provides a more reliable basis for target constant false alarm detection.
The technical solution of the invention is as follows: a local goodness-of-fit inspection method based on a variable window comprises the following steps:
(1) Extracting sea clutter amplitude sequence X = { X } from sea clutter echo data measured by radar k K =1,2,3, a e (x l ) Wherein N is the length of the sea clutter amplitude sequence, L =1,2,3, …, L;
(2) According to the sea clutter amplitude sequence, calculating the amplitude statistical distribution model parameter of the sequence by adopting a moment estimation method, and fitting the sea clutter theoretical probability distribution function p of the statistical model by utilizing a common fitting method t (x l )；
(3) Determining the number, position and length of local observation windows;
(4) The theoretical probability distribution function p of the sea clutter, which is fitted by aiming at each fitting method t (x l ) Calculating the local goodness-of-fit test result of each observation window, and obtaining the sea clutter theoretical probability distribution function p fitted by each fitting method through weighting processing t (x l ) And the actual probability density distribution p e (x l ) The final fitting result of (2);
(5) The minimum value of the final fitting result corresponds to the theoretical probability distribution function p of the sea clutter t (x l ) Namely the statistical distribution model of the amplitude sequence of the sea clutter with the best fitting.
In the step (1), the sea clutter amplitude sequence is equally divided into L intervals to obtain the actual probability density distribution p of sea clutter amplitude statistics e (x l ) The method comprises the following steps:
equally dividing the sea clutter amplitude sequence into L intervals according to the maximum value and the minimum value in the sea clutter amplitude sequence, counting the number of the sea clutter amplitudes falling into each interval, and obtaining the actual probability density distribution function p of sea clutter amplitude statistics e (x l )，p e (x l ) And (= number of sea clutter amplitudes falling in the l-th interval)/N. Wherein, the interval L>20。
In the step (2), the number of the local observation windows is equal to the number of the radar-interested local amplitude sequence areas, and the position [ N ] of the ith local observation window i1 ，N i2 ]Coinciding with the i-th local amplitude sequence region of interest of the radar, N i1 ,N i2 Respectively represents the starting position and the ending position of the ith local observation window, and the length len of the ith local observation window i ＝N i2 -N i1 ≥4，N i2 、N i1 Are all positive integers.
In the step (4), the goodness-of-fit test result of the ith local observation window is calculated by using the following formula:
in the step (4), the final fitting result D of the j fitting method is calculated by using the following formula lmsd ：
D lmsd ＝α 1 D lmsd,1 +α 2 D lmsd,2 +…+α i D lmsd,i +…+α M D lmsd,M
0<α i < 1, and alpha 1 +α 2 +…+α i +…+α M &And (lt) 1,M is the total number of local observation windows.
α i Is proportional to the sea state grade.
Compared with the prior art, the invention has the following beneficial effects:
the method solves the problem that the amplitude observation interval is single in the traditional method, provides the idea of windowing, can select the number, the position and the length of the local observation windows according to the actual observation requirement of the radar, matches the most suitable distribution model by calculating the inspection result of goodness of fit after windowing, and can inspect the fitting condition in the sea clutter local amplitude observation interval under different sea conditions. When the radar needs to take the fitting results of multiple groups of different intervals (multiple local intervals or multiple local intervals and the whole amplitude interval) into consideration, the combined fitting results of the multiple groups of intervals can be weighted and evaluated, the comprehensiveness of the statistical model fitting goodness test result is enhanced, and a powerful support is provided for the design of the sea surface target constant false alarm detection algorithm.
Drawings
FIG. 1 is a flow chart of a method for goodness of fit of variable window based local statistics;
FIG. 2 is a statistical fit result of an actually measured sea clutter amplitude sequence;
fig. 3 shows the goodness-of-fit test results of the actually measured sea clutter data, (a) shows MSD test results of four common distributions, (b) shows MMSD test results of four common distributions, and (c) shows local goodness-of-fit test results of four common distributions.
Detailed Description
In consideration of model fitting conditions under different sea conditions, the invention provides a variable window-based local goodness of fit (LMSD) method, which overcomes the problem that the existing method is single in amplitude observation area, and can calculate the goodness of fit inspection result of a local amplitude observation interval through windowing according to the observation requirement of a radar; when a plurality of observation intervals with different amplitudes are interested, the amplitude statistical goodness-of-fit test result under the combined action of the plurality of intervals can be calculated through weighting processing of the plurality of observation intervals, so that the reliability and the comprehensiveness of the goodness-of-fit test are enhanced, and a more reliable basis is provided for judgment of a sea clutter amplitude statistical model and design of a sea surface target constant false alarm detection method.
As shown in fig. 1, the main implementation steps of the present invention are as follows:
step 1, extracting a sea clutter amplitude sequence X = { X } from sea clutter echo data actually measured by a radar k K =1,2,3,.., N }, where N is the length of the sea clutter amplitude sequence. Equally dividing the sea clutter amplitude sequence into L intervals according to the maximum value and the minimum value in the sea clutter amplitude sequence, counting the occurrence frequency of the sea clutter amplitude in each interval, and obtaining the actual probability density distribution function p of sea clutter amplitude statistics e (x l ). Where L is not chosen too small, and a smaller L results in a fitting junction of the amplitude distribution functionIf the fruit is inaccurate, therefore, L is generally chosen>20。l＝1,2,3,…,L。
The interval between two adjacent intervals is (maximum value-minimum value in amplitude sequence)/L, p e (x l ) And = dividing the number of the sea clutter amplitudes falling into the l-th interval by the total number N of the sea clutter amplitudes.
Step 2, sea clutter amplitude statistical fitting:
according to the sea clutter amplitude sequence, calculating the amplitude statistical distribution model parameter of the sequence by adopting a moment estimation method, and fitting the sea clutter theoretical probability distribution function p of the statistical model by utilizing a common fitting method t (x l )。
Step 3, determining the number, position and length of the local observation windows:
3.1 select the number M of observation windows. If the radar is only interested in a single local region of the sea clutter amplitude statistics, a single rectangular window is selected, i.e. M =1. If the radar is interested in a plurality of local areas of sea clutter amplitude statistics, a rectangular window is selected for each area, namely M >1, and then weighting processing is carried out on a plurality of local area fitting results. The invention mainly takes the fitting result of one group or two groups of observation intervals as an example for analysis, if more observation intervals are interested, M can also be selected to be a larger value, and the analysis method is similar to the condition of M = 2.
3.2 after the number of the observation windows is determined according to the number 3.1, the position of the observation windows can be selected according to the interested area of the radar. When the radar is interested in the statistical distribution of the local amplitude of a certain part of the sea clutter, a rectangular window can be added to any observation interval; a rectangular window may be added over the entire observation interval when the radar is interested in a global fitting statistical distribution of sea clutter.
3.3 selection of the local observation window length. Let the position of the ith local observation window be [ N ] i1 ，N i2 ]，N i1 ,N i2 Respectively representing the start and end positions, N, of the ith local observation window i2 、N i1 Are all positive integers, and the length len of the ith local observation window i ＝N i2 -N i1 Not less than 4, i.e. the amplitude range of the windowed sequence should cover 4N/L of the amplitude of the whole sequenceTherefore, enough amplitude samples can be guaranteed to participate in operation, and the influence of random errors caused by too few noise and samples is reduced.
Step 4, calculating the local goodness of fit (LMSD) result of each observation window
For the jth fitting method, the result D of goodness-of-fit test of the ith local observation window lmsd,i ：
Calculating the final fitting result D of the jth fitting method by using the following formula lmsd ：
D lmsd ＝α 1 D lmsd,1 +α 2 D lmsd,2 +…+α i D lmsd,i +…+α M D lmsd,M
0<α i < 1, and alpha 1 +α 2 +…+α i +…+α M &And lt 1,M is the total number of local observation windows.
α i Is proportional to the level of the sea state. Alpha is selected according to actual observation needs i The higher the sea state grade is, the higher the weight of the large-amplitude sequence can be increased; the lower the sea state level is, the more the weight of the amplitude sequence can be reduced.
The choice of M weights is analyzed for the example of M =2, and so on for the case of M >2, i.e.:
D lmsd ＝α 1 D lmsd,1 +α 2 D lmsd,2 ,0<α 1 <1,0<α 2 <1
wherein alpha is 1 ,α 2 Respectively representing the weight values of the two groups of goodness-of-fit results, and satisfying alpha 1 +α 2 ＝1。
For the weight value alpha 1 ,α 2 Can be selected according to the actual radar observation needs. Under the condition of high sea, the weight of the large-amplitude observation interval can be increased; in the case of low sea conditions, canAnd increasing the weight of the small-amplitude observation interval.
Suppose D lmsd,1 And D lmsd,2 Respectively representing the fitting test result of the global amplitude and the fitting test result of the local amplitude, and if the radar pays attention to the result of the statistical combined action of the amplitude of the sea clutter and the global amplitude, selecting alpha 1 ＝α 2 =0.5; if the radar is more focused on global fit test results, α may be selected 1 ＝0.7,α 2 =0.3; if the radar is more focused on local fitting test results, α may be selected 1 ＝0.3,α 2 =0.7. Wherein alpha is 1 ,α 2 The proportion of (A) can be properly adjusted according to actual needs.
Step 5, judging a statistical distribution model:
obtaining a local goodness-of-fit test result D of the sea clutter amplitude according to the step 4 lmsd Similar to conventional MSD and MMSD, D lmsd A smaller value of (a) indicates a better fit over this interval. By comparing different statistical distribution models lmsd As a result, the set of sea clutter data can be determined to fit the best statistical distribution model.
The effect of the invention is demonstrated by the test of measured data:
simulation conditions
Extracting sea clutter echo amplitude sequence X = { X = from sea clutter echo data actually measured by certain shore-based S-band radar k ,k＝1,2,3,...,N}。
(II) simulation content
FIG. 2 is a statistical fit result of the actually measured sea clutter amplitude sequence. The test divides the sea clutter amplitude sequence into L (L = 50) observation intervals, and the histogram is the actual probability density distribution p of the sea clutter amplitude statistics e (x l ) The four curves respectively correspond to the sea clutter theoretical probability distribution function p obtained by fitting the four fitting methods t (x l )。
Aiming at the fitting result of the measured data in fig. 2, the MSD test, the MMSD test and the LMSD test proposed by the present invention are respectively adopted to evaluate the goodness and badness of the fitting result.
Fig. 3 (a) shows the results of four common distributions of MSD tests, which show that: for the global statistical fitting of the sea clutter amplitude, the fitting effect of the K distribution is the best, the Weibull distribution and the Rayleigh distribution are the second, and the fitting effect of the lognormal distribution is the worst.
FIG. 3 (b) shows the results of MMSD test for four common distributions, which can be seen: for the statistical fitting of the sea clutter large amplitude sequence (tailing fitting result), the fitting effect of Weibull distribution is the best, and the fitting effect of K distribution, rayleigh distribution and lognormal distribution is the worst.
FIG. 3 (c) shows the local goodness-of-fit test results for four common distributions, where two observation windows are selected, the first window being [1, 50 ], taking the weighted fit result as an example]Calculating a global fitting test result of the amplitude; the second observation window is [40, 50 ]]And calculating a fitting test result of the amplitude tail. In addition, the weight of two groups of fitting test results is selected as follows: alpha is alpha 1 ＝α 2 =0.5. As can be seen in fig. 3 (c): for the statistical fitting of the overall sea clutter amplitude and the combined action of the tail, the fitting effect of K distribution is the best, and then Weibull distribution and Rayleigh distribution are carried out, and the fitting effect of log-normal distribution is the worst.
The MSD and the MMSD only consider partial information of sea clutter amplitude, however, the local goodness-of-fit inspection method provided by the invention integrates global and local information, and is more comprehensive than the number of samples adopted from the whole, tail and other partial intervals, so that the obtained conclusion is more accurate and reliable.
Those skilled in the art will appreciate that the details of the invention not described in detail in the specification are within the skill of those skilled in the art.
Claims (6)
1. A local goodness-of-fit inspection method based on a variable window is characterized by comprising the following steps:
(1) Extracting a sea clutter amplitude sequence X = { X } from sea clutter echo data actually measured by radar k K =1,2,3, the.., N }, equally dividing the sea clutter amplitude sequence into L intervals to obtain the actual probability density distribution p of the sea clutter amplitude statistics e (x l ) Wherein N isLength of sea clutter amplitude sequence, L =1,2,3, …, L;
(2) According to the sea clutter amplitude sequence, calculating the amplitude statistical distribution model parameter of the sequence by adopting a moment estimation method, and fitting the sea clutter theoretical probability distribution function p of the statistical model by utilizing a common fitting method t (x l )；
(3) Determining the number, position and length of local observation windows;
(4) Sea clutter theoretical probability distribution function p fitted by aiming at each fitting method t (x l ) Calculating the local fitting goodness test result of each observation window, and obtaining the sea clutter theoretical probability distribution function p fitted by each fitting method through weighting processing t (x l ) And the actual probability density distribution p e (x l ) The final fitting result of (1);
(5) The minimum value of the final fitting result corresponds to the theoretical probability distribution function p of the sea clutter t (x l ) Namely the statistical distribution model of the amplitude sequence of the sea clutter with the best fitting.
2. The method of claim 1, wherein the local goodness-of-fit test based on the variable window is as follows: in the step (1), the sea clutter amplitude sequence is equally divided into L intervals to obtain the actual probability density distribution p of sea clutter amplitude statistics e (x l ) The method comprises the following steps:
equally dividing the sea clutter amplitude sequence into L intervals according to the maximum value and the minimum value in the sea clutter amplitude sequence, counting the number of the sea clutter amplitudes falling into each interval, and obtaining the actual probability density distribution function p of sea clutter amplitude statistics e (x l )，p e (x l ) And = N number of sea clutter amplitudes falling in the l-th interval. Wherein, the interval L>20。
3. The method of claim 1, wherein the variable window-based local goodness-of-fit test comprises: in the step (2), the number of the local observation windows is equal to the number of the local amplitude sequence region interested by the radar, and the ith local observation window is positionedN i1 ，N i2 ]Coinciding with the i-th local amplitude sequence region of interest of the radar, N i1 ,N i2 Respectively represents the starting position and the ending position of the ith local observation window, and the length len of the ith local observation window i ＝N i2 -N i1 ≥4，N i2 、N i1 Are all positive integers.
4. The method of claim 3, wherein the local goodness-of-fit test based on the variable window is as follows: in the step (4), the goodness-of-fit test result of the ith local observation window is calculated by using the following formula:
5. the method of claim 4, wherein the local goodness-of-fit test based on the variable window is as follows: in the step (4), the final fitting result D of the j fitting method is calculated by using the following formula lmsd ：
D lmsd ＝α 1 D lmsd,1 +α 2 D lmsd,2 +…+α i D lmsd,i +…+α M D lmsd,M
0<α i < 1, and alpha 1 +α 2 +…+α i +…+α M &And lt 1,M is the total number of local observation windows.
6. The method of claim 5, wherein the local goodness-of-fit test based on the variable window is as follows: alpha is alpha i Is proportional to the sea state grade.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710762845.8A CN107728121B (en) | 2017-08-30 | 2017-08-30 | Local goodness-of-fit inspection method based on variable window |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710762845.8A CN107728121B (en) | 2017-08-30 | 2017-08-30 | Local goodness-of-fit inspection method based on variable window |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107728121A true CN107728121A (en) | 2018-02-23 |
| CN107728121B CN107728121B (en) | 2020-04-10 |
Family
ID=61205556
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710762845.8A Active CN107728121B (en) | 2017-08-30 | 2017-08-30 | Local goodness-of-fit inspection method based on variable window |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107728121B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111580091A (en) * | 2020-05-17 | 2020-08-25 | 西北工业大学 | Sea surface weak target detection method based on AR spectrum singular intensity function |
| US11347152B2 (en) | 2018-01-17 | 2022-05-31 | Asml Netherlands B.V. | Scan signal characterization diagnostics |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104992159A (en) * | 2015-07-13 | 2015-10-21 | 北京化工大学 | SAR image ship detection method based on Gaussian function fitting of sea clutter |
| CN105046706A (en) * | 2015-07-13 | 2015-11-11 | 北京化工大学 | Rational polynomial function fitting sea clutter based SAR image ship detection method |
| CN105069775A (en) * | 2015-07-13 | 2015-11-18 | 北京化工大学 | SAR image ship detection method based on Fourier series |
-
2017
- 2017-08-30 CN CN201710762845.8A patent/CN107728121B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104992159A (en) * | 2015-07-13 | 2015-10-21 | 北京化工大学 | SAR image ship detection method based on Gaussian function fitting of sea clutter |
| CN105046706A (en) * | 2015-07-13 | 2015-11-11 | 北京化工大学 | Rational polynomial function fitting sea clutter based SAR image ship detection method |
| CN105069775A (en) * | 2015-07-13 | 2015-11-18 | 北京化工大学 | SAR image ship detection method based on Fourier series |
Non-Patent Citations (1)
| Title |
|---|
| YIFEI FAN ET AL.: ""Weak target detection in sea clutter background using local-multifractal spectrum with adaptive window length"", 《IET RADAR, SONAR & NAVIGATION》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11347152B2 (en) | 2018-01-17 | 2022-05-31 | Asml Netherlands B.V. | Scan signal characterization diagnostics |
| CN111580091A (en) * | 2020-05-17 | 2020-08-25 | 西北工业大学 | Sea surface weak target detection method based on AR spectrum singular intensity function |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107728121B (en) | 2020-04-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111722199B (en) | Radar signal detection method based on convolutional neural network | |
| Cho et al. | Identification and removal of ground echoes and anomalous propagation using the characteristics of radar echoes | |
| CN104898103B (en) | Low velocity target detection method based on multichannel clutter map | |
| CN110208806B (en) | Marine radar image rainfall identification method | |
| CN109143195B (en) | Radar target detection method based on full KL divergence | |
| CN103760542A (en) | MMVI-CFAR target detection method | |
| CN110658508B (en) | K distribution sea clutter parameter estimation method based on characteristic quantity | |
| CN103995259B (en) | Radar target adaptive-filtering fusion detection method under intensive interference environment | |
| CN109212503B (en) | Offshore target detection method and device based on AR spectrum fractal correlation characteristics | |
| Newman et al. | Range-correcting azimuthal shear in Doppler radar data | |
| Richardson et al. | Bragg scatter detection by the WSR-88D. Part I: Algorithm development | |
| CN112859012B (en) | Radar spoofing interference identification method based on cascade convolution neural network | |
| Toledo et al. | Cluster analysis: A new approach applied to lidar measurements for atmospheric boundary layer height estimation | |
| Caracciolo et al. | Precipitation classification at mid-latitudes in terms of drop size distribution parameters | |
| Kuster et al. | Rapid-update radar observations of ZDR column depth and its use in the warning decision process | |
| CN107728121B (en) | Local goodness-of-fit inspection method based on variable window | |
| CN104198998B (en) | Clustering treatment based CFAR (Constant False Alarm Rate) detection method under non-uniform background | |
| CN112327293A (en) | Sea surface target detection method based on fractal feature intelligent learning | |
| Zou et al. | A method of radar echo extrapolation based on TREC and Barnes filter | |
| Williams et al. | Cluster analysis techniques to separate air motion and hydrometeors in vertical incident profiler observations | |
| CN117787081A (en) | Hydrological model parameter uncertainty analysis method based on Morris and Sobol methods | |
| CN116879899A (en) | Method based on aerial precipitation particle spectrum inversion | |
| CN106526545A (en) | Method for detecting robust CFAR detector | |
| Liu et al. | Using wavelet analysis to detect tornadoes from Doppler radar radial-velocity observations | |
| CN113296072B (en) | Method and system for automatically identifying thunderstorm strong wind based on YOLOv3 model |
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 |