Ship navigation dimension and traffic flow detection method, device, equipment and medium

Abstract

The present application relates to the field of ship navigation, and provides a ship navigation scale and traffic flow detection method, device, equipment and medium, wherein the method comprises: acquiring original laser point cloud data; performing point cloud processing on the original laser point cloud data to obtain ship point cloud data of a to-be-detected ship; based on a main direction of the ship point cloud data, calculating the size of the minimum circumscribed cuboid of the ship point cloud data, performing smoothing processing on the size of the minimum circumscribed cuboid to obtain the ship scale of the to-be-detected ship; based on the ship point cloud data, calculating the distance between the to-be-detected ship and a specified laser radar at a current time and a previous time respectively, and performing traffic flow detection based on the distance. The method, device, equipment and medium provided by the present application can not only accurately measure the ship scale, but also obtain traffic flow data in real time through laser point cloud data, thereby improving the detection efficiency and reducing the equipment cost and deployment complexity.

Description

Technical Field

[0001] This invention relates to the field of ship navigation technology, and in particular to a method, apparatus, equipment and medium for detecting ship navigation dimensions and traffic flow. Background Technology

[0002] In recent years, my country's inland waterway shipping industry has developed rapidly. With the continuous increase in the scale of cargo transportation, the trend of larger ships has become increasingly obvious. The size and traffic density of ships are getting larger and larger, which increases the risk of collision when they pass through structures such as bridges and locks, posing an unprecedented challenge to navigation safety.

[0003] In related technologies, the dimensions of navigable vessels and traffic flow are typically detected separately using different techniques. Vessel dimensions are primarily measured using video surveillance analysis and infrared ultra-high-altitude detection, but the accuracy of these measurements is unsatisfactory. Vessel traffic flow is mainly based on AIS (Automatic Identification System) statistics to determine cross-sectional vessel flow in river sections; however, factors such as human error (AIS shutdown, malfunction) can interfere with the statistical accuracy. Furthermore, separating the detection of vessel dimensions and traffic flow not only reduces detection efficiency but also increases equipment costs and deployment complexity.

[0004] Therefore, it is particularly urgent to provide a technical method and device that can intensively, efficiently, accurately, and simultaneously detect ship dimensions and traffic flow density. Summary of the Invention

[0005] This invention provides a method, apparatus, equipment, and medium for detecting ship navigation dimensions and traffic flow, in order to solve the shortcomings of existing technologies in detecting the dimensions and traffic flow of navigable ships, such as poor accuracy, low detection efficiency, high equipment cost, and high deployment complexity.

[0006] This invention provides a method for detecting ship navigation dimensions and traffic flow, comprising:

[0007] Acquire raw laser point cloud data;

[0008] The original laser point cloud data is processed to obtain the ship point cloud data of the ship under test. The point cloud processing includes at least one of point cloud filtering, coordinate system transformation and point cloud filtering.

[0009] Based on the main direction of the ship point cloud data, the size of the minimum bounding box of the ship point cloud data is calculated, and the size of the minimum bounding box is smoothed to obtain the ship dimensions of the ship under test.

[0010] Based on the ship point cloud data, the distance between the ship under test and the designated lidar at the current time and the previous time are calculated, and traffic flow detection is performed based on the distance.

[0011] According to the ship navigation dimensions and traffic flow detection method provided by the present invention, the step of calculating the size of the minimum circumscribed cuboid of the ship point cloud data based on the principal direction of the ship point cloud data includes:

[0012] Calculate the centroid of the ship point cloud data, and determine the main direction of the ship point cloud data based on the centroid;

[0013] The ship point cloud data is projected onto the main direction to obtain projected point cloud data;

[0014] The dimensions of the minimum circumscribed cuboid are determined based on the minimum and maximum coordinates of the projected point cloud data in each direction.

[0015] According to the ship navigation dimensions and traffic flow detection method provided by the present invention, determining the main direction of the ship point cloud data based on the centroid includes:

[0016] Based on the centroid and the coordinates of each point cloud in the ship point cloud data, a covariance matrix is ​​calculated, which reflects the distribution of the point cloud in various directions.

[0017] The covariance matrix is ​​decomposed into eigenvalues ​​to obtain eigenvalues ​​and corresponding eigenvectors in each direction;

[0018] The eigenvector corresponding to the largest eigenvalue is taken as the main direction of the ship point cloud data.

[0019] According to the ship navigation dimensions and traffic flow detection method provided by the present invention, the traffic flow detection based on the distance includes:

[0020] If the current distance is less than or equal to a preset distance threshold, and the ships to be measured have not yet been counted, then the magnitude between the current distance and the previous distance is determined, and traffic flow statistics are performed based on the magnitude between the current distance and the previous distance.

[0021] If the current distance is greater than the preset distance threshold, and the vessel to be tested has been counted, then the state of the vessel to be tested is reset;

[0022] Wherein, the current distance refers to the distance between the vessel under test and the designated lidar at the current moment, and the previous distance refers to the distance between the vessel under test and the designated lidar at the previous moment.

[0023] According to the ship navigation dimensions and traffic flow detection method provided by the present invention, the step of performing traffic flow statistics based on the difference between the current distance and the previous distance includes:

[0024] If the current distance is less than the previous distance, it is determined that the vessel under test is approaching the designated water section;

[0025] If the current distance is greater than the previous distance, it is determined that the vessel under test has passed the designated water section, and the vessel under test is marked as counted.

[0026] According to the ship navigation dimensions and traffic flow detection method provided by the present invention, after obtaining the ship dimensions of the ship to be tested, the method further includes:

[0027] Based on the radar ship position indicated by the ship point cloud data, the radar ship position is converted into the standard coordinate system of Automatic Identification System (AIS).

[0028] Obtain the AIS vessel position reported in the AIS data;

[0029] The radar vessel position at the same time is matched with the AIS vessel position, and the identity of the vessel under test is verified based on the matching result.

[0030] According to the ship navigation dimensions and traffic flow detection method provided by the present invention, the step of performing point cloud processing on the original laser point cloud data to obtain the ship point cloud data of the ship to be tested includes:

[0031] Based on the intensity values ​​of each point cloud in the original laser point cloud data, the initial ship point cloud data of the ship to be tested is obtained by filtering.

[0032] The initial ship point cloud data is transformed from the lidar coordinate system to the world coordinate system to obtain the transformed ship point cloud data;

[0033] After performing multi-level directional filtering, grid sampling, representative point selection, and outlier removal on the converted ship point cloud data, the ship point cloud data of the ship under test is obtained.

[0034] The present invention also provides a device for detecting ship navigation dimensions and traffic flow, comprising:

[0035] Point cloud acquisition unit, used to acquire raw laser point cloud data;

[0036] A point cloud processing unit is used to process the original laser point cloud data to obtain the ship point cloud data of the ship under test. The point cloud processing includes at least one of point cloud filtering, coordinate system transformation, and point cloud filtering.

[0037] The scale calculation unit is used to calculate the size of the minimum bounding box of the ship point cloud data based on the main direction of the ship point cloud data, and to smooth the size of the minimum bounding box to obtain the ship scale of the ship under test.

[0038] The traffic flow detection unit is used to calculate the distance between the vessel under test and the designated lidar at the current time and the previous time, respectively, based on the vessel point cloud data, and to perform traffic flow detection based on the distance.

[0039] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the ship navigation dimensions and traffic flow detection method as described above.

[0040] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the ship navigation dimensions and traffic flow detection method as described above.

[0041] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the ship navigation dimensions and traffic flow detection method as described above.

[0042] The present invention provides a method, apparatus, equipment and medium for detecting ship navigation dimensions and traffic flow. By using laser point cloud data, it enables simultaneous monitoring of ship dimensions and detection of traffic flow. This technology can not only accurately measure ship dimensions, but also acquire traffic flow data in real time, significantly improving detection efficiency while reducing equipment costs and deployment complexity.

[0043] Furthermore, by continuously scanning the ship with lidar and combining it with advanced point cloud data processing algorithms, high-precision ship dimension data (length, width, and height) can be calibrated and output in real time, effectively solving the problem of insufficient accuracy in real-time measurement using traditional methods. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in this 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 this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0045] Figure 1 This is a flowchart illustrating the ship navigation dimensions and traffic flow detection method provided by the present invention.

[0046] Figure 2This is a schematic diagram of the transmitted ship point cloud image provided by the present invention.

[0047] Figure 3 This is a schematic diagram of the structure of the ship navigation dimensions and traffic flow detection device provided by the present invention.

[0048] Figure 4 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0050] During navigation on inland waterways, bridges, locks, and other structures impose certain restrictions on the height, width, and even traffic flow density of vessels. Failure to strictly adhere to these restrictions and deviating from the prescribed procedures can pose significant safety hazards to both the vessels themselves and the bridges and locks. Relevant technologies include methods for real-time measurement of vessel dimensions such as:

[0051] 1. Video Surveillance and Image Analysis Method. By installing cameras at key locations on the bridge, the size, speed, and heading of vessels are analyzed to promptly identify potential risks. Specifically, a binocular vision measurement algorithm uses two cameras to capture images of the same scene from different angles, utilizing parallax information to calculate the depth and 3D position of objects. This allows for real-time monitoring of vessel height and position, ensuring safe passage across the bridge. However, video surveillance is highly susceptible to environmental factors; lighting conditions and weather (such as fog, rain, and snow) can affect image quality, thus impacting measurement accuracy. Furthermore, image processing and analysis algorithms are complex, especially binocular vision measurement, which requires significant computational resources.

[0052] 2. LiDAR Measurement Method. LiDAR emits laser pulses and detects their reflection from target objects to obtain information such as the target's position and velocity. The control system calculates the object's distance and orientation based on the speed of light and the reflection time difference, generating a 3D environment model and calculating the ship's dimensions, including length, width, and height. The calculation accuracy of the LiDAR method is higher than that of video surveillance and image analysis methods. LiDAR can maintain stable operation at night or in low-light conditions and has some penetration capability through fog and smoke. However, the generated 3D point cloud data requires complex algorithms for processing and analysis.

[0053] In recent years, with the development of drone technology, drones equipped with high-definition cameras or lidar devices have been applied in ship dimension measurement. In addition, multi-sensor data fusion has become a hot topic in related fields. By utilizing information from different types of sensors, fusion algorithms can achieve complementarity and synergy between data from different sensors.

[0054] To address the problems of poor detection accuracy, low detection efficiency, high equipment cost, and high deployment complexity in related technologies, this invention proposes a method for detecting ship navigation dimensions and traffic flow. This method first processes the raw laser point cloud data to obtain the ship's point cloud data, where point cloud processing includes at least one of point cloud filtering, coordinate system transformation, and point cloud filtering. Then, by calculating the dimensions of the minimum bounding cube of the ship's point cloud data, the ship's dimensions are obtained, achieving millimeter-level measurement accuracy. Simultaneously, based on the ship's point cloud data, the distances between the ship and a designated laser radar at the current time and the previous time are calculated, and traffic flow detection is performed based on these distances.

[0055] This invention innovatively utilizes laser point cloud data to simultaneously monitor ship dimensions and detect traffic flow. This technology can not only accurately measure ship dimensions but also acquire traffic flow data in real time, significantly improving detection efficiency while reducing equipment costs and deployment complexity.

[0056] Furthermore, by continuously scanning the ship with lidar and combining it with advanced point cloud data processing algorithms, high-precision ship dimension data (length, width, and height) can be calibrated and output in real time, effectively solving the problem of insufficient accuracy in real-time measurement using traditional methods.

[0057] This invention can be applied to scenarios requiring the detection of ship navigation dimensions and traffic flow. For example, it can provide early warnings and intervention for oversized ships preparing to pass through bridges or locks, and can also statistically analyze the number of ships passing through within a selected period, achieving reliable real-time monitoring of ship navigation dimensions and traffic flow. The execution subject of this method can be electronic equipment such as terminal devices, computers, servers, server clusters, or specially designed detection equipment, or it can be a detection device installed in such electronic equipment, which can be implemented through software, hardware, or a combination of both.

[0058] Figure 1 This is a flowchart illustrating the ship navigation dimensions and traffic flow detection method provided by the present invention, as shown below. Figure 1 As shown, the method includes the following steps:

[0059] Step 110: Obtain the raw laser point cloud data.

[0060] Specifically, raw laser point cloud data refers to the 3D point cloud data of the surrounding environment, including the ship, collected by a 3D lidar sensor. LiDAR acquires 3D point cloud data of the surrounding environment by emitting laser beams and receiving reflected signals. Each raw laser point cloud data point typically contains the following information:

[0061] x, y, z: The coordinates of a point in three-dimensional space.

[0062] Intensity: Reflection intensity, representing the intensity of the laser beam reflected back.

[0063] Step 120: Perform point cloud processing on the original laser point cloud data to obtain the ship point cloud data of the ship to be tested. Point cloud processing includes at least one of point cloud filtering, coordinate system transformation, and point cloud filtering.

[0064] Specifically, considering that the raw laser point cloud data acquired by lidar typically contains noise and invalid points, point cloud processing is required to improve the accuracy of ship size detection. This process yields the ship point cloud data for the vessel under test. Here, "ship under test" refers to the vessel whose dimensions need to be measured. The ship point cloud data for the vessel under test is obtained by removing point cloud noise and containing only the point cloud data of the vessel's surface.

[0065] Point cloud filtering refers to extracting a subset of data that meets specific conditions from the original laser point cloud data based on geometric range, attribute thresholds, or semantic rules.

[0066] Coordinate system transformation refers to transforming a point cloud from the local coordinate system of the acquisition device (such as the lidar coordinate system) to a global reference system (such as the world coordinate system).

[0067] Point cloud filtering refers to removing noise and retaining salient features through local weighted averaging or geometric simplification algorithms.

[0068] In this embodiment, at least one of point cloud screening, coordinate system transformation, and point cloud filtering can be performed on the original laser point cloud data to improve the accuracy of subsequent ship dimension inspection.

[0069] In some embodiments, step 120 specifically includes:

[0070] Step 121: Based on the intensity values ​​of each point cloud in the original laser point cloud data, the initial ship point cloud data of the ship to be tested is obtained by filtering.

[0071] Step 122: Transform the initial ship point cloud data from the lidar coordinate system to the world coordinate system to obtain the transformed ship point cloud data;

[0072] Step 123: After performing multi-level directional filtering, grid sampling, representative point selection, and outlier removal on the converted ship point cloud data, the ship point cloud data of the ship to be tested is obtained.

[0073] Specifically, the first step is to filter the raw laser point cloud data. The purpose of this step is to separate different objects from the raw laser point cloud data. Each point in the raw laser point cloud is traversed, and its intensity value is used to determine whether to retain it. An intensity threshold can be preset, for example, set to 25. If the intensity value of a point is less than 25, the point is filtered out; if the intensity value is greater than or equal to 25, the point is retained and assigned a color value. Through intensity filtering, some invalid points (such as water surface reflection points) can be removed, retaining the point cloud data of the target object (such as a ship), thus obtaining the initial ship point cloud data of the ship under test.

[0074] In step 122, considering that the point cloud data acquired by the lidar is usually represented in the lidar's own coordinate system, it is necessary to transform this point cloud data to a world coordinate system, such as the geocentric rectangular coordinate system, for subsequent processing and analysis. This can be achieved by rotating and translating the point cloud.

[0075] Rotation: Rotate the point cloud from the attitude of the lidar coordinate system to the attitude of the target coordinate system.

[0076] Translation: The point cloud is translated from the origin of the lidar coordinate system to the origin of the target coordinate system.

[0077] Define a 4x4 transformation matrix T, and set a rotation matrix R and a translation vector t. Iterate through each point in the initial input ship point cloud, transforming it from the lidar coordinate system to the world coordinate system.

[0078]

[0079] Where R is a 3x3 rotation matrix representing the rotation from the lidar coordinate system to the target coordinate system. t is a 3x1 translation vector representing the translation from the lidar coordinate system to the target coordinate system. R and t are obtained from the precise position and attitude of the lidar in the world coordinate system.

[0080] The coordinate system transformation formula for point clouds is:

[0081]

[0082] Represent each point as homogeneous coordinates (X... L Y L Z L ,1), and multiply with the transformation matrix T to obtain the transformed point (X). W YW Z W 1). The converted points are stored in the output point cloud to obtain the converted ship point cloud data.

[0083] In step 123, since the point cloud data acquired by the lidar usually contains noise and invalid points, the point cloud is filtered. This can be achieved through multi-level directional filtering, grid sampling, selection of representative points, and outlier removal.

[0084] Multi-level directional filtering: Filter the point cloud along the Z-axis, retaining points within a specified range; filter the point cloud along the Y-axis, retaining points within a specified range; filter the point cloud along the Z-axis, retaining points within a specified range. The specified range can be preset.

[0085] Mesh sampling: Determine the side length of each mesh cell (voxel) to be 0.01; based on the defined voxel size, divide the entire point cloud space into several cubic meshes. For each voxel, find all points falling within it.

[0086] Select representative points: Calculate the centroid (i.e., the average coordinates of all points) of all points within each voxel, use the centroid as the representative point of that voxel, and construct a new point cloud using the selected representative points.

[0087] Outlier removal: Outliers are removed using statistical filtering or radius filtering methods.

[0088] If the number of sampled points is less than 50% of the original number of points, statistical filtering is used. The average distance is estimated by considering the 50 nearest neighbors. If the distance between a point and its neighbors deviates from 0.7 times the standard deviation of the average distance, it is considered an outlier.

[0089] If the number of sampled points is between 50% and 80% of the original number of points, radius filtering is applied, and a search radius of 2 meters is specified. The number of neighbors for each point within its specified 2-meter search radius is checked. If there are not enough neighbors within its search radius (e.g., less than the set minimum number of neighbors of 8), it is considered an outlier.

[0090] If the number of sampled points is greater than 80%, change the radius filtering parameters; for example, specify a search radius of 3 meters. Check the number of neighbors for each point within its specified 3-meter search radius. If there are not enough neighbors within its search radius (e.g., less than the set minimum number of neighbors of 6), it is considered an outlier.

[0091] The method provided in this invention, through an advanced point cloud preprocessing algorithm, can obtain more accurate point cloud data of the ship under test, thereby improving the accuracy of subsequent ship dimension calculations.

[0092] Step 130: Based on the main direction of the ship point cloud data, calculate the size of the minimum bounding cuboid of the ship point cloud data, smooth the size of the minimum bounding cuboid, and obtain the ship dimensions of the ship to be measured.

[0093] Specifically, the principal direction of ship point cloud data can be calculated using Principal Component Analysis (PCA) to determine the main extension direction of the point cloud data in three-dimensional space, typically corresponding to the ship's length, width, and height. The minimum bounding cuboid is the smallest cuboid that can completely enclose the ship point cloud data; its size reflects the distribution range of the point cloud in three-dimensional space. The calculated size of the minimum bounding cuboid is then filtered to reduce the impact of noise or outliers on the results, improving the stability and reliability of the data, thereby obtaining the ship dimensions of the vessel under test.

[0094] In some embodiments, step 130, based on the principal direction of the ship point cloud data, calculates the size of the minimum bounding cuboid of the ship point cloud data, specifically including:

[0095] Step 131: Calculate the centroid of the ship point cloud data and determine the main direction of the ship point cloud data based on the centroid;

[0096] Step 132: Project the ship point cloud data onto the main direction to obtain projected point cloud data;

[0097] Step 133: Determine the size of the minimum circumscribed cuboid based on the minimum and maximum coordinates of the projected point cloud data in each direction.

[0098] Specifically, the centroid refers to the geometric center of the ship's point cloud data, which can be obtained by calculating the mean coordinates of all points. The formula for calculating the centroid of a point cloud is as follows:

[0099]

[0100] in It is the first in point cloud There are N points, where N is the total number of points.

[0101] The main orientation of the ship point cloud data is then determined based on the centroid. This includes the following steps:

[0102] Step 131-1: Based on the coordinates of each point cloud in the centroid and ship point cloud data, calculate the covariance matrix. The covariance matrix reflects the distribution of the point cloud in various directions.

[0103] Step 131-2: Perform eigenvalue decomposition on the covariance matrix to obtain the eigenvalues ​​and corresponding eigenvectors in each direction;

[0104] Step 131-3: Use the eigenvector corresponding to the largest eigenvalue as the main direction of the ship point cloud data.

[0105] Specifically, by obtaining the coordinates of the centroid and each point in the ship's point cloud data, the covariance matrix can be calculated. The covariance matrix can be expressed by the formula:

[0106]

[0107] Eigenvalue decomposition of the covariance matrix yields three eigenvalues ​​λ1, λ2, λ3 and corresponding eigenvectors v1, v2, v3:

[0108]

[0109] in, It is an eigenvalue. It is the corresponding feature vector.

[0110] The eigenvector corresponding to the largest eigenvalue is the main direction of the point cloud (i.e., the direction in which the point cloud is most widely distributed).

[0111] After obtaining the main direction of the ship point cloud data, the ship point cloud data is projected onto the main direction to obtain the projected point cloud data.

[0112] Construct a 3x3 matrix whose column vectors are the eigenvectors of the covariance matrix, and sort the eigenvectors by their eigenvalues. Generate a third eigenvector by cross-producting the first and second eigenvectors, ensuring it is orthogonal to the first two directions and satisfies the right-hand rule. Transpose the eigenvector matrix to obtain the rotation matrix R'.

[0113] Projecting ship point cloud data onto the main direction can be expressed by the following formula:

[0114]

[0115] Calculate the minimum and maximum coordinates of the projected point cloud:

[0116]

[0117] Calculate the dimensions of the smallest circumscribed cuboid:

[0118]

[0119] in, , Let these be the maximum and minimum values ​​of the point cloud along its length. , These are the maximum and minimum values ​​of the point cloud in the width direction. , These are the maximum and minimum values ​​of the point cloud in the height direction.

[0120] Based on the dimensions of the minimum circumscribed cuboid, the dimensions of the minimum circumscribed cuboid are smoothed to reduce the impact of noise, and the smoothed dimensions are used as the ship dimensions of the ship under test.

[0121] In some embodiments, smoothing the dimensions of the minimum circumscribed cuboid can be achieved using Kalman filtering.

[0122] Kalman filtering is a recursive filtering algorithm that can effectively estimate the state of a system. Kalman filtering consists of two steps: prediction and update. It achieves optimal estimation by dynamically adjusting the weights of the predicted and measured values.

[0123] Prediction steps:

[0124]

[0125] State prediction: based on the estimate from the previous time step. Predict the state at the current moment. A=1 indicates that the state remains unchanged (such as in a uniform motion model).

[0126] Covariance prediction: prediction error covariance Reflecting the uncertainty of the prediction, Q is the process noise covariance, which represents the uncertainty of the model (such as acceleration disturbances).

[0127] Update steps:

[0128] Calculate the Kalman gain:

[0129]

[0130] Kalman gain The weights of the measurements are determined. R'' is the measurement noise covariance, which represents the noise level of the sensor.

[0131] Status Update:

[0132]

[0133] Fusion Predictions Given the measured value zk, output the optimal estimate. .

[0134] If the measurement noise is low (R''→0), the measured values ​​are more trusted; if the model noise is low (Q→0), the predicted values ​​are more trusted.

[0135] Covariance update:

[0136]

[0137] Update the estimated error covariance This reflects the uncertainty of the current state.

[0138] The ship's length, width, and height, obtained after smoothing by Kalman filtering, are sent to a remote server via TCP protocol. Figure 2 This is a schematic diagram of the transmitted ship point cloud image provided by the present invention.

[0139] Step 140: Based on the ship point cloud data, calculate the distance between the ship under test and the designated lidar at the current time and the previous time, and perform traffic flow detection based on the distance.

[0140] Specifically, this embodiment, through the processing of laser point cloud data, can not only accurately measure the dimensions of ships, but also acquire traffic flow data in real time, accurately count the number, movement status and direction of ships in transit, overcome the problem of inaccurate ship count acquisition in traditional statistical methods, and provide reliable data support for traffic flow analysis.

[0141] For traffic flow detection, this is achieved by analyzing the distances between the vessel under test and a designated lidar at the current moment and the previous moment. These distances can be calculated using the Euclidean distance formula. A specific point (such as the vessel's geometric center or a specific marker point) is selected as a reference point in the vessel point cloud data. The distance between the reference point and the lidar position is then calculated using the Euclidean distance formula.

[0142] Analyze the trend of distance changes to determine the vessel's motion status (e.g., approaching, moving away, or stationary). Based on the distance change analysis, calculate traffic flow parameters such as the vessel's speed, acceleration, and whether it has passed through a designated waterway section. In this step, traffic flow detection refers to monitoring vessel traffic conditions within the waterway by analyzing parameters such as the distance change between the vessel and a designated lidar sensor, the vessel's speed, and its direction of travel.

[0143] In some embodiments, step 140, which involves traffic flow detection based on distance, specifically includes:

[0144] Step 141: If the current distance is less than or equal to the preset distance threshold and the ships to be measured have not yet been counted, then determine the size between the current distance and the previous distance, and perform traffic flow statistics based on the size between the current distance and the previous distance.

[0145] Step 142: If the current distance is greater than the preset distance threshold and the ship to be tested has been counted, then reset the state of the ship to be tested.

[0146] The current distance refers to the distance between the vessel under test and the designated lidar at the current moment, while the previous distance refers to the distance between the vessel under test and the designated lidar at the previous moment.

[0147] Specifically, the distance threshold is a preset value used to determine whether a ship is approaching or moving away from the lidar. When the current distance is less than or equal to the preset distance threshold, it may indicate that the ship is approaching the lidar; conversely, it may indicate that the ship is moving away. The distance threshold can be preset according to actual needs.

[0148] The determination of a vessel's status includes whether it has been counted. If a vessel has not been counted and its current distance is less than or equal to a preset distance threshold, it may indicate that a new vessel has entered the monitoring area (such as a designated water section), and monitoring needs to continue, such as counting and updating the vessel's status to "counted". If a vessel has been counted and its current distance is greater than a preset distance threshold, it may indicate that the vessel has left the monitoring area, and its status needs to be reset.

[0149] Here, the traffic flow statistics method includes making judgments based on the magnitude of the distance between the current distance and the previous distance. Analyzing the magnitude of the current distance and the previous distance determines the movement status of the vessel (e.g., approaching, moving away, or stationary). If the current distance is less than the previous distance, it may indicate that the vessel is approaching the lidar; conversely, it indicates that the vessel has passed through the designated water area. In this way, the vessel flow rate can be calculated, that is, the number of vessels passing through the monitored area per unit time.

[0150] Preferably, in step 141, traffic flow statistics are performed based on the difference between the current distance and the previous distance, including:

[0151] If the current distance is less than the previous distance, it is determined that the vessel under test is approaching the designated water section;

[0152] If the current distance is greater than the previous distance, it is determined that the vessel under test has passed the designated water section, and the vessel under test is marked as counted.

[0153] In some embodiments, the method for traffic flow statistics includes:

[0154] 1. Based on ship point cloud data, store the coordinate trajectory of each ship according to the timestamp.

[0155] 2. Extract the last two positions from the position record, representing the ship's position at the previous time and the current time, respectively, denoted as loc_prev and loc_curr.

[0156] 3. Calculate the planar distance from each location point to the known LiDAR using Euclidean distance. `dist_prev` and `dist_curr` represent the distances from the previous and current locations to the LiDAR, respectively.

[0157] 4. Determine whether a vessel has passed through a section of designated waterway:

[0158] If the current distance dist_curr is less than the set threshold and the vessel has not yet been counted, further checks are performed (close to the water section and not counted).

[0159] If the current distance dist_curr is smaller than the previous distance dist_prev, it means the vessel is approaching the water section. If the vessel is approaching the water section from a specified direction and loc_curr > loc_prev, the vessel is marked as counted, indicating that it has just passed the water section, and the flow count is incremented by 1.

[0160] If the current distance dist_curr is greater than the threshold, it means that the ship has left the waterway. At this time, the ship's status is reset so that it can be counted again the next time it passes through.

[0161] Based on any of the above embodiments, after obtaining the ship dimensions of the vessel to be tested in step 130, the method further includes:

[0162] Based on the radar ship position indicated by ship point cloud data, the radar ship position is converted into the standard coordinate system of Automatic Identification System (AIS).

[0163] Obtain the AIS vessel position reported in the AIS data;

[0164] The radar ship position at the same time is matched with the AIS ship position, and the identity of the ship to be tested is verified based on the matching result.

[0165] Specifically, AIS base station equipment is responsible for collecting, processing, and identifying AIS data of ships within a certain surrounding range, mainly serving functions such as ship positioning and identification. AIS data includes dynamic and static data. Dynamic data includes: ship's nine-digit code, ship positioning information (latitude and longitude), heading, speed, and equipment time; static data includes the nine-digit code, ship's English name, port of origin, and port of destination.

[0166] AIS relies on data autonomously reported by ships (such as position, speed, and length / width), which carries the risk of data falsification, signal delays, or equipment malfunctions. For example, a vessel violating regulations may disable AIS or falsify information. LiDAR, on the other hand, directly measures the ship's true size and precise position through point clouds, without relying on the ship's active cooperation, and can verify the authenticity of AIS data. Furthermore, even when the AIS signal is lost, LiDAR can still provide position and size. Therefore, this embodiment fuses AIS data with LiDAR data to authenticate the target ship, improving the reliability of target identification and forming a complete profile of the target vessel.

[0167] First, based on the radar ship position indicated by the ship point cloud data, the radar ship position is converted into the standard coordinate system of the Automatic Identification System (AIS).

[0168] WGS-84 (World Geodetic System 1984) is the standard reference coordinate system for AIS latitude and longitude data.

[0169]

[0170] Where r is the Earth's radius, this method converts the radar-detected ship position from the geocentric rectangular coordinate system to WGS-84 latitude and longitude.

[0171] AIS data includes the vessel's MMSI (Maritime Mobile Service Identity), position (latitude and longitude), speed, heading, length, and beam. LiDAR data includes the vessel's geometric center latitude and longitude, length, beam, height, and heading (calculated via PCA). Here, AIS vessel position refers to the vessel's position in the AIS data, and radar vessel position refers to the vessel's position indicated in the LiDAR data.

[0172] Align AIS and LiDAR data at the same time using timestamps. Match the radar vessel position with the AIS vessel position at the same time. For example, calculate the distance between the radar vessel position detected by LiDAR and the AIS vessel position reported by AIS. If the distance is less than a threshold (e.g., 50 meters), the position is considered to be a successful match, and the vessel is regarded as the same target, thus authentication is successful. If the distance is greater than the threshold, the position is considered to be a failed match, and authentication is unsuccessful.

[0173] AIS's latitude and longitude coordinates and semantic data (such as vessel type and destination), combined with LiDAR's geometric data, can cross-verify target identity, improve the reliability of target identification, and form a complete profile of the target vessel. Through data fusion, managers can achieve a paradigm shift from "passive response" to "proactive early warning," and from "local monitoring" to "global intelligence."

[0174] The method provided in this invention solves the limitation of lidar in being unable to identify ship identities by accurately fusing point cloud data with AIS information, and achieves the unification of ship size detection and identity recognition, providing comprehensive technical support for ship early warning, scheduling and other application scenarios.

[0175] The following describes the ship navigation dimensions and traffic flow detection device provided by the present invention. The ship navigation dimensions and traffic flow detection device described below can be referred to in correspondence with the ship navigation dimensions and traffic flow detection method described above.

[0176] Figure 3 This is a schematic diagram of the structure of the ship navigation dimensions and traffic flow detection device provided by the present invention, as shown below. Figure 3 As shown, the device includes:

[0177] Point cloud acquisition unit 310 is used to acquire raw laser point cloud data;

[0178] The point cloud processing unit 320 is used to perform point cloud processing on the original laser point cloud data to obtain the ship point cloud data of the ship under test. The point cloud processing includes at least one of point cloud filtering, coordinate system transformation, and point cloud filtering.

[0179] The scale calculation unit 330 is used to calculate the size of the minimum bounding box of the ship point cloud data based on the main direction of the ship point cloud data, and to smooth the size of the minimum bounding box to obtain the ship scale of the ship under test.

[0180] The traffic flow detection unit 340 is used to calculate the distance between the vessel under test and the designated lidar at the current time and the previous time, respectively, based on the vessel point cloud data, and to perform traffic flow detection based on the distance.

[0181] The device provided in this invention, based on laser point cloud data, enables simultaneous monitoring of ship dimensions and traffic flow detection. This technology can not only accurately measure ship dimensions but also acquire traffic flow data in real time, significantly improving detection efficiency while reducing equipment costs and deployment complexity.

[0182] Furthermore, by continuously scanning the ship with lidar and combining it with advanced point cloud data processing algorithms, high-precision ship dimension data (length, width, and height) can be calibrated and output in real time, effectively solving the problem of insufficient accuracy in real-time measurement using traditional methods.

[0183] Based on any of the above embodiments, the scale calculation unit is specifically used for:

[0184] Calculate the centroid of the ship point cloud data, and determine the main direction of the ship point cloud data based on the centroid;

[0185] The ship point cloud data is projected onto the main direction to obtain projected point cloud data;

[0186] The dimensions of the minimum circumscribed cuboid are determined based on the minimum and maximum coordinates of the projected point cloud data in each direction.

[0187] Based on any of the above embodiments, the scale calculation unit is specifically used for:

[0188] Based on the centroid and the coordinates of each point cloud in the ship point cloud data, a covariance matrix is ​​calculated, which reflects the distribution of the point cloud in various directions.

[0189] The covariance matrix is ​​decomposed into eigenvalues ​​to obtain eigenvalues ​​and corresponding eigenvectors in each direction;

[0190] The eigenvector corresponding to the largest eigenvalue is taken as the main direction of the ship point cloud data.

[0191] Based on any of the above embodiments, the traffic flow detection unit is specifically used for:

[0192] If the current distance is less than or equal to a preset distance threshold, and the ships to be measured have not yet been counted, then the magnitude between the current distance and the previous distance is determined, and traffic flow statistics are performed based on the magnitude between the current distance and the previous distance.

[0193] If the current distance is greater than the preset distance threshold, and the vessel to be tested has been counted, then the state of the vessel to be tested is reset;

[0194] Wherein, the current distance refers to the distance between the vessel under test and the designated lidar at the current moment, and the previous distance refers to the distance between the vessel under test and the designated lidar at the previous moment.

[0195] Based on any of the above embodiments, the traffic flow detection unit is specifically used for:

[0196] If the current distance is less than the previous distance, it is determined that the vessel under test is approaching the designated water section;

[0197] If the current distance is greater than the previous distance, it is determined that the vessel under test has passed the designated water section, and the vessel under test is marked as counted.

[0198] Based on any of the above embodiments, the device further includes an authentication unit, used for:

[0199] Based on the radar ship position indicated by the ship point cloud data, the radar ship position is converted into the standard coordinate system of Automatic Identification System (AIS).

[0200] Obtain the AIS vessel position reported in the AIS data;

[0201] The radar vessel position at the same time is matched with the AIS vessel position, and the identity of the vessel under test is verified based on the matching result.

[0202] Based on any of the above embodiments, the point cloud processing unit is specifically used for:

[0203] Based on the intensity values ​​of each point cloud in the original laser point cloud data, the initial ship point cloud data of the ship to be tested is obtained by filtering.

[0204] The initial ship point cloud data is transformed from the lidar coordinate system to the world coordinate system to obtain the transformed ship point cloud data;

[0205] After performing multi-level directional filtering, grid sampling, representative point selection, and outlier removal on the converted ship point cloud data, the ship point cloud data of the ship under test is obtained.

[0206] Based on any of the above embodiments, the device further includes an electronic navigation chart module. The electronic navigation chart module integrates and accesses electronic navigation chart data from the S-57 / S-52 and CJ-57 / CJ-52 standards to realize the display function of the monitoring area on the system's main interface, and provides functions such as map zooming, panning, ship target overlay display, and ship target point selection query.

[0207] Figure 4 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 4 As shown, the electronic device may include a processor 410, a communication interface 420, a memory 430, and a communication bus 440. The processor 410, communication interface 420, and memory 430 communicate with each other via the communication bus 440. The processor 410 can call logical instructions in the memory 430 to execute a ship navigation scale and traffic flow detection method. This method includes: acquiring raw laser point cloud data; performing point cloud processing on the raw laser point cloud data to obtain ship point cloud data of the ship to be tested, wherein the point cloud processing includes at least one of point cloud filtering, coordinate system transformation, and point cloud filtering; calculating the size of the minimum bounding cuboid of the ship point cloud data based on the principal direction of the ship point cloud data, smoothing the size of the minimum bounding cuboid to obtain the ship scale of the ship to be tested; and calculating the distance between the ship to be tested and a designated lidar at the current time and the previous time, respectively, based on the ship point cloud data, and performing traffic flow detection based on the distance.

[0208] Furthermore, the logical instructions in the aforementioned memory 430 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0209] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the ship navigation dimensions and traffic flow detection methods provided by the above methods. The method includes: acquiring raw laser point cloud data; performing point cloud processing on the raw laser point cloud data to obtain ship point cloud data of the ship to be tested, wherein the point cloud processing includes at least one of point cloud filtering, coordinate system transformation, and point cloud filtering; calculating the size of the minimum bounding cuboid of the ship point cloud data based on the principal direction of the ship point cloud data, smoothing the size of the minimum bounding cuboid to obtain the ship dimensions of the ship to be tested; and calculating the distance between the ship to be tested and a designated lidar at the current time and the previous time, respectively, based on the ship point cloud data, and performing traffic flow detection based on the distance.

[0210] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the ship navigation dimensions and traffic flow detection methods provided by the above methods. The method includes: acquiring raw laser point cloud data; performing point cloud processing on the raw laser point cloud data to obtain ship point cloud data of the ship to be tested, wherein the point cloud processing includes at least one of point cloud filtering, coordinate system transformation, and point cloud filtering; calculating the size of the minimum bounding cube of the ship point cloud data based on the principal direction of the ship point cloud data, smoothing the size of the minimum bounding cube to obtain the ship dimensions of the ship to be tested; and calculating the distances between the ship to be tested and a designated lidar at the current time and the previous time, respectively, based on the ship point cloud data, and performing traffic flow detection based on the distances.

[0211] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0212] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0213] 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 of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for detecting ship navigation dimensions and traffic flow, characterized in that, include: Acquire raw laser point cloud data; The original laser point cloud data is processed to obtain the ship point cloud data of the ship under test. The point cloud processing includes at least one of point cloud filtering, coordinate system transformation and point cloud filtering. Calculate the centroid of the ship point cloud data, determine the principal direction of the ship point cloud data based on the centroid, calculate the size of the minimum bounding box of the ship point cloud data based on the principal direction of the ship point cloud data, and smooth the size of the minimum bounding box to obtain the ship dimensions of the ship under test. The step of determining the main direction of the ship point cloud data based on the centroid includes: calculating a covariance matrix based on the centroid and the coordinates of each point cloud in the ship point cloud data, wherein the covariance matrix reflects the distribution of the point cloud in each direction; performing eigenvalue decomposition on the covariance matrix to obtain eigenvalues ​​and corresponding eigenvectors in each direction; and taking the eigenvector corresponding to the largest eigenvalue as the main direction of the ship point cloud data. Based on the ship point cloud data, the distance between the ship under test and the designated lidar at the current time and the previous time are calculated, and traffic flow detection is performed based on the distance. The traffic flow detection based on the distance includes: if the current distance is less than or equal to a preset distance threshold and the vessel to be tested has not yet been counted, then the magnitude between the current distance and the previous distance is determined, and traffic flow statistics are performed based on the magnitude between the current distance and the previous distance; if the current distance is greater than the preset distance threshold and the vessel to be tested has been counted, then the state of the vessel to be tested is reset. Wherein, the current distance refers to the distance between the vessel under test and the designated lidar at the current moment, and the previous distance refers to the distance between the vessel under test and the designated lidar at the previous moment. The traffic flow statistics based on the difference between the current distance and the previous distance include: if the current distance is less than the previous distance, it is determined that the vessel under test is approaching the designated water area section; if the current distance is greater than the previous distance, it is determined that the vessel under test has passed the designated water area section, and the vessel under test is marked as counted. After obtaining the ship dimensions of the vessel to be tested, the method further includes: converting the radar ship position indicated by the ship point cloud data into the standard coordinate system of Automatic Identification System (AIS); obtaining the AIS ship position reported in the AIS data; matching the radar ship position and the AIS ship position at the same time; and verifying the identity of the vessel to be tested based on the matching result.

2. The method for detecting ship navigation dimensions and traffic flow according to claim 1, characterized in that, The step of calculating the size of the minimum bounding box of the ship point cloud data based on the main direction of the ship point cloud data includes: The ship point cloud data is projected onto the main direction to obtain projected point cloud data; The dimensions of the minimum circumscribed cuboid are determined based on the minimum and maximum coordinates of the projected point cloud data in each direction.

3. The method for detecting ship navigation dimensions and traffic flow according to any one of claims 1 to 2, characterized in that, The step of performing point cloud processing on the original laser point cloud data to obtain the ship point cloud data of the vessel under test includes: Based on the intensity values ​​of each point cloud in the original laser point cloud data, the initial ship point cloud data of the ship to be tested is obtained by filtering. The initial ship point cloud data is transformed from the lidar coordinate system to the world coordinate system to obtain the transformed ship point cloud data; After performing multi-level directional filtering, grid sampling, representative point selection, and outlier removal on the converted ship point cloud data, the ship point cloud data of the ship under test is obtained.

4. A device for detecting ship navigation dimensions and traffic flow, characterized in that, include: Point cloud acquisition unit, used to acquire raw laser point cloud data; A point cloud processing unit is used to process the original laser point cloud data to obtain the ship point cloud data of the ship under test. The point cloud processing includes at least one of point cloud filtering, coordinate system transformation, and point cloud filtering. The scale calculation unit is used to calculate the centroid of the ship point cloud data, determine the main direction of the ship point cloud data based on the centroid, calculate the size of the minimum bounding cuboid of the ship point cloud data based on the main direction of the ship point cloud data, and smooth the size of the minimum bounding cuboid to obtain the ship scale of the ship under test. The scale calculation unit is specifically used for: Based on the centroid and the coordinates of each point cloud in the ship point cloud data, a covariance matrix is ​​calculated, which reflects the distribution of the point cloud in each direction. The covariance matrix is ​​then decomposed into eigenvalues ​​to obtain eigenvalues ​​and corresponding eigenvectors in each direction. The eigenvector corresponding to the largest eigenvalue is taken as the main direction of the ship point cloud data. The traffic flow detection unit is used to calculate the distance between the vessel under test and the designated lidar at the current time and the previous time, respectively, based on the vessel point cloud data, and to perform traffic flow detection based on the distance. The traffic flow detection unit is specifically used for: If the current distance is less than or equal to a preset distance threshold and the vessel to be tested has not yet been counted, then the difference between the current distance and the previous distance is determined, and traffic flow statistics are performed based on the difference between the current distance and the previous distance; if the current distance is greater than the preset distance threshold and the vessel to be tested has already been counted, then the status of the vessel to be tested is reset. Wherein, the current distance refers to the distance between the vessel under test and the designated lidar at the current moment, and the previous distance refers to the distance between the vessel under test and the designated lidar at the previous moment. If the current distance is less than the previous distance, it is determined that the vessel under test is approaching the designated water section; if the current distance is greater than the previous distance, it is determined that the vessel under test has passed the designated water section, and the vessel under test is marked as counted. The device further includes an authentication unit for: Based on the radar vessel position indicated by the vessel point cloud data, the radar vessel position is converted into the standard coordinate system of Automatic Identification System (AIS); the AIS vessel position reported in the AIS data is obtained; the radar vessel position and the AIS vessel position at the same time are matched, and the identity of the vessel under test is verified based on the matching result.

5. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the ship navigation scale and traffic flow detection method as described in any one of claims 1 to 3.

6. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the ship navigation scale and traffic flow detection method as described in any one of claims 1 to 3.

Images