Marine boundary monitoring system and method based on image analysis and electronic chart

Abstract

The invention relates to a system and a method for marine boundary monitoring. In the prior art, video images are generally obtained through a fixed camera installed on the land, monitoring personnel conduct monitoring for 24 hours all day long, the obtaining and utilization efficiency of image information is low and limited, a large number of ocean boundary monitoring personnel are needed, and meanwhile the risk of a monitoring blind area is possibly caused due to fatigue accumulation of each monitoring personnel. In addition, when an unidentified ship appears, a person needs to be dispatched to go out on site for face-to-face confirmation, and the problem of low efficiency in the aspects of manpower and cost is caused. In order to solve the problem, the invention provides the ocean boundary monitoring system and method based on the image analysis and the electronic chart, by automatically analyzing camera and radar images, the intervention of ocean monitoring personnel is minimized, when an unidentified ship appears, an unmanned ship and an unmanned aerial vehicle are used for obtaining a video image of the ship, and the video image of the ship is displayed. And identification is carried out through a visual search method, so that on-site operation and face-to-face confirmation are reduced. On the basis, efficient operation can be achieved with a simpler structure and lower cost.

Description

Technical Field

[0001] The present invention relates to a system and method for marine boundary surveillance. More specifically, it relates to a marine boundary surveillance system and method based on image analysis and electronic nautical charts. In the prior art, video images are usually obtained by fixed cameras installed on land and monitored by monitoring personnel 24 hours a day. Therefore, the acquisition and utilization efficiency of image information is low and limited, resulting in a large number of marine boundary monitoring personnel being required. At the same time, due to the cumulative fatigue of each monitoring personnel, there may be a risk of monitoring blind spots. In addition, when an unidentified ship appears, personnel need to be dispatched to the scene for face-to-face confirmation, which causes great inefficiency in terms of manpower and cost and becomes the main limitation of traditional marine boundary surveillance systems and methods. To solve the above problems, the present invention provides a marine boundary surveillance system and method based on image analysis and electronic nautical charts, which automatically analyzes camera and radar images to minimize the intervention of marine boundary monitoring personnel. When an unidentified ship appears, an unmanned ship and an unmanned aerial vehicle are used to obtain video images of the target ship and identify it through visual search methods to reduce on-site dispatch and face-to-face confirmation, thereby realizing the efficient operation of the system.

[0002] In addition, as described above, in the prior art, since the video images obtained by the camera are monitored manually, it is easy to cause monitoring blind spots, and when an unidentified ship appears, personnel need to be dispatched to the scene for face-to-face confirmation, thus there are technical limitations of low efficiency. To solve the above problems, the present invention provides a marine boundary surveillance system and method based on image analysis and electronic nautical charts, which includes: a global surveillance unit that roughly monitors the entire boundary surveillance target area through a camera; a local surveillance unit that analyzes in detail a partial area or a target of interest in the boundary surveillance target area through a camera; an image analysis unit that analyzes the images of the camera and the radar and fuses information from various sensors and monitoring devices (including cameras, radars, AIS, V-pass, e-Nav, etc.) to automatically perform target detection and tracking; an electronic nautical chart surveillance unit that superimposes and displays information on the surveillance area, information on various sensors and monitoring devices, sensor fusion information of the image analysis unit, and target recognition results, etc. on the electronic nautical chart; an unmanned ship and unmanned aerial vehicle control unit that, when there is an unidentified target, uses the unmanned ship and the unmanned aerial vehicle to take close-up shots of the unidentified target for identification. Through the above configuration, the images of the camera and the radar can be automatically analyzed, minimizing the intervention of marine boundary monitoring personnel. When an unidentified ship appears, the unmanned ship and the unmanned aerial vehicle are used to obtain video images and perform visual recognition to reduce on-site dispatch and face-to-face confirmation, thereby realizing the efficient operation of the system. Background Art

[0003] Traditionally, maritime surveillance has usually been carried out by large ships (such as fishery guidance ships or coastal guard ships), or the video images captured by cameras are observed by monitoring personnel for 24 hours to identify ships suspected of illegal activities or in distress.

[0004] In this process, if monitoring personnel manually monitor the video images captured by cameras, they need to compare the ship positions received through the Automatic Identification System (AIS) with the ship positions captured by the cameras and identify the ships manually. This manual identification method not only results in a slow identification speed but also relies on manual judgment, so the accuracy of the identification results cannot be guaranteed.

[0005] In other words, the existing maritime boundary surveillance systems and methods rely on fixed cameras installed on land to obtain video images and are monitored manually for 24 hours. Therefore, the acquisition and use efficiency of video information is low and limited, resulting in a large number of maritime boundary monitoring personnel being required. Moreover, due to the accumulation of fatigue of the monitoring personnel, blind spots may occur.

[0006] In addition, for the existing maritime boundary surveillance systems and methods, when an unidentified ship appears, personnel need to be dispatched to the scene for face-to-face confirmation, which is inefficient in terms of both personnel and cost.

[0007] Therefore, to solve the limitations of the above traditional technologies, the preferred solution is to propose a new maritime boundary surveillance system and method. By automatically analyzing camera and radar images, the intervention of maritime boundary monitoring personnel is reduced, and when an unidentified ship appears, on-site dispatch and face-to-face confirmation are reduced, thereby realizing the efficient operation of the system. However, so far, no devices and methods that can meet these requirements have been proposed. Summary of the Invention

[0008] Problems to be Solved by the Invention

[0009] The present invention aims to solve the problems in the above-mentioned traditional technologies. Therefore, the object of the present invention is to solve the following problems: In the prior art, video images are usually obtained by fixed cameras installed on land and monitored by monitoring personnel 24 hours a day, resulting in low and limited efficiency in the acquisition and utilization of image information. As a result, a large number of marine boundary monitoring personnel are required. At the same time, due to the accumulation of fatigue of each monitoring personnel, there may be a risk of monitoring blind spots. In addition, when an unidentified ship appears, personnel need to be dispatched to the scene for face-to-face confirmation, resulting in inefficient problems in terms of manpower and cost. To overcome these problems existing in the traditional technology, the present invention proposes a marine boundary monitoring system and method based on image analysis and electronic nautical charts, which automatically analyzes camera and radar images to minimize the intervention of marine boundary monitoring personnel. When an unidentified ship appears, an unmanned ship and an unmanned aerial vehicle are used to obtain video images of the ship and identify it through a visual search method, so as to reduce on-site dispatch and face-to-face confirmation, thereby realizing the efficient operation of the system.

[0010] Another object of the present invention is to solve the following problems of the prior art marine boundary monitoring system and method: Since the images obtained by the camera are monitored manually, it is easy to cause monitoring blind spots, and when an unidentified ship appears, personnel need to be dispatched to the scene for face-to-face confirmation, thus there are technical limitations of low efficiency. To solve the above problems, the present invention proposes a marine boundary monitoring system and method based on image analysis and electronic nautical charts, which includes: a global monitoring unit that roughly monitors the overall boundary monitoring target area through a camera; a local monitoring unit that analyzes in detail a partial area or a concerned object of the boundary monitoring target area through a camera; an image analysis unit that analyzes camera and radar images, and fuses information from various sensors and monitoring devices (including cameras, radars, AIS, V-pass, e-Nav, etc.) to automatically perform target detection and tracking; an electronic nautical chart monitoring unit that superimposes and displays information on the monitoring area, information on various sensors and monitoring devices, sensor fusion information of the image analysis unit, and target recognition results, etc. on the electronic nautical chart; an unmanned ship and unmanned aerial vehicle control unit that, when there is an unidentified target, uses the unmanned ship and the unmanned aerial vehicle to take close-up photos of the unidentified target for identification. Based on this, the present invention automatically analyzes camera and radar images to minimize the intervention of marine boundary monitoring personnel. When an unidentified ship appears, an unmanned ship and an unmanned aerial vehicle are used to obtain images and identify them through a visual search method, so as to reduce on-site dispatch and face-to-face confirmation, thereby realizing the efficient operation of the system.

[0011] Means for solving the problems

[0012] To achieve the above object, according to the present invention, there is provided an ocean boundary monitoring system based on video analysis and electronic charts, comprising: a global monitoring unit, including a global camera for photographing the entire preset boundary monitoring target area, and used for monitoring the entire boundary monitoring target area; a local monitoring unit, including a local camera for detailed photographing of a partial area or a concerned object in the boundary monitoring area, and used for monitoring a specific area or object according to a preset configuration or user input; an image analysis unit, which fuses the images received through the global camera and the area camera, as well as the images and various information from a radar, various sensors and monitoring devices pre-installed for ocean monitoring, and automatically detects and tracks targets through image analysis; an electronic chart monitoring unit, which, according to a preset configuration, superimposes and displays the sensor fusion information and target recognition results of the image analysis unit, as well as various data received from various sensors and monitoring devices on the electronic chart; and an unmanned ship and unmanned aerial vehicle control unit, when there is an unrecognized target in the target recognition result of the image analysis unit, uses an unmanned ship or an unmanned aerial vehicle to take a close-up photograph of the unrecognized target.

[0013] Herein, the system is characterized by further comprising: an output unit, used for displaying various information such as the processing content and processing results processed by the system through a display; a communication unit, which performs data transceiver with external devices including a server through at least one of wired or wireless communication methods, and the processing content includes the transceiver of various data such as information received from various sensors and monitoring devices and target recognition results; and a control unit, used for controlling the overall operation of the system.

[0014] In addition, the global monitoring unit is characterized by including at least one camera, which is used for photographing the entire preset boundary monitoring target area and transmitting the photographed image to the image analysis unit.

[0015] At this time, when the global monitoring unit includes more than two cameras, the images of each camera are presented in a panoramic form through an image stitching technique, and through PTZ (Pan-Tilt-Zoom) control, the linkage and synchronous operation of each camera are realized.

[0016] In addition, the local monitoring unit includes at least one camera, which, according to the detection and tracking results of the image analysis unit or according to user designation, performs detailed photographing of a partial area or a concerned object in the boundary monitoring target area and transmits the photographed image to the image analysis unit.

[0017] In addition, the image analysis unit is characterized in that it uses a deep learning (Deep Learning) or machine learning (Machine Learning)-based artificial intelligence (AI) target recognition model trained with visual search capabilities to fuse the information of camera and radar images, various sensors, and monitoring devices, so as to automatically detect and track targets according to preset configurations and transmit the generated sensor fusion information to the global monitoring unit, the local monitoring unit, and the electronic chart monitoring unit respectively.

[0018] In addition, the image analysis unit uses the close-range images and metadata information received through the unmanned ship and UAV control unit to identify the unrecognized targets captured and transmits the recognition results of the unrecognized targets to the global monitoring unit, the local monitoring unit, and the electronic chart monitoring unit respectively.

[0019] In addition, the electronic chart monitoring unit is characterized in that according to preset configurations, it superimposes and displays various types of data such as information about the boundary monitoring target area, images of the global cameras and local cameras, information of various sensors and monitoring devices such as radar, AIS, Vpass, e-Nav, and the sensor fusion information and target recognition results of the image analysis unit on the electronic chart.

[0020] Furthermore, when there are unrecognized targets, the electronic chart monitoring unit generates path planning and control information for the unmanned ship and UAV to obtain images of the unrecognized ships according to the information received from the global monitoring unit and the local monitoring unit and transmits it to the unmanned ship and UAV control unit.

[0021] In addition, the unmanned ship and the UAV include an optical camera for image capture, or, in addition to the optical camera, also include a thermal imaging camera, a lidar (LiDAR), a searchlight, and a flash.

[0022] In addition, a method for monitoring a marine boundary using the above marine boundary monitoring system according to the present invention is characterized by comprising the following steps: a monitoring step of taking images of a preset boundary monitoring target area through the global monitoring unit and the local monitoring unit of the marine boundary monitoring system; a detection step of fusing the images collected in real time in the monitoring step and the images and various types of information from radars, various sensors, and monitoring devices pre-installed for marine monitoring through the image analysis unit of the marine boundary monitoring system, and automatically detecting and tracking targets through image analysis; a display step of superimposing and displaying the sensor fusion information and target recognition results of the image analysis unit, as well as various types of data received from various sensors and monitoring devices, on an electronic nautical chart according to a preset configuration through the electronic nautical chart monitoring unit of the marine boundary monitoring system; and a re-detection step of using an unmanned ship and an unmanned aerial vehicle control unit of the marine boundary monitoring system to take a close-up photo of an unrecognized target when there is an unrecognized target in the target recognition result of the analysis step, and re-identifying the unrecognized target based on the close-up image through the image analysis unit.

[0023] Advantages of the Invention

[0024] As described above, according to the present invention, there is provided a marine boundary monitoring system and method based on image analysis and an electronic nautical chart, which includes: a global monitoring unit for roughly monitoring the whole of a boundary monitoring target area through a camera; a local monitoring unit for detailed analysis of a partial area or an object of interest in the boundary monitoring target area through a camera; an image analysis unit for analyzing camera and radar images and fusing information from various sensors and monitoring devices such as cameras, radars, AIS, V-pass, e-Nav, etc., and automatically performing target detection and tracking; an electronic nautical chart monitoring unit for superimposing and displaying information on the monitoring area, information on various sensors and monitoring devices, sensor fusion information of the image analysis unit, target recognition results, etc. on the electronic nautical chart; and an unmanned ship and unmanned aerial vehicle control unit for taking a close-up photo of an unrecognized target using an unmanned ship and an unmanned aerial vehicle when there is an unrecognized target for identification. Based on the above system and method, the present invention automatically analyzes camera and radar images to minimize the intervention of marine boundary monitoring personnel. When an unrecognized ship appears, images are obtained using an unmanned ship and an unmanned aerial vehicle and identified through a visual retrieval method to reduce on-site deployments and face-to-face confirmations, thereby enabling more efficient and effective marine boundary monitoring.

[0025] In addition, according to the present invention, there is provided an ocean boundary monitoring system and method based on image analysis and electronic nautical charts as described above, which automatically analyzes camera and radar images to minimize the intervention of ocean boundary monitoring personnel. When an unidentified ship appears, an unmanned ship and an unmanned aerial vehicle are used to obtain images of the ship, and identification is performed through a visual search method to reduce on-site deployment and face-to-face confirmation, thereby enabling more efficient and effective operation. Based on this, the following problems existing in the prior art ocean boundary monitoring system and method can be solved: Usually, video images are obtained through fixed cameras installed on land and monitored by monitoring personnel 24 hours a day, resulting in low and limited efficiency in the acquisition and utilization of image information, and thus a large number of ocean boundary monitoring personnel are required. At the same time, due to the accumulation of fatigue of each monitoring personnel, there may be a risk of monitoring blind spots. In addition, when an unidentified ship appears, personnel need to be dispatched to the scene for face-to-face confirmation, resulting in inefficiencies in terms of manpower and costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 FIG. is a schematic block diagram of the overall structure of an ocean boundary monitoring system based on image analysis and electronic nautical charts according to an embodiment of the present invention.

[0027] Figure 2 is Figure 1 FIG. is a schematic diagram of the overall processing flow of an ocean boundary monitoring system based on image analysis and electronic nautical charts according to an embodiment of the present invention as shown.

[0028] Figure 3 is using Figure 1 FIG. is a conceptual schematic diagram of ocean boundary monitoring using an ocean boundary monitoring system based on image analysis and electronic nautical charts according to an embodiment of the present invention as shown. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0029] Hereinafter, specific embodiments of an ocean boundary monitoring system and method based on image analysis and electronic nautical charts according to the present invention will be described with reference to the accompanying drawings.

[0030] It should be noted that the following content is only one embodiment of implementing the present invention, and the present invention is not limited to the content of the following embodiments.

[0031] In addition, in the following description of the embodiments of the present invention, parts that are the same as or similar to the prior art content, or parts that can be easily understood and implemented by those skilled in the art, are omitted from detailed description to simplify the relevant description.

[0032] Next, the specific content of an ocean boundary monitoring system and method based on image analysis and electronic nautical charts according to the present invention will be described with reference to the accompanying drawings.

[0033] More specifically, first refer to Figure 1 andFigure 2 , Figure 1 is a schematic block diagram of the overall structure of the marine boundary monitoring system 10 based on image analysis and electronic chart according to an embodiment of the present invention. Figure 2 is Figure 1 a schematic diagram of the overall processing flow of the marine boundary monitoring system 10 based on image analysis and electronic chart according to an embodiment of the present invention as shown.

[0034] Here, Figure 2 in, the blue font represents the information transmitted from the left side to the right side of the figure, and the green font also represents the information transmitted from the left side to the right side of the figure.

[0035] As Figure 1 and Figure 2 shown, the marine boundary monitoring system 10 based on image analysis and electronic chart according to an embodiment of the present invention generally includes the following parts: a global monitoring unit 11 that roughly shoots the overall boundary monitoring target area through a camera; a local monitoring unit 12 that shoots a partial area or an object of interest in the boundary monitoring target area in detail through a camera; an image analysis unit 13 that analyzes camera and radar images and fuses information from various sensors and monitoring devices such as cameras, radars, AIS, V-pass, and e-Nav, and automatically performs automatic detection and tracking of targets; an electronic chart monitoring unit 14 that superimposes and displays information on the monitoring area, information on various sensors and monitoring devices, sensor fusion information of the image analysis unit, and target recognition results, etc. on the electronic chart; an unmanned ship and unmanned aerial vehicle control unit 15 that, when there is an unrecognized target, uses the unmanned ship and unmanned aerial vehicle to take a close-up shot of the unrecognized target for identification.

[0036] Here, although not shown in the figure, the system 10 may further include the following components: an output unit for displaying various information related to the overall operation of the system 10 such as the processing content and processing results processed by the above-mentioned units through devices such as a display; a communication unit for performing data transceiver with external devices including a server through at least one of wired or wireless communication methods, and the processing content includes the transceiver of various data such as information received from various sensors and monitoring devices and target recognition results; and a control unit: for controlling the overall operation of the above-mentioned units and the system 10.

[0037] More specifically, the global monitoring unit 11 may include one or more cameras to shoot the overall boundary monitoring target area and achieve the function of overall monitoring.

[0038] When there are more than two cameras, through image stitching technology, the images between cameras can be naturally stitched into a panorama form. Through PTZ (Pan-Tilt-Zoom) control of the cameras, the linkage and synchronous operation of all cameras can be achieved.

[0039] If there is no overlapping area between camera images due to excessive zooming, the stitching function can be set not to execute.

[0040] In addition, the global monitoring unit 11 can also receive sensor fusion information and target recognition results from the image analysis unit 13 and superimpose and display them on the camera image in the form of augmented reality.

[0041] At the same time, the local monitoring unit 12 can include one or more cameras to take more detailed pictures of part of the boundary monitoring target area or the object of interest. For this purpose, the target area information that needs to be investigated in detail can be specified by the user through the global monitoring unit 11 or the electronic chart monitoring unit 14 and transmitted as a camera control signal.

[0042] If the object of investigation is a target (object), the detection and tracking results of the image analysis unit 13 are transmitted as a camera control signal. That is, each local camera operates independently. When no specific area or target (object) is specified, it can automatically move in a loop along the pre-set monitoring path.

[0043] In addition, the local monitoring unit 12 can also receive sensor fusion information and target recognition results from the image analysis unit 13 and superimpose and display them on the camera image in the form of augmented reality, just like the global monitoring unit 11.

[0044] Next, the image analysis unit 13 realizes the automatic detection and tracking of targets by analyzing camera and radar images, and further fuses the information of various sensors and monitoring devices to more accurately detect and track targets.

[0045] More specifically, the image analysis unit 13 can combine various types of information that are considered to belong to the same target, so as to specifically identify the ship. The sensor fusion information generated through this process will be transmitted to the global monitoring unit 11, the local monitoring unit 12, and the electronic chart monitoring unit 14 for transmission to the user.

[0046] In addition, the image analysis unit 13 can also use the captured video and metadata information transmitted through the unmanned ship and UAV control unit 15 to identify the captured target (such as a ship), and transmit the recognition results to the global monitoring unit 11, the local monitoring unit 12, and the electronic chart monitoring unit 14.

[0047] The function of identifying targets can adopt deep learning (Deep Learning) or artificial intelligence (AI) target recognition models based on machine learning (Machine Learning) that are trained to have visual search capabilities, and the model will be continuously updated during the process of unrecognized targets being recognized.

[0048] Regarding the processing process of ship identification using the pre-trained artificial intelligence algorithm as described above, the specific content can refer to data processing methods or artificial intelligence target recognition algorithms in the prior art, and those skilled in the art can appropriately construct it accordingly. Therefore, for the sake of simplicity of description, the content known to those skilled in the art or easily understood and implemented through prior art literature will not be elaborated in this invention.

[0049] In addition, the electronic chart monitoring unit 14 can superimpose and display the information of the monitored area, the information of various sensors and monitoring devices such as cameras, radars, AIS, V-pass, e-Nav, etc., and the sensor fusion information and target recognition results of the image analysis unit 13 on the two-dimensional electronic chart.

[0050] Furthermore, when no specific area or object is specified for the local camera, the electronic chart monitoring unit 14 can automatically execute the mobile monitoring path input by the user for the camera. In addition, when it is difficult to identify a target (such as a ship) through the local camera, the path planning of the unmanned ship and the unmanned aerial vehicle can be generated and transmitted, enabling them to go to obtain the video of the unrecognized ship, and generating and transmitting control information based on the navigation information of the unmanned ship and the unmanned aerial vehicle.

[0051] That is, when identifying a target by fusing the information of various sensors and monitoring devices but there are still unrecognized targets, the unmanned ship and unmanned aerial vehicle control unit 15 can take a close-up photo of the unrecognized target through the unmanned ship and the unmanned aerial vehicle for identification.

[0052] At this time, the path planning and control information of the unmanned ship and the unmanned aerial vehicle are generated and transmitted by the electronic chart monitoring unit 14 based on the selected target information of the global monitoring unit 11 and the local monitoring unit 12, so as to achieve automatic video shooting of unrecognized targets.

[0053] In addition, the unmanned ship and the unmanned aerial vehicle can be appropriately used according to factors such as marine meteorological conditions, time periods, and the distance from the identified target. In addition to optical cameras, devices such as thermal imaging cameras, lidar (LiDAR), searchlights, and flashlights can also be equipped. Therefore, this invention can be appropriately modified and changed by those skilled in the art according to needs without departing from its spirit and essence.

[0054] As described above, the marine boundary monitoring system 10 and method based on image analysis and electronic nautical chart according to the embodiments of the present invention can be conveniently implemented. Specifically, referring to Figure 3 , Figure 3 is a conceptual schematic diagram of marine boundary monitoring using the marine boundary monitoring system 10 based on image analysis and electronic nautical chart according to the embodiments of the present invention shown in Figure 1 .

[0055] As Figure 3 shown, according to the present invention, by collecting various types of data about ships, a training database for artificial intelligence algorithm learning is constructed, and the AI model is trained based on the content of the constructed database. Using the AI model trained in this way, a series of processes are carried out to identify ships through real-time collected videos and various monitoring information, display the identification information, and send it to the outside according to the preset configuration.

[0056] In addition, when an unidentified ship that cannot be identified by the AI model appears, a request for deployment can be sent to the drone or unmanned ship, controlling it to go to the location of the unidentified ship, and using the video of the unidentified ship taken by the drone or unmanned ship to re-identify the ship.

[0057] According to the above structure, the present invention analyzes the camera and radar images by using the visual search method of artificial intelligence, and can automatically identify ships, thereby minimizing the intervention of marine boundary monitoring personnel. At the same time, when an unidentified ship appears, the close-range images of the ship can be obtained by using the unmanned ship and drone, and identified by the visual search method, thereby greatly reducing the on-site deployment and face-to-face confirmation. Compared with the previous method of relying on monitoring personnel to monitor the videos obtained by land-fixed cameras for 24 hours, the present invention can achieve more efficient and effective marine boundary monitoring with a simpler structure and lower cost.

[0058] Therefore, the marine boundary monitoring system and method based on image analysis and electronic nautical chart according to the embodiments of the present invention can be implemented as described above.

[0059] The above has detailed the content of the marine boundary monitoring system and method based on image analysis and electronic nautical chart through the embodiments of the present invention, but the present invention is not limited to the content recorded in the above embodiments. Therefore, the present invention can be variously modified, changed, combined, and replaced by those of ordinary skill in the art according to design needs and other various factors, which is natural.

[0060] Description of Reference Numerals

[0061] 11: Global Monitoring Unit;

[0062] 12: Local Monitoring Unit;

[0063] 13: Image analysis unit;

[0064] 14: Electronic chart monitoring unit;

[0065] 15: Unmanned ship and UAV control unit.

Claims

1. A marine boundary monitoring system based on image analysis and electronic nautical chart, characterized in that the marine boundary monitoring system based on image analysis and electronic nautical chart includes: a global monitoring unit, including a global camera for photographing the whole of a preset boundary monitoring target area, and used for monitoring the whole of the boundary monitoring target area; a local monitoring unit, including a local camera for detailed photographing of a partial area or a concerned object in the boundary monitoring target area, and used for monitoring a specific area or object according to a preset configuration or user input; an image analysis unit, which fuses the images received through the global camera and the local camera, as well as the images and various information from radars, various sensors and monitoring devices pre-installed for marine monitoring, and automatically detects and tracks targets through image analysis; an electronic nautical chart monitoring unit, which, according to a preset configuration, superimposes and displays the sensor fusion information and target recognition results of the image analysis unit, as well as various data received from various sensors and monitoring devices on the electronic nautical chart; and an unmanned ship and unmanned aerial vehicle control unit, when there is an unrecognized target in the target recognition result of the image analysis unit, uses an unmanned ship or an unmanned aerial vehicle to take a close-up photograph of the unrecognized target.

2. The marine boundary monitoring system based on image analysis and electronic nautical chart according to claim 1, characterized in that the marine boundary monitoring system based on image analysis and electronic nautical chart further includes: an output unit, used for displaying various information such as the processing content and processing results processed by the system through a display; a communication unit, which performs data transceiver with external devices including a server through at least one of wired or wireless communication methods, and the processing content includes the transceiver of various data such as information and target recognition results received from various sensors and monitoring devices; and a control unit, used for controlling the overall operation of the system.

3. The marine boundary monitoring system based on image analysis and electronic nautical chart according to claim 1, characterized in that the global monitoring unit includes at least one camera, which is used for photographing the whole of the preset boundary monitoring target area and transmitting the photographed image to the image analysis unit.

4. The marine boundary monitoring system based on image analysis and electronic nautical chart according to claim 3, characterized in that when the global monitoring unit includes more than two cameras, the images of each camera are presented in a panoramic form through image stitching technology, and through PTZ (Pan-Tilt-Zoom) control, the linkage and synchronous operation of each camera are realized.

5. The marine boundary monitoring system based on image analysis and electronic nautical chart according to claim 1, characterized in that the local monitoring unit includes at least one camera, which, according to the detection and tracking results of the image analysis unit, or according to user designation, takes a detailed photograph of a partial area or a concerned object in the boundary monitoring target area and transmits the photographed image to the image analysis unit.

6. The marine boundary monitoring system based on image analysis and electronic chart according to claim 1, wherein the image analysis unit uses a deep learning (Deep Learning) or machine learning (Machine Learning) based artificial intelligence (AI) target recognition model trained with visual search capabilities, fuses camera and radar images and information from various sensors and monitoring devices, automatically detects and tracks targets according to preset configurations, and transmits the generated sensor fusion information to the global monitoring unit, the local monitoring unit, and the electronic chart monitoring unit respectively.

7. The marine boundary monitoring system based on image analysis and electronic chart according to claim 6, wherein the image analysis unit uses the close-range images and metadata information received through the unmanned ship and UAV control unit to identify the unrecognized targets in the captured images, and transmits the identification results of the unrecognized targets to the global monitoring unit, the local monitoring unit, and the electronic chart monitoring unit respectively.

8. The marine boundary monitoring system based on image analysis and electronic chart according to claim 1, wherein the electronic chart monitoring unit, according to preset configurations, superimposes and displays on the electronic chart various data such as information on the boundary monitoring target area, images of each global camera and local camera, information from various sensors and monitoring devices such as radar, AIS, V-pass, e-Nav, as well as the sensor fusion information and target recognition results of the image analysis unit.

9. The marine boundary monitoring system based on image analysis and electronic chart according to claim 1, wherein when there are unrecognized targets, the electronic chart monitoring unit generates path planning and control information for the unmanned ship and UAV to obtain images of the unrecognized ships according to the information received from the global monitoring unit and the local monitoring unit, and transmits it to the unmanned ship and UAV control unit.

10. The marine boundary monitoring system based on image analysis and electronic chart according to claim 1, wherein the unmanned ship and the UAV include an optical camera for image capture, or, in addition to the optical camera, also include a thermal imaging camera, a lidar (LiDAR), a searchlight, and a flash.

11. A marine boundary monitoring method using the marine boundary monitoring system according to any one of claims 1 to 10, wherein the marine boundary monitoring method of the marine boundary monitoring system includes the following steps: A monitoring step of taking images of a preset boundary monitoring target area through the global monitoring unit and the local monitoring unit of the marine boundary monitoring system; Detection step: Through the image analysis unit of the marine boundary monitoring system, fuse the images collected in real time during the monitoring step, as well as the images and various information from the radar, various sensors and monitoring devices pre-installed for marine monitoring, and automatically detect and track targets through image analysis; Display step: Through the electronic chart monitoring unit of the marine boundary monitoring system, according to the preset configuration, superimpose and display the sensor fusion information and target recognition results of the image analysis unit, as well as various data received from various sensors and monitoring devices, on the electronic chart; and Redetection step: Through the unmanned ship and UAV control unit of the marine boundary monitoring system, when there are unrecognized targets in the target recognition results of the analysis step, use the unmanned ship or UAV to take close-up photos of the unrecognized targets, and re-recognize the unrecognized targets based on the close-up images through the image analysis unit.

Images