Ship wake detection and identification method and related device

Abstract

The application discloses a ship wake detection and identification method and related equipment, and belongs to the technical field of ship detection. The method comprises the following steps: acquiring a remote sensing image of a target ship; identifying and detecting a ship wake area on the sea surface based on the remote sensing image and outputting a wake detection result; obtaining the motion parameters of the target ship through a physical model and a data-driven model based on the wake detection result; and determining the ship wake detection abnormality identification result of the target ship by performing consistency comparison on the basis of the motion parameters of the target ship and the data reported by the ship automatic identification system of the target ship. The physical model is constructed based on the wake fluid mechanics characteristics. The data-driven model is constructed based on a convolutional neural network model. The application can effectively identify the authenticity of AIS data and accurately detect abnormal behaviors of ships by performing wake detection on the remote sensing image, determining the ship motion parameters through a double-path model, and then cross-comparing the parameters reported by the AIS.

Description

Technical Field

[0001] This application relates to the field of ship inspection technology, and in particular to a method and related equipment for detecting and identifying ship wakes. Background Technology

[0002] Currently, with the increasing number and density of ships at sea, there is a need for deeper analysis and processing of ship trajectory data in order to enhance the safety of maritime traffic.

[0003] Among related technologies, the Automatic Identification System (AIS) is a core tool for maritime traffic control and vessel tracking. It typically utilizes AIS data for trajectory analysis and is widely used in scenarios such as vessel activity monitoring, track analysis, and maritime law enforcement. However, in practical applications, it has been found that traditional detection methods cannot handle scenarios where AIS is disabled or falsified, leading to difficulties in identifying unauthorized vessels and AIS deception.

[0004] In summary, the technical problems existing in the relevant technologies need to be improved. Summary of the Invention

[0005] This application provides a method and related equipment for detecting and identifying ship wakes, which can effectively identify the authenticity of AIS data and accurately detect abnormal ship behavior.

[0006] On the one hand, embodiments of this application provide a method for detecting and identifying ship wakes, the method comprising the following steps: Acquire remote sensing images of the target vessel; Based on the remote sensing image of the target vessel, identify and detect the vessel's wake region on the sea surface and output the wake detection results; Based on the wake detection results, the motion parameters of the target vessel are obtained by inversion using a physical model and a data-driven model. Based on the motion parameters of the target vessel, and combined with the data reported by the target vessel's Automatic Identification System, a consistency comparison is performed to determine the abnormal identification result of the target vessel's wake detection. The physical model is constructed based on the hydrodynamic properties of the wake; the data-driven model is constructed based on the convolutional neural network model.

[0007] Optionally, the step of obtaining the motion parameters of the target vessel by inverting the wake detection results through a physical model and a data-driven model includes: Based on the geometric morphological features of the wake detection results, the physical model and the data-driven model are operated synchronously and independently to obtain the calculation results of the physical model and the data-driven model. Based on the calculation results of the physical model and the data-driven model, combined with the wake detection confidence, calculation result deviation and sea state interference, the motion parameters and confidence level of the target ship are output; wherein, the motion parameters include heading and speed.

[0008] Optionally, the step of obtaining the calculation results of the physical model and the data-driven model based on the geometric morphological features of the wake detection results through synchronous and independent calculations of the physical model and the data-driven model includes: Based on the geometric morphological features of the wake detection results, the data is input into a physical model constructed from the Kelvin wake hydrodynamic characteristics. The heading is determined by the extension direction of the wake rotation bounding box or the main axis direction of the pixel mask distribution. The speed is determined by combining the wake shear wave wavelength, half angle, and timing features. Based on the geometric features of the wake detection results, a wake feature map is extracted and input into a data-driven model constructed by a convolutional neural network model to obtain the heading and speed results output by the data-driven model. By combining the heading and speed results output by the physical model and the data-driven model, the calculation results of the physical model and the data-driven model are obtained.

[0009] Optionally, the step of identifying and detecting the wake region of a vessel on the sea surface and outputting the wake detection result based on the remote sensing image of the target vessel includes: The remote sensing image of the target ship is input into the rotating target detection network model, so that the rotating target detection network model can extract features from the remote sensing image, identify and detect the ship's wake region on the sea surface, and output the wake detection result. The rotating target detection network model is built based on the YOLOv8 architecture and includes a multi-directional perceptual convolutional module and a frequency-space fusion attention module. The multi-directional perceptual convolutional module includes multiple sets of convolutional kernels with different orientations for capturing trail texture features from multiple angles. The frequency-space fusion attention module adaptively fuses frequency domain and spatial domain features through a multi-head attention mechanism.

[0010] Optionally, the step of performing a consistency comparison based on the motion parameters of the target vessel and the data reported by the target vessel's Automatic Identification System to determine the abnormal identification result of the target vessel's wake detection includes: Based on the motion parameters of the target vessel, and according to the time window and spatial range, the target is matched with the automatic identification system for ships. If the motion parameters of the target vessel fail to match the target of the Automatic Identification System (AIS), the abnormal identification result of the target vessel's wake detection is determined to be that the AIS is shut down as a dark vessel. If the motion parameters of the target vessel successfully match the target of the Automatic Identification System (AIS), and the deviation between the motion parameters of the target vessel and the data reported by the AIS is greater than a preset deviation threshold, the abnormal identification result of the target vessel's wake detection is determined to be data fraud. If the motion parameters of the target vessel successfully match the target of the Automatic Identification System (AIS), and the deviation between the motion parameters of the target vessel and the data reported by the AIS is less than or equal to the preset deviation threshold, the motion parameters of the target vessel are detected and matched based on preset abnormal behavior rules, thereby determining the abnormal identification result of the target vessel's wake detection.

[0011] Optionally, after acquiring the remote sensing image of the target ship, the method further includes: Perform radiometric calibration, geometric correction, and cropping normalization on the remote sensing images of the target ship; When the remote sensing image of the target ship is a synthetic aperture radar image, speckle suppression and multi-view processing are performed on the remote sensing image; When the remote sensing image of the target ship is an optical remote sensing image, atmospheric correction and cloud area masking are performed on the remote sensing image.

[0012] On the other hand, embodiments of this application provide a ship wake detection and identification device, the device comprising: The image acquisition module is used to acquire remote sensing images of the target ship; The wake detection module is used to identify and detect the wake region of the target vessel on the sea surface and output the wake detection results based on the remote sensing image of the target vessel. The parameter inversion module is used to obtain the motion parameters of the target ship by inverting the wake detection results through a physical model and a data-driven model. An anomaly detection module is used to perform a consistency comparison based on the motion parameters of the target vessel and the data reported by the target vessel's Automatic Identification System to determine the anomaly detection result of the target vessel's wake. The physical model is constructed based on the hydrodynamic properties of the wake; the data-driven model is constructed based on the convolutional neural network model.

[0013] On the other hand, embodiments of this application provide an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the above-described method.

[0014] On the other hand, embodiments of this application provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method.

[0015] On the other hand, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0016] This application embodiment detects wakes in remote sensing images and determines ship motion parameters through parameter inversion using a dual-path model. These parameters are then cross-compared with parameters reported by AIS to effectively identify the authenticity of AIS data and accurately detect abnormal ship behavior. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the implementation environment of a ship wake detection and identification method provided in an embodiment of this application; Figure 2 This is a flowchart illustrating a method for detecting and identifying ship wakes provided in an embodiment of this application; Figure 3 This is a schematic diagram of a remote sensing image wake detection and recognition process provided in an embodiment of this application; Figure 4 This is a schematic flowchart illustrating the process of detecting and retrieving motion parameters from a wake, as provided in an embodiment of this application. Figure 5 This is a schematic diagram of a process for identifying anomalies in ship wake detection provided in an embodiment of this application; Figure 6 This is a schematic diagram of the structure of a ship wake detection and identification device provided in an embodiment of this application; Figure 7 This is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit it. In the following description, when referring to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with those of this application; they are merely examples of apparatuses and methods consistent with some aspects of the embodiments of this application as detailed in the appended claims.

[0019] It is understood that the terms “first,” “second,” etc., used in this application may be used herein to describe various concepts, but unless otherwise stated, these concepts are not limited by these terms. These terms are only used to distinguish one concept from another. For example, without departing from the scope of the embodiments of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the words “if,” “when,” or “in response to a determination” as used herein may be interpreted as “when…” or “when…” or “in response to a determination.”

[0020] As used in this application, the terms "at least one", "multiple", "each", "any", etc., "at least one" includes one, two or more, "multiple" includes two or more, "each" refers to each of the corresponding multiples, and "any" refers to any one of the multiples.

[0021] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0022] Currently, with the increasing number and density of ships at sea, there is a need for deeper analysis and processing of ship trajectory data in order to enhance the safety of maritime traffic.

[0023] Among related technologies, the Automatic Identification System (AIS) is a core tool for maritime traffic control and vessel tracking. It typically utilizes AIS data for trajectory analysis and is widely used in scenarios such as vessel activity monitoring, track analysis, and maritime law enforcement. However, in practical applications, it has been found that traditional detection methods cannot handle scenarios where AIS is disabled or falsified, leading to difficulties in identifying unauthorized vessels and AIS deception.

[0024] In view of this, this application provides a method and related equipment for detecting and identifying ship wakes. By detecting wakes in remote sensing images and determining ship motion parameters through parameter inversion using a dual-path model, the method can effectively identify the authenticity of AIS data and accurately detect abnormal ship behavior by cross-comparing the parameters reported by AIS.

[0025] It should be noted that in all specific embodiments of this application, when processing data related to user identity or characteristics, such as user information, user behavior data, user historical data, and user location information, user permission or consent is obtained first. Furthermore, the collection, use, and processing of this data comply with relevant laws, regulations, and standards. In addition, when embodiments of this application require access to sensitive personal information of users, separate permission or consent from the user is obtained through pop-ups or redirection to confirmation pages. Only after obtaining the user's separate permission or consent is the necessary user-related data required for the proper functioning of these embodiments acquired.

[0026] The specific implementation methods of the embodiments of this application will be described in detail below with reference to the accompanying drawings. First, a method for detecting and identifying ship wakes provided in the embodiments of this application will be described with reference to the accompanying drawings.

[0027] Please refer to Figure 1 , Figure 1 This is a schematic diagram of the implementation environment for a ship wake detection and identification method provided in this application embodiment. In this implementation environment, the main hardware and software components involved include a terminal processor 110 and a server 120.

[0028] Specifically, the terminal processor 110 may be equipped with a control program for a ship wake detection and identification method, and the server 120 serves as the backend server for this control program. The terminal processor 110 and the backend server 120 are connected in communication. The ship wake detection and identification method provided in this embodiment can be executed on the terminal processor 110 side.

[0029] Server 120 can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms.

[0030] In addition, server 120 can also be a node server in a blockchain network.

[0031] The terminal processor 110 and the server 120 can establish a communication connection via a wireless network. This wireless network uses standard communication technologies and / or protocols. The network can be the Internet or any other network, including but not limited to a Local Area Network (LAN), Metropolitan Area Network (MAN), Wide Area Network (WAN), mobile, or any combination of wireless networks, private networks, or virtual private networks. Furthermore, these hardware and software components can use the same or different communication connection methods; this application does not impose specific limitations in this regard.

[0032] Of course, this is understandable. Figure 1 The implementation environment described in this application is only one of the optional application scenarios for the ship wake detection and identification method provided in this embodiment. The actual application is not fixed. Figure 1 The software and hardware environment shown is not specifically limited in this application.

[0033] like Figure 2 As shown, Figure 2 This is a flowchart illustrating a method for detecting and identifying ship wakes provided in an embodiment of this application, specifically including but not limited to steps 100 to 400.

[0034] Step 100: Obtain remote sensing images of the target ship.

[0035] In this embodiment, remote sensing images of the target vessel, including its wake, can be obtained by accessing various image data sources such as satellites, drones, and shore-based observation stations. The target vessel is the vessel to be detected and identified.

[0036] Specifically, as an optional implementation, after acquiring the remote sensing image of the target ship, the method further includes: Perform radiometric calibration, geometric correction, and cropping normalization on the remote sensing images of the target ship; When the remote sensing image of the target ship is a synthetic aperture radar image, speckle suppression and multi-view processing are performed on the remote sensing image; When the remote sensing image of the target ship is an optical remote sensing image, atmospheric correction and cloud area masking are performed on the remote sensing image.

[0037] In practical applications, the remote sensing images of the target ship can include different types of remote sensing images such as Synthetic Aperture Radar (SAR) satellite images, optical satellite images, UAV images, and shore-based station monitoring images.

[0038] Furthermore, radiometric calibration, geometric correction, and cropping normalization can be performed on the remote sensing images of the target ship to complete the preliminary preprocessing of the original remote sensing images.

[0039] Among them, radiometric calibration can adjust the image grayscale or image brightness, geometric correction can correct the positional deviation in the original remote sensing image caused by imaging angle, platform attitude, etc., and cropping normalization can adjust the original remote sensing image to a uniform size to facilitate subsequent model detection and processing.

[0040] Furthermore, for SAR images, speckle suppression and multi-view processing can be performed to reduce speckle noise in remote sensing images; for optical remote sensing images, atmospheric correction and cloud masking can be performed to reduce cloud and atmospheric interference and improve image quality.

[0041] Therefore, this application performs a unified image preprocessing process after acquiring the original remote sensing image of the target ship, transforming the original image into a format more suitable for subsequent algorithm analysis, ensuring that the image quality meets the detection requirements, and thus improving the accuracy of subsequent detection and recognition.

[0042] In practical applications, by being compatible with various remote sensing data sources such as SAR satellites, optical satellites, drones, and shore-based stations, it can support high-precision conventional deployment and lightweight edge deployment, improve adaptability to multiple scenarios, and facilitate expansion and deployment.

[0043] Step 200: Based on the remote sensing image of the target vessel, identify and detect the vessel wake region on the sea surface and output the wake detection result.

[0044] In this embodiment of the application, based on the acquired remote sensing image containing ship wakes, the ship wake region on the sea surface is detected and identified and the wake detection result is output as the data basis for the subsequent dual-path model inversion parameters.

[0045] In practical applications, deep learning models can be used to capture and extract features related to ship wakes on the sea surface, thereby locating the wake region in remote sensing images, determining the wake position and direction, and outputting the wake's bounding box or code as the wake detection result. This allows background data in the remote sensing image to be filtered out, facilitating the subsequent dual-path model to invert motion parameters based on the wake detection result.

[0046] Specifically, as an optional implementation, the step of identifying and detecting the wake region of a vessel on the sea surface and outputting the wake detection result based on the remote sensing image of the target vessel includes: The remote sensing image of the target ship is input into the rotating target detection network model, so that the rotating target detection network model can extract features from the remote sensing image, identify and detect the ship's wake region on the sea surface, and output the wake detection result. The rotating target detection network model is built based on the YOLOv8 architecture and includes a multi-directional perceptual convolutional module and a frequency-space fusion attention module. The multi-directional perceptual convolutional module includes multiple sets of convolutional kernels with different orientations for capturing trail texture features from multiple angles. The frequency-space fusion attention module adaptively fuses frequency domain and spatial domain features through a multi-head attention mechanism.

[0047] In this embodiment of the application, the remote sensing image of the target ship is input into a pre-trained rotating target detection network model. The rotating target detection network model extracts features from the remote sensing image of the target ship, identifies and detects the ship's wake region on the sea surface, and outputs the wake detection result.

[0048] Ship wakes typically appear as strip-shaped, arc-shaped, linear, or diffused structures on the sea surface and can point in any direction. Therefore, rotating target detection network models can effectively capture ship wake features in remote sensing images.

[0049] Alternatively, a rotating object detection network model can be built based on the Transformer architecture or an improved YOLOv8 architecture. For example, the rotating target detection network model includes a feature extraction layer, a feature fusion layer, and a prediction output layer. When constructing a rotating target detection network model based on the improved YOLOv8 architecture, the feature extraction layer can be used as the backbone network of the rotating target detection network model to perform multi-scale feature extraction. Then, the feature fusion layer fuses the wake feature information at multiple scales (e.g., deep and shallow) to obtain a fused feature map. Finally, the prediction head performs classification branch prediction based on the fused feature map to determine and output the sea surface ship wake region in the remote sensing image. Furthermore, a multi-directional perceptual convolution module and a frequency-space fusion attention module can be set for the rotating target detection network model constructed based on the YOLOv8 architecture. In the feature extraction process, multiple sets of convolution kernels with different orientations are used to capture wake texture features from multiple angles to extract texture features from different orientations, thereby improving the multi-scale feature extraction capability and accuracy. In the feature fusion process, a multi-head attention mechanism is used to adaptively fuse frequency domain and spatial domain features, while using frequency domain enhancement features and spatial context features to improve wake recognition capability.

[0050] Specifically, please refer to Figure 3 , Figure 3This is a schematic diagram of a remote sensing image wake detection and recognition process provided in an embodiment of this application. When constructing a rotating target detection network model based on the Transformer architecture, after inputting a preprocessed remote sensing image, the input image is first encoded into a feature map by the encoder in the feature extraction layer, and multi-scale feature extraction is performed. Then, the wake features are located by the decoder. Next, cross-scale fusion is performed by using a feature pyramid network structure in the feature fusion layer, so that the model can accurately identify significant size differences in the image at the same time. Finally, in the prediction output layer, classification and regression output are achieved by decoupling the prediction head. The wake detection confidence is predicted by the classification branch, and the wake position information is located by the regression branch. The wake detection rotation box data is output, which may specifically include position, size, and orientation.

[0051] In practical applications, multiple remote sensing images can be selected as training data, and the areas of ship wakes on the sea surface can be labeled as tags to construct a training dataset. Then, the rotating target detection network model can be pre-trained using the training dataset, enabling the rotating target detection network model to learn to extract wake texture features.

[0052] Therefore, this application identifies and detects the sea surface ship wake region in remote sensing images by completing a pre-trained rotating target detection network model, which can accurately capture wake texture features and realize wake classification confidence prediction and rotating bounding box regression output.

[0053] Step 300: Based on the wake detection results, obtain the motion parameters of the target ship by inversion through a physical model and a data-driven model; wherein, the physical model is constructed based on the hydrodynamic characteristics of the wake; and the data-driven model is constructed based on a convolutional neural network model.

[0054] In this embodiment, the wake detection results are used as the data basis, and the motion parameters of the target ship are inverted by independent calculations through the physical model and the data-driven model, thereby inferring the motion state of the target ship through the ship's wake.

[0055] Specifically, the physical model is constructed based on the hydrodynamic characteristics of the wake, and the motion parameters of the target ship are calculated by inversion based on the laws of hydrodynamics. The data-driven model can be constructed based on the convolutional neural network model. The pre-trained convolutional neural network model learns the correlation between the wake, heading and speed from the sample data. Based on the input wake map, the wake features are extracted through the convolutional layer, and then the heading and speed are directly output through the fully connected layer as the motion parameters of the target ship.

[0056] Optionally, the step of obtaining the motion parameters of the target vessel by inverting the wake detection results through a physical model and a data-driven model includes: Based on the geometric morphological features of the wake detection results, the physical model and the data-driven model are operated synchronously and independently to obtain the calculation results of the physical model and the data-driven model. Based on the calculation results of the physical model and the data-driven model, combined with the wake detection confidence, calculation result deviation and sea state interference, the motion parameters and confidence level of the target ship are output; wherein, the motion parameters include heading and speed.

[0057] In the embodiments of this application, please refer to Figure 4 , Figure 4 This is a flowchart illustrating a wake detection and motion parameter inversion process provided in this application embodiment. Based on the input wake detection result, geometric features such as wake principal axis direction, wake extension direction, transverse wave wavelength, half angle, and multi-frame continuous change trend are obtained. The physical model and the data-driven model are then operated synchronously and independently to obtain the calculation results of the physical model and the data-driven model.

[0058] Furthermore, based on the calculation results of the physical model and the data-driven model, combined with the wake detection confidence, calculation result deviation and sea state interference, the motion parameters and confidence markers of the target ship are output through collaborative redundant calculation. The motion parameters of the target ship specifically include heading and speed.

[0059] Specifically, the output motion parameters can be classified into three levels of confidence (high, medium, and low) and labeled based on a combination of wake detection confidence, dual-path model parameter deviation, and sea state interference. For example, if the wake detection confidence of the output detected by the rotating target detection network model is greater than a preset threshold (e.g., it can be set to 0.8, or preset according to specific scenario requirements), the deviation between the calculation results of the physical model and the calculation results of the data-driven model is less than or equal to a preset deviation value (e.g., heading deviation ≤ 8°, speed deviation ≤ 2 knots), and there is no significant sea state interference, the output motion parameters can be determined to be of high confidence level. Furthermore, if the wake detection confidence is within a preset range (e.g., 0.6 to 0.8), the dual-path model calculation results have a large deviation, and there is slight sea state interference, the output motion parameters can be determined to be of medium confidence level. If the wake detection confidence is low, the wake is incomplete or strongly interfered with, and the dual-path model calculation results have a large deviation, it can be labeled as low confidence level.

[0060] Therefore, by labeling the output motion parameters with multi-level confidence levels, the reliability of the output results can be quickly determined.

[0061] In practical applications, when outputting the motion parameters of a target vessel through collaborative redundant calculation, if the sea state is good and the wake is intact, the physical model result can be used as the main parameter. The average of the calculation results of the physical model and the data-driven model can be used as the motion parameters of the target vessel. If the wake is incomplete or there is interference from high sea state, the calculation result of the data-driven model can be selected as the motion parameters of the target vessel. Alternatively, the two models can be fused in real time according to their confidence levels. If either model fails, the calculation result of the other model can be used directly as the motion parameters of the target vessel.

[0062] Therefore, this application obtains the calculation results by independently calculating through a physical model, which can effectively utilize the physical model's strength in interpreting and stabilizing scenarios. It also obtains the calculation results by independently calculating through a data-driven model, which can effectively utilize the data-driven model's strength in complex and weak prior scenarios. Furthermore, it performs collaborative redundant calculations through a dual-path model, which can achieve complementarity between the two paths and improve the adaptability and accuracy of the target ship's motion parameters.

[0063] Furthermore, this application constructs a physical model and a data-driven model to invert ship motion parameters, which serve as benchmark values ​​independent of AIS data. This transforms AIS data from a single source of information into an object to be verified, enabling a shift from passive reception to active verification. The actual ship motion parameters can be calculated without relying on AIS signals, effectively detecting ships that have turned off or blocked AIS. This addresses the inherent shortcomings of traditional AIS monitoring and improves the accuracy and efficiency of identifying ships that have turned off or blocked AIS.

[0064] Specifically, as an optional implementation, the step of obtaining the calculation results of the physical model and the data-driven model by synchronously and independently calculating the geometric morphological features based on the wake detection results includes: Based on the geometric morphological features of the wake detection results, the data is input into a physical model constructed from the Kelvin wake hydrodynamic characteristics. The heading is determined by the extension direction of the wake rotation bounding box or the main axis direction of the pixel mask distribution. The speed is determined by combining the wake shear wave wavelength, half angle, and timing features. Based on the geometric features of the wake detection results, a wake feature map is extracted and input into a data-driven model constructed by a convolutional neural network model to obtain the heading and speed results output by the data-driven model. By combining the heading and speed results output by the physical model and the data-driven model, the calculation results of the physical model and the data-driven model are obtained.

[0065] In this embodiment, the geometric features of the wake detection results (such as the wake principal axis direction, the wake extension direction, the shear wave wavelength, the half angle, and the continuous change trend of multiple frames) can be input into the physical model constructed by the Kelvin wake hydrodynamic characteristics. Since the overall extension direction of the wake can approximately reflect the ship's heading, the heading can be determined by the extension direction of the wake rotating bounding box or the principal axis direction of the pixel mask distribution. The wake wave spacing, shape, and angle can be used to further estimate the speed. Therefore, the speed can be determined by combining the wake shear wave wavelength, half angle, and timing features.

[0066] Furthermore, the wake feature map is extracted from the geometric morphological features of the wake detection results and input into the data-driven model that has been pre-trained by the convolutional neural network model. The wake features are extracted through the convolutional layer and then the heading and speed are directly output as the output results through the fully connected layer.

[0067] In practical applications, multiple sets of samples can be constructed by using multiple different historical wake detection results as samples and labeling the actual heading and speed as sample labels. These samples serve as training datasets, which are then used to pre-train the data-driven model until the model converges.

[0068] Therefore, this application achieves the reverse inference of the ship's motion state from the wake detection results by using a dual-path model to independently calculate the wake detection results and dynamically fuse the calculation results based on the wake detection confidence, sea state environment, and calculation result deviation. This serves as the data basis for subsequent cross-comparison with AIS data.

[0069] Step 400: Based on the motion parameters of the target vessel, and combined with the data reported by the Automatic Identification System of the target vessel, a consistency comparison is performed to determine the abnormal identification result of the target vessel's wake detection.

[0070] In this embodiment of the application, the motion parameters of the target ship obtained by inversion calculation are compared with the data reported by the target ship's Automatic Identification System (AIS) for consistency comparison. This can accurately identify abnormal behaviors such as fake course and fake speed, effectively overcoming the shortcomings of traditional AIS monitoring that cannot adapt to scenarios where AIS is turned off, blocked, or forged.

[0071] Therefore, this application detects wakes in remote sensing images and determines ship motion parameters through parameter inversion using a dual-path model. These parameters are then cross-referenced with parameters reported by AIS to effectively identify the authenticity of AIS data and accurately detect abnormal ship behavior.

[0072] Specifically, as an optional implementation, the step of performing a consistency comparison based on the motion parameters of the target vessel and the data reported by the target vessel's Automatic Identification System (AIS) to determine the abnormal identification result of the target vessel's wake detection includes: Based on the motion parameters of the target vessel, and according to the time window and spatial range, the target is matched with the automatic identification system for ships. If the motion parameters of the target vessel fail to match the target of the Automatic Identification System (AIS), the abnormal identification result of the target vessel's wake detection is determined to be that the AIS is shut down as a dark vessel. If the motion parameters of the target vessel successfully match the target of the Automatic Identification System (AIS), and the deviation between the motion parameters of the target vessel and the data reported by the AIS is greater than a preset deviation threshold, the abnormal identification result of the target vessel's wake detection is determined to be data fraud. If the motion parameters of the target vessel successfully match the target of the Automatic Identification System (AIS), and the deviation between the motion parameters of the target vessel and the data reported by the AIS is less than or equal to the preset deviation threshold, the motion parameters of the target vessel are detected and matched based on preset abnormal behavior rules, thereby determining the abnormal identification result of the target vessel's wake detection.

[0073] In the embodiments of this application, please refer to Figure 5 , Figure 5 This is a flowchart illustrating the process of detecting and identifying anomalies in ship wakes provided in this application. By accessing the Automatic Identification System (AIS) to obtain the reported parameters, this application considers that the reported parameters of AIS carry latitude, longitude, speed, and heading, while the remote sensing image is located at the target ship in the image. Therefore, it is necessary to use a customized time window and spatial range to achieve the association and matching between the remote sensing wake target and the AIS ship target.

[0074] Furthermore, if the motion parameters of the target vessel obtained through inversion are valid but cannot be successfully matched with the target of the Automatic Identification System (AIS), it can be determined that the target vessel's AIS system is turned off, and the abnormal identification result of the target vessel's wake detection is determined to be a dark vessel with the AIS system turned off.

[0075] Furthermore, after spatiotemporal correlation matching, the target vessel's motion parameters are successfully matched with the target of the Automatic Identification System (AIS), establishing a correspondence. The retrieved heading and speed data are then compared with the heading and speed data reported by AIS to calculate the deviation. The relationship between this deviation and a preset deviation threshold is further compared. The preset deviation threshold can be pre-set according to specific usage requirements, for example, a heading deviation threshold of 5°–15° and a speed deviation threshold of 2–5 knots. This threshold can be updated adaptively based on different sea areas (nearshore / offshore) and vessel types (cargo ships / fishing vessels / speedboats).

[0076] Furthermore, if the deviation between the target vessel's motion parameters and the AIS-reported data is greater than a preset deviation threshold, the abnormal identification result of the target vessel's wake detection is determined to be data fraud; if the deviation between the target vessel's motion parameters and the AIS-reported data is less than or equal to the preset deviation threshold, the motion parameters of the target vessel can be further detected and matched according to preset abnormal behavior rules, thereby determining the abnormal identification result of the target vessel's wake detection.

[0077] Specifically, the pre-defined abnormal behavior rules can be defined based on typical abnormal behavior characteristics, such as frequent small-angle turns in the ship's course, ship speed less than a preset threshold, and can be further combined with the target ship's geographical location information to determine that the ship may be engaging in abnormal fishing behavior; ship course drifting randomly with ocean currents and deviating from the main channel, ship speed less than a preset threshold, can determine that the ship may be engaging in distress abnormal behavior.

[0078] Furthermore, if the deviation between the ship's motion parameters and the data reported by AIS is small, and no preset abnormal behavior rule is matched, it can be determined that the abnormal identification result of the target ship's wake detection is no abnormal behavior.

[0079] Finally, the system outputs the results of the detected ship wake anomaly identification and can issue alarms based on specific abnormal behaviors, so that relevant management personnel can conduct secondary confirmation and anomaly handling.

[0080] Therefore, this application uses the motion parameters obtained by wake inversion as an independent true value benchmark, cross-compares them with the parameters reported by AIS, and combines them with a preset deviation threshold to accurately identify abnormal behaviors such as heading fraud and speed fraud. Furthermore, it can achieve intelligent identification of multiple types of abnormal behaviors based on preset abnormal behavior rules, effectively meeting the needs of marine management and control and reducing the cost of manual judgment.

[0081] In practical applications, during the anomaly detection process, the time-series filtering unit can cache multiple consecutive frames of ship motion parameters and detection confidence to construct a time-series feature sequence. Then, the time-series filtering can be used to identify the detection results of multiple frames. An alarm result is output only when the anomaly features are maintained for a preset number of frames (e.g., 3 frames), which can effectively suppress false alarms caused by single-frame noise.

[0082] For example, by acquiring SAR remote sensing images of the target vessel and performing image preprocessing to remove cloud interference and noise, a rotating target detection network model is used to detect sea surface wake targets and filter low-confidence samples. These images are then input into a physical model and a data-driven model, where independent computation is used to retrieve the vessel's true heading and speed. Furthermore, AIS data is accessed to perform spatiotemporal correlation matching. The retrieved parameters are compared with the parameters reported by AIS, and combined with the vessel's geographical location information, it is determined that the target vessel is engaging in abnormal fishing activities. After time-series filtering confirms that the abnormality has lasted for four frames, an alarm is triggered, and relevant data is pushed to the management platform, achieving accurate identification and rapid processing of abnormal behavior.

[0083] The following section provides a detailed explanation and description of the ship wake detection and identification method provided in this invention, in conjunction with a specific application implementation process: This application provides a method for detecting and identifying ship wakes. This method can be applied to stage lighting control scenarios. By detecting wakes in remote sensing images and determining ship motion parameters through parameter inversion using a dual-path model, the method can effectively identify the authenticity of AIS data and accurately detect abnormal ship behavior.

[0084] Specifically, remote sensing images of the target vessel, including its wake, can be obtained by accessing various image data sources such as satellites, drones, and shore-based observation stations. Radiometric calibration, geometric correction, and cropping normalization are then performed on the remote sensing images of the target vessel to complete the preliminary preprocessing of the original remote sensing images.

[0085] Furthermore, for SAR images, speckle suppression and multi-view processing can be performed to reduce speckle noise in remote sensing images; for optical remote sensing images, atmospheric correction and cloud masking can be performed to reduce cloud and atmospheric interference and improve image quality.

[0086] Furthermore, by inputting the remote sensing image of the target ship into a pre-trained rotating target detection network model, the rotating target detection network model extracts features from the remote sensing image of the target ship, identifies and detects the wake region of the ship on the sea surface, and outputs the wake detection results. Optionally, the rotating target detection network model can be constructed based on the Transformer architecture or an improved YOLOv8 architecture.

[0087] Furthermore, based on the input wake detection results, geometric features are acquired, such as wake principal axis direction, wake extension direction, shear wave wavelength, half angle, and multi-frame continuous change trend, etc. These features are then processed synchronously and independently by the physical model and the data-driven model to obtain the calculation results of the physical model and the data-driven model. Further, based on the calculation results of the physical model and the data-driven model, and combined with the wake detection confidence, calculation result deviation, and sea state interference level, collaborative redundant calculation is used to output the motion parameters and confidence level of the target vessel. The motion parameters of the target vessel specifically include heading and speed.

[0088] The geometric features of the wake detection results (such as the wake's main axis direction, the wake's extension direction, the shear wave wavelength, the half angle, and the continuous change trend over multiple frames) can be input into a physical model constructed from the Kelvin wake hydrodynamic characteristics. Since the overall extension direction of the wake can approximately reflect the ship's heading, the heading can be determined by the extension direction of the wake's rotating bounding box or the main axis direction of the pixel mask distribution. The wake wave spacing, shape, and angle can be used to further estimate the speed. Therefore, the speed can be determined by combining the wake's shear wave wavelength, half angle, and timing features.

[0089] Furthermore, the wake feature map is extracted from the geometric morphological features of the wake detection results and input into the data-driven model that has been pre-trained by the convolutional neural network model. The wake features are extracted through the convolutional layer and then the heading and speed are directly output as the output results through the fully connected layer.

[0090] Finally, by accessing the Automatic Identification System (AIS) to obtain the reported parameters, this application takes into account that the AIS reported parameters are reports of targets carrying latitude, longitude, speed, and heading, while the remote sensing images are located on the target ships in the images. Therefore, it is necessary to use a customized time window and spatial range to achieve the association and matching between the remote sensing wake targets and the AIS ship targets.

[0091] Furthermore, if the motion parameters of the target vessel obtained through inversion are valid but cannot be successfully matched with the target of the Automatic Identification System (AIS), it can be determined that the target vessel's AIS system is turned off, and the abnormal identification result of the target vessel's wake detection is determined to be a dark vessel with the AIS system turned off.

[0092] Furthermore, after spatiotemporal correlation matching, the target vessel's motion parameters are successfully matched with the target of the Automatic Identification System (AIS), establishing a correspondence. The retrieved heading and speed data are then compared with the heading and speed data reported by AIS to calculate the deviation. The relationship between this deviation and a preset deviation threshold is further compared. The preset deviation threshold can be pre-set according to specific usage requirements, for example, a heading deviation threshold of 5°–15° and a speed deviation threshold of 2–5 knots. This threshold can be updated adaptively based on different sea areas (nearshore / offshore) and vessel types (cargo ships / fishing vessels / speedboats).

[0093] Furthermore, if the deviation between the target vessel's motion parameters and the AIS-reported data is greater than a preset deviation threshold, the abnormal identification result of the target vessel's wake detection is determined to be data fraud; if the deviation between the target vessel's motion parameters and the AIS-reported data is less than or equal to the preset deviation threshold, the motion parameters of the target vessel can be further detected and matched according to preset abnormal behavior rules, thereby determining the abnormal identification result of the target vessel's wake detection.

[0094] Furthermore, during the anomaly detection process, the timing filter unit can cache multiple consecutive frames of ship motion parameters and detection confidence to construct a timing feature sequence. Then, the timing filter can identify the detection results of multiple frames. The alarm result is output only when the anomaly features are maintained for a preset number of frames (e.g., 3 frames), which can effectively suppress false alarms caused by single-frame noise.

[0095] Please see Figure 6 , Figure 6 This is a schematic diagram of a ship wake detection and identification device provided in an embodiment of this application. This application also provides a ship wake detection and identification device that can implement the above-mentioned ship wake detection and identification method. The device includes: Image acquisition module 610 is used to acquire remote sensing images of the target ship; The wake detection module 620 is used to identify and detect the wake region of the target vessel on the sea surface and output the wake detection result based on the remote sensing image of the target vessel. The parameter inversion module 630 is used to obtain the motion parameters of the target ship by inverting the wake detection results through a physical model and a data-driven model. The anomaly detection module 640 is used to perform a consistency comparison based on the motion parameters of the target vessel and the data reported by the target vessel's Automatic Identification System to determine the anomaly detection result of the target vessel's wake. The physical model is constructed based on the hydrodynamic properties of the wake; the data-driven model is constructed based on the convolutional neural network model.

[0096] It is understood that the content of the above method embodiments is applicable to the present device embodiments. The specific functions implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0097] Please see Figure 7 , Figure 7 This is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of this application. The electronic device includes: The processor 701 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application. The memory 702 can be implemented as a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 702 can store the operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 702 and is called and executed by the processor 701 using the methods described in the embodiments of this application. The input / output interface 703 is used to implement information input and output; The communication interface 704 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.). Bus 705 transmits information between various components of the device (e.g., processor 701, memory 702, input / output interface 703, and communication interface 704); The processor 701, memory 702, input / output interface 703, and communication interface 704 are connected to each other within the device via bus 705.

[0098] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method.

[0099] It is understood that the content of the above method embodiments is applicable to this storage medium embodiment. The specific functions implemented in this storage medium embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0100] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0101] It is understood that the content of the above method embodiments is applicable to the embodiments of this program product. The specific functions implemented by the embodiments of this program product are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0102] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0103] This application provides a method and related equipment for detecting and identifying ship wakes. It detects wakes in remote sensing images, determines ship motion parameters through parameter inversion using a dual-path model, and then cross-compares these parameters with those reported by AIS to effectively identify the authenticity of AIS data and accurately detect abnormal ship behavior.

[0104] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

[0105] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0106] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0107] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0108] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0109] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0110] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0111] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0112] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0113] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0114] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.

Claims

1. A method for detecting and identifying ship wakes, characterized in that, The method includes the following steps: Acquire remote sensing images of the target vessel; Based on the remote sensing image of the target vessel, identify and detect the vessel's wake region on the sea surface and output the wake detection results; Based on the wake detection results, the motion parameters of the target vessel are obtained by inversion using a physical model and a data-driven model. Based on the motion parameters of the target vessel, and combined with the data reported by the target vessel's Automatic Identification System, a consistency comparison is performed to determine the abnormal identification result of the target vessel's wake detection. The physical model is constructed based on the hydrodynamic properties of the wake; the data-driven model is constructed based on the convolutional neural network model.

2. The method according to claim 1, characterized in that, The process of obtaining the motion parameters of the target vessel by inverting the wake detection results through a physical model and a data-driven model includes: Based on the geometric morphological features of the wake detection results, the physical model and the data-driven model are operated synchronously and independently to obtain the calculation results of the physical model and the data-driven model. Based on the calculation results of the physical model and the data-driven model, combined with the wake detection confidence, calculation result deviation and sea state interference, the motion parameters and confidence level of the target ship are output; wherein, the motion parameters include heading and speed.

3. The method according to claim 2, characterized in that, The geometric morphological features based on the wake detection results are obtained through synchronous and independent computation using a physical model and a data-driven model, including: Based on the geometric morphological features of the wake detection results, the data is input into a physical model constructed from the Kelvin wake hydrodynamic characteristics. The heading is determined by the extension direction of the wake rotation bounding box or the main axis direction of the pixel mask distribution. The speed is determined by combining the wake shear wave wavelength, half angle, and timing features. Based on the geometric features of the wake detection results, a wake feature map is extracted and input into a data-driven model constructed by a convolutional neural network model to obtain the heading and speed results output by the data-driven model. By combining the heading and speed results output by the physical model and the data-driven model, the calculation results of the physical model and the data-driven model are obtained.

4. The method according to claim 1, characterized in that, The process of identifying and detecting the wake region of a vessel on the sea surface based on the remote sensing image of the target vessel and outputting the wake detection result includes: The remote sensing image of the target ship is input into the rotating target detection network model, so that the rotating target detection network model can extract features from the remote sensing image, identify and detect the ship's wake region on the sea surface, and output the wake detection result. The rotating target detection network model is built based on the YOLOv8 architecture and includes a multi-directional perceptual convolutional module and a frequency-space fusion attention module. The multi-directional perceptual convolutional module includes multiple sets of convolutional kernels with different orientations for capturing trail texture features from multiple angles. The frequency-space fusion attention module adaptively fuses frequency domain and spatial domain features through a multi-head attention mechanism.

5. The method according to claim 1, characterized in that, The step of performing a consistency comparison based on the motion parameters of the target vessel and the data reported by the target vessel's Automatic Identification System (AIS) to determine the abnormal identification result of the target vessel's wake detection includes: Based on the motion parameters of the target vessel, and according to the time window and spatial range, the target is matched with the automatic identification system for ships. If the motion parameters of the target vessel fail to match the target of the Automatic Identification System (AIS), the abnormal identification result of the target vessel's wake detection is determined to be that the AIS is shut down as a dark vessel. If the motion parameters of the target vessel successfully match the target of the Automatic Identification System (AIS), and the deviation between the motion parameters of the target vessel and the data reported by the AIS is greater than a preset deviation threshold, the abnormal identification result of the target vessel's wake detection is determined to be data fraud. If the motion parameters of the target vessel successfully match the target of the Automatic Identification System (AIS), and the deviation between the motion parameters of the target vessel and the data reported by the AIS is less than or equal to the preset deviation threshold, the motion parameters of the target vessel are detected and matched based on preset abnormal behavior rules, thereby determining the abnormal identification result of the target vessel's wake detection.

6. The method according to claim 1, characterized in that, After acquiring the remote sensing image of the target ship, the method further includes: Perform radiometric calibration, geometric correction, and cropping normalization on the remote sensing images of the target ship; When the remote sensing image of the target ship is a synthetic aperture radar image, speckle suppression and multi-view processing are performed on the remote sensing image; When the remote sensing image of the target ship is an optical remote sensing image, atmospheric correction and cloud area masking are performed on the remote sensing image.

7. A device for detecting and identifying ship wakes, characterized in that, The device includes: The image acquisition module is used to acquire remote sensing images of the target ship; The wake detection module is used to identify and detect the wake region of the target vessel on the sea surface and output the wake detection results based on the remote sensing image of the target vessel. The parameter inversion module is used to obtain the motion parameters of the target ship by inverting the wake detection results through a physical model and a data-driven model. An anomaly detection module is used to perform a consistency comparison based on the motion parameters of the target vessel and the data reported by the target vessel's Automatic Identification System to determine the anomaly detection result of the target vessel's wake. The physical model is constructed based on the hydrodynamic properties of the wake; the data-driven model is constructed based on the convolutional neural network model.

8. An electronic device, characterized in that, The electronic device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method according to any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 6.

Images