Ship channel identification method and device, electronic equipment and storage medium
The navigation beacon and ship characteristics were extracted through the target detection model, combined with the least squares method to fit the channel curve and the color mapping method to visualize the channel boundary, solving the problem of inaccurate channel identification in the existing technology, real-time visualization of the channel boundary, and improving navigation safety sex.
Patent Information
- Application Number
- CN202411092888.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-09
- Publication Date
- 2024-12-06
AI Technical Summary
The lack of a method for visualizing the waterway boundary based on the real-time location of the navigation beacon in the prior art leads to inaccurate identification of waterways in the bridge area and increases the risk of navigation accidents.
The target detection model is used to extract the features of the water beacons and ship pictures, fit the channel curves through the least squares method, and use the color mapping method to fill and overlap the channel boundaries to realize the visualization of the channel.
It improves the accuracy of channel identification and navigation safety, reduces the occurrence of abnormal navigation ships, and enhances the monitoring capabilities of bridge waters.
Smart Images

Figure CN119091400A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of machine vision technology, and in particular to a ship channel identification method, device, electronic equipment and storage medium. Background Art
[0002] In real life, the waterways in some bridge areas are narrow and curved, and ships entering the waterways are prone to deviate from the waterway, leading to accidents such as grounding or collision with navigation marks and bridge piers. Therefore, it is very necessary to comprehensively and accurately detect navigation marks and ships, realize automatic identification and visualization of waterways, and detect and warn abnormal ships in advance, improve navigation safety, and reduce casualties and property losses.
[0003] At present, the channel identification method fails to visualize the channel boundary based on the real-time position of the navigation mark. Therefore, the prior art lacks a method for visualizing the channel boundary based on the real-time position of the navigation mark. Summary of the invention
[0004] In view of this, it is necessary to provide a ship channel identification method, device, electronic device and storage medium to solve the problem in the prior art that there is a lack of a method for visualizing channel boundaries based on the real-time position of navigation marks.
[0005] In order to solve the above technical problems, on the one hand, the present invention provides a ship channel identification method, comprising: Collect basic information of the target waters; basic information includes: pictures of waters navigation marks and ships; Based on the target detection model, feature extraction is performed on the basic information to obtain multiple target feature information; the target feature information includes: feature data of the target navigation mark and feature data of the target ship; The characteristic data of multiple target ships are fitted based on the least square method to obtain the channel curve; Based on the characteristic data of the channel curve and multiple target navigation marks, the position is identified to obtain the ship channel boundary; Based on the color mapping method, the ship channel boundary is filled with color to obtain a colored ship channel boundary, and the colored ship channel boundary is weighted superimposed to obtain a visualized ship channel.
[0006] In a possible implementation, the target detection model includes: an input layer, a backbone network layer, a neck network layer, and an output layer connected in series in sequence; The input layer is used to process the basic information and obtain the cropped image; The backbone network layer is used to extract features from the cropped image and obtain multiple cropped image features; The neck network layer is used to fuse the features of multiple cropped images to obtain a feature fused image; The output layer is used to fuse images based on features to obtain multiple target feature information.
[0007] In a possible implementation, performing image processing on the basic information to obtain a cropped image includes: The basic information is randomly cropped, scaled, and spliced to obtain a cropped image.
[0008] In a possible implementation, the feature extraction is performed on the cropped image to obtain multiple cropped image features, including: The cropped image is subjected to feature extraction by performing convolution calculation to obtain multiple cropped image features.
[0009] In a possible implementation, the feature-based fusion image is used to obtain multiple target feature information, including: Based on the feature fusion image, detection frames of different sizes are obtained, based on the detection frames of different sizes, a target detection frame is obtained, and based on the target detection frame, multiple target feature information is obtained; the target detection frame is: a detection frame of a target ship and a target navigation mark.
[0010] In a possible implementation, obtaining the target detection frame based on the detection frames of different sizes includes: Based on detection frames of different sizes, non-maximum suppression filtering redundancy processing is performed to obtain the target detection frame.
[0011] In a possible implementation, the performing of position identification based on the channel curve and the characteristic data of a plurality of target navigation marks to obtain the ship channel boundary includes: Based on the channel curve and the characteristic data of multiple target navigation marks, the positions of the left and right sides are identified to obtain the ship channel boundary.
[0012] On the other hand, the present invention also provides a ship channel identification device, comprising: The acquisition module is used to collect basic information of the target waters; the basic information includes: pictures of waters navigation marks and ships; A feature extraction module is used to extract features from basic information based on a target detection model to obtain multiple target feature information; the target feature information includes: feature data of target navigation marks and feature data of target ships; The channel curve module is used to fit the characteristic data of multiple target ships based on the least square method to obtain the channel curve; A position identification module is used to identify the position based on the channel curve and the characteristic data of multiple target navigation marks to obtain the ship channel boundary; The display module is used to fill the ship channel boundary with color based on the color mapping method to obtain a colored ship channel boundary, and to weight-superimpose the colored ship channel boundary to obtain a visualized ship channel.
[0013] On the other hand, the present invention also provides an electronic device, including a memory and a processor, wherein: The memory is used to store programs; The processor is coupled to the memory and is used to execute the program stored in the memory to implement the steps in a ship channel identification method described in any of the above implementations.
[0014] On the other hand, the present invention also provides a computer-readable storage medium for storing computer-readable programs or instructions, which, when executed by a processor, can implement the steps of a ship channel identification method described in any of the above-mentioned implementation methods.
[0015] The beneficial effects of the present invention are as follows: the ship channel identification method provided by the present invention includes: collecting basic information of the target water area; the basic information includes: pictures of water area navigation marks and ships; extracting features of the basic information based on the target detection model to obtain multiple target feature information, the purpose of which is to obtain target feature information based on the target detection model; the target feature information includes: feature data of the target navigation mark and feature data of the target ship; fitting multiple target feature information based on the least squares method to obtain a channel curve, the purpose of which is to fit the target feature information obtained based on the target detection model based on the least squares method to obtain a channel curve; position identification based on the channel curve and the feature data of multiple target navigation marks to obtain a ship channel boundary; color filling of the ship channel boundary based on the color mapping method to obtain a colored ship channel boundary, and weight superposition of the colored ship channel boundary to obtain a visualized ship channel curve, the purpose of which is to color map the channel curve and the ship channel boundary based on the color mapping method to obtain a visualized ship channel. The present application creatively comes up with a method for ship channel identification. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.
[0017] Figure 1 A schematic diagram of a flow chart of an embodiment of a ship channel identification method provided by the present invention; Figure 2A visualized channel schematic diagram of an embodiment of the ship channel identification method provided by the present invention; Figure 3 A fitted track point curve diagram of an embodiment of the ship channel identification method provided by the present invention; Figure 4 A navigation mark classification curve diagram of an embodiment of the ship channel identification method provided by the present invention; Figure 5 A schematic diagram of the structure of an embodiment of a ship channel identification device provided by the present invention; Figure 6 A schematic structural diagram of an embodiment of an electronic device provided by the present invention. DETAILED DESCRIPTION
[0018] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.
[0019] In the description of the embodiments of the present application, unless otherwise specified, “plurality” means two or more than two.
[0020] The terms "including" and "having" and any variations thereof in the embodiments of the present invention are intended to cover non-exclusive inclusions. For example, a process, method, apparatus, product or equipment comprising a series of steps or modules is not necessarily limited to those steps or modules clearly listed, but may include other steps or modules not clearly listed or inherent to these processes, methods, products or equipment.
[0021] The naming or numbering of the steps in the embodiments of the present invention does not mean that the steps in the method flow must be executed in the time / logical sequence indicated by the naming or numbering. The execution order of the named or numbered process steps can be changed according to the technical purpose to be achieved, as long as the same or similar technical effects can be achieved.
[0022] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.
[0023] The present invention provides a ship channel identification method, device, electronic device and storage medium, which are described below respectively.
[0024] Figure 1 A schematic diagram of an embodiment of a ship channel identification method provided by the present invention and Figure 2 A visualized channel schematic diagram of an embodiment of the ship channel identification method provided by the present invention includes: S101. Collect basic information of the target waters; the basic information includes: pictures of waters navigation marks and ships; S102, extracting features from the basic information based on the target detection model to obtain multiple target feature information; the target feature information includes: feature data of the target navigation mark and feature data of the target ship; S103, fitting characteristic data of multiple target ships based on the least square method to obtain a channel curve; S104, performing position recognition based on the channel curve and characteristic data of multiple target navigation marks to obtain the ship channel boundary; S105, filling the ship channel boundary with color based on the color mapping method to obtain a colored ship channel boundary, and weighting the colored ship channel boundary to obtain a visualized ship channel curve.
[0025] It can be understood that the present invention is mainly based on the YOLOv5 (You Only Look Once version 5) target detection algorithm to identify and characterize navigation marks and ships, and obtain the characteristics of navigation marks and ships and track target ships. Secondly, a method based on the clustering of ship trajectory points is proposed, the least squares method is used to fit the channel curve, the left and right navigation marks are identified based on the channel curve and the relative position of the navigation mark, the left and right navigation marks are connected in series to obtain the channel boundary, and the color mapping method is used to visualize the channel.
[0026] In one embodiment of the present invention, a camera installed at the bottom of the upstream bridge of Bridge A is used to obtain real-time video data of the bridge area from 18:00 to 20:00 on a certain day, and the detection and visualization effects are tested using the Python platform. Finally, the warning effect of abnormal ship navigation in the bridge area is analyzed.
[0027] In order to verify the adaptive visualization effect of the channel, 1801 pictures were randomly collected during this period, and the number of ship and navigation mark samples was manually counted to be 5403 and 12607 respectively. The trained yolov5 target detection model was used to detect navigation marks and ships in the pictures. The detection effects of single camera and linkage model were compared, and the results are shown in Table 1 below.
[0028] Table 1: Comparison of model target detection performance
[0029] Define the detection accuracy to evaluate the detection effect, as shown in the following formula:
[0030] In the formula, is the number of samples detected by the model, The total number of samples is manually counted.
[0031] The linkage model achieved 89.2% and 90.3% accuracy in detecting navigation marks and ships, respectively, both higher than the accuracy of a single camera. The linkage model compensates for the shortcomings of a single camera in terms of sensing distance and range, and improves the accuracy of target detection in the bridge area.
[0032] In some embodiments of the present invention, the target detection model includes: an input layer, a backbone network layer, a neck network layer and an output layer connected in series in sequence; The input layer is used to process the basic information and obtain the cropped image; The backbone network layer is used to extract features from the cropped image and obtain multiple cropped image features; The neck network layer is used to fuse the features of multiple cropped images to obtain a feature fused image; The output layer is used to fuse images based on features to obtain multiple target feature information.
[0033] It can be understood that the present invention mainly uses the YOLOv5 target detection algorithm to process visual information, identify and characterize navigation marks and track ships.
[0034] It needs to be further explained that YOLOv5 is a real-time target detection algorithm model, which consists of four parts: input layer, backbone network, neck network, and output layer. The input layer performs data enhancement, randomly crops, scales, and splices the input image to increase data diversity and improve the generalization of the model. The backbone network is CSPDarknet, which slides the convolution kernel on the input image to perform convolution calculation to extract image features, that is, at each position, the data in the convolution kernel is multiplied element by element with the data in the input image and summed. The output matrix is the extracted image features. The convolution kernel is a one-dimensional weight matrix, and its weight value is regressed and optimized according to the difference between the output value and the input value of the calculation network. The role of the neck network is to fuse the image features of different scales output by the backbone network and improve the network's detection ability for targets of different scales. The role of the output layer is to generate detection frames of different sizes based on the feature image fused by the neck network, and filter redundant detection frames using the non-maximum suppression (NMS) algorithm.
[0035] In some embodiments of the present invention, the step of performing image processing on the basic information to obtain a cropped image includes: The basic information is randomly cropped, scaled, and spliced to obtain a cropped image.
[0036] It is understandable that the input layer performs data augmentation, randomly crops, scales, and splices the input images to increase data diversity and improve the generalization of the model.
[0037] In some embodiments of the present invention, the step of extracting features from the cropped image to obtain a plurality of cropped image features includes: The cropped image is subjected to feature extraction by performing convolution calculation to obtain multiple cropped image features.
[0038] It can be understood that the backbone network is CSPDarknet, which slides the convolution kernel on the input image to perform convolution calculation to extract image features, that is, at each position, the data in the convolution kernel is multiplied element by element with the data in the input image and the sum is taken. The output matrix is the extracted image features. The convolution kernel is a one-dimensional weight matrix, and its weight value is regressed and optimized according to the difference between the output value and the input value of the computing network.
[0039] In some embodiments of the present invention, the step of obtaining multiple target feature information based on the feature-fused image includes: Based on the feature fusion image, detection frames of different sizes are obtained, based on the detection frames of different sizes, a target detection frame is obtained, and based on the target detection frame, multiple target feature information is obtained; the target detection frame is: a detection frame of a target ship and a target navigation mark.
[0040] It can be understood that the role of the output layer is to generate detection frames of different sizes based on the feature image fused by the neck network, and filter redundant detection frames using the non-maximum suppression (NMS) algorithm.
[0041] In some embodiments of the present invention, The object detection frame is obtained based on the detection frames of different sizes, including: Based on detection frames of different sizes, non-maximum suppression filtering redundancy processing is performed to obtain the target detection frame.
[0042] It can be understood that the role of the output layer is to generate detection frames of different sizes based on the feature image fused by the neck network, and filter redundant detection frames using the non-maximum suppression (NMS) algorithm.
[0043] Figure 3 A fitted track point curve diagram of an embodiment of the ship channel identification method provided by the present invention and Figure 4 The navigation mark classification curve diagram of an embodiment of the ship channel identification method provided by the present invention includes: Based on the channel curve and the characteristic data of multiple target navigation marks, the positions of the left and right sides are identified to obtain the ship channel boundary.
[0044] It is understandable that inland waterways are marked by navigation marks that are available for ships to navigate. The side marks are divided into left and right marks, which respectively indicate the left and right boundaries of the channel and guide ships to navigate in the channel. In some waterways in bridge areas, the left and right navigation marks may be unevenly distributed and not in a straight line. In order to identify the channel, after identifying the position of the navigation mark, the left and right navigation marks need to be connected in series respectively. Based on the historical trajectory points of the ship Clustering to obtain the channel trend curve . According to the navigation mark and curve The left and right navigation marks are identified by relative position and connected in series to indicate the channel boundary. The channel is visualized using color mapping and displayed on the monitor interface.
[0045] It needs further explanation that, according to the relevant laws and regulations, when a motor vessel is sailing, an upstream vessel should sail along the slow current or one side of the channel, and a downstream vessel should sail along the mainstream or the middle of the channel. However, in tidal river sections, lakes, reservoirs, and flat flow areas, any ship should sail along the channel on the starboard side of the ship as much as possible. Most ships will sail along the direction of the channel, so based on the density clustering algorithm of the ship's trajectory points, after filtering out the trajectory points outside the channel, the remaining trajectory points can reflect the direction of the channel and identify the navigation marks on both sides.
[0046] It can be further understood that the irregularly distributed trajectory points are divided into core points, noise points, and boundary points, which are the in-channel, out-channel, and start-end trajectory points, respectively.
[0047] Core Points The formula (5) must be satisfied, as shown below: (5) For The number of trajectory points in a circle with point as the center and r as the radius, let is a point in the field, then , Distance As shown in formula (6), it is as follows: (6) Boundary Points The formula (7) must be satisfied, as shown below: (7) Less than threshold , The point does not meet the definition of a core point, but Points are within the domain of other core points. Point is a boundary point.
[0048] Noise point: a point that does not meet the definition of core point or boundary point.
[0049] It should be noted that the specific implementation method of navigation mark classification based on trajectory point clustering is as follows: Filter out the noise points outside the channel and use the least squares method to fit the track points in the channel to generate a curve , which can reflect the general direction of the waterway.
[0050] The filtered trajectory point set T and curve are as described in equations (8)-(9).
[0051] (8) (9) In the formula, , , , and are the horizontal and vertical coordinates of the trajectory points, is the number of trajectory points.
[0052] To make the curve More closely fitting the track points, calculate the squared loss as the loss function to calculate the difference , and solve The minimum value of .
[0053] (10) right , Find the partial derivative so that its value is 0.
[0054] (11) (12) (13) Simplify equations (11)-(13) into matrix form as shown in equation (14).
[0055]
[0056] Substitute the coordinate points in T into equation (14) and solve , , and get the curve expression.
[0057] Navigational aids The coordinates of the center point of the detection box are ,judge and Taking the monitoring waters upstream of the bridge area as an example, if Greater than ,but is the left label, otherwise it is the right label.
[0058]
[0059] In the formula, and They are respectively the pixel coordinate point sets of the right and left beacons.
[0060] It should be further explained that the left and right side mark of the channel boundary have different attributes such as color, top mark shape, light color and light quality. During navigation, the driver uses the attributes of the navigation mark to determine the channel boundary and direction to ensure that the ship is sailing within the channel. Onshore supervisors use video surveillance systems to monitor whether the ship is sailing within the channel. Traditional shore-based video surveillance systems have difficulty in clearly displaying and identifying long-distance navigation marks, and do not clearly connect navigation marks and mark channel boundaries. It is relatively difficult to identify and warn ships with abnormal navigation.
[0061] In order to improve the efficiency of identifying and warning ships with abnormal navigation, it is necessary to accurately mark the boundaries of the channel and navigable waters, that is, use green and red lines to connect the left and right navigation marks in the video surveillance screen, and visualize the channel through color mapping.
[0062] (1) Locate navigation marks, filter out-of-channel trajectory points, fit the channel curve, identify the left and right navigation marks and record their pixel coordinates.
[0063] (2) Generate a screen mask of the same size as the monitoring video interface and map the beacon coordinate points.
[0064] (3) Sort the coordinate points according to their ordinates, and connect the left and right navigation points in series to form green polylines. and red line .
[0065] (4) , The closed area formed by two polylines The inner fill color.
[0066] (5) The screen and the video image are superimposed based on the weight and displayed on the monitoring interface.
[0067] It needs to be further explained that the adaptive visualization experiment of the waterway, such as Figure 3 and4 As shown, in Figure 3 Obtain the ship's trajectory points within this time period and filter out the trajectory points outside the channel to fit the channel trend curve. The blue points are the tracked and recorded trajectory points, and the red points are the fitted curve. Figure 4 In the figure, based on the fitting curve, the left and right navigation marks are identified and represented by green and red respectively.
[0068] Figure 5 A schematic structural diagram of an embodiment of a ship channel identification device provided by the present invention includes: The acquisition module 501 is used to acquire basic information of the target waters; the basic information includes: pictures of waters navigation marks and pictures of ships; The feature extraction module 502 is used to extract features from the basic information based on the target detection model to obtain multiple target feature information; the target feature information includes: feature data of the target navigation mark and feature data of the target ship; A channel curve module 503 is used to fit the characteristic data of multiple target ships based on the least square method to obtain a channel curve; A position identification module 504 is used to perform position identification based on the channel curve and multiple target navigation marks to obtain the ship channel boundary; The display module 505 is used to fill the channel curve and the ship channel boundary with color based on the color mapping method to obtain colored ship channel boundaries and channel curves, and to weight-superimpose the colored ship channel boundaries and channel curves to obtain a visualized ship channel.
[0069] The ship channel identification device 500 provided in the above embodiment can implement the technical solution described in the above ship channel identification method embodiment. The specific implementation principles of the above modules or units can refer to the corresponding contents in the above ship channel identification method embodiment, which will not be repeated here.
[0070] like Figure 6 As shown, the present invention also provides an electronic device 600. The electronic device 600 includes a processor 601, a memory 602 and a display 603. Figure 6 Only some components of the electronic device 600 are shown, but it should be understood that it is not required to implement all of the components shown, and more or fewer components may be implemented instead.
[0071] In some embodiments, the processor 601 may be a central processing unit (CPU), a microprocessor or other data processing chip, used to run program codes or process data stored in the memory 602, such as the ship channel identification method of the present invention.
[0072] In some embodiments, the processor 601 may be a single server or a server group. The server group may be centralized or distributed. In some embodiments, the processor 601 may be local or remote. In some embodiments, the processor 601 may be implemented in a cloud platform. In one embodiment, the cloud platform may include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an internal cloud, a multi-cloud, etc., or any combination thereof.
[0073] In some embodiments, the memory 602 may be an internal storage unit of the electronic device 600, such as a hard disk or memory of the electronic device 600. In other embodiments, the memory 602 may also be an external storage device of the electronic device 600, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc., equipped on the electronic device 600.
[0074] Furthermore, the memory 602 may include both an internal storage unit of the electronic device 600 and an external storage device. The memory 602 is used to store application software installed in the electronic device 600 and various data.
[0075] In some embodiments, the display 603 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch device, etc. The display 606 is used to display information on the electronic device 600 and to display a visual user interface. The components 601-603 of the electronic device 600 communicate with each other via a system bus.
[0076] In one embodiment, when the processor 601 executes the ship channel identification program in the memory 602, the following steps may be implemented: Collect basic information of the target waters; basic information includes: pictures of waters navigation marks and ships; Based on the target detection model, feature extraction is performed on the basic information to obtain multiple target feature information; the target feature information includes: feature data of the target navigation mark and feature data of the target ship; The characteristic data of multiple target ships are fitted based on the least square method to obtain the channel curve; Based on the characteristic data of the channel curve and multiple target navigation marks, the position is identified to obtain the ship channel boundary; Based on the color mapping method, the ship channel boundary is filled with color to obtain a colored ship channel boundary, and the colored ship channel boundary is weighted superimposed to obtain a visualized ship channel.
[0077] It should be understood that: when the processor 601 executes the ship channel identification program in the memory 602, in addition to the above functions, other functions can also be implemented. For details, please refer to the description of the corresponding method embodiment above.
[0078] Furthermore, the embodiment of the present invention does not specifically limit the type of the electronic device 600 mentioned, and the electronic device 600 may be a portable electronic device such as a mobile phone, a tablet computer, a personal digital assistant (PDA), a wearable device, a laptop computer, etc. Exemplary embodiments of portable electronic devices include but are not limited to portable electronic devices equipped with IOS, Android, Microsoft or other operating systems. The above-mentioned portable electronic devices may also be other portable electronic devices, such as a laptop computer with a touch-sensitive surface (e.g., a touch panel). It should also be understood that in some other embodiments of the present invention, the electronic device 600 may not be a portable electronic device, but a desktop computer with a touch-sensitive surface (e.g., a touch panel).
[0079] Accordingly, an embodiment of the present application also provides a computer-readable storage medium, which is used to store computer-readable programs or instructions. When the program or instructions are executed by a processor, the steps or functions of the ship channel identification method provided in the above-mentioned method embodiments can be implemented.
[0080] Those skilled in the art will appreciate that all or part of the processes of the above-mentioned embodiments can be implemented by instructing related hardware (such as a processor, a controller, etc.) through a computer program, and the computer program can be stored in a computer-readable storage medium, wherein the computer-readable storage medium is a disk, an optical disk, a read-only storage memory, or a random access memory, etc.
[0081] The ship channel identification method, device, electronic device and storage medium provided by the present invention are introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for technical personnel in this field, according to the idea of the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.
Claims
1. A ship channel identification method, characterized in that: include: Collect basic information of target waters; Basic information includes: pictures of water area navigation marks and ships; Based on the target detection model, feature extraction is performed on the basic information to obtain multiple target feature information; the target feature information includes: feature data of the target navigation mark and feature data of the target ship; The characteristic data of multiple target ships are fitted based on the least square method to obtain the channel curve; Based on the characteristic data of the channel curve and multiple target navigation marks, the position is identified to obtain the ship channel boundary; Based on the color mapping method, the ship channel boundary of the channel curve is filled with color to obtain a colored ship channel boundary, and the colored ship channel boundary is weighted superimposed to obtain a visualized ship channel.
2. The ship channel identification method according to claim 1, characterized in that: The target detection model includes: an input layer, a backbone network layer, a neck network layer and an output layer connected in series in sequence; The input layer is used to process the basic information and obtain the cropped image; The backbone network layer is used to extract features from the cropped image and obtain multiple cropped image features; The neck network layer is used to fuse the features of multiple cropped images to obtain a feature fused image; The output layer is used to fuse images based on features to obtain multiple target feature information.
3. The ship channel identification method according to claim 2, characterized in that: The step of performing image processing on the basic information to obtain a cropped image includes: The basic information is randomly cropped, scaled, and spliced to obtain a cropped image.
4. The ship channel identification method according to claim 2, characterized in that: The feature extraction of the cropped image is performed to obtain multiple cropped image features, including: The cropped image is subjected to feature extraction by performing convolution calculation to obtain multiple cropped image features.
5. The ship channel identification method according to claim 2, characterized in that: The feature-based fusion image is used to obtain multiple target feature information, including: Based on the feature fusion image, detection frames of different sizes are obtained, based on the detection frames of different sizes, a target detection frame is obtained, and based on the target detection frame, multiple target feature information is obtained; the target detection frame is: a detection frame of a target ship and a target navigation mark.
6. The ship channel identification method according to claim 5, characterized in that: The object detection frame is obtained based on the detection frames of different sizes, including: Based on detection frames of different sizes, non-maximum suppression filtering redundancy processing is performed to obtain the target detection frame.
7. The ship channel identification method according to claim 1, characterized in that: The position identification is performed based on the channel curve and the characteristic data of multiple target navigation marks to obtain the channel curve of the ship channel boundary, including: Based on the channel curve and the characteristic data of multiple target navigation marks, the positions of the left and right sides are identified to obtain the channel curve of the ship's channel boundary.
8. A ship channel identification device, characterized in that: include: Collection module, used to collect basic information of target waters; Basic information includes: pictures of water area navigation marks and ships; A feature extraction module is used to extract features from basic information based on a target detection model to obtain multiple target feature information; the target feature information includes: feature data of a target navigation mark and feature data of a target ship; The channel curve module is used to fit the characteristic data of multiple target ships based on the least square method to obtain the channel curve; A position identification module is used to identify the position based on the channel curve and the characteristic data of multiple target navigation marks to obtain the ship channel boundary; The display module is used to fill the ship channel boundary with color based on the color mapping method to obtain a colored ship channel boundary, and to weight-superimpose the colored ship channel boundary to obtain a visualized ship channel.
9. An electronic device, characterized in that: comprising a memory and a processor, wherein: The memory is used to store programs; The processor is coupled to the memory and is used to execute the program stored in the memory to implement the steps in the ship channel identification method as described in any one of claims 1 to 7 above.
10. A computer-readable storage medium, characterized in that: Used to store computer-readable programs or instructions, which, when executed by a processor, can implement the steps of the ship channel identification method described in any one of claims 1 to 7.
Citation Information
Cited By
Navigation channel electronic fence marking method and device
CN121505925A
A method and device for marking an electronic channel fence
CN121505925B
