CN110940971B - Radar target point trace recording method and device and storage medium - Google Patents
Radar target point trace recording method and device and storage medium Download PDFInfo
- Publication number
- CN110940971B CN110940971B CN201911076994.4A CN201911076994A CN110940971B CN 110940971 B CN110940971 B CN 110940971B CN 201911076994 A CN201911076994 A CN 201911076994A CN 110940971 B CN110940971 B CN 110940971B
- Authority
- CN
- China
- Prior art keywords
- data
- track
- radar
- point
- prediction model
- 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
Images
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
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
- G01S13/50—Systems of measurement based on relative movement of target
- G01S13/58—Velocity or trajectory determination systems; Sense-of-movement determination systems
-
- 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)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention discloses a radar target point trace recording method, which comprises the following steps: reading radar data of N periods, wherein N is a natural number, and processing the radar data of the N periods by using a logic starting method to establish temporary track data; inputting the temporary track data into a track prediction model to generate predicted track data, wherein the track prediction model comprises a track prediction model trained by a Long Short-Term Memory network (LSTM) in a Recurrent Neural Network (RNN); and comparing the predicted track data with the real point track data in the radar scanning period, and selecting the point track data of which the distance between the predicted track data and the real point track meets the preset condition to realize point track recording. The method can realize intelligent accurate admission of radar points by applying a deep learning technology under the condition of low CFAR threshold.
Description
Technical Field
The invention relates to the field of computer application, in particular to radar point trace data, which realizes intelligent and accurate recording of radar point traces by applying a deep learning technology under the condition of a low CFAR threshold value.
Background
The traditional method for radar target point trace recording adopts a digital signal processing technology, adopts methods such as pulse compression, moving target display (MTI), Moving Target Detection (MTD) and the like to suppress clutter, adopts a Constant False Alarm Rate (CFAR) radar signal detection method to reduce non-target signal interference, and has the inevitable interference with the identification of target point traces in the radar detection process due to the complex conditions of random noise, ground clutter and meteorological clutter. In order to extract signals in strong interference, the acquired signals are required to have a certain signal-to-noise ratio, and constant false alarm processing must be performed on the signals. Constant false alarm refers to that in order to ensure the identification precision and efficiency of the radar, in the radar signal detection, when the external interference intensity changes, the radar can automatically adjust the sensitivity thereof, so that the false alarm probability of the radar remains unchanged, and the characteristic is called as constant false alarm rate characteristic. The false alarm probability refers to the probability that a target is judged to be present when no target actually exists due to the ubiquitous and fluctuating noise in the radar detection process by adopting a threshold detection method.
Since the data processor will process the result to have a large number of false targets and display them on the radar terminal simultaneously with the moving target, it will make the quick finding and observation of the target difficult. At present, the basic method for suppressing false targets in radar data processing is to perform fast starting and filtering tracking of target tracks by data association and using a logic method. For example, a method for processing radar data is disclosed in chinese patent application No. 201910330735.3, which includes: acquiring radar trace data and preprocessing the radar trace data to obtain to-be-selected trace data corresponding to-be-selected traces, wherein the to-be-selected trace data comprises trace attribute information; performing first-layer and second-layer associated screening on the point trace data to be selected by utilizing at least two wave gates to obtain target point trace data, and forming a temporary flight path by using a target point trace corresponding to the target point trace data; performing track starting judgment on the temporary track by using a logic method, and starting the temporary track as a real track if the temporary track meets the condition; and based on the real track, tracking and filtering by using a Kalman filter. When a similar data processing method is adopted, the selection of a high constant false alarm rate can easily cause the increase of the false-positive rate, and the selection of a lower constant false alarm rate threshold condition can easily cause the reduction of the accuracy of the dotting and the increase of the calculation load of hardware.
With the development of neural network technology, people are also researching technology for predicting tracks by using a neural network to process radar data. For example, in chinese patent application (application No. 201810040500.6), a target track prediction method based on a recurrent neural network is disclosed, which includes collecting radar measurement point tracks and tracking track data of the same model in multiple scenes, simultaneously collecting a target track by using a cooperative target information receiving device, removing and correcting the data to form a track original data set, then constructing a target track prediction recurrent neural network, setting training sample feature vectors to generate a track training set, and finally training and optimizing the target track prediction recurrent neural network based on the cooperative track training set and the radar track training set to generate a target track prediction method matched with a radar.
Disclosure of Invention
The invention aims to overcome the defects in the prior art and provide a radar target trace recording method, a device and a storage medium, which can reduce the false alarm rate in radar target trace recording and improve the recording accuracy.
In order to achieve the above purpose, the invention provides the following technical scheme:
a radar target point trace recording method comprises the following steps:
reading radar data of N periods, wherein N is a natural number, and processing the radar data of the N periods by using a logic starting method to establish temporary track data;
inputting the temporary track data into a track prediction model to generate predicted track data, wherein the track prediction model comprises a track prediction model trained by a Long Short-Term Memory network (LSTM) in a Recurrent Neural Network (RNN);
and comparing the predicted track data with the real point track data in the radar scanning period, and selecting the point track data of which the distance between the predicted track data and the real point track meets the preset condition to realize point track recording.
In some embodiments, the meeting of the predetermined condition is that the distance is less than or equal to 2 meters.
Some preferred technical solutions of the present invention are as follows:
preferably, the threshold of the constant false alarm rate value is adjusted down when the radar data of N periods is read.
Preferably, the threshold of the constant false alarm rate value is set to 3 or less when N periods of radar data are read. This threshold is typically below the normal threshold value of 5.
After the Constant False Alarm Rate (CFAR) threshold is adjusted down or set low, a large number of trace signals can enter a data processing stage, more target signals are allowed to pass through, and the possibility of false alarm is reduced.
Preferably, the trajectory prediction model training comprises data screening, feature selection, label labeling and model training, wherein the data screening comprises: firstly, extracting track attribute data, wherein the attribute data comprises time, distance, pitch, azimuth and signal intensity, and selecting the track attribute data value smaller than a preset threshold vector as same-target data.
Preferably, the time, distance, azimuth, pitch, and radar signal echo strength in the track attribute data are selected as relevant features of the track prediction model.
Preferably, a sliding window mode is used in label labeling to determine a label, the window width is n, the window step size is 1, the sequence length is L, when L > n, the window slides forward in 1 step size, when L < n, and the insufficient part is occupied by null values, where n is a natural number and L is a natural number greater than 3.
The present invention also provides an electronic device, comprising:
a processor, a memory, and a computer program; wherein the computer program is stored in the memory and configured to be executed by the processor, the computer program comprising instructions for performing the radar target spot logging method described above.
The invention also provides a computer readable storage medium, which stores a computer program, and when the computer program is executed, the method for realizing the radar target point admission method is realized.
Compared with the prior art, the invention has the beneficial effects that: and the intelligent and accurate admission of radar points is realized on the basis of lower hardware overhead by applying a deep learning technology. In some preferred embodiments, intelligent accurate recording of radar points can be realized under the condition of low CFAR threshold.
Description of the drawings:
FIG. 1 is a flow chart of the dot blots in an embodiment of the present invention;
FIG. 2 is a flow chart of a predictive model training process in an embodiment of the invention;
FIG. 3 is a flow chart of data screening according to an embodiment of the present invention;
FIG. 4 is a schematic representation of track data characteristics according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of a track tag in an embodiment of the invention;
FIG. 6 is a schematic diagram of a dynamic track tag manufacturing process according to an embodiment of the present invention;
FIG. 7 is a flow chart of a model training process in an embodiment of the present invention;
FIG. 8 is a diagram illustrating initial parameters of an LSTM model according to an embodiment of the present invention;
FIG. 9 is a flowchart for building a computation graph according to an embodiment of the present invention;
fig. 10 is a schematic diagram of a dot blot in an embodiment of the invention.
Detailed Description
The present invention will be described in further detail with reference to test examples and specific embodiments. It should be understood that the scope of the above-described subject matter is not limited to the following examples, and any techniques implemented based on the disclosure of the present invention are within the scope of the present invention.
In the radar signal processing stage, a lower CFAR threshold is set or reduced to be 3 or 2 or lower, for example, so that a large number of point track signals enter a data processing stage, a temporary track is established by using a logic starting method in the stage, temporary track data in N periods are sent to a track accurate prediction model to generate predicted next-time track data, and after a next radar scanning period comes, the predicted track data and real point track data are compared through a discrimination algorithm to select reasonable point track data, and accurate point track recording is realized.
As shown in fig. 1, firstly, using a logic method to temporarily build a navigation for the trace point data of a radar scanning period, it is first necessary to wait N (for example, 3 or more) radar scanning periods, where the scanning period is set to M, that is, every M seconds (for example, 1 second or more) radar rotates one circle (N and M are natural numbers).
Then, after the track initiation is completed, a set of track data sequence _ dot is formed, and the set of sequence _ dot is respectively input into a track prediction model, and the model is realized by using a function _ prediction as a carrier. The function _ prediction is a model obtained through training; at this point the function _ predict will return the extrapolated trace-predict _ dot for that track. The track prediction model comprises a track prediction model trained by a Long Short-Term Memory network (LSTM) in a Recurrent Neural Network (RNN).
And finally, performing correlation matching on the extrapolated trace predicate _ dot and real-time trace data real _ time _ dot which arrives at the next scanning period, and selecting the distance measure as a matching dimension.
In another embodiment of the present invention, the above-mentioned track prediction model is preferably implemented by the following steps:
the prediction model training process based on the LSTM for the flight path prediction model comprises four steps, as shown in FIG. 2, data screening, feature selection, labeling and model training. The method comprises the following specific steps:
A. data screening
The trace point data refers to a group of data which is output by the data recording equipment and contains parameters such as position coordinates of echo points and the like after meeting the signal detection criterion. The track data is a curve formed by continuous point tracks of the same target. The training data is selected as flight path data, and the batch number in the flight path data is an integer, so that the minimum available batch number principle is adopted to recover when the radar runs for a long time, and the situation that different targets are in the same batch number exists in the flight path data.
Through statistics, the difference values of the time, the distance, the pitch, the direction, the signal intensity and other attribute values between the adjacent track points of the same batch of numbers and the same target are all within a certain threshold range. Therefore, a threshold vector H is respectively determined as a threshold of each attribute through screening and statistics of the track data, and when the threshold vector H is smaller than the vector H, the same target is determined, otherwise, the different targets are determined. The data screening process is shown in FIG. 3.
B. Feature selection
The track data includes attributes such as time, lot number, distance, bearing, pitch, and intensity, pitch intensity, bearing intensity, channel number, wave position number, time quality, pitch intensity, bearing phase difference, noise intensity, CFAR threshold, etc. The most representative time, distance, orientation, pitch, and intensity are selected as relevant features of the prediction model, as shown in fig. 4.
C. Label labeling
Description of basic concept of label labeling:
(1) formalized description of flight path
sequence _ dot = { dot _1, dot _2, dot _3, dot _4, … dot _ i }, (i =1,2,3 … w) (formula one)
Wherein dot _ i is trace point information, i is a natural number, and a sequence formed by the trace point information is called a flight path.
(2) Data tag structure
dat _ label = < data, label > (formula two)
data represents data required to participate in training, label represents a label of the training data, and dat _ label represents a data label association structure.
And (3) track marking process:
and (4) determining the label by selecting a sliding window mode, wherein the window width n is the step length of the window is 1. An exemplary process of track data and tag is shown in FIG. 5 below.
(1) Initial track marking
Wherein data _ label = < sequence _ dot (1, n), dot _ n +1>, which means that a flight path sequence with the length of n participates in training, and a label is n +1 th point path data.
(2) Track sliding label
The dynamic labeling process of the track label is shown in fig. 6, where the window length is n, the sequence length is L (L > 3), when L > n, the window slides forward by L steps, sequence _ dot becomes sequence _ dot (2, n + L), label also slides forward to become dot _ n +2, a new data _ label pair is formed, and when L < n and L >3, the insufficient part is filled with null occupation. Wherein L and n are natural numbers.
D. Model training
The model training process comprises three steps of model building, data training, model parameter tuning and the like, and is shown in the following figure 7.
The initial parameters set up by the LSTM model include 5 parameters such as an input data array, each batch of data specification, an input data sequence length, the number of hidden layers, and a dimension of each element in an input sequence, as shown in fig. 8. After the initial parameters are determined, the LSTM model building is completed by calling a library function in the TenSoFlow.
After the LSTM model is built, a calculation graph is continuously built, the building process of the calculation graph comprises 6 steps of determining an input parameter format, a label format and an example LSTM model, building a full connection layer, outputting prediction, automatically adjusting parameters and the like, and the code schematic and the flow are shown in FIG. 9.
The two steps comprise the building of the LSTM training model and the completion process of the case prediction. And matching the input training data and the labels to the prediction model, and completing the flight path prediction model training after K-round iterative computation.
After model training is completed, track-point recording can be performed, track-point data are accumulated through M rounds of radar scanning periods, track data are generated by using a logic method and are sent to a track prediction model for prediction, and predicted track data at the next moment are generated. As shown in fig. 10, because the track point data is a point in a three-dimensional space, according to a distance measure, an appropriate H _ Radius value is selected as a Radius (for example, 2 meters or 1 meter, or a longer or shorter distance is selected according to specific situations), a threshold space of a Sphere is formed in the three-dimensional space, and the track point data at the Next moment is recorded if the track point data is within a Sphere, for example, a point Y _ Next, and if the track point data is not within the Sphere, the recording is abandoned if the track point data is not within the Sphere, for example, a point Y _ Next _1, and a point Y _ Next _2, and after a plurality of iterations, the accurate recording of all the track points is completed.
Yet another embodiment of the present invention provides a computer-readable storage medium storing instructions that, when executed, cause a computer to perform a method as described in the above-described embodiments of the present invention.
It should be understood that the Processor involved in the present invention may be a Central Processing Unit (CPU), other general purpose processors, Digital Signal Processors (DSP), Application Specific Integrated Circuit (ASIC), etc. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of a method disclosed in connection with the embodiments of the present application may be directly implemented by a hardware processor, or may be implemented by a combination of hardware and software modules in a processor.
Those of ordinary skill in the art will understand that: all or a portion of the steps of implementing the above-described method embodiments may be performed by hardware associated with program instructions. The program may be stored in a computer-readable storage medium. When executed, the program performs steps comprising the method embodiments described above; and the aforementioned storage medium includes: read-only memory (ROM), RAM, flash memory, hard disk, solid state disk, magnetic tape (magnetic tape), floppy disk (optical disk), and any combination thereof.
Claims (4)
1. A radar target point recording method comprises the following steps:
reading radar data of N periods, wherein N is a natural number, and processing the radar data of the N periods by using a logic starting method to establish temporary track data;
inputting the temporary track data into a track prediction model to generate predicted track data, wherein the track prediction model comprises a track prediction model trained by a long-short term memory network in a recurrent neural network;
comparing the predicted track data with real track point data in a radar scanning period where the predicted track data is located, selecting track point data with a distance between the predicted track data and the real track point meeting a preset condition, and realizing track point recording;
when the radar data of N periods are read, setting the threshold of the constant false alarm rate value to be less than or equal to 3;
the flight path prediction model training comprises data screening, feature selection, label marking and model training, wherein the data screening comprises the following steps: firstly, extracting track attribute data, wherein the attribute data comprises time, distance, pitch, azimuth and signal intensity, and selecting a track attribute data value smaller than a preset threshold vector as same-target data;
selecting time, distance, azimuth, elevation and radar signal echo intensity in the track attribute data as relevant characteristics of a track prediction model;
and determining a label in a sliding window mode in the label labeling, wherein the width n of the window is n, the step length of the window is 1, the sequence length is L, when L is greater than n, the window slides forwards in 1 step length, when L is less than n, the insufficient part is occupied and filled by null values, wherein n is a natural number, and L is a natural number greater than 3.
2. The radar target spot admission method of claim 1, characterized by: and selecting the point trace data of which the distance between the predicted flight path data and the real point trace meets the preset condition, wherein the distance of the preset condition is less than or equal to 2 meters.
3. An electronic device, characterized by comprising: a processor, a memory, and a computer program; wherein the computer program is stored in the memory and configured to be executed by the processor, the computer program comprising instructions for performing the radar target spot-recording method of one of claims 1 to 2.
4. A computer-readable storage medium storing a computer program which, when executed, implements the radar target spot logging method of one of claims 1 to 2.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911076994.4A CN110940971B (en) | 2019-11-06 | 2019-11-06 | Radar target point trace recording method and device and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911076994.4A CN110940971B (en) | 2019-11-06 | 2019-11-06 | Radar target point trace recording method and device and storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110940971A CN110940971A (en) | 2020-03-31 |
CN110940971B true CN110940971B (en) | 2022-05-31 |
Family
ID=69906637
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911076994.4A Active CN110940971B (en) | 2019-11-06 | 2019-11-06 | Radar target point trace recording method and device and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110940971B (en) |
Families Citing this family (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP7463820B2 (en) | 2020-04-14 | 2024-04-09 | 日本電気株式会社 | Target tracking device, target tracking system, target tracking method, and target tracking program |
CN112816982A (en) * | 2020-12-31 | 2021-05-18 | 中国电子科技集团公司第十四研究所 | Radar target detection method |
CN113109796A (en) * | 2021-03-03 | 2021-07-13 | 四川九洲防控科技有限责任公司 | Target detection method and device |
CN113325413B (en) * | 2021-04-25 | 2022-11-22 | 清华大学 | Radar target locking tracking method and device |
CN113486960A (en) * | 2021-07-12 | 2021-10-08 | 安徽耀峰雷达科技有限公司 | Unmanned aerial vehicle tracking method and device based on long-time memory neural network, storage medium and computer equipment |
CN114740465A (en) * | 2022-03-18 | 2022-07-12 | 四川九洲防控科技有限责任公司 | Radar track rapid batching method and device, storage medium and electronic equipment |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH0943346A (en) * | 1995-07-31 | 1997-02-14 | Nec Corp | Tracking processor |
CN101183494A (en) * | 2007-12-20 | 2008-05-21 | 四川川大智胜软件股份有限公司 | Air Traffic Control radar data automatic identification method |
CN107219504A (en) * | 2016-08-10 | 2017-09-29 | 深圳市亿威尔信息技术股份有限公司 | The date collection system of radar system |
Family Cites Families (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2013190398A (en) * | 2012-03-15 | 2013-09-26 | Toshiba Corp | Radar display processor, radar display processing method, and radar display device |
CN106249232B (en) * | 2016-08-24 | 2018-08-21 | 中国电子科技集团公司第二十八研究所 | Method for tracking target based on target state of motion information data associating policy |
US10989791B2 (en) * | 2016-12-05 | 2021-04-27 | Trackman A/S | Device, system, and method for tracking an object using radar data and imager data |
CN107679557B (en) * | 2017-09-19 | 2020-11-27 | 平安科技(深圳)有限公司 | Driving model training method, driver identification method, device, equipment and medium |
CN108254741B (en) * | 2018-01-16 | 2021-02-09 | 中国人民解放军海军航空大学 | Target track prediction method based on cyclic neural network |
CN109543715A (en) * | 2018-10-23 | 2019-03-29 | 武汉理工大学 | A kind of ship air route is extracted and the method for track deviation detection |
CN109655822A (en) * | 2018-11-09 | 2019-04-19 | 上海无线电设备研究所 | A kind of improved track initiation method |
CN109740742A (en) * | 2019-01-14 | 2019-05-10 | 哈尔滨工程大学 | A kind of method for tracking target based on LSTM neural network |
CN110187318B (en) * | 2019-04-23 | 2021-07-06 | 四川九洲防控科技有限责任公司 | Radar data processing method |
CN110133610B (en) * | 2019-05-14 | 2020-12-15 | 浙江大学 | Ultra-wideband radar action identification method based on time-varying distance-Doppler diagram |
CN110297222B (en) * | 2019-06-20 | 2021-04-06 | 西安电子科技大学 | Track starting method based on multi-dimensional measurement information |
-
2019
- 2019-11-06 CN CN201911076994.4A patent/CN110940971B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH0943346A (en) * | 1995-07-31 | 1997-02-14 | Nec Corp | Tracking processor |
CN101183494A (en) * | 2007-12-20 | 2008-05-21 | 四川川大智胜软件股份有限公司 | Air Traffic Control radar data automatic identification method |
CN107219504A (en) * | 2016-08-10 | 2017-09-29 | 深圳市亿威尔信息技术股份有限公司 | The date collection system of radar system |
Non-Patent Citations (2)
Title |
---|
Arindam Sengupta ; Feng Jin ; Siyang Cao.《A DNN-LSTM based Target Tracking Approach using mmWave Radar and Camera Sensor Fusion》.《2019 IEEE National Aerospace and Electronics Conference (NAECON)》.2019, * |
权波 ; 杨博辰.《基于LSTM的船舶航迹预测模型》.《计算机科学》.2018, * |
Also Published As
Publication number | Publication date |
---|---|
CN110940971A (en) | 2020-03-31 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110940971B (en) | Radar target point trace recording method and device and storage medium | |
CN113866742B (en) | Method for point cloud processing and target classification of 4D millimeter wave radar | |
CN109087510B (en) | Traffic monitoring method and device | |
CN101794437B (en) | Method for detecting abnormal target in hyperspectral remotely sensed image | |
CN109188430B (en) | Target extraction method based on ground monitoring radar system | |
CN106468771A (en) | A kind of multi-target detection and tracking method under the high clutter conditions of low Observable | |
CN105787950A (en) | Infrared image sea-sky-line detection algorithm based on line gradient accumulation | |
Xu et al. | Fast ship detection combining visual saliency and a cascade CNN in SAR images | |
CN115761534A (en) | Method for detecting and tracking small target of infrared unmanned aerial vehicle under air background | |
CN108414992B (en) | Target detection method based on phase information clutter map | |
CN115546705A (en) | Target identification method, terminal device and storage medium | |
CN112802108A (en) | Target object positioning method and device, electronic equipment and readable storage medium | |
CN108983194B (en) | Target extraction and condensation method based on ground monitoring radar system | |
CN117075097B (en) | Maritime radar target tracking method and system based on expanded target cluster division | |
CN110515044A (en) | Low slow small radar null tone domain four-dimension clutter suppression method | |
CN115600101B (en) | Priori knowledge-based unmanned aerial vehicle signal intelligent detection method and apparatus | |
CN115220002B (en) | Multi-target data association tracking method and related device for fixed single station | |
CN114415123B (en) | Non-coherent neighborhood based weighting pulse accumulation processing method and system | |
Gao et al. | Intelligent multiframe detection aided by Doppler information and a deep neural network | |
CN115980689A (en) | Point cloud detection-based radiation source signal sorting method, device, equipment and medium | |
CN110502979B (en) | Laser radar waveform signal classification method based on decision tree | |
Lei et al. | Multi-feature fusion sonar image target detection evaluation based on particle swarm optimization algorithm | |
CN113625266A (en) | Method, device, storage medium and equipment for detecting low-speed target by using radar | |
CN116263735A (en) | Robustness assessment method, device, equipment and storage medium for neural network | |
CN112731388A (en) | Target detection method based on effective scattering point energy accumulation |
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 |