A method and system for producing a time-frequency domain detection dataset for high-frequency radar ship targets

Abstract

The present invention discloses a method and system for producing a time-frequency domain detection dataset for high-frequency radar ship targets. The method comprises the following steps: obtaining the start and end times, radar position, and range-slow time spectrum of the field data based on high-frequency radar reception data; calculating the range-Doppler spectrum and the time-frequency spectrum of each range element based on the range-slow time spectrum; and calculating the first-order spectrum and the Doppler range of ground clutter based on the range-Doppler spectrum; obtaining ship AIS information based on the start and end times of the high-frequency radar data, and calculating the radial range and radial velocity of the AIS ship target to obtain the target range element and the target Doppler frequency, thereby establishing an AIS target database; generating a pre-selected annotation box centered on the target Doppler frequency in the corresponding range element time-frequency spectrum based on the AIS ship target database information, and saving a target annotation file; and completing target annotation through manual review based on the target time-frequency spectrum and the target annotation file, thereby producing a time-frequency domain detection dataset for high-frequency radar ship targets.

Description

Technical Field

[0001] The present invention relates to the field of high-frequency radar ship target time-frequency domain detection, and in particular to a method and system for preparing a high-frequency radar ship target time-frequency domain detection data set. Background Art

[0002] High-frequency radar, with its advantages of 24 / 7, real-time, all-weather operation, and over-the-horizon monitoring range, has become a key technology for observing ship dynamics in maritime areas. However, high-frequency radar typically uses long coherent integration times for detection. However, the scattering cross-section of echoes from ship targets fluctuates. Furthermore, due to the influence of the ocean environment, it is difficult for ships to maintain a constant speed during navigation, resulting in non-stationary echoes from ship targets.

[0003] The non-stationary nature of ship target echoes causes energy diffusion in the range-Doppler spectrum. Combined with the dense interference and noise of high-frequency signals, constant false alarm (CFAR) detectors (CFARs) perform poorly in this scenario. Given the large range bins of high-frequency radars, a target typically travels no more than one bin within a single radar data frame. Consequently, target echoes exhibit continuous or discontinuous ridges within the time-frequency spectrum of a single bin. Consequently, many researchers, both domestically and internationally, have attempted to improve the performance of high-frequency radar non-stationary target detection by transforming the detection task from the range-Doppler domain to the time-frequency domain. Currently, time-frequency domain target detection in high-frequency radars primarily relies on graphics and image processing techniques, transforming the target detection task into an image edge extraction problem. However, such methods require extensive parameter setting based on manual experience.

[0004] In recent years, deep learning-based target networks have demonstrated excellent performance in image object detection, providing new insights into the task of detecting ridges in target time-frequency spectra. It's important to note that deep learning networks are data-driven, end-to-end target detection methods, and to ensure optimal detection performance, they require a large amount of high-quality annotated data. Because high-frequency radars cannot image targets, labeling target ridges requires both ensuring that the time-frequency ridges belong to ships and requiring extensive manual effort and time. Summary of the Invention

[0005] To overcome the shortcomings of the above-mentioned existing technologies, the present invention provides a method and system for producing a high-frequency radar ship target detection dataset in the time-frequency domain. The method uses AIS information to generate pre-selected annotation boxes in the high-frequency radar time-frequency domain, providing a reference for constructing a target detection dataset in the time-frequency domain, supporting the application of deep neural network models in high-frequency radar target detection, and thus improving the detection performance of high-frequency radar systems for ship targets.

[0006] According to one aspect of the present invention, a method for preparing a high-frequency radar ship target time-frequency domain detection dataset is provided, comprising:

[0007] According to the high-frequency radar receiving data, the start and end time of the high-frequency radar data, the radar position, and the distance-slow time spectrum are obtained;

[0008] According to the distance-slow time spectrum, a time spectrum of each distance element is calculated;

[0009] According to the start and end time of the high-frequency radar data session, the ship AIS information within the current start and end time period is obtained, and the range element and Doppler element information of the ship target are calculated based on the ship AIS information and the radar position, and a ship AIS target database is established;

[0010] Reading the distance element and Doppler element information in the ship AIS target database, obtaining the time-frequency spectrum of the ship target according to the distance element, and generating a pre-selected annotation box centered on the Doppler frequency of the ship target according to the Doppler element of the ship target, to obtain a primary screening annotation file;

[0011] Reviewing the primary screening and marking file to form a secondary screening and marking file;

[0012] The suspected target mark box area that does not contain the pre-selected mark box in the time spectrum of the ship target is processed to form a new time spectrum, and the suspected target mark box that does not contain the pre-selected mark box in the secondary screening mark file is deleted to obtain a final screening mark file.

[0013] As a further technical solution, the method further includes: saving the final screened and labeled file and the corresponding time-frequency spectrum to a high-frequency radar ship target time-frequency domain detection data set.

[0014] As a further technical solution, after obtaining the distance-slow time spectrum, the method further includes:

[0015] Generate range-Doppler spectrum using fast Fourier transform;

[0016] According to the range-Doppler spectrum and in combination with the first-order Bragg scattering theorem of sea clutter, a signal-to-noise ratio threshold is set to calculate the Doppler range of the first-order spectrum of sea clutter;

[0017] According to the range-Doppler spectrum and in combination with the stationary characteristics of the ground clutter, the ground clutter Doppler range is obtained.

[0018] As a further technical solution, the range element and Doppler element information of the ship target are calculated based on the ship AIS information and the radar position, including:

[0019] Calculating a radial distance of the ship target according to the radar position and the ship target position, and calculating a range element of the ship target according to the radial distance;

[0020] The radial velocity of the ship target is calculated according to the radar position and the speed and heading of the ship target, and the Doppler element of the ship target is calculated according to the radial velocity.

[0021] As a further technical solution, the generation of the initial screening annotation file includes:

[0022] Read the distance element and Doppler element information of the ship target from the established ship AIS target database;

[0023] According to the distance element of the ship target, read the time spectrum corresponding to the distance element;

[0024] According to the Doppler element of the ship target, a pre-selected marking frame for ship target detection is generated, a category attribute of the pre-selected marking frame is set, and information of the pre-selected marking frame is saved to a primary screening marking file.

[0025] As a further technical solution, the initial screening and annotation files are reviewed, including:

[0026] Determine whether there is a ship target time-frequency ridge in the preselected marked frame of the time-frequency spectrum, and if so, adjust the size of the marked frame so that the ship target time-frequency ridge line falls entirely within the marked frame;

[0027] It is determined whether there is a suspected target time-frequency ridge in the time-frequency spectrum that does not contain a pre-selected labeling box. If so, a target labeling box is generated and category attributes are set.

[0028] As a further technical solution, after obtaining the secondary screening and marking file, it also includes:

[0029] Setting the area of ​​the time-frequency spectrum marked with a non-ship category to 0 to obtain a new time-frequency spectrum;

[0030] The annotation boxes whose annotation box categories are non-ship are deleted to obtain the final filtered annotation file.

[0031] According to one aspect of the present invention, a system for preparing a high-frequency radar ship target time-frequency domain detection dataset is provided, comprising:

[0032] The first calculation module is used to obtain the start and end time of the high-frequency radar data, the radar position, and the range-slow time spectrum based on the high-frequency radar received data;

[0033] A second calculation module is used to calculate the time-frequency spectrum of each range element according to the range-slow time spectrum;

[0034] The third calculation module is used to obtain the ship AIS information within the current start and end time period based on the start and end time of the high-frequency radar data session, and calculate the range element and Doppler element information of the ship target based on the ship AIS information and the radar position, and establish a ship AIS target database;

[0035] The first screening module is configured to read the distance element and Doppler element information in the ship AIS target database, obtain the time-frequency spectrum of the ship target based on the distance element, and generate a pre-selected annotation box centered on the Doppler frequency of the ship target based on the Doppler element of the ship target to obtain a primary screening annotation file;

[0036] A second screening module is used to review the primary screening and marking file to form a secondary screening and marking file;

[0037] The third screening module is used to process the suspected target marking box area that does not contain the pre-selected marking box in the time-frequency spectrum of the ship target to form a new time-frequency spectrum, and delete the suspected target marking box that does not contain the pre-selected marking box in the secondary screening marking file to obtain a final screening marking file.

[0038] According to one aspect of the present invention, there is provided an electronic device comprising: at least one processor, at least one memory and a communication interface; wherein the processor, memory and communication interface communicate with each other; the memory stores program instructions to be executed by the processor, and the processor calls the program instructions to execute the described method.

[0039] According to one aspect of the present disclosure, a non-transitory computer-readable storage medium is provided, wherein the non-transitory computer-readable storage medium stores computer instructions, wherein the computer instructions cause the computer to execute the method described above.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] The present invention first uses AIS information to generate preselected annotation boxes for labeling high-frequency radar ship time-frequency ridges. Then, combined with manual experience, the time-frequency ridges of ship targets are verified to construct a high-frequency radar time-frequency target detection dataset. The method designed in this invention is highly accurate. Because the time-frequency spectrum of high-frequency radar echoes contains a large amount of noise and interference, using AIS information as a reference can improve the accuracy of target time-frequency ridge labeling. It is also easy to operate, and the preselected annotation boxes generated based on AIS information can reduce the workload of manual labeling, providing strong support for the generation of high-quality high-frequency radar time-frequency target detection datasets. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, a brief introduction will be given below to the drawings used in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0043] Figure 1 A flowchart of a method for preparing a high-frequency radar ship target time-frequency domain detection dataset provided by an embodiment of the present invention.

[0044] Figure 2 A schematic structural diagram of a system for producing a high-frequency radar ship target time-frequency domain detection dataset provided by an embodiment of the present invention.

[0045] Figure 3 This is a schematic diagram of the initial screening provided by an embodiment of the present invention.

[0046] Figure 4 A schematic diagram of secondary screening provided by an embodiment of the present invention.

[0047] Figure 5 A schematic structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0048] It should be noted that:

[0049] The Automatic Identification System (AIS) provides information such as a ship's MMSI (Maritime Mobile Service Identity) code, location, speed, and heading. It is often used as a reference to test radar system target detection performance, providing a basis for labeling high-frequency radar target detection in the time and frequency domains. Based on this, the present invention designs a method for producing a high-frequency radar ship target time and frequency domain detection dataset using AIS information. This method reduces the labor and time costs of manual labeling while providing a labeling reference, facilitating target labeling in the time and frequency domains of high-frequency radars and providing strong support for the production of high-frequency radar ship target time and frequency domain detection datasets.

[0050] The terms "including" and "having" and any variations thereof in the description and claims of the present invention and the above-mentioned drawings are intended to cover non-exclusive inclusions, for example, a process, method, system, product or apparatus that includes a series of steps or units is not necessarily limited to the steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to these processes, methods, products or apparatuses.

[0051] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically separate entities. That is, these functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices. The flowcharts shown in the accompanying drawings are merely illustrative and do not necessarily include all content and operations / steps, nor do they necessarily need to be executed in the order described. For example, some operations / steps may be further decomposed, while others may be combined or partially combined, so the actual execution order may vary depending on the actual situation.

[0052] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. In addition, the technical features in the various embodiments or single embodiments provided by the present invention are arbitrarily combined with each other to form a new technical solution. This combination is not restricted by the sequence of steps and / or structural composition mode, but must be based on the ability of ordinary technicians in this field to implement it. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that this combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0053] The embodiment of the present invention provides a method for preparing a high-frequency radar ship target time-frequency domain detection data set. Figure 1 As shown, including:

[0054] Step 1: Based on the high-frequency radar received data, obtain the start and end time of the high-frequency radar data, radar position, and distance-slow time spectrum.

[0055] Preferably, the distance-slow time spectrum is defined as , whose dimension is Dimension, where is the maximum distance element to be observed, is the number of sweep cycles, in the embodiment , .

[0056] Step 2: Calculate the time-frequency spectrum of each range element according to the range-slow time spectrum.

[0057] Preferably, based on the range-slow-time spectrum obtained in step 1, a time-frequency spectrum of each range element is first generated using a multiple synchronous squeezing algorithm; secondly, a range-Doppler spectrum is generated using a fast Fourier transform; thirdly, based on the generated range-Doppler spectrum, a signal-to-noise ratio threshold is set according to the first-order Bragg scattering theorem of sea clutter to calculate the Doppler range of the first-order spectrum region of sea clutter; finally, based on the stationary characteristics of the ground clutter, the ground clutter Doppler range is obtained.

[0058] Furthermore, after obtaining the first-order spectrum region of the sea clutter and the Doppler range of the ground clutter, the first-order spectrum region of the sea clutter and the ground clutter region of the time-frequency spectrum are set to 0.

[0059] In the method described in the embodiment of the present invention, the time-frequency spectrum is defined as ,in yes dimension, is the number of sampling points in the time dimension of the time spectrum, is the number of sampling points of the time-spectral Doppler frequency dimension, Represents the observation distance element number, , N is the maximum distance element to be observed. In the embodiment, .

[0060] Preferably, the signal-to-noise ratio threshold in step 2 is positioned as .

[0061] Step 3: Obtain the ship AIS information within the current start and end time period based on the start and end time of the high-frequency radar data session, and calculate the range element and Doppler element information of the ship target based on the ship AIS information combined with the radar position to establish a ship AIS target database.

[0062] Preferably, according to the start and end time of the high-frequency radar data session obtained in step 1, the AIS data within the start and end time period is read, including the ship's MMSI, speed, position, heading and other information.

[0063] Furthermore, the establishment of the ship AIS target database provided by the embodiment of the present invention includes:

[0064] Calculate the radial distance of the ship target based on the radar position obtained in step 1 and the ship target position obtained based on the ship AIS information;

[0065] Obtain radial velocity information of the ship target based on the radar position in step 1 and the speed and heading of the ship target obtained based on the ship AIS information;

[0066] Calculating the range element and Doppler element information of the ship target according to the radial distance and radial velocity of the ship target;

[0067] The MMSI, target range element, and target Doppler metadata of the ship target are stored in the AIS target information database.

[0068] It should be noted that within the start and end time period of a radar data session, multiple ships' AIS signals may be included. These multiple ship's AIS signals are processed separately to obtain the range element and Doppler element information of multiple ship targets and store them in the ship AIS target database.

[0069] Step 4: read the distance element and Doppler element information in the ship AIS target database, obtain the time spectrum of the ship target based on the distance element, and generate a pre-selected annotation box centered on the Doppler frequency of the ship target based on the Doppler element of the ship target to obtain the initial screening annotation file.

[0070] Preferably, according to the AIS target information database obtained in step 3, the distance meta information and Doppler meta information of the ship AIS are read one by one.

[0071] Furthermore, according to the target distance element information, the time-frequency spectrum corresponding to the target distance element obtained in step 2 is read, and then according to the target Doppler information, a target detection pre-selected annotation box is generated with the target Doppler as the center, and the annotation box category attribute is set to "vessel", such as Figure 3 As shown, the pre-selected annotation box information is saved to the initial screening annotation file.

[0072] As a preferred embodiment, the size of the pre-selected annotation box generated in step 4 is , where L is the number of sampling points in the time dimension of the time spectrum, is the maximum number of target Doppler migration points. In the embodiment, .

[0073] Step 5: Review the primary screening and marking file to form a secondary screening and marking file.

[0074] Preferably, according to the time-frequency spectrum obtained in step 2 and the initial screening annotation file in step 4, first observe whether there is a ship echo time-frequency ridge in the pre-selected annotation box in the time-frequency spectrum. If there is a target time-frequency ridge in the pre-selected box, adjust the size of the annotation box so that all the target time-frequency ridges fall into the annotation box. If there is no target time-frequency ridge in the pre-selected box, delete the pre-selected annotation box; then, observe whether there are other suspected target time-frequency ridges in the time-frequency spectrum that do not contain the pre-selected annotation box. If so, manually generate a target annotation box and set the annotation box category to "clear". Figure 4 As shown; finally, save it to the secondary screening annotation file.

[0075] Step 6: Process the suspected target labeled areas in the ship target's time-frequency spectrum that do not contain the pre-selected labeled areas to form a new time-frequency spectrum. The suspected target labeled areas that do not contain the pre-selected labeled areas in the secondary screening are deleted to obtain a final screening and labeling file. Suspected targets that do not contain pre-selected labeled areas can be understood as those that resemble a ship target's time-frequency ridge but do not contain AIS information.

[0076] Preferably, based on the secondary screening annotation file obtained in step 5 and the Doppler range of the sea clutter first-order spectrum region and the ground clutter obtained in step 2, first, the region with the annotation box category of "clear" in the time-frequency spectrum is set to 0, and then the sea clutter first-order spectrum region and the ground clutter region of the time-frequency spectrum are set to 0. The time-frequency spectrum with the 0-setting operation is saved as the preprocessed time-frequency spectrum; then, the annotation box with the category of "clear" is deleted, and finally, the new annotation file is saved as the final screening annotation file.

[0077] As a preferred embodiment, the method for preparing a high-frequency radar ship target time-frequency domain detection dataset provided by the embodiment of the present invention further includes:

[0078] In step 7, based on the final filtered annotation file obtained in step 6 and the preprocessed time-frequency spectrum, determine whether the annotation file contains the target annotation box (i.e., the annotation box with the category "vessel"). If so, save the final annotation file and the corresponding preprocessed time-frequency spectrum to the dataset. Otherwise, delete the final annotation file and the corresponding preprocessed time-frequency spectrum.

[0079] The implementation of each embodiment of the present invention is based on programmed processing performed by a device with processor functionality. Therefore, in practical engineering, the technical solutions and functions of each embodiment of the present invention are packaged into various modules. Based on this reality, and in addition to the aforementioned embodiments, an embodiment of the present invention provides a system for generating a high-frequency radar ship target detection dataset in the time-frequency domain. This system is used to execute one of the methods for generating a high-frequency radar ship target detection dataset in the time-frequency domain described in the aforementioned method embodiments.

[0080] See also Figure 2The system includes: a first calculation module for obtaining the start and end time of the high-frequency radar data, the radar position, and the range-slow time spectrum based on the high-frequency radar received data; a second calculation module for calculating the time-frequency spectrum of each distance element based on the range-slow time spectrum; a third calculation module for obtaining the ship AIS information within the current start and end time period based on the start and end time of the high-frequency radar data, and calculating the range element and Doppler element information of the ship target based on the ship AIS information and the radar position, and establishing a ship AIS target database; a first screening module for reading the ship AIS target database The distance element and Doppler element information of the ship target are obtained, and the time spectrum of the ship target is obtained according to the distance element of the ship target, and the annotation pre-selection box is generated with the Doppler frequency of the ship target as the center according to the Doppler element of the ship target to obtain a primary screening annotation file; the second screening module is used to review the primary screening annotation file to form a secondary screening annotation file; the third screening module is used to process the suspected target annotation box area in the time spectrum of the ship target that does not contain the pre-selected annotation box to form a new time spectrum, and delete the suspected target annotation box in the secondary screening annotation file that does not contain the pre-selected annotation box to obtain a final screening annotation file.

[0081] The embodiment of the present invention provides a high-frequency radar ship target time-frequency domain detection dataset production system. It solves the problem that high-frequency radar cannot perform target imaging. When labeling target ridges, it is necessary to determine whether the time-frequency ridges belong to ship targets, and it takes a lot of manpower and time to manually label them. Figure 2 Several modules in the system use AIS information to generate pre-selected annotation boxes in the time and frequency domain of high-frequency radar, providing a reference for constructing a target detection dataset in the time and frequency domain, supporting the application of deep neural network models in high-frequency radar target detection, and thus improving the detection performance of high-frequency radar systems for ship targets.

[0082] It should be noted that the system embodiments provided by the present invention are not only used to implement the methods in the above-mentioned method embodiments, but also used to implement the methods in other method embodiments provided by the present invention. The only difference is the setting of corresponding functional modules. The principles thereof are basically the same as those of the above-mentioned system embodiments provided by the present invention. As long as those skilled in the art refer to the specific technical solutions in other method embodiments on the basis of the above-mentioned system embodiments, obtain corresponding technical means and technical solutions composed of these technical means by combining technical features, and ensure the practicality of the technical solutions, they can improve the equipment in the above-mentioned system embodiments to obtain corresponding system-type embodiments, and obtain corresponding system-type embodiments for implementing the methods in other method-type embodiments. For example:

[0083] Based on the content of the above system embodiment, as a preferred embodiment, a high-frequency radar ship target time-frequency domain detection data set production system provided in the embodiment of the present invention further includes:

[0084] The storage module is used to save the final screened and labeled file and its corresponding time-frequency spectrum to the high-frequency radar ship target time-frequency domain detection dataset.

[0085] Based on the content of the above system embodiment, as a preferred embodiment, in a high-frequency radar ship target time-frequency domain detection data set production system provided in an embodiment of the present invention, the second calculation module is further used to execute the following instructions:

[0086] Generate range-Doppler spectrum using fast Fourier transform;

[0087] According to the range-Doppler spectrum and in combination with the first-order Bragg scattering theorem of sea clutter, a signal-to-noise ratio threshold is set to calculate the Doppler range of the first-order spectrum of sea clutter;

[0088] According to the range-Doppler spectrum and in combination with the stationary characteristics of the ground clutter, the ground clutter Doppler range is obtained.

[0089] Furthermore, after obtaining the first-order spectrum region of the sea clutter and the Doppler range of the ground clutter, the first-order spectrum region of the sea clutter and the ground clutter region of the time-frequency spectrum are set to 0.

[0090] Based on the content of the above system embodiment, as a preferred embodiment, in a high-frequency radar ship target time-frequency domain detection data set production system provided in an embodiment of the present invention, the third calculation module is further used to execute the following instructions:

[0091] Calculating a radial distance of the ship target according to the radar position and the ship target position, and calculating a range element of the ship target according to the radial distance;

[0092] The radial velocity of the ship target is calculated according to the radar position and the speed and heading of the ship target, and the Doppler element of the ship target is calculated according to the radial velocity.

[0093] Based on the content of the above system embodiment, as a preferred embodiment, in a high-frequency radar ship target time-frequency domain detection data set production system provided in an embodiment of the present invention, the first screening module is further used to execute the following instructions:

[0094] Read the distance element and Doppler element information of the ship target from the established ship AIS target database;

[0095] According to the distance element of the ship target, read the time spectrum corresponding to the distance element;

[0096] According to the Doppler element of the ship target, a pre-selected marking frame for ship target detection is generated, a category attribute of the pre-selected marking frame is set, and information of the pre-selected marking frame is saved to a primary screening marking file.

[0097] Based on the content of the above system embodiment, as a preferred embodiment, in a high-frequency radar ship target time-frequency domain detection data set production system provided in an embodiment of the present invention, the second screening module is further used to execute the following instructions:

[0098] Determine whether there is a ship target time-frequency ridge in the preselected marked frame of the time-frequency spectrum, and if so, adjust the size of the marked frame so that the ship target time-frequency ridge line falls entirely within the marked frame;

[0099] It is determined whether there is a suspected target time-frequency ridge in the time-frequency spectrum that does not contain a pre-selected labeling box. If so, a target labeling box is generated and category attributes are set.

[0100] Based on the content of the above system embodiment, as a preferred embodiment, in a high-frequency radar ship target time-frequency domain detection data set production system provided in an embodiment of the present invention, the third screening module is further used to execute the following instructions:

[0101] Setting the area of ​​the time-frequency spectrum marked with a non-ship category to 0 to obtain a new time-frequency spectrum;

[0102] The annotation boxes whose annotation box categories are non-ship are deleted to obtain the final filtered annotation file.

[0103] The method of the embodiment of the present invention is implemented by electronic devices, so it is necessary to introduce the relevant electronic devices. Based on this purpose, the embodiment of the present invention provides an electronic device, such as Figure 5 As shown, the electronic device includes: at least one processor, a communications interface, at least one memory, and a communications bus, wherein the at least one processor, the communications interface, and the at least one memory communicate with each other via the communications bus. The at least one processor invokes logic instructions in the at least one memory to execute all or part of the steps of the methods provided in the aforementioned method embodiments.

[0104] In addition, when the logic instructions in the at least one memory are implemented in the form of a software functional unit and sold or used as an independent product, they are stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or the part of the technical solution is embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (a personal computer, a server, or a network device) to execute all or part of the steps of the method described in each method embodiment of the present invention. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, various media for storing program codes.

[0105] The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, located in one place or distributed across multiple network units. Depending on practical needs, some or all of these modules may be selected to achieve the objectives of this embodiment. Persons of ordinary skill in the art will understand and implement these embodiments without inventive effort.

[0106] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0107] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0108] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0109] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0110] In summary, the present invention obtains the start and end time, radar position, and range-slow time spectrum of the field data based on the high-frequency radar received data; calculates the range-Doppler spectrum and the time-frequency spectrum of each range element based on the range-slow time spectrum; calculates the first-order spectrum and the Doppler range of the ground clutter based on the range-Doppler spectrum; obtains the ship AIS information based on the start and end time of the high-frequency radar data, and calculates the radial distance and radial velocity of the AIS ship target, thereby calculating the target range element and the target Doppler frequency, and establishing an AIS target database; based on the target database information of the AIS ship, generates a pre-selected box in the corresponding range element time-frequency spectrum with the target Doppler as the center and saves the target annotation file; completes the target annotation through manual review based on the target time-frequency spectrum and the target annotation file, thereby realizing the production of a high-frequency radar ship target time-frequency domain detection data set. The present invention can reduce the manpower and time costs of manual annotation while ensuring the accuracy of the annotated targets.

[0111] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the technical solutions of the embodiments of the present invention.

Claims

1. A method for preparing a high-frequency radar ship target time-frequency domain detection dataset, characterized in that: include: According to the high-frequency radar receiving data, the start and end time of the high-frequency radar data, the radar position, and the distance-slow time spectrum are obtained; According to the distance-slow time spectrum, a time spectrum of each distance element is calculated; According to the start and end time of the high-frequency radar data session, the ship AIS information within the current start and end time period is obtained, and the range element and Doppler element information of the ship target are calculated based on the ship AIS information and the radar position, and a ship AIS target database is established; Reading the distance element and Doppler element information in the ship AIS target database, obtaining the time-frequency spectrum of the ship target according to the distance element, and generating a pre-selected annotation box centered on the Doppler frequency of the ship target according to the Doppler element of the ship target, to obtain a primary screening annotation file; Reviewing the primary screening and marking file to form a secondary screening and marking file; The suspected target mark box area that does not contain the pre-selected mark box in the time spectrum of the ship target is processed to form a new time spectrum, and the suspected target mark box that does not contain the pre-selected mark box in the secondary screening mark file is deleted to obtain a final screening mark file.

2. The method for preparing a high-frequency radar ship target time-frequency domain detection dataset according to claim 1, characterized in that: The method further includes: saving the final screened and labeled file and the corresponding time-frequency spectrum to a high-frequency radar ship target time-frequency domain detection data set.

3. The method for preparing a high-frequency radar ship target time-frequency domain detection dataset according to claim 1, characterized in that: After obtaining the distance-slow time spectrum, the method further includes: Generate range-Doppler spectrum using fast Fourier transform; According to the range-Doppler spectrum and in combination with the first-order Bragg scattering theorem of sea clutter, a signal-to-noise ratio threshold is set to calculate the Doppler range of the first-order spectrum of sea clutter; According to the range-Doppler spectrum and in combination with the stationary characteristics of the ground clutter, the ground clutter Doppler range is obtained.

4. The method for preparing a high-frequency radar ship target time-frequency domain detection dataset according to claim 1, characterized in that: The range element and Doppler element information of the ship target are calculated based on the ship AIS information and the radar position, including: Calculating a radial distance of the ship target according to the radar position and the ship target position, and calculating a range element of the ship target according to the radial distance; The radial velocity of the ship target is calculated according to the radar position and the speed and heading of the ship target, and the Doppler element of the ship target is calculated according to the radial velocity.

5. The method for preparing a high-frequency radar ship target time-frequency domain detection dataset according to claim 1, characterized in that: The generation of the initial screening annotation file includes: Read the distance element and Doppler element information of the ship target from the established ship AIS target database; According to the distance element of the ship target, read the time spectrum corresponding to the distance element; According to the Doppler element of the ship target, a pre-selected marking frame for ship target detection is generated, a category attribute of the pre-selected marking frame is set, and information of the pre-selected marking frame is saved to a primary screening marking file.

6. The method for preparing a high-frequency radar ship target time-frequency domain detection dataset according to claim 5, characterized in that: Review the initial screening and annotation documents, including: Determine whether there is a ship target time-frequency ridge in the preselected marked frame of the time-frequency spectrum, and if so, adjust the size of the marked frame so that the ship target time-frequency ridge line falls entirely within the marked frame; It is determined whether there is a suspected target time-frequency ridge in the time-frequency spectrum that does not contain a pre-selected labeling box. If so, a target labeling box is generated and category attributes are set.

7. The method for preparing a high-frequency radar ship target time-frequency domain detection dataset according to claim 3, characterized in that: After obtaining the secondary screening and annotation file, it also includes: Setting the area of ​​the time-frequency spectrum marked with a non-ship category to 0 to obtain a new time-frequency spectrum; The annotation boxes whose annotation box categories are non-ship are deleted to obtain the final filtered annotation file.

8. A high-frequency radar ship target time-frequency domain detection data set production system, characterized by: include: The first calculation module is used to obtain the start and end time of the high-frequency radar data, the radar position, and the range-slow time spectrum based on the high-frequency radar received data; A second calculation module is used to calculate the time-frequency spectrum of each range element according to the range-slow time spectrum; The third calculation module is used to obtain the ship AIS information within the current start and end time period based on the start and end time of the high-frequency radar data session, and calculate the range element and Doppler element information of the ship target based on the ship AIS information and the radar position, and establish a ship AIS target database; The first screening module is configured to read the distance element and Doppler element information in the ship AIS target database, obtain the time-frequency spectrum of the ship target based on the distance element, and generate a pre-selected annotation box centered on the Doppler frequency of the ship target based on the Doppler element of the ship target to obtain a primary screening annotation file; A second screening module is used to review the primary screening and marking file to form a secondary screening and marking file; The third screening module is used to process the suspected target marking box area that does not contain the pre-selected marking box in the time-frequency spectrum of the ship target to form a new time-frequency spectrum, and delete the suspected target marking box that does not contain the pre-selected marking box in the secondary screening marking file to obtain a final screening marking file.

9. An electronic device, characterized in that: include: At least one processor, at least one memory and a communication interface; wherein the processor, memory and communication interface communicate with each other; the memory stores program instructions executed by the processor, and the processor calls the program instructions to execute the method described in any one of claims 1 to 7.

10. A non-transitory computer readable storage medium, characterized in that The non-transitory computer-readable storage medium stores computer instructions, and the computer instructions cause the computer to execute the method of any one of claims 1 to 7.

Images