Sea radar track start judgment method and system, control server and storage medium
Feature extraction and judgment of radar echo videos through multi-channel convolutional neural networks, solving the problem of false tracks of traditional radars in sea clutter environments, and achieving the effectiveness and quality improvement of the starting judgment of sea radar tracks .
Patent Information
- Application Number
- CN202410983548.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-22
- Publication Date
- 2024-11-22
AI Technical Summary
Traditional radar methods for sea target track start are prone to false tracks and discontinuities in sea clutter and cloud and rain clutter environments, resulting in erroneous associations and a large number of false tracks.
The track start judgment method based on multi-channel convolutional neural network is adopted, and the radar echo video is preprocessed and feature extraction is performed to construct candidate track data sets, and the multi-channel convolutional neural network is used to train the judge to achieve the Valid judgment on the sea radar track.
Significantly reduce the generation of false tracks, improve the starting quality of tracks, especially in high-density clutter environments, and reduce subsequent network processing pressure.
Smart Images

Figure CN119001645A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the field of radar data processing, and in particular relates to a method, a system, a control server and a storage medium for determining the start of a sea radar track. Background Art
[0002] As a long-range detection sensor, radar can monitor sea areas over a large area and plays an important role in areas such as maritime target detection and tracking.
[0003] Traditional radar sea target track initiation methods are mainly based on statistical models, such as logic method and multiple hypothesis method. These methods extract and utilize fewer radar target features, mainly processed point information. In sea detection application scenarios, due to the large number of sea clutter and cloud and rain clutter points, traditional methods are prone to false associations, resulting in a large number of false tracks and discontinuous track states. Summary of the invention
[0004] The present invention proposes a method for determining the start of a sea radar track, which can effectively suppress the generation of false radar tracks in a sea clutter scenario.
[0005] The technical solution to achieve the purpose of the present invention is: a method for determining the start of a sea radar track, comprising the following steps:
[0006] Step 1: Preprocess the sea detection radar echo and point tracks to obtain candidate tracks and their corresponding point tracks;
[0007] Step 2: Based on the candidate tracks and their points, obtain the radar original echo video corresponding to the multiple frames of the candidate tracks, build a candidate track dataset for sea radar, extract samples from the dataset and divide it into training set and test set according to the specific usage scenarios and requirements;
[0008] Step 3: Construct a classification network based on a multi-channel convolutional neural network, and use the multi-channel convolutional classification network to train the radar candidate track samples in the training set to obtain a multi-channel convolutional decision device;
[0009] Step 4: Use the test set samples to test the multi-channel convolution decision device obtained in step 3 and record the test indicators;
[0010] Step 5: Check whether the number of iterations has reached the maximum value. If not, adjust the parameters and jump to step 3. If it has reached the maximum value, go to step 6.
[0011] Step 6: Based on the recorded test indicators, output the decision maker with the best performance indicators in all iterations;
[0012] Step 7: Obtain the original echo video images of the set number of frames corresponding to the candidate track in real time, use the decision device with the best index to make a decision, and obtain the decision result of the start of the sea radar track.
[0013] The present invention also proposes a sea radar track initiation judgment system based on a multi-channel convolutional neural network, comprising:
[0014] A preprocessing module is used to preprocess the echo and trace of the sea detection radar;
[0015] The data set generation module is used to obtain the radar original echo video corresponding to the multiple frames of candidate track points according to the candidate track and its point track, build the candidate track data set of the sea radar, extract samples from the data set and divide it into training set and test set according to the specific usage scenarios and requirements;
[0016] The classification network training and testing module uses a multi-channel convolutional classification network to train the radar candidate track samples in the training set to obtain the optimal decision device;
[0017] The real-time judgment module is used to make a judgment using a judge with the best index to obtain the judgment result of the start of the sea radar track.
[0018] Compared with the prior art, the present invention has the following significant advantages: the present invention uses a convolutional neural network to extract complete information from radar echo images, and efficiently captures effective correlation features between candidate track echo frames through a multi-channel structure, providing support for suppressing false tracks in sea detection clutter scenarios. The present invention obtains candidate tracks based on a logic method, which can reduce the generation of false tracks, efficiently extracts potential target point track information through knowledge-driven, reduces subsequent network processing pressure, improves the quality of track initiation, and improves the quality of track initiation, especially in radar high-density clutter environments, where the track initiation effect is better. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] Figure 1 It is a schematic diagram of the processing flow of the present invention.
[0020] Figure 2 It is a schematic diagram of the multi-channel convolution structure in the present invention. DETAILED DESCRIPTION
[0021] As an embodiment, a method for determining the start of a sea radar track is shown in the processing flow chart. Figure 1 As shown, in combination with the embodiments, the specific implementation is as follows:
[0022] Step 1: Preprocess the sea detection radar echo and point traces, and obtain candidate tracks and their corresponding point traces based on the logic method. Assume that the maximum speed of the target is v max, the radar period is T, and a circular initial correlation gate is established with the free point track as the track head, and its radius is R = v max T, select the next scan point track that falls into the initial correlation gate and is closest to the track head to establish a possible track. Extrapolate each possible track, take the extrapolated point as the center, and use the track extrapolation error covariance matrix to establish a subsequent correlation gate. Select the point track closest to the extrapolated point in the correlation gate as the associated point track, so as to sequentially update the track sequence. If there are more than 5 successful associated points within 7 radar scan times, a candidate track is generated. This step is based on the logic method to obtain candidate tracks, which mainly involves the processing of multiple frames of continuous original observation data of the radar, classifying the data through a hierarchical clustering algorithm, and then screening to obtain candidate tracks. The logical method is used to obtain candidate tracks. The track initiation effect is better in the high-density clutter environment of the radar, which can reduce the generation of false tracks. The potential target point track information is efficiently extracted through knowledge-driven, which reduces the subsequent network processing pressure and improves the track initiation quality.
[0023] Step 2: Based on the candidate tracks and their points, obtain the radar original echo video corresponding to multiple frames of candidate tracks, construct a candidate track dataset for sea radar, and extract samples from the dataset and divide them into training and test sets according to specific usage scenarios and requirements. When constructing a candidate track dataset for sea radar, the candidate track data is constructed by superimposing the original echoes corresponding to the 7 frames of point tracks of the candidate track, and obtaining the minimum interception rectangle containing the 7 frames of point tracks as the data interception range. Each frame of data is the echo corresponding to the candidate track of the current frame within the interception range. Remove other echoes of the current frame within the interception range, and 7 frames of data constitute a candidate track training or test data.
[0024] Step 3: Construct a classification network based on a multi-channel convolutional neural network. Figure 2 As shown in the figure, a classification network based on a multi-channel convolutional neural network is used to train the radar candidate track samples in the training set. The classification results are obtained after the data passes through the convolution layer and the fully connected layer of the network. The convolution layer is a multi-channel convolution, which includes calculation processes such as convolution, activation function and pooling; the fully connected layer includes calculation processes such as linear multiplication and addition and activation function. The convolution network uses the ReLU activation function
[0025] ReLU(a)=max(0,a)
[0026] Step 4: Use the test set samples to test the multi-channel convolutional decision maker obtained through training and record the test indicators.
[0027] Step 5: Check whether the number of iterations has reached the maximum value. If not, adjust the network parameters and jump to step 3. If it has reached the maximum value, go to step 6.
[0028] Step 6: According to the recorded test indicators, output the decision maker with the best performance indicators in all iterations. The decision maker includes a multi-channel convolution classification network and corresponding network parameters.
[0029] Step 7: Obtain 7 frames of original echo video images corresponding to the candidate track in real time, use the decision device with the best index to make a decision, and obtain the decision result of the start of the sea radar track.
[0030] The traditional radar sea target track initiation method extracts and utilizes fewer radar target features, mainly the processed point track information. In the sea detection application scenario, due to the influence of sea clutter and cloud and rain clutter, it is easy to produce false associations, resulting in a large number of false tracks or discontinuous tracks. The present invention uses a convolutional neural network to extract complete information from the radar echo image, and through a multi-channel structure, efficiently captures the data-related features between candidate track echo frames, which can improve the radar false track suppression capability in sea clutter scenarios.
[0031] As an embodiment, an adaptive radar display enhancement system includes:
[0032] A radar data preprocessing module, used for preprocessing radar data;
[0033] An image pyramid construction module is used to construct an image pyramid based on the preprocessed radar data;
[0034] A layer selection coefficient determination module is used to generate a layer selection coefficient according to a display distance and a data viewing range;
[0035] The display layer switching module is used to adaptively switch the image pyramid display layer according to the layer selection coefficient.
[0036] As an embodiment, a control server is presented in the form of a general-purpose computing device, including but not limited to: one or more processors or processing units, a system memory, and a bus connecting the system memory and the processing unit.
[0037] The term "bus" refers to one or more of several types of bus structures, including a memory bus or a memory controller peripheral bus, an accelerated graphics port, a processor, or a local bus using any of a variety of bus architectures. Examples of these architectures include, but are not limited to, an Industry Standard Architecture (ISA) bus, a Micro Channel Architecture (MAC) bus, an Enhanced ISA bus, a Video Electronics Standards Association (VESA) local bus, and a Peripheral Component Interconnect (PCI) bus.
[0038] The control server includes a variety of computer system readable media. These media can be any available media that can be accessed by the control server, including volatile and non-volatile media, removable and non-removable media.
[0039] The system memory may include computer system readable media in the form of volatile memory, such as random access memory (RAM) and / or cache memory. The control server may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, the storage system may be used to read and write non-removable, non-volatile magnetic media (commonly referred to as "hard drives"). A disk drive for reading and writing removable non-volatile disks (such as "floppy disks"), and an optical disk drive for reading and writing removable non-volatile optical disks (such as CD-ROMs, DVD-ROMs or other optical media) may be provided. In these cases, each drive may be connected to the bus via one or more data medium interfaces. The system memory may include at least one program product having a set (e.g., at least one) of program modules that are configured to perform the functions of various embodiments of the present invention.
[0040] A program / utility having a set (at least one) of program modules may be stored, for example, in a system memory, such program modules including but not limited to an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment. The program modules generally perform the functions and / or methods of the embodiments described herein.
[0041] The control server may also communicate with one or more external devices, such as keyboards, pointing devices, displays, etc., and may also communicate with one or more devices that enable users to interact with the device, and / or communicate with any device (such as a network card, a modem, etc.) that enables the control server to communicate with one or more other computing devices. Such communication may be performed through an input / output (I / O) interface. In addition, the server may also communicate with one or more networks (such as a local area network (LAN), a wide area network (WAN) and / or a public network, such as the Internet) through a network adapter. The network adapter communicates with other modules of the control server through a bus. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the control server, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems. The processing unit executes various functional applications and data processing by running programs stored in the system memory, such as implementing a method for determining the start of a sea radar track provided in an embodiment of the present invention.
[0042] As an embodiment, a computer-readable storage medium stores a computer program, which, when executed by a processor, can implement any of the sea radar track initiation determination methods in the above embodiments.
[0043] The computer storage medium of the embodiment of the present invention can adopt any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, but not limited to: an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only storage (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium can be any tangible medium containing or storing a program, which can be used by an instruction execution system, device or device or used in combination with it.
[0044] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.
[0045] Program code embodied on a computer-readable medium may be transmitted using any appropriate medium, including but not limited to:
[0046] Wireless, wire, optical cable, RF, etc., or any suitable combination of the above.
[0047] Computer program code for performing the operations of the present invention may be written in one or more programming languages or a combination thereof, including object-oriented programming languages such as Java, Smalltalk, C++, and conventional procedural programming languages such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).
Claims
1. A method for determining the start of a sea radar track, characterized in that: The steps include: Step 1: Preprocess the sea detection radar echo and point tracks to obtain candidate tracks and their corresponding point tracks; Step 2: Based on the candidate tracks and their points, obtain the radar original echo video corresponding to multiple frames of candidate track points, build a candidate track dataset for sea-surface radar, extract samples from the dataset and divide it into training and test sets according to specific usage scenarios and requirements; Step 3: Construct a classification network based on a multi-channel convolutional neural network. Use the multi-channel convolutional classification network to train the radar candidate track samples in the training set to obtain a multi-channel convolutional decision maker. Step 4: Use the test set samples to test the multi-channel convolution decision maker obtained in step 3 and record the test indicators; Step 5: Check whether the number of iterations has reached the maximum value. If not, adjust the parameters and jump to step 3. If it has reached the maximum value, go to step 6. Step 6: Based on the recorded test indicators, output the decision maker with the best performance indicators in all iterations; Step 7: Obtain the set number of original echo video images corresponding to the candidate track in real time, use the decision device with the best index to make a decision, and obtain the starting decision result of the sea radar track.
2. The method for determining the start of a sea radar track according to claim 1, wherein: The specific method for preprocessing the sea detection radar echo and point tracks to obtain candidate tracks and their corresponding point tracks is as follows: A circular initial correlation gate is established with the free point track as the track head; Select the next scan point that falls into the initial correlation gate and is closest to the track head to establish a possible track; Extrapolate each possible track, take the extrapolation point as the center, use the track extrapolation error covariance matrix to establish the correlation gate, select the track closest to the extrapolation point in the correlation gate as the associated track, and sequentially update the track sequence; If the number of successfully associated tracks within the set radar scanning time reaches the set threshold, a candidate track is generated.
3. The method for determining the start of a sea radar track according to claim 2, wherein: The radius of the circular initial correlation gate is: R=v max T Where T is the radar period, v max is the target maximum speed.
4. The method for determining the start of a sea radar track according to claim 2, wherein: If the number of successful associated tracks reaches 5 within 7 radar scan times, a candidate track is generated.
5. The method for determining the start of a sea radar track according to claim 1, wherein: The method for constructing candidate track data in the sea radar candidate track dataset is as follows: The original echoes corresponding to the 7 frames of candidate tracks are superimposed together, and the minimum interception rectangle containing the 7 frames of tracks is obtained as the data interception range; Each frame of data is the echo corresponding to the candidate track of the current frame within the interception range. After removing other echoes of the current frame within the interception range, 7 frames of data constitute a candidate track training or testing data.
6. The method for determining the start of a sea radar track according to claim 1, wherein: The classification network based on the multi-channel convolutional neural network includes a convolution layer and a fully connected layer. The convolution layer is a multi-channel convolution, which includes convolution, activation function and pooling calculation process.
7. The method for determining the start of a sea radar track according to claim 1, wherein: The decision maker includes a multi-channel convolutional classification network and corresponding network parameters.
8. A multi-channel convolutional neural network-based sea radar track initiation judgment system according to any one of claims 1 to 7, characterized in that: include: Preprocessing module, used to preprocess the echo and trace of sea detection radar; The dataset generation module is used to obtain the radar raw echo video corresponding to multiple frames of candidate track points based on the candidate track points and their points, and construct a candidate track dataset for the sea radar. Based on the specific usage scenario and requirements, samples in the dataset are extracted and divided into training and test sets. The classification network training and testing module uses a multi-channel convolutional classification network to train radar candidate track samples in the training set to obtain the optimal decision maker; The real-time judgment module is used to make a judgment using a judge with the best index to obtain the judgment result of the starting point of the sea radar track.
9. A control server, characterized in that: include: one or more processors; a storage device for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the method for determining the start of a sea radar track as described in any one of claims 1 to 7.
10. A storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method for determining the start of a sea radar track as claimed in any one of claims 1 to 7 is implemented.
