Image recognition-based marine safety evaluation method, device, equipment and medium

Abstract

The application discloses a kind of marine safety evaluation methods, devices and equipment based on image recognition and medium, the method comprises: from the multi-view sea state image data detected to be identified object, and the object to be identified is matched with preset object, then the first satellite positioning information of current ship equipment is obtained, and the preset authority information of the object to be identified, according to relative position and first satellite positioning information, determine the second satellite positioning information of the object to be identified, if the second satellite positioning information matches the first preset sea area, based on multi-view sea area image data, determine the motion state information of the object to be identified, evaluate motion state information, obtain the safe navigation value of the object to be identified, if safety level is less than set level threshold, then based on ship equipment sends early warning alarm. So as to reduce the labor cost increased by manual detection, and improve the safety detection efficiency.

Description

Technical Field

[0001] This invention relates to the field of image recognition technology, specifically to a marine safety assessment method, apparatus, equipment, and medium based on image recognition. Background Technology

[0002] Currently, traditional marine safety protection solutions mainly rely on manual monitoring of sensor data or ordinary video surveillance. The deployed marine monitoring facilities can only provide data information, lacking a direct understanding of the situation in case of anomalies. Video surveillance is primarily "passive," requiring constant monitoring by on-duty personnel. Most of the time, it's only suitable for reviewing video footage for event tracing and cannot trigger alarms immediately upon the occurrence of danger, preventing timely response from relevant personnel. Furthermore, the safety supervision of some small facilities is inadequate. For monitoring scenarios covering a wide geographical area, marine safety inspections require significant manpower and effort. Summary of the Invention

[0003] To address the high cost of maritime safety inspections in existing technologies, this invention provides a method, apparatus, equipment, and medium for maritime safety assessment based on image recognition.

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] According to a first aspect of the present invention, a marine safety assessment method based on image recognition is provided, applied to ship equipment, the method comprising:

[0006] Acquire multi-view sea state image data within a set range of the ship equipment collected by the multi-view camera sensor; if an object to be identified is detected from the multi-view sea state image data, determine whether the object to be identified matches a preset object.

[0007] If the object to be identified matches the preset object, then the current first satellite positioning information of the ship equipment and the preset permission information of the object to be identified are obtained. The preset permission indicated by the preset permission information includes the permission of the object to be identified to enter the first preset sea area corresponding to the first satellite positioning information. The preset sea area authorized by the preset object is a part of the preset total sea area.

[0008] In response to the preset permission information, the relative position of the ship equipment and the object to be identified is identified based on the multi-view sea state image data, and the second satellite positioning information of the object to be identified is determined based on the relative position and the first satellite positioning information;

[0009] If the second satellite positioning information matches the first preset sea area, the motion state information of the object to be identified is determined based on the multi-view sea area image data; the motion state information is evaluated to obtain the safe navigation value of the object to be identified; and the safety level of the object to be identified is determined by the safe navigation value and the preset weight value of the object to be identified.

[0010] If the safety level is less than the set threshold, an early warning alarm will be issued based on the ship's equipment.

[0011] Optionally, determining the motion state information of the object to be identified based on the multi-view sea area image data includes:

[0012] Based on the multi-view marine image data, the type and attitude features of the object to be identified are identified, and a finite element model of the object to be identified is generated. The finite element model is used to simulate the motion features of the object to be identified.

[0013] Based on the finite element model and the multi-dimensional modeling module installed in the ship equipment, the motion characteristics of the object to be identified in the current marine scene are simulated to generate motion state information of the object to be identified relative to the ship equipment.

[0014] Optionally, evaluating the motion state information to obtain a safe navigation value for the object to be identified includes:

[0015] Based on the preset safety evaluation module installed in the ship equipment and the relative position, the motion state information of the simulated ship equipment is evaluated for navigation stability and navigation safety, so as to generate navigation stability information of the object to be identified.

[0016] Based on the motion state information and the navigation stability information, the motion trajectory of the object to be identified within a preset time range is predicted, and the predicted trajectory information of the object to be identified is generated.

[0017] Based on the predicted trajectory information, the safe navigation value of the object to be identified is generated.

[0018] Optionally, after acquiring the multi-view sea state image data within the set range of the ship's equipment collected by the multi-view camera sensor, the method further includes:

[0019] Based on the BEV bird's-eye view module installed in the ship's equipment, the multi-view sea state image data is converted into BEV space to generate BEV data centered on the ship's equipment.

[0020] By using a preset feature detection model, the feature objects in the BEV data are identified to determine whether the object to be identified exists in the BEV data.

[0021] Optionally, the step of identifying feature objects in the BEV data using a preset feature detection model to determine whether the object to be identified exists in the BEV data includes:

[0022] Based on the preset shooting angle of the multi-view camera sensor corresponding to the BEV data, the BEV data is fused to generate BEV image frames of the ship equipment within the set range.

[0023] The preset feature detection module determines multiple grayscale differences between adjacent pixels in the BEV image frame.

[0024] If multiple target grayscale differences among the multiple grayscale differences are greater than a preset threshold, then it is determined that the object to be identified exists in the multi-view sea state image data;

[0025] If all the grayscale differences are less than or equal to the preset threshold, then it is determined that the object to be identified does not exist in the multi-view sea state image data.

[0026] Optionally, determining multiple grayscale differences between adjacent pixels in the BEV image frame using the preset feature detection module includes:

[0027] The BEV image frame is denoised using the preset feature detection module and ocean image feature parameters to generate an initial image frame.

[0028] The plurality of grayscale differences are generated based on the gradient information and / or brightness difference information between each adjacent pixel in the initial image frame.

[0029] Optionally, determining the safety level of the object to be identified using the safe navigation values ​​and the preset weight values ​​of the object to be identified includes:

[0030] Determine the object type of the object to be identified;

[0031] Determine the preset weight value that matches the object type from the preset weight mapping table;

[0032] The target safety value is generated by multiplying the safe navigation value by the preset weight value.

[0033] Based on a preset security level mapping table, the security level corresponding to the target security value is determined.

[0034] According to a second aspect of the present invention, an image recognition-based marine safety assessment device is provided, applied to ship equipment, the device comprising:

[0035] The acquisition module is used to acquire multi-view sea state image data within a set range of the ship equipment collected by the multi-view camera sensor. If an object to be identified is detected from the multi-view sea state image data, it is determined whether the object to be identified matches a preset object.

[0036] The first determination module is used to obtain the current first satellite positioning information of the ship equipment and the preset permission information of the object to be identified if the object to be identified matches the preset object. The preset permission indicated by the preset permission information includes the permission of the object to be identified to enter the first preset sea area corresponding to the first satellite positioning information. The preset sea area authorized by the preset object is a part of the preset total sea area.

[0037] The determination module is used to respond to the preset permission information, identify the relative position of the ship equipment and the object to be identified based on the multi-view sea state image data, and determine the second satellite positioning information of the object to be identified based on the relative position and the first satellite positioning information;

[0038] The second determination module is used to determine the motion state information of the object to be identified based on the multi-view sea area image data if the second satellite positioning information matches the first preset sea area; evaluate the motion state information to obtain the safe navigation value of the object to be identified; and determine the safety level of the object to be identified by the safe navigation value and the preset weight value of the object to be identified.

[0039] The execution module is used to issue a warning alarm based on the ship's equipment if the safety level is less than a set level threshold.

[0040] According to a third aspect of the present invention, an electronic device is provided, comprising:

[0041] A memory on which computer programs are stored;

[0042] A processor is configured to execute the computer program in the memory to implement the steps of the image recognition-based marine safety assessment method according to any one of the first aspects of this disclosure.

[0043] According to a fourth aspect of the present invention, a computer-readable storage medium is provided having computer program instructions stored thereon, which, when executed by a processor, implement the steps of the method described in any one of the first aspects of the present disclosure.

[0044] This invention provides a marine safety assessment method, apparatus, equipment, and medium based on image recognition, which has the following advantages compared with existing technologies:

[0045] Using the above method, multi-view sea state image data within a set range of the ship's equipment is acquired from the multi-view camera sensor. If an object to be identified is detected in the multi-view sea state image data, it is determined whether the object to be identified matches a preset object. If the object to be identified matches the preset object, the ship's current first satellite positioning information and the preset permission information of the object to be identified are acquired. The preset permissions indicated by the preset permission information include the permission of the object to be identified to enter the first preset sea area corresponding to the first satellite positioning information. The preset sea area authorized by the preset object is a part of the preset total sea area. In response to the preset permission information, the relative position of the ship's equipment and the object to be identified is identified based on the multi-view sea state image data. Based on the relative position and the first satellite positioning information, the second satellite positioning information of the object to be identified is determined. If the second satellite positioning information matches the first preset sea area, the motion state information of the object to be identified is determined based on the multi-view sea area image data. The motion state information is evaluated to obtain the safe navigation value of the object to be identified. The safety level of the object to be identified is determined by the safe navigation value and the preset weight value of the object to be identified. If the safety level is less than the set level threshold, an early warning alarm is issued based on the ship's equipment. This allows for the automatic detection of objects around ship equipment, and the automatic issuance of warnings when the safety level of an object is below a threshold, reducing the manpower costs associated with manual inspection and improving safety inspection efficiency. Attached Figure Description

[0046] Figure 1 This is a flowchart illustrating an image recognition-based marine safety assessment method according to an exemplary embodiment.

[0047] Figure 2 This is a flowchart illustrating a method for detecting an object to be identified according to an exemplary embodiment.

[0048] Figure 3 This is an image recognition-based marine safety assessment device illustrated according to an exemplary embodiment. Detailed Implementation

[0049] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0050] Figure 1 This is a flowchart illustrating an image recognition-based marine safety assessment method according to an exemplary embodiment, such as... Figure 1 As shown, this method is applied to ship equipment and includes the following steps.

[0051] Step S11: Obtain multi-view sea state image data within the set range of the ship equipment collected by the multi-view camera sensor, and determine whether the object to be identified in the multi-view sea state image data matches the preset object.

[0052] For example, multi-view sea state image data within a set range of the ship's equipment is acquired by a multi-view camera sensor. If an object to be identified is detected from the multi-view sea state image data, it is determined whether the object to be identified matches a preset object.

[0053] In some implementations, camera sensors can be set up at multiple perspectives of the ship to collect sea state image data from different perspectives. Each perspective can cover the ship's omnidirectional angles, thereby determining the environmental information of the current environment of the ship based on the multi-view sea state image data.

[0054] Step S12: If the object to be identified matches the preset object, then obtain the current first satellite positioning information of the ship equipment and the preset permission information of the object to be identified.

[0055] For example, if the object to be identified matches the preset object, then the current first satellite positioning information of the ship's equipment and the preset permission information of the object to be identified are obtained. The preset permission information indicates preset permissions including the permission for the object to be identified to enter the first preset sea area corresponding to the first satellite positioning information. The preset sea area authorized by the preset object is a part of the preset total sea area. For example, in some embodiments, image feature information of the preset object can be pre-stored in the ship's equipment. An image recognition model is used to identify the object to be identified, generating an image feature vector of the object. The image feature information and the image feature vector are compared to determine the similarity between the two, thereby determining whether the object to be identified matches the preset object.

[0056] Step S13: Determine the second satellite positioning information of the object to be identified based on the relative position and the first satellite positioning information.

[0057] For example, in response to the preset permission information, the relative position of the ship equipment and the object to be identified is identified based on the multi-view sea state image data, and the second satellite positioning information of the object to be identified is determined based on the relative position and the first satellite positioning information.

[0058] Step S14: Determine the safety level of the object to be identified by using the safe navigation values ​​and the preset weight values ​​of the object to be identified.

[0059] For example, if the second satellite positioning information matches the first preset sea area, the motion state information of the object to be identified is determined based on the multi-view sea area image data; the motion state information is evaluated to obtain the safe navigation value of the object to be identified; and the safety level of the object to be identified is determined by the safe navigation value and the preset weight value of the object to be identified. For example, a mapping table between safety level and safe navigation score can be set in the ship's equipment. This table sets multiple safe navigation score ranges and mapping relationships between multiple safety levels. By multiplying the safe navigation value by the preset weight value, a safe navigation score is generated. The safety level of the ship's equipment is determined by consulting this table.

[0060] Optionally, in some embodiments, step S14 above includes:

[0061] Determine the object type of the object to be identified;

[0062] Determine the preset weight value that matches the object type from the preset weight mapping table;

[0063] The target safety value is generated by multiplying the safe navigation value by the preset weight value.

[0064] Based on a preset security level mapping table, the security level corresponding to the target security value is determined.

[0065] Optionally, in some embodiments, step S14 above includes:

[0066] Based on the multi-view marine image data, the type and attitude features of the object to be identified are identified, and a finite element model of the object to be identified is generated. The finite element model is used to simulate the motion features of the object to be identified.

[0067] Based on the finite element model and the multi-dimensional modeling module installed in the ship equipment, the motion characteristics of the object to be identified in the current marine scene are simulated to generate motion state information of the object to be identified relative to the ship equipment.

[0068] Optionally, in some embodiments, the above steps of evaluating the motion state information to obtain the safe navigation value of the object to be identified may further include:

[0069] Based on the preset safety evaluation module installed in the ship equipment and the relative position, the motion state information of the simulated ship equipment is evaluated for navigation stability and navigation safety, so as to generate navigation stability information of the object to be identified.

[0070] Based on the motion state information and the navigation stability information, the motion trajectory of the object to be identified within a preset time range is predicted, and the predicted trajectory information of the object to be identified is generated.

[0071] Based on the predicted trajectory information, the safe navigation value of the object to be identified is generated.

[0072] Step S15: If the safety level is less than the set level threshold, then issue a warning alarm based on the ship's equipment.

[0073] Using the above method, multi-view sea state image data within a set range of the ship's equipment is acquired from the multi-view camera sensor. If an object to be identified is detected in the multi-view sea state image data, it is determined whether the object to be identified matches a preset object. If the object to be identified matches the preset object, the ship's current first satellite positioning information and the preset permission information of the object to be identified are acquired. The preset permissions indicated by the preset permission information include the permission of the object to be identified to enter the first preset sea area corresponding to the first satellite positioning information. The preset sea area authorized by the preset object is a part of the preset total sea area. In response to the preset permission information, the relative position of the ship's equipment and the object to be identified is identified based on the multi-view sea state image data. Based on the relative position and the first satellite positioning information, the second satellite positioning information of the object to be identified is determined. If the second satellite positioning information matches the first preset sea area, the motion state information of the object to be identified is determined based on the multi-view sea area image data. The motion state information is evaluated to obtain the safe navigation value of the object to be identified. The safety level of the object to be identified is determined by the safe navigation value and the preset weight value of the object to be identified. If the safety level is less than the set level threshold, an early warning alarm is issued based on the ship's equipment. This allows for the automatic detection of objects around ship equipment, and the automatic issuance of warnings when the safety level of an object is below a threshold, reducing the manpower costs associated with manual inspection and improving safety inspection efficiency.

[0074] Figure 2 This is a flowchart illustrating a method for detecting an object to be identified according to an exemplary embodiment, such as... Figure 2 As shown, the detection method includes the following steps.

[0075] Step S21: Based on the BEV bird's-eye view module installed in the ship's equipment, the multi-view sea state image data is converted into BEV space to generate BEV data centered on the ship's equipment.

[0076] For example, in this embodiment, to detect whether there is an object to be identified within a set range of the ship's equipment, detection can be performed based on a BEV (Browser Active Vehicle) bird's-eye view to determine the information of the object to be identified. Therefore, the multi-view sea state image data collected can be converted based on the set positions of each multi-view camera sensor in the ship's equipment, converting the multi-view sea state image data into BEV space to generate ship-centered BEV data.

[0077] Step S22: Identify the feature objects in the BEV data using a preset feature detection model to determine whether there are any objects to be identified in the BEV data.

[0078] Optionally, in some embodiments, step S22 above includes:

[0079] Based on the preset shooting angle of the multi-view camera sensor corresponding to the BEV data, the BEV data is fused to generate BEV image frames of the ship equipment within the set range.

[0080] The preset feature detection module determines multiple grayscale differences between adjacent pixels in the BEV image frame.

[0081] If multiple target grayscale differences among the multiple grayscale differences are greater than a preset threshold, then it is determined that the object to be identified exists in the multi-view sea state image data;

[0082] If all the grayscale differences are less than or equal to the preset threshold, then it is determined that the object to be identified does not exist in the multi-view sea state image data.

[0083] Optionally, in some embodiments, the above steps, through the preset feature detection module, determine multiple grayscale differences between adjacent pixels in the BEV image frame, including:

[0084] The BEV image frame is denoised using the preset feature detection module and ocean image feature parameters to generate an initial image frame.

[0085] The plurality of grayscale differences are generated based on the gradient information and / or brightness difference information between each adjacent pixel in the initial image frame.

[0086] Using the above method, the target objects within a preset range are detected based on ocean image frames in the BEV space of the transformation value, and the presence of the target objects in the corresponding image is determined based on the gray-scale difference between adjacent pixels in the image frame, thereby improving the detection accuracy of the target objects.

[0087] Figure 3This is an exemplary embodiment illustrating a marine safety assessment device based on image recognition, such as... Figure 3 As shown, the device 100 includes: an acquisition module 110, a first determination module 120, a determination module 130, a second determination module 140, and an execution module 150.

[0088] The acquisition module 110 is used to acquire multi-view sea state image data within a set range of the ship equipment collected by the multi-view camera sensor. If an object to be identified is detected from the multi-view sea state image data, it is determined whether the object to be identified matches a preset object.

[0089] The first determination module 120 is used to obtain the current first satellite positioning information of the ship equipment and the preset permission information of the object to be identified if the object to be identified matches the preset object. The preset permission information indicates the preset permission including the permission of the object to be identified to enter the first preset sea area corresponding to the first satellite positioning information. The preset sea area authorized by the preset object is a part of the preset total sea area.

[0090] The determination module 130 is used to respond to the preset permission information, identify the relative position of the ship equipment and the object to be identified based on the multi-view sea state image data, and determine the second satellite positioning information of the object to be identified based on the relative position and the first satellite positioning information;

[0091] The second determination module 140 is used to determine the motion state information of the object to be identified based on the multi-view sea area image data if the second satellite positioning information matches the first preset sea area; evaluate the motion state information to obtain the safe navigation value of the object to be identified; and determine the safety level of the object to be identified by the safe navigation value and the preset weight value of the object to be identified.

[0092] The execution module 150 is used to issue a warning alarm based on the ship's equipment if the safety level is less than a set level threshold.

[0093] Optionally, the second determination module 140 includes:

[0094] The first generation submodule is used to identify the type and attitude features of the object to be identified based on the multi-view sea area image data, and generate a finite element model of the object to be identified. The finite element model is used to simulate the motion features of the object to be identified.

[0095] The second generation submodule is used to simulate the motion characteristics of the object to be identified in the current marine scene based on the finite element model and the multi-dimensional modeling module installed in the ship equipment, so as to generate motion state information of the object to be identified relative to the ship equipment.

[0096] Optionally, the second determination module 140 further includes:

[0097] The third generation submodule is used to evaluate the navigation stability and navigation safety of the simulated motion state information of the ship equipment based on the preset safety evaluation module installed in the ship equipment and the relative position, so as to generate the navigation stability information of the object to be identified.

[0098] The fourth generation submodule is used to predict the motion trajectory of the object to be identified within a preset time range based on the motion state information and the navigation stability information, and generate the predicted trajectory information of the object to be identified.

[0099] The fifth generation submodule is used to generate the safe navigation value of the object to be identified based on the predicted trajectory information.

[0100] Optionally, the device 100 further includes a conversion module, which includes:

[0101] The sixth generation submodule is used to convert the multi-view sea state image data into BEV space based on the BEV bird's-eye view module installed in the ship equipment, and generate BEV data centered on the ship equipment.

[0102] The determination submodule is used to identify feature objects in the BEV data using a preset feature detection model, and to determine whether the object to be identified exists in the BEV data.

[0103] Optionally, submodules are defined, including:

[0104] The generation unit is used to fuse the BEV data based on the preset shooting angle of the multi-view camera sensor corresponding to the BEV data, and generate BEV image frames of the ship equipment within the set range.

[0105] The determining unit is used to determine multiple grayscale differences between adjacent pixels in the BEV image frame through the preset feature detection module.

[0106] The first determination unit is used to determine that the object to be identified exists in the multi-view sea state image data if multiple target gray-scale differences among the multiple gray-scale differences are greater than a preset threshold.

[0107] The second determination unit is used to determine that the object to be identified does not exist in the multi-view sea state image data if all of the multiple grayscale differences are less than or equal to the preset threshold.

[0108] Optionally, the determining unit is used for:

[0109] The BEV image frame is denoised using the preset feature detection module and ocean image feature parameters to generate an initial image frame.

[0110] The plurality of grayscale differences are generated based on the gradient information and / or brightness difference information between each adjacent pixel in the initial image frame.

[0111] Optionally, the second determination module 140 is used for:

[0112] Determine the object type of the object to be identified;

[0113] Determine the preset weight value that matches the object type from the preset weight mapping table;

[0114] The target safety value is generated by multiplying the safe navigation value by the preset weight value.

[0115] Based on a preset security level mapping table, the security level corresponding to the target security value is determined.

[0116] Using the above method, multi-view sea state image data within a set range of the ship's equipment is acquired from the multi-view camera sensor. If an object to be identified is detected in the multi-view sea state image data, it is determined whether the object to be identified matches a preset object. If the object to be identified matches the preset object, the ship's current first satellite positioning information and the preset permission information of the object to be identified are acquired. The preset permissions indicated by the preset permission information include the permission of the object to be identified to enter the first preset sea area corresponding to the first satellite positioning information. The preset sea area authorized by the preset object is a part of the preset total sea area. In response to the preset permission information, the relative position of the ship's equipment and the object to be identified is identified based on the multi-view sea state image data. Based on the relative position and the first satellite positioning information, the second satellite positioning information of the object to be identified is determined. If the second satellite positioning information matches the first preset sea area, the motion state information of the object to be identified is determined based on the multi-view sea area image data. The motion state information is evaluated to obtain the safe navigation value of the object to be identified. The safety level of the object to be identified is determined by the safe navigation value and the preset weight value of the object to be identified. If the safety level is less than the set level threshold, an early warning alarm is issued based on the ship's equipment. This allows for the automatic detection of objects around ship equipment, and the automatic issuance of warnings when the safety level of an object is below a threshold, reducing the manpower costs associated with manual inspection and improving safety inspection efficiency.

[0117] Based on the above-described preferred embodiments according to this application, and through the foregoing description, those skilled in the art can make various changes and modifications without departing from the technical concept of this application. The technical scope of this application is not limited to the contents of the specification.

[0118] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A marine safety assessment method based on image recognition, characterized in that, Applied to marine equipment, the method includes: Acquire multi-view sea state image data within a set range of the ship equipment collected by the multi-view camera sensor; if an object to be identified is detected from the multi-view sea state image data, determine whether the object to be identified matches a preset object. If the object to be identified matches the preset object, then the current first satellite positioning information of the ship equipment and the preset permission information of the object to be identified are obtained. The preset permission indicated by the preset permission information includes the permission of the object to be identified to enter the first preset sea area corresponding to the first satellite positioning information. The preset sea area authorized by the preset object is a part of the preset total sea area. In response to the preset permission information, the relative position of the ship equipment and the object to be identified is identified based on the multi-view sea state image data, and the second satellite positioning information of the object to be identified is determined based on the relative position and the first satellite positioning information; If the second satellite positioning information matches the first preset sea area, the motion state information of the object to be identified is determined based on the multi-view sea area image data; the motion state information is evaluated to obtain the safe navigation value of the object to be identified; and the safety level of the object to be identified is determined by the safe navigation value and the preset weight value of the object to be identified. If the safety level is less than the set threshold, an early warning alarm will be issued based on the ship's equipment. The step of determining the motion state information of the object to be identified based on the multi-view sea area image data includes: Based on the multi-view marine image data, the type and attitude features of the object to be identified are identified, and a finite element model of the object to be identified is generated. The finite element model is used to simulate the motion features of the object to be identified. Based on the finite element model and the multi-dimensional modeling module installed in the ship equipment, the motion characteristics of the object to be identified in the current marine scene are simulated to generate motion state information of the object to be identified relative to the ship equipment; The step of evaluating the motion state information to obtain the safe navigation value of the object to be identified includes: Based on the preset safety evaluation module installed in the ship equipment and the relative position, the motion state information of the simulated ship equipment is evaluated for navigation stability and navigation safety, so as to generate navigation stability information of the object to be identified. Based on the motion state information and the navigation stability information, the motion trajectory of the object to be identified within a preset time range is predicted, and the predicted trajectory information of the object to be identified is generated. Based on the predicted trajectory information, the safe navigation value of the object to be identified is generated.

2. The method according to claim 1, characterized in that, After acquiring the multi-view sea state image data within the set range of the ship's equipment collected by the multi-view camera sensor, the method further includes: Based on the BEV bird's-eye view module installed in the ship's equipment, the multi-view sea state image data is converted into BEV space to generate BEV data centered on the ship's equipment. By using a preset feature detection model, the feature objects in the BEV data are identified to determine whether the object to be identified exists in the BEV data.

3. The method according to claim 2, characterized in that, The step of identifying feature objects in the BEV data using a preset feature detection model to determine whether the object to be identified exists in the BEV data includes: Based on the preset shooting angle of the multi-view camera sensor corresponding to the BEV data, the BEV data is fused to generate BEV image frames of the ship equipment within the set range. The preset feature detection module determines multiple grayscale differences between adjacent pixels in the BEV image frame. If multiple target grayscale differences among the multiple grayscale differences are greater than a preset threshold, then it is determined that the object to be identified exists in the multi-view sea state image data; If all the grayscale differences are less than or equal to the preset threshold, then it is determined that the object to be identified does not exist in the multi-view sea state image data.

4. The method according to claim 3, characterized in that, The step of determining multiple grayscale differences between adjacent pixels in the BEV image frame through the preset feature detection module includes: The BEV image frame is denoised using the preset feature detection module and ocean image feature parameters to generate an initial image frame. The plurality of grayscale differences are generated based on the gradient information and / or brightness difference information between each adjacent pixel in the initial image frame.

5. The method according to any one of claims 1-4, characterized in that, The process of determining the safety level of the object to be identified using the safe navigation values ​​and the preset weight values ​​of the object to be identified includes: Determine the object type of the object to be identified; Determine the preset weight value that matches the object type from the preset weight mapping table; The target safety value is generated by multiplying the safe navigation value by the preset weight value. Based on a preset security level mapping table, the security level corresponding to the target security value is determined.

6. A marine safety assessment device based on image recognition, characterized in that, Applied to marine equipment, the device includes: The acquisition module is used to acquire multi-view sea state image data within a set range of the ship equipment collected by the multi-view camera sensor. If an object to be identified is detected from the multi-view sea state image data, it is determined whether the object to be identified matches a preset object. The first determination module is used to obtain the current first satellite positioning information of the ship equipment and the preset permission information of the object to be identified if the object to be identified matches the preset object. The preset permission indicated by the preset permission information includes the permission of the object to be identified to enter the first preset sea area corresponding to the first satellite positioning information. The preset sea area authorized by the preset object is a part of the preset total sea area. The determination module is used to respond to the preset permission information, identify the relative position of the ship equipment and the object to be identified based on the multi-view sea state image data, and determine the second satellite positioning information of the object to be identified based on the relative position and the first satellite positioning information; The second determination module is used to determine the motion state information of the object to be identified based on the multi-view sea area image data if the second satellite positioning information matches the first preset sea area; evaluate the motion state information to obtain a safe navigation value for the object to be identified; and determine the safety level of the object to be identified based on the safe navigation value and a preset weight value for the object to be identified. The second determination module is further used to: identify the type and attitude features of the object to be identified based on the multi-view sea area image data; generate a finite element model of the object to be identified; the finite element model is used to simulate the motion features of the object to be identified. The second determination module is also used to: based on the ship... The system utilizes a preset safety evaluation module installed in the ship's equipment and the relative position to assess the navigation stability and safety of the simulated motion state information of the ship's equipment, thereby generating navigation stability information for the object to be identified. Based on the motion state information and the navigation stability information, the system predicts the motion trajectory of the object to be identified within a preset time range, generating predicted trajectory information for the object to be identified. Based on the predicted trajectory information, the system generates a safe navigation value for the object to be identified. Finally, based on the finite element model and the multi-dimensional modeling module installed in the ship's equipment, the system simulates the motion characteristics of the object to be identified in the current ocean scene, thereby generating motion state information of the object to be identified relative to the ship's equipment. The execution module is used to issue a warning alarm based on the ship's equipment if the safety level is less than a set level threshold.

7. An electronic device, characterized in that, include: A memory on which computer programs are stored; A processor for executing the computer program in the memory to implement the steps of the image recognition-based marine safety assessment method according to any one of claims 1-5.

8. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When the program instructions are executed by the processor, they implement the steps of the method according to any one of claims 1-5.

Images