A method for identifying a fishing net position marker based on group motion features

Abstract

The application provides a fishing net marker recognition method based on group motion characteristics, comprising the following steps: acquiring AIS data, extracting static information and dynamic information of the AIS data; performing primary recognition on the static information in the AIS data, and if the primary recognition is passed, the current AIS data is a fishing net marker; and performing secondary recognition on the AIS data which fails to pass the primary recognition by using the dynamic information, wherein the secondary recognition is used for classifying AIS trajectory sequences with the same motion state, and AIS trajectory sequences with the same category are determined as fishing net markers on a towed fishing boat. The application can intelligently recognize the fishing net markers, has significant application value, and can effectively improve the work efficiency of maritime supervision personnel and ensure maritime traffic safety.

Description

Technical Field

[0001] This invention relates to the field of intelligent shipping technology, and more particularly to a method for identifying fishing net position markers based on swarm movement characteristics. Background Technology

[0002] "Automatic Identification System (AIS) Channel Protection" is a long-term issue led by the International Maritime Organization (IMO), calling on competent authorities and researchers in various countries to conduct relevant research to protect AIS channel resources and ensure the safety of ship navigation. Among the various AIS channel resource occupancy issues, my country uses a large number of fishing net positioning beacons based on AIS technology, with nearly 100,000 entering the domestic market each year. Their application is particularly widespread in the waters of fishing-developed provinces such as the Shandong and Zhejiang coasts, leading to a significant problem of AIS channel resource occupancy.

[0003] Meanwhile, because fishing net beacons do not yet have clearly assigned corresponding messages, existing AIS fishing net beacons often misuse Category B AIS Station Message 18 or 19 and AIS Navigation Mark Message 21, causing identification difficulties, interfering with safe navigation, and consequently leading to various maritime accidents. Therefore, timely and accurate identification of fishing net beacons among a large number of ship AIS signals is of great economic and social significance for ensuring the safety of maritime navigation. Summary of the Invention

[0004] To address the aforementioned technical problem that existing AIS fishing net beacons often spoof Class B AIS station messages, causing identification difficulties and interfering with safe navigation, this invention provides a fishing net beacon identification method based on swarm movement characteristics. This invention primarily utilizes static and dynamic information from AIS data to identify fishing net beacons.

[0005] The technical means employed in this invention are as follows:

[0006] A method for identifying fishing net position markers based on swarm movement characteristics, comprising:

[0007] Acquire AIS data and extract static and dynamic information from the AIS data;

[0008] The static information in the AIS data is identified once. If the identification is successful, it indicates that the current AIS data is a fishing net positioning beacon. The identification is used to determine the vessel name in the static information.

[0009] For AIS data that fails the first identification, dynamic information is used for a second identification. The second identification is used to classify AIS trajectories under the same motion state. AIS trajectories of the same category are determined to be fishing net markers on a towed fishing boat.

[0010] Furthermore, the static and dynamic information of the AIS data is extracted, including:

[0011] Extract static information: maritime mobile service identifier and vessel name;

[0012] Extract dynamic information: latitude, longitude, speed, heading, and data reception time.

[0013] Furthermore, extracting the static and dynamic information of the AIS data also includes: storing the AIS data in a preset format as follows:

[0014] AIS=[MMSI,shipname,latitude,longitude,SOG,COG,t]

[0015] Wherein, MMSI is the vessel mobility service identifier, shipname is the vessel name, latitude is the latitude, longitude is the longitude, SOG is the speed, COG is the heading, and t is the data reception time.

[0016] Furthermore, extracting the static and dynamic information of the AIS data also includes:

[0017] Based on the vessel mobility service identifier data, extract vessel data according to the data reception time series and generate AIS trajectory;

[0018] Delete AIS trajectory sequences with a length less than 20 and generate processed AIS trajectories.

[0019] Furthermore, for AIS data that fails the first identification, a second identification is performed on the dynamic information within the AIS data, including:

[0020] The AIS track is classified into three categories based on the speed information in the dynamic information: Category 1 data, Category 2 data, and Category 3 data. Category 1 data represents abnormal navigation status data, Category 2 data represents anchorage status data, and Category 3 data represents navigation status data. AIS tracks identified as Category 1 data are identified as suspected fishing net position markers and sent for manual judgment.

[0021] Several AIS trajectories are randomly extracted from the two types of data and subjected to secondary classification. The secondary classification is used to classify the AIS trajectory sequences based on the DBSCAN algorithm of Mahalanobis distance. After classification, AIS trajectory sequences belonging to the same category represent fishing net markers on the same towed fishing vessel. The predefined clustering is constructed based on the fishing net markers obtained through a first identification.

[0022] Furthermore, for AIS data that fails the first identification, a second identification is performed on the dynamic information within the AIS data, which also includes:

[0023] Several AIS trajectory sequences are randomly extracted from three types of data and subjected to secondary classification. The secondary classification is used to classify the AIS trajectory sequences based on the DBSCAN algorithm of Mahalanobis distance. After classification, AIS trajectory sequences belonging to the same category represent fishing net markers on the same towed fishing vessel. The predefined clustering is constructed based on the fishing net markers obtained through a first identification.

[0024] Compared with the prior art, the present invention has the following advantages:

[0025] 1. This invention collects a large amount of AIS data, and through preprocessing analysis, identifies data that conforms to the static characteristics of fishing net positioning markers;

[0026] 2. This invention extracts dynamic information of fishing net positioning marks, proposes the group movement characteristics of fishing net positioning marks, and classifies AIS data based on its group movement characteristics.

[0027] 3. The present invention has a stable recognition effect and an accuracy rate of over 95%.

[0028] In summary, the algorithm proposed in this invention demonstrates for the first time intelligent identification of fishing net markers, which has significant application value and can effectively improve the work efficiency of maritime regulatory personnel and ensure maritime traffic safety. Attached Figure Description

[0029] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0030] Figure 1 This is a flowchart of the fishing net positioning mark identification method based on group motion characteristics according to the present invention.

[0031] Figure 2 This is a flowchart illustrating the identification of fishing net position markers in the embodiment.

[0032] Figure 3 This is a flowchart illustrating data classification based on Mahalanobis distance in an embodiment. Detailed Implementation

[0033] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. 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 should fall within the scope of protection of the present invention.

[0034] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention 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 the invention 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 a 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.

[0035] This invention addresses the problem of inconsistent information on fishing net position markers causing identification difficulties and thus affecting ship navigation safety. It proposes an intelligent identification method for fishing net position markers based on swarm motion characteristics, such as... Figure 1-3 As shown, the main steps include:

[0036] S1. Obtain AIS data and extract the static and dynamic information of the AIS data.

[0037] In this invention, the extracted static information from AIS data includes the Maritime Mobile Service Identifier and the vessel name. The extracted dynamic information includes latitude, longitude, speed, heading, and data reception time.

[0038] Store AIS data in the following format according to the preset format:

[0039] AIS=[MMSI,shipname,latitude,longitude,SOG,COG,t] T

[0040] Wherein, MMSI is the vessel mobility identifier, shipname is the vessel name, latitude is the latitude, longitude is the longitude, SOG is the speed, COG is the heading, and t is the time when the data was received.

[0041] Furthermore, the acquired AIS data undergoes preprocessing, primarily including the formation of finite AIS trajectory sequences based on ship mobility identifier data according to time series; AIS trajectory sequences with a length less than 20 are deleted. Additionally, trajectory information with a reception interval greater than 15 minutes can also be deleted.

[0042] In this invention, the mathematical expression for the AIS trajectory point x is preferably defined as x = [MMSI, shipname, latitude, longitude, SOG, COG, t]. T The mathematical expression for the AIS trajectory is traj. i ={x1,x2...x j}

[0043] S2. Perform a first identification on the static information in the AIS data. If the first identification is successful, it indicates that the current AIS data is a fishing net positioning mark. The first identification is used to determine the ship name information in the static information.

[0044] This invention analyzes and processes extracted AIS data to generate an algorithm-usable dataset, which then determines whether the vessel name conforms to the naming rules. If it does, it is directly identified as a fishing net location marker. Specifically, data in the static information data where the vessel name is "AAAA-BB-CC%" is directly identified as a fishing net location marker. Here, AAAA represents the name of the fishing net location marker, BB represents the serial number of the fishing net location marker on that fishing net, and CC represents the remaining battery power of the fishing net location marker. The above vessel name format is the most common naming format for fishing net location markers in practical applications. It is set at the factory and is used directly in most cases. In a few cases, the naming rules for fishing net location markers are manually modified. The application scenario of this invention is set so that at least some fishing net location markers use the above conventional naming rules.

[0045] S3. For AIS data that fails the first identification, the dynamic information in the AIS data is identified a second time. The second identification is used to classify the AIS trajectory. AIS trajectories of the same category are determined to be fishing net markers on a towed fishing boat.

[0046] A fishing net typically has 2 to 20 fishing net position markers installed, which move with the main vessel and auxiliary vessels during operations. Therefore, the speed and direction of movement of the fishing net position markers on the same fishing net and between the main and auxiliary vessels are strongly correlated. When the motion states are the same, the fishing net position markers of the same towed fishing vessel exhibit similar speed and heading change patterns and similar trajectory shapes. Specifically, this includes a primary classification of AIS trajectories. This primary classification is used to divide AIS trajectories into three categories based on speed information in the dynamic data: Category I data represents abnormal navigation status data, Category II data represents anchored status data, and Category III data represents navigation status data. AIS trajectories identified as Category I data are then identified as suspected fishing net position markers and sent for manual judgment.

[0047] Specifically: traj i AIS trajectory points i Classify them. The classification criteria are as follows:

[0048]

[0049] x→x i =[MMSI,shipname,latitude,longitude,SOG,COG,t,l];

[0050] traj i Classified as follows:

[0051]

[0052] Here, Class 1 is identified as a suspected fishing net beacon, indicating that its movement status is constantly changing, which is abnormal data and needs to be sent to the higher-level monitoring system for manual judgment. Class 2 is identified as anchored while in motion, and Class 3 is identified as sailing while in motion. Both require further judgment to confirm whether they represent fishing net beacons. Therefore, secondary classification is necessary.

[0053] Furthermore, the second type of data is further classified, specifically:

[0054] Several AIS trajectories are randomly extracted from the two types of data and then classified in a secondary manner. The secondary classification is used to classify the AIS trajectory data based on the DBSCAN algorithm of Mahalanobis distance. After classification, AIS trajectory data belonging to the same category represent fishing net markers on the same towed fishing vessel. The predefined clustering is constructed based on the fishing net markers obtained through a first identification.

[0055] Traditional DBSCAN suffers from high computational costs and only considers local features when identifying clusters. Mahalanobis distance addresses these drawbacks by improving computational efficiency and considering the correlation between points within the entire cluster. Mahalanobis distance describes the distance between a point and a group of points. The expression for the Mahalanobis distance from a data point x to a point with a mean μ and a covariance matrix S is given by...

[0056]

[0057] μ = [μ lat ,μ lon ,μ sog ,μ cog ]

[0058] S ij =cov(x i ,x j )

[0059] In this invention, an AIS trajectory sequence from Class2 is extracted, and its Mahalanobis distance to a pre-defined cluster is calculated. The distance is then compared with preset clustering parameters to determine whether the sequence belongs to that cluster. The specific operation is as follows: Input ε, minPts, and a predefined cluster. Each trajectory point is iterated, and the Mahalanobis distance D to the predefined cluster is calculated. M This distance will be compared with the input ε. If D M If the value is less than ε, the trajectory point is identified as belonging to this cluster, and the clustering parameters are updated. If D M If the distance is greater than ε, then the trajectory point is an outlier. The algorithm updates the clusters with newly added trajectory points and begins the second clustering iteration, continuing until no outlier is closer to any cluster than ε. Here, ε is the radius of a point relative to its neighborhood, and minPts determines whether a point is a core point. If at least minPts points are within a distance of ε, then that point is defined as a core point, and points within ε will be in the same cluster as the core point.

[0060] Furthermore, the steps for secondary classification of the three types of data are similar to those for the two types of data, specifically:

[0061] Several AIS trajectories are randomly extracted from the three types of data and subjected to secondary classification. The secondary classification uses the DBSCAN algorithm based on Mahalanobis distance to classify the AIS trajectory data. After classification, AIS trajectory data belonging to the same category represent fishing net position markers on the same towed fishing vessel. The predefined clustering is constructed based on the fishing net position markers obtained through the first identification. For details, please refer to the above content, which will not be repeated here.

[0062] The following specific application examples will further illustrate the solution and effects of the present invention.

[0063] In this embodiment, it is assumed that the input AIS data is Data = {AIS1, AIS2, AIS3, ... AIS...} n The AIS contains both static and dynamic information for a VDL message.

[0064] Step 1: Extract some static and dynamic information of the ship:

[0065] AIS=[MMSI,shipname,latitude,longitude,SOG,COG,t]

[0066] In this data set, MMSI is the vessel mobility identifier, shipname is the vessel name, latitude is the latitude, longitude is the longitude, SOG is the speed, COG is the heading, and t is the time the data was received.

[0067] Step 2: Data Preprocessing: Based on MMSI, extract ship data according to time t and generate AIS trajectory traj. i ={AIS1,AIS2...AIS j}, Delete traj data where j is less than 20, that is, delete AIS data with a trajectory length of less than 20 within the receiving time.

[0068] Step 3: Vessels with the shipname "AAAA-BB-CC%" are directly identified as fishing net position markers. Here, AAAA represents the name of the fishing net position marker, BB represents the serial number of the fishing net position marker on that fishing net, and CC represents the remaining battery power of the fishing net position marker.

[0069] Step 4: Analyze the dynamic information of the fishing net position markers identified in Step 3: When the motion states are the same, the fishing net position markers of the same towed fishing vessel have the characteristics of the same speed and heading change patterns and similar trajectory shapes.

[0070] Step 5: Based on the analysis of the movement characteristics of the fishing net marker group in Step 4, traj i AIS trajectory points i Classify them. The classification criteria are as follows:

[0071]

[0072] AIS i →AIS i =[MMSI,shipname,latitude,longitude,SOG,COG,t,l];

[0073] Step Six: Transfer the traj file i Classify

[0074]

[0075] Step 7: Identify Class 1 from Step 6 as a suspected fishing net marker, and extract 15 tracks each from Class 2 and Class 3 to perform the operation in Step 8.

[0076] Step 8: Input clustering parameters ε, minpts, and predefined clusters;

[0077] Step 9: AIS of trajectory points in data Class 2 2i Calculate the Mahalanobis distance D M ;

[0078] Step 10: If D M <ε,AIS 1i Belongs to a predefined cluster, otherwise AIS 1i Outlier;

[0079] Step 11: Update clustering parameters;

[0080] Step 12: Repeat steps 9, 10 and 11 until no outlier is closer to any cluster than ε, otherwise repeat step 11.

[0081] The output is:

[0082] shipnet={(traj1,traj2,traj3,traj4),(traj5,traj6,traj7),(traj 10 ,traj 11 ),(traj 14 ,traj 15 )}.

[0083] This indicates that traj1, traj2, traj3, and traj4 have similar trajectories, belong to the same cluster, and are identified as fishing net position markers on a towed fishing vessel. 10 traj 11 and traj 14 traj 15 Similarly.

[0084] The steps for Class3 are the same as those for Class2.

[0085] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for identifying fishing net position markers based on group movement characteristics, characterized in that, include: Acquire AIS data and extract static and dynamic information from the AIS data; The static information in the AIS data is identified once. If the identification is successful, it indicates that the current AIS data is a fishing net location marker. The identification is used to determine the name of the vessel in the static information. For AIS data that fails the first identification, dynamic information is used for a second identification. The second identification is used to classify AIS trajectories under the same motion state. AIS trajectories of the same category are determined to be fishing net markers on a towed fishing boat. Secondary identification of dynamic information in AIS data, including: The AIS track is first classified, which involves dividing the AIS track into three categories based on the speed information in the dynamic data: Category I data represents abnormal navigation status data, Category II data represents anchored status data, and Category III data represents navigation status data. AIS tracks identified as Category I data are then identified as suspected fishing net position markers and sent for manual judgment. Several AIS trajectories are randomly extracted from the two types of data and subjected to secondary classification. The secondary classification is used to classify the AIS trajectory sequences based on the DBSCAN algorithm using Mahalanobis distance. After classification, AIS trajectory sequences belonging to the same category represent fishing net position markers on the same towed fishing vessel. Predefined clusters are constructed based on the fishing net position markers obtained through a single identification. Several AIS trajectory sequences are randomly extracted from three types of data and subjected to secondary classification. The secondary classification is used to classify the AIS trajectory sequences based on the DBSCAN algorithm of Mahalanobis distance. After classification, AIS trajectory sequences belonging to the same category represent fishing net markers on the same towed fishing vessel. Predefined clusters are constructed based on the fishing net markers obtained through a single identification.

2. The method for identifying fishing net positioning markers based on group movement characteristics according to claim 1, characterized in that, Extracting the static and dynamic information of the AIS data, including: Extract static information: maritime mobile service identifier and vessel name; Extract dynamic information: latitude, longitude, speed, heading, and data reception time.

3. The method for identifying fishing net position markers based on group movement characteristics according to claim 2, characterized in that, Extracting the static and dynamic information of the AIS data further includes: storing the AIS data in the following format according to a preset format: in, MMSI For vessel movement business identifiers, shipname The name of the ship. latitude Latitude longitude Longitude SOG For speed, COG For the course, t This refers to the data reception time.

4. The method for identifying fishing net position markers based on group movement characteristics according to claim 3, characterized in that, Extracting the static and dynamic information of the AIS data also includes: Based on the vessel mobility service identifier data, extract vessel data according to the data reception time series and generate AIS trajectory; Delete AIS trajectory sequences with a length less than 20 and generate processed AIS trajectories.

Images