Method and system for identifying ship in navigation channel, electronic equipment and storage medium

Through the combination of deep learning models and radar data, ships in the waterway are identified and monitored, and the problem of poor management of ships entering and leaving the jurisdiction in the existing technology is solved, and real-time and accurate ship supervision and safety management are achieved.

CN118942057APending Publication Date: 2024-11-12WATER TRANSPORT PLANNING & DESIGN INST +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411363518.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-27
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

The existing technology is difficult to effectively manage ships entering and leaving the waters in the jurisdiction, and lacks all-around supervision methods, resulting in the inability to grasp ship information and status in real time and the inability to deal with safety hazards in a timely manner.

Method used

Deep learning model is used to identify ship image data in the waterway, match the position data collected by the radar to determine the target position of the ship, and measure the freeboard value and scale through laser scanning equipment to determine whether there is overload or overlimit behavior. Generate a result of a violation judgment.

Benefits of technology

It realizes the precise location and identity identification of ships entering and leaving the waters in the jurisdiction, provides real-time active supervision basis, improves the efficiency and safety of waterway management, and solves the problem of full-region coverage of ship management.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118942057A_ABST
    Figure CN118942057A_ABST
Patent Text Reader

Abstract

The invention discloses a method and system for identifying a ship in a navigation channel, electronic equipment and a storage medium, and relates to the technical field of navigation channel monitoring or other related technical fields, and the method for identifying the ship in the navigation channel comprises the steps: carrying out the uninterrupted scanning of the navigation channel, and collecting the image data of the ship passing through the navigation channel; identifying the image data of the ship by using a deep learning model to obtain a name and features of the ship; the name and the features of the ship are matched with position data collected by a radar, the target position of the ship is determined, and the position data comprise position information of the ship within a preset range around image collection equipment for collecting image data; and determining whether the ship at the target position violates rules or not, and generating a ship violation judgment result. According to the invention, the technical problem of lack of effective management of ships entering and exiting a jurisdiction water area in the prior art is solved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of waterway monitoring technology or other related fields, and in particular to a method and system for identifying ships in a waterway, an electronic device and a storage medium. Background Art

[0002] Waterway management is an important part of the transportation system and the comprehensive utilization of water resources. At present, when promoting the construction of smart ports and shipping, it is necessary to automatically identify ships in trunk waterways and key waters, improve the ship recognition rate, implement real-time monitoring of key ships and key docks, ferries, bridges, and waterways, and strengthen ship safety supervision and management.

[0003] Currently, when managing ships sailing in waterways, ship monitoring is often carried out through radars deployed around the waterways. There is a lack of means to supervise the entire waterway. At the same time, when monitoring waterways with a large width or area, the picture is not comprehensive due to the area of ​​the waterway, resulting in the inability to effectively manage ships entering and leaving the jurisdiction.

[0004] With the development of social economy, higher requirements are put forward for navigation safety management. How to effectively manage the ships entering and leaving the jurisdiction, grasp the information and status of the ships, and deal with the ships with safety hazards in time has become an urgent problem to be solved by the port and shipping management departments.

[0005] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0006] The embodiments of the present invention provide a method and system for identifying ships in a waterway, an electronic device and a storage medium, so as to at least solve the technical problem in the related art of lacking effective management of ships entering and leaving the waters under the jurisdiction.

[0007] According to one aspect of an embodiment of the present invention, a method for identifying ships in a waterway is provided, comprising: continuously scanning the waterway to collect image data of ships passing through the waterway; using a deep learning model to identify the image data of the ship to obtain the name and features of the ship; matching the name and features of the ship with position data collected by radar to determine a target position of the ship, wherein the position data includes position information of ships within a preset range around an image acquisition device that collects the image data; determining whether the ship at the target position has violated regulations, and generating a ship violation determination result.

[0008] Optionally, the deep learning model is trained by the following method: obtaining a training data set, wherein the training data set includes: images of ships passing through the waterway collected within a historical time period and labels corresponding to the images, and the labels include: the name and characteristics of the ship; training an initial neural network model based on the training data set to obtain the deep learning model.

[0009] Optionally, the name and features of the ship are matched with position data collected by radar to determine the target position of the ship, including: obtaining position information of the target ship corresponding to the name from a ship automatic identification system; and determining the position information of the target ship as the target position.

[0010] Optionally, after using a deep learning model to identify the image data of the ship and obtain the name of the ship, the method also includes: connecting the name of the ship to an interface of key tracking ship data; if there is a ship with the same name as the ship in the key tracking ship data, marking the name of the ship and generating a reminder message.

[0011] Optionally, the step of determining whether the ship at the target position is in violation includes: measuring the freeboard value and scale of the ship at the target position; comparing the freeboard value with a minimum safe freeboard threshold, and comparing the scale with a preset scale threshold to obtain a comparison result; when the comparison result indicates that the freeboard value is less than the minimum safe freeboard threshold, determining that the ship is overloaded; or, when the comparison result indicates that the scale is greater than the preset scale threshold, determining that the ship is over-limit.

[0012] Optionally, the method for identifying a ship in a waterway further includes: when it is determined that the ship is overloaded, and / or when it is determined that the ship is out of limit, sending the name of the ship and the target position of the ship to a monitoring terminal.

[0013] Optionally, the method for identifying ships in a waterway further includes: generating a trigger instruction when a ship is detected passing through the waterway; sending the trigger instruction to an image acquisition device and a laser scanning device arranged on both sides of the waterway, wherein the trigger instruction is used to trigger the image acquisition device to acquire image data of the ship passing through the waterway, or to trigger the laser scanning device to acquire the freeboard value and scale of the ship passing through the waterway.

[0014] According to another aspect of an embodiment of the present invention, a ship identification system in a waterway is also provided, comprising: a waterway image acquisition unit, for continuously scanning the waterway and collecting image data of ships passing through the waterway; an image recognition unit, for identifying the image data of the ship using a deep learning model to obtain the name and features of the ship; a ship position matching unit, for matching the name and features of the ship with the position data collected by radar to determine the target position of the ship, wherein the position data includes position information of ships within a preset range around an image acquisition device that collects the image data; a ship violation determination unit, for determining whether the ship at the target position has violated any regulations and generating a ship violation determination result.

[0015] Optionally, when training the deep learning model, the ship identification system in the waterway includes: a training set acquisition unit, used to acquire a training data set, wherein the training data set includes: images of ships passing through the waterway collected within a historical time period and labels corresponding to the images, and the labels include: the name and characteristics of the ship; and a model training unit, used to train an initial neural network model based on the training data set to obtain the deep learning model.

[0016] Optionally, the ship position matching unit includes: a ship position acquisition module, used to acquire the position information of the target ship corresponding to the name from the ship automatic identification system; and a position determination module, used to determine the position information of the target ship as the target position.

[0017] Optionally, the ship identification system in the waterway also includes: a name access unit, which is used to identify the image data of the ship using a deep learning model, and after obtaining the name of the ship, connect the name of the ship to the interface of the key tracking ship data; a ship marking unit, which is used to mark the name of the ship and generate a reminder message when there is a ship with the same name as the ship in the key tracking ship data.

[0018] Optionally, the ship violation judgment unit includes: a numerical measurement module, used to measure the freeboard value and scale of the ship at the target position; a numerical comparison module, used to compare the freeboard value with the minimum safe freeboard threshold, and compare the scale with a preset scale threshold to obtain a comparison result; a ship violation judgment module, used to determine that the ship is overloaded when the comparison result indicates that the freeboard value is less than the minimum safe freeboard threshold; or, when the comparison result indicates that the scale is greater than the preset scale threshold, determine that the ship is overloaded.

[0019] Optionally, the ship identification system in the waterway also includes: a position sending unit, which is used to send the name of the ship and the target position of the ship to the monitoring terminal when it is determined that the ship is overloaded and / or when it is determined that the ship is out of limit.

[0020] Optionally, the ship identification system in the waterway also includes: a trigger instruction generating unit, used to generate a trigger instruction when a ship is detected passing through the waterway; and an instruction sending unit, used to send the trigger instruction to an image acquisition device and a laser scanning device arranged on both sides of the waterway, wherein the trigger instruction is used to trigger the image acquisition device to collect image data of the ship passing through the waterway, or to trigger the laser scanning device to collect the freeboard value and scale of the ship passing through the waterway.

[0021] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is further provided, wherein the computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute any one of the above-mentioned methods for identifying ships in a waterway.

[0022] According to another aspect of an embodiment of the present invention, there is also provided an electronic device, comprising one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement any one of the above-mentioned methods for identifying ships in a waterway.

[0023] According to another aspect of an embodiment of the present invention, there is further provided a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the steps of any one of the above-mentioned methods for identifying a ship in a waterway are implemented.

[0024] In the present disclosure, uninterrupted scanning of the waterway is adopted to collect image data of ships passing through the waterway; the image data of the ship is identified by using a deep learning model to obtain the name and characteristics of the ship; the name and characteristics of the ship are matched with the position data collected by the radar to determine the target position of the ship, wherein the position data is the position information of the ship within a preset range around the image acquisition device used to collect the image data; a method for determining whether the ship at the target position has violated the regulations is adopted, and the image data of the ship is processed by using a deep learning model to realize the recognition of the name and characteristics of the ship, and at the same time, the recognition data and the position data are intelligently matched, so as to achieve the purpose of determining the precise position and identity of the ship while capturing the image, and providing a basis for real-time active supervision, thereby achieving the technical effect of realizing effective management of ships entering and leaving the waters under the jurisdiction, thereby solving the technical problem of lack of effective management of ships entering and leaving the waters under the jurisdiction in the related technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0026] Figure 1 is a flow chart of an optional method for identifying a ship in a waterway according to an embodiment of the present invention;

[0027] Figure 2 is a schematic diagram of an optional system for monitoring waterway dynamic data according to an embodiment of the present invention;

[0028] Figure 3 The present invention is a hardware structure block diagram of an electronic device (or mobile device) for a method for identifying ships in a waterway according to an embodiment of the present invention. DETAILED DESCRIPTION

[0029] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the 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 ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0030] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0031] To facilitate those skilled in the art to understand the present invention, some terms or nouns involved in the embodiments of the present invention are explained below:

[0032] Automatic Identification System, AIS for short, is a technical system used for communication and monitoring between ships. The AIS system collects information such as the ship's location, navigation status, heading and speed through satellites and ground base stations, and transmits this information to other ships, ship traffic control centers and coastal monitoring stations to improve the safety and navigation efficiency of ships.

[0033] The following embodiments of the present invention can be applied to systems / applications / devices for monitoring dynamic data of various waterways. The present invention can process the image data of ships traveling in the waterway through a deep learning model to recognize the name and features of the ship, and intelligently match the recognition data with the location data to achieve effective management of ships entering and leaving the waters under the jurisdiction.

[0034] The present invention is described in detail below in conjunction with various embodiments.

[0035] Embodiment 1

[0036] According to an embodiment of the present invention, an embodiment of a method for identifying ships in a waterway is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in an order different from that shown here.

[0037] Figure 1 is a flow chart of an optional method for identifying ships in a waterway according to an embodiment of the present invention. Figure 1 As shown, the method comprises the following steps:

[0038] Step S101, continuously scan the waterway to collect image data of ships passing through the waterway.

[0039] Step S102: Use the deep learning model to identify the image data of the ship to obtain the name and features of the ship.

[0040] Step S103, matching the name and features of the ship with the position data collected by the radar to determine the target position of the ship, wherein the position data includes the position information of the ship within a preset range around the image acquisition device that collects the image data.

[0041] Step S104, determining whether the ship at the target location has violated any regulations, and generating a ship violation determination result.

[0042] Through the above steps, the waterway can be scanned uninterruptedly to collect image data of ships passing through the waterway; the image data of the ship can be identified using a deep learning model to obtain the name and features of the ship; the name and features of the ship can be matched with the position data collected by the radar to determine the target position of the ship, wherein the position data includes the position information of the ship within a preset range around the image acquisition device that collects the image data; determine whether the ship at the target position has violated the regulations, and generate a ship violation judgment result. In this embodiment, the image data of the ship is processed by using a deep learning model to realize the recognition of the name and features of the ship, and at the same time, the recognition data and the position data are intelligently matched, so as to achieve the purpose of determining the precise location and identity of the ship while capturing and providing a basis for real-time active supervision, thereby achieving the technical effect of effectively managing ships entering and leaving the waters of the jurisdiction, thereby solving the technical problem of the lack of effective management of ships entering and leaving the waters of the jurisdiction in the related technology.

[0043] The embodiment of the present invention is described in detail below in combination with the above steps.

[0044] Step S101, continuously scan the waterway to collect image data of ships passing through the waterway.

[0045] According to an optional embodiment of the present application, when executing step S101, a multi-laser sensor is used as a means of collecting ship data, and the adjustment and rotation of the laser ranging head are utilized to achieve uninterrupted scanning of the waterway. After the collected data is converted into a coordinate system, the characteristics of the ship are identified.

[0046] Laser sensor is a sensor that uses laser technology for measurement. It consists of a laser, a laser detector and a measuring circuit. Laser sensor is a new type of measuring instrument. Its advantages are that it can achieve non-contact long-distance measurement, fast speed, high accuracy, large range, and strong anti-photoelectric interference ability.

[0047] In step S101, the collected image data of the ship includes photos of the ship and video data of the ship. By using the laser sensor to continuously scan the ship, it is possible to avoid the phenomenon of missing the scanning of some ships passing through the waterway.

[0048] Step S102: Use the deep learning model to identify the image data of the ship to obtain the name and features of the ship.

[0049] In this step, ship identification and analysis technology is used, based on the basic data of the ship and the video capture images, and a deep learning model is used to recognize the ship's name and characteristics.

[0050] It should be noted that the ship name mentioned in this embodiment refers to the ship name of a type of ship such as a fishing boat, a cargo ship or a cruise ship.

[0051] According to an optional embodiment of the present application, the deep learning model in step S102 is trained by the following method: obtaining a training data set, wherein the training data set includes: images of ships passing through the waterway collected during a historical time period and labels corresponding to the images, and the labels include: the name and characteristics of the ship; training an initial neural network model based on the training data set to obtain a deep learning model.

[0052] In the embodiments provided in the present application, support vector machine (SVM) is used to train the above-mentioned deep learning model. Support vector machine is a generalized linear classifier that performs binary classification of data in a supervised learning manner, and its decision boundary is the maximum margin hyperplane solved for the learning sample. SVM uses hinge loss function to calculate empirical risk and adds regularization term to the solution system to optimize structural risk. It is a classifier with sparsity and robustness. SVM can perform nonlinear classification by kernel method, which is one of the common kernel learning methods.

[0053] The training process of a deep learning model consists of the following steps:

[0054] 1) Data preprocessing.

[0055] The collected ship image data is subjected to image scale normalization and data enhancement. The image size set by image scale normalization is not limited and can be adjusted according to the image processing requirements and the image processing capabilities of the equipment. For example, the image size is set to 256×256 pixels. Data enhancement is to flip the ship image, flip it upside down, flip it left to right, and flip it diagonally, respectively, to increase the amount of ship image sample data.

[0056] After preprocessing the ship image data, the preprocessed ship image data is divided into a training set and a test set. In the embodiment provided in the present application, for example, the image ratio of the training set to the test set can be set to 7:3.

[0057] 2) Feature vector extraction.

[0058] The Histogram of Oriented Gradient (HOG) feature extraction is performed on the ship images in the training set and the test set, respectively, to obtain the HOG feature vector set of the ship images in the training set and the HOG feature vector set of the ship images in the test set. HOG feature is a feature description widely used in computer vision and image processing, mainly for target detection. HOG feature constructs feature vectors by calculating the gradient directional histogram of the local area of ​​the image, and these feature vectors can be used for image recognition.

[0059] Specifically, HOG feature extraction includes the following steps:

[0060] 2.1) Grayscale the image.

[0061] Image grayscale is to convert a color image into a grayscale image. The color image can be processed directly, that is, the color channel RGB (Red, Green, Blue) components of each pixel in the image are weighted averaged according to the sensitivity of the human eye to the three components, and the grayscale value of the pixel is obtained according to a certain weight, and the grayscale value is normalized to obtain the normalized image matrix I. The image grayscale processing formula is as follows:

[0062] Gray=0.3R+0.59G+0.11B;

[0063] Among them, R, G, and B represent the red, green, and blue color channels respectively, and 0.3, 0.59, and 0.11 represent weights.

[0064] 2.2) Gamma processing.

[0065] Gamma processing is performed on each pixel value I(x, y) in the image matrix I to increase or decrease the overall brightness of the image, adjust the image contrast, reduce the impact of light on the image, and reduce the impact of image illumination non-uniformity. The gamma processing formula is as follows:

[0066]

[0067] Among them, I ′ (x, y) is the pixel value of the image after Gamma processing, I(x, y) is the pixel value of the image before Gamma processing, C Gamma As the correction value, take CGamma =0.5.

[0068] 2.3) Gradient calculation.

[0069] Calculate the horizontal and vertical gradients of each pixel in the image matrix after Gamma processing, obtain the gradient direction and gradient amplitude of each pixel, and thus obtain the contour information of the image. The image can be filtered by the horizontal template K = [-1, 0, 1] and its transpose, so that the calculation of the horizontal and vertical gradients can be easily completed.

[0070] Calculate pixel I ′ Horizontal gradient at (x,y): g x =I ′ (x+1,y)-I ′ (x-1,y).

[0071] Calculate pixel I ′ The vertical gradient at (x,y): g x =I ′ (x,y+1)-I ′ (x,y-1).

[0072] Calculate pixel I ′ Gradient magnitude at (x,y):

[0073] Calculate pixel I ′ Gradient direction at (x,y):

[0074] 2.4) Construct a directional gradient histogram for each cell unit.

[0075] The image is decomposed into several cells. Cells are the smallest statistical and processing units in the process of algorithm decomposition and processing of images. Each cell contains m×m pixels. The gradient direction is quantized into 9 histogram channels, representing 9 different gradient directions. Based on these 9 histogram channels, each pixel in the cell is weighted to vote. The size of the gradient is used as a projection, and the projection is superimposed on the histogram represented by the 9 gradient directions in turn. For example, if the size of each cell is set to 8×8 pixels, the image has a total of 32×32 cell units.

[0076] 2.5) HOG feature integration.

[0077] Extract the features of each cell unit of the ship image, and connect the features of each cell unit in series to obtain the overall features of the image. First, it is necessary to combine n×n cell units into large, spatially connected intervals (blocks), and connect the gradients of the 9 histogram channels in each cell unit in series to obtain a multidimensional vector with a dimension of n×n×9 as the local feature of the interval. For example, if the size of the interval is set to 2×2 cell units, the dimension of the feature vector of the interval is 2×2×9. Secondly, collect local features of all overlapping intervals in the detection window (i.e., the image with a size normalized to 256*256), and combine the collected local features into the HOG feature vector of the ship image for use by the classifier.

[0078] 3) Model training.

[0079] The label information representing the ship image data in the training set and the HOG feature vector set of the ship images in the training set are imported into the SVM classifier for training to obtain the SVM multi-classification model.

[0080] Among them, SVM is a binary classification model. Its basic model is a linear classifier with the largest margin defined in the feature space. The largest margin makes it different from the perceptron. SVM also includes kernel techniques, which makes it a nonlinear classifier in essence. The core idea of ​​the SVM algorithm is to map the vector into a high-dimensional space and establish a segmentation hyperplane with the largest margin in this space.

[0081] After the model training is completed, the label information and HOG feature vector of a ship image in the test set are respectively input into the SVM multi-classification model. The SVM multi-classification model automatically counts the ship category to which the ship image belongs in the form of voting, and takes the ship category with the most votes as the category of the ship image in the test set. The category of the ship image is matched with the label information of the ship image to determine whether the image recognition is correct. If the match is successful, the image recognition is correct, and the category of the ship image is automatically saved.

[0082] There are many types of ships entering the monitored waterway every day, for example, fishing boats, container ships used to transport goods, and cruise ships. Considering the above application scenarios, the technical solution provided by this application uses SVM to train a deep learning model to identify the image data of ships, which can achieve more accurate identification and distinction of various types of ships. And by training a deep learning model to identify ships, it can not only reduce the error rate of manual identification, but also save labor costs and improve identification efficiency.

[0083] After the names and features of ships passing through the waterway are identified through the deep learning model, in this step, the location information of the target ship corresponding to the identified ship name is found from the ship automatic identification system, and then the location information is used as the location information of the identified ship.

[0084] Through the above technical solution, the precise location of the ship can be determined using the AIS system, and the ship flow and uplink and downlink conditions passing through the section can be accurately counted around the clock, solving the problem of poor images and difficulty in searching videos in the waterway video surveillance system at night.

[0085] It should be noted that the deep learning model mentioned in this embodiment can be a visual module plug-in multimodal retrieval model MARVEL, which can perform ship feature search on multimodal information (including monitored ship images and tracked ship information documents) to extract image semantic information, including extraction of ship name, shape, type, length, width, height, draft depth and other information.

[0086] As an optional embodiment of the present application, after the image data of the ship is recognized by a deep learning model and the name of the ship is obtained, the name of the ship is connected to the interface of the key tracking ship data; if there is a ship with the same name as the ship in the key tracking ship data, the name of the ship is marked and a reminder message is generated.

[0087] In terms of dynamic control, the system recognizes the identity of the ship (including name and characteristics) through a deep learning model, accesses the system's key tracking data interface for ships, and implements automatic ship identity tagging and intelligent reminders, reducing tedious manual operations.

[0088] Step S103, matching the name and features of the ship with the position data collected by the radar to determine the target position of the ship, wherein the position data includes the position information of the ship within a preset range around the image acquisition device that collects the image data.

[0089] According to another optional embodiment of the present application, step S103 is executed to match the name and characteristics of the ship with the position data collected by the radar to determine the target position of the ship, which is achieved by the following method: obtaining the position information of the target ship corresponding to the name from the ship automatic identification system; and determining the position information of the target ship as the target position.

[0090] According to an optional embodiment of the present application, the ship automatic identification system AIS and radar data are combined to obtain real-time position data of ships around the capture camera, and the identification data and position data are intelligently matched, so as to determine the precise position and identity of the ship at the same time as the capture, providing a basis for real-time active supervision.

[0091] Step S104, determining whether the ship at the target location has violated any regulations, and generating a ship violation determination result.

[0092] After determining the location of the ship, it is possible to further determine whether the ship has committed any violation, such as overloading or exceeding the limit.

[0093] In some optional embodiments of the present application, step S104 is executed to determine whether the ship at the target position has violated the regulations, which is achieved by the following method: measuring the freeboard value and scale of the ship at the target position; comparing the freeboard value with the minimum safe freeboard value and comparing the scale with a preset scale; if some freeboard values ​​are less than the minimum safe freeboard value, determining that the ship is overloaded; and / or if the scale is greater than the preset scale, determining that the ship is over-limit.

[0094] In terms of electronic supervision, radar laser panoramic scanning is used to detect the ship's size in real time, identify the ship's freeboard, and provide basic data support for data analysis to determine overloading and over-limit phenomena.

[0095] Freeboard refers to the vertical distance from the fully loaded waterline to the upper edge of the deck in the middle of the ship. The minimum freeboard of a ship is the limit that ensures that the ship floats safely on the water. All ships have the minimum freeboard specified by the pre-determined ship load line. If the ship is overloaded and the freeboard is reduced to less than the specified limit, the ship cannot float safely on the water, so the minimum freeboard is also called the safety freeboard. The size of the freeboard is of great significance to the stability of the ship at large angles. The increase in freeboard can improve the ability of the hull to recover to its initial state from a large angle of heel, thereby preventing the ship from capsizing. At the same time, the freeboard can effectively reduce the probability of waves on the deck. If the freeboard is too small, the waves will continue to impact the deck, causing damage to the deck structure and equipment, and posing a threat to the safety of the crew on the deck.

[0096] In an embodiment of the present application, a laser scanning device is used to measure the freeboard value of a ship, and the measured freeboard value is compared with the minimum safe freeboard value. If the measured freeboard value is smaller than the minimum safe freeboard value, it indicates that the ship is overloaded.

[0097] Detecting whether a ship is over-limit mainly involves detecting whether the length and width of the ship (i.e. the dimensions of the ship mentioned above) exceed the preset length and preset width respectively. If the length of the ship exceeds the preset length, or the width of the ship exceeds the preset width, it is determined that the ship is over-limit.

[0098] As an optional embodiment of the present application, when it is determined that the ship is overloaded and / or exceeds the limit, the name of the ship and the target location of the ship are sent to the monitoring terminal.

[0099] The collected and calculated ship data is associated with the ship files to determine whether the ships entering the monitoring area are in violation of regulations. When violations such as overloading and over-limit occur, the system intelligently pushes early warning information to the supervisors to achieve the effect of active supervision.

[0100] In other optional embodiments of the present application, when the presence of a ship passing through a waterway is detected, a trigger instruction is generated; the trigger instruction is sent to an image acquisition device and a laser scanning device arranged on both sides of the waterway, wherein the trigger instruction is used to trigger the image acquisition device to acquire image data of the ship passing through the waterway, and to trigger the laser scanning device to acquire the freeboard value and scale of the ship passing through the waterway.

[0101] In the embodiment of the present application, the freeboard of the ship is measured by laser scanning equipment, and the channel scene is recorded by closed-circuit television monitoring system (CCTV) equipment, and a high-definition intelligent card-entry system based on video detection is built. When the ship passes by, the front-end equipment is triggered to capture and laser scan the freeboard value of the ship, and 6-8 images are captured and 2-3 15-30 seconds of video are shot automatically. In the image acquisition method, megapixel high-resolution digital image acquisition technology, intelligent visual analysis, and pattern recognition technology are used to improve the image clarity.

[0102] CCTV system is an important part of the security technology prevention system. It is an advanced and comprehensive system with strong prevention capabilities. Its front end is generally composed of cameras and lenses, pan / tilt heads, encoders, dust covers, etc., which are used to obtain images of the monitored area. Its transmission part is generally composed of feeders, video cable compensators, video amplifiers, etc., and its function is to feed the video (sometimes including audio) signals output by the camera to the central computer room or other monitoring points. Its terminals generally include monitors, various control devices, and recording devices, etc., which are used for display and recording, video processing, output control signals, and receiving signals from the front end.

[0103] The above-mentioned technical solution provided in the embodiment of the present application processes the image data of the ship by utilizing a deep learning model to realize the recognition of the ship's name and features, and intelligently matches the recognition data with the location data, thereby achieving the purpose of determining the precise location and identity of the ship while capturing it, and providing a basis for real-time active supervision, thereby achieving the technical effect of effectively managing the ships entering and leaving the jurisdiction's waters.

[0104] Through the above technical solution, the purpose of taking pictures of every ship in an unmanned situation can be achieved, avoiding the phenomenon of missing ships.

[0105] By adopting the above-mentioned technical solution proposed in this application, it is possible to deploy smart checkpoints in key areas to form key defense lines, which have multiple functions such as public security management, water rescue, pollution prevention and control, and navigation safety. The system runs stably, has good recognition effect, and has broad application prospects.

[0106] The following is a detailed description in conjunction with another embodiment.

[0107] Embodiment 2

[0108] A waterway dynamic data monitoring system provided in this embodiment includes multiple implementation units, each implementation unit corresponding to each implementation step in the above-mentioned embodiment 1.

[0109] Figure 2 is a schematic diagram of an optional waterway dynamic data monitoring system according to an embodiment of the present invention, such as Figure 2 As shown, the monitoring system for waterway dynamic data may include: a waterway image acquisition unit 21, a recognition unit 22, a ship position matching unit 23, and a ship violation determination unit 24.

[0110] The channel image acquisition unit 21 is used to continuously scan the channel and acquire image data of ships passing through the channel.

[0111] The recognition unit 22 is used to recognize the image data of the ship using a deep learning model to obtain the name and characteristics of the ship.

[0112] The ship position matching unit 23 is used to match the name and characteristics of the ship with the position data collected by the radar to determine the target position of the ship, wherein the position data includes the position information of the ship within a preset range around the image acquisition device that collects the image data.

[0113] The ship violation determination unit 24 is used to determine whether the ship at the target location has violated any regulations and generate a ship violation determination result.

[0114] The above-mentioned waterway dynamic data monitoring system can continuously scan the waterway through the waterway image acquisition unit 21, collect image data of ships passing through the waterway, and use the recognition unit 22 to recognize the image data of the ship using a deep learning model to obtain the name and characteristics of the ship. The ship position matching unit 23 matches the name and characteristics of the ship with the position data collected by the radar to determine the target position of the ship, wherein the position data includes the position information of the ship within a preset range around the image acquisition device that collects the image data. The ship violation determination unit 24 determines whether the ship at the target position has violated the regulations and generates a ship violation determination result. In this embodiment, the image data of the ship is processed by using a deep learning model to realize the recognition of the name and characteristics of the ship, and the recognition data and the position data are intelligently matched, so as to achieve the purpose of determining the precise position and identity of the ship while capturing and providing a basis for real-time active supervision, thereby achieving the technical effect of effectively managing ships entering and leaving the waters of the jurisdiction, thereby solving the technical problem of the lack of effective management of ships entering and leaving the waters of the jurisdiction in the related technology.

[0115] Optionally, when training a deep learning model, the ship identification system in the waterway includes: a training set acquisition unit, used to acquire a training data set, wherein the training data set includes: images of ships passing through the waterway collected within a historical time period and labels corresponding to the images, and the labels include: the name and characteristics of the ship; and a model training unit, used to train an initial neural network model based on the training data set to obtain a deep learning model.

[0116] Optionally, the ship position matching unit includes: a ship position acquisition module, used to obtain the position information of the target ship corresponding to the name from the ship automatic identification system; and a position determination module, used to determine the position information of the target ship as the target position.

[0117] Optionally, the ship identification system in the waterway also includes: a name access unit, which is used to identify the image data of the ship using a deep learning model, and after obtaining the name of the ship, connect the name of the ship to the interface of the key tracking ship data; a ship marking unit, which is used to mark the name of the ship and generate a reminder message when there is a ship with the same name as the ship in the key tracking ship data.

[0118] Optionally, the ship violation judgment unit includes: a numerical measurement module, used to measure the freeboard value and scale of the ship at the target position; a numerical comparison module, used to compare the freeboard value with the minimum safe freeboard threshold, and compare the scale with a preset scale threshold to obtain a comparison result; a ship violation judgment module, used to determine that the ship is overloaded when the comparison result indicates that the freeboard value is less than the minimum safe freeboard threshold; or determine that the ship is over-limit when the comparison result indicates that the scale is greater than the preset scale threshold.

[0119] Optionally, the ship identification system in the waterway also includes: a position sending unit, which is used to send the name of the ship and the target position of the ship to the monitoring terminal when it is determined that the ship is overloaded and / or when it is determined that the ship is out of limit.

[0120] Optionally, the ship identification system in the waterway also includes: a trigger instruction generating unit, used to generate a trigger instruction when a ship is detected passing through the waterway; and an instruction sending unit, used to send the trigger instruction to an image acquisition device and a laser scanning device arranged on both sides of the waterway, wherein the trigger instruction is used to trigger the image acquisition device to collect image data of a ship passing through the waterway, or to trigger the laser scanning device to collect the freeboard value and scale of the ship passing through the waterway.

[0121] The above-mentioned waterway dynamic data monitoring system may also include a processor and a memory. The above-mentioned waterway image acquisition unit 21, recognition unit 22, ship position matching unit 23, ship violation determination unit 24, etc. are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize corresponding functions.

[0122] The processor includes a kernel, which retrieves the corresponding program unit from the memory. One or more kernels can be provided, and the kernel parameters can be adjusted to determine whether the ship sailing in the waterway violates the regulations.

[0123] The above-mentioned memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash memory (fl ash RAM), and the memory includes at least one storage chip.

[0124] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is further provided, the computer-readable storage medium including a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the method for identifying ships in a waterway of any one of the above-mentioned embodiments one.

[0125] According to another aspect of an embodiment of the present invention, there is also provided an electronic device, comprising one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by one or more processors, the one or more processors implement the method for identifying ships in a waterway of any one of the above-mentioned embodiments one.

[0126] The present application also provides a computer program product, including a computer program, which, when executed by a processor, implements the steps of the method for identifying a ship in a waterway described in each embodiment of the present application.

[0127] The present application also provides a computer program product, including a non-volatile computer-readable storage medium, wherein the non-volatile computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method for identifying a ship in a waterway described in each embodiment of the present application are implemented.

[0128] Figure 3 1 is a hardware structure block diagram of an electronic device (or mobile device) for a method for identifying ships in a waterway according to an embodiment of the present invention. Figure 3 As shown, the electronic device may include one or more ( Figure 3 302a, 302b, ..., 302n are used to illustrate) processor (the processor may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 304 for storing data. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the I / O interface), a network interface, a keyboard, a power supply and / or a camera. It can be understood by those skilled in the art that Figure 3 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 3 More or fewer components as shown, or with Figure 3 Different configurations shown.

[0129] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0130] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0131] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0132] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0133] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0134] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes.

[0135] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for identifying ships in a waterway, characterized in that: include: Scanning the waterway continuously to collect image data of ships passing through the waterway; Using a deep learning model to identify the image data of the ship, and obtain the name and features of the ship; Matching the name and features of the ship with the position data collected by radar to determine the target position of the ship, wherein the position data includes the position information of the ship within a preset range around the image acquisition device that collects the image data; Determine whether the ship at the target location has violated regulations, and generate a ship violation determination result.

2. The ship identification method according to claim 1, characterized in that: The deep learning model is trained by the following method: Acquire a training data set, wherein the training data set includes: images of ships passing through the waterway collected during a historical period and labels corresponding to the images, wherein the labels include: names and features of the ships; The initial neural network model is trained based on the training data set to obtain the deep learning model.

3. The ship identification method according to claim 1, characterized in that: Matching the name and features of the ship with the position data collected by radar to determine the target position of the ship, including: Acquire the location information of the target ship corresponding to the name from the ship automatic identification system; The position information of the target ship is determined as the target position.

4. The ship identification method according to claim 3, characterized in that: After the image data of the ship is recognized by using the deep learning model to obtain the name of the ship, the method further includes: The name of the vessel is connected to the interface of key tracking vessel data; If there is a ship with the same name as the ship in the key tracking ship data, the name of the ship is marked and a reminder message is generated.

5. The ship identification method according to claim 1, characterized in that: The step of determining whether the vessel at the target location has violated regulations comprises: Measuring the freeboard value and size of the ship at the target position; Comparing the freeboard value with a minimum safety freeboard threshold, and comparing the scale with a preset scale threshold, to obtain a comparison result; When the comparison result indicates that the freeboard value is less than the minimum safety freeboard threshold, determining that the ship is overloaded; or, When the comparison result indicates that the dimension is greater than the preset dimension threshold, it is determined that the ship has exceeded the limit.

6. The ship identification method according to claim 5, characterized in that: Also includes: When it is determined that the ship is overloaded, and / or when it is determined that the ship is out of limit, the name of the ship and the target location of the ship are sent to the monitoring terminal.

7. The ship identification method according to any one of claims 1 to 6, characterized in that: Also includes: When a ship is detected passing through the waterway, a trigger instruction is generated; The trigger instruction is sent to the image acquisition device and the laser scanning device arranged on both sides of the waterway, wherein the trigger instruction is used to trigger the image acquisition device to collect image data of the ship passing through the waterway, or to trigger the laser scanning device to collect the freeboard value and scale of the ship passing through the waterway.

8. A ship identification system in a waterway, characterized in that: include: A channel image acquisition unit, used to continuously scan the channel and collect image data of ships passing through the channel; An image recognition unit, used to recognize the image data of the ship using a deep learning model to obtain the name and features of the ship; a ship position matching unit, used to match the name and features of the ship with the position data collected by the radar to determine the target position of the ship, wherein the position data includes the position information of the ship within a preset range around the image acquisition device that collects the image data; The ship violation determination unit is used to determine whether the ship at the target location has violated any regulations and generate a ship violation determination result.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the method for identifying ships in a waterway according to any one of claims 1 to 7.

10. An electronic device, characterized in that: It includes one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the ship identification method in the waterway as described in any one of claims 1 to 7.