Water area monitoring and early warning system and method

Abstract

The invention relates to the field of water area monitoring and early warning, in particular to a water area monitoring and early warning system and method. The system comprises a data acquisition unit, a data processing unit, an application unit and an early warning implementation unit. The method comprises the following steps: detecting a dangerous object target; a tracking mode and selected equipment are determined, and if only one detection target exists, a camera closest to the target is selected for key tracking and linkage; if the camera has an initial tracking task and a plurality of targets appear at the same time, starting a dynamic task allocation algorithm, re-evaluating the priority of the tracking task according to the features of the targets, performing task insertion, replacement or deletion, and displaying the moving speed and volume of the tracking targets in sequence; identification is carried out based on multi-dimensional parameter estimation obtained through tracking, a state identification matrix is established for parameter results of the camera, the radar, the liquid level meter and the sonar, when the radar or sound wave expelling equipment or the liquid level meter or the sonar gives an alarm, the camera automatically starts a video tracking function at the moment, and if the camera rechecks that the alarm really exists, the state identification matrix is established. At the moment, the alarm signal is actually generated in the application server; if the camera recheck alarm does not exist, the alarm signal is not generated in the application server; and expelling the target. The method is more accurate in identification, the dynamic disturbance degree is minimum, and system energy consumption can be saved to the maximum extent.

Description

Technical Field

[0001] The present invention relates to the field of water area monitoring and early warning, and in particular to a water area monitoring and early warning system and method. Background Art

[0002] In recent years, there have been a number of shutdown accidents at domestic nuclear power plants due to blockage of cooling water intakes, resulting in the inability of units to obtain sufficient cooling water. Therefore, it is necessary to monitor and respond to alarms on the water surface and objects in the water in a timely manner. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide a water area monitoring and early warning system and method, which has more accurate identification, minimum dynamic disturbance and can save system energy consumption to the maximum extent.

[0004] The present invention provides a water area monitoring and early warning system, comprising:

[0005] A data acquisition unit, used to monitor and collect information about dangerous objects in real time, and transmit the information to a data processing unit; the data acquisition unit includes a camera and other non-visual acquisition devices;

[0006] A data processing unit, used to perform data algorithm calculations on dangerous object information and form a database;

[0007] The data processing unit includes an application server and a data server.

[0008] The application server performs algorithmic operations, including dynamic task allocation algorithms and state recognition matrix calculations;

[0009] The dynamic task allocation algorithm is that when the camera has an initial tracking task and multiple targets appear at the same time, the priority of the tracking task is evaluated according to the target information, and the task is inserted, replaced or deleted;

[0010] The state recognition matrix calculation is to establish a state recognition matrix for the parameters detected by the non-visual acquisition device and the camera, and when the non-visual acquisition device alarms, the camera is matched to verify the alarm and the alarm is comprehensively calculated to obtain an alarm signal; when the non-visual acquisition device alarms, if the camera does not issue an alarm, a non-alarm signal is calculated;

[0011] The data server includes a database, which stores real-time monitoring data collected by the data collection unit, data after algorithm calculation by the application server, and system authority information;

[0012] Application unit, used to display the results of data processing and realize human-computer interaction;

[0013] The early warning implementation unit is used to receive the alarm signal sent by the data processing unit to drive away the target.

[0014] In a specific embodiment of the present invention, other non-visual acquisition devices of the data acquisition unit include radar, liquid level meter, sonar and externally connected data acquisition device.

[0015] In a specific embodiment of the present invention, the application server performs identification based on multi-dimensional parameter estimation, establishes a state identification matrix for the parameter results of the camera, radar, level meter, and sonar, and identifies or detects each state at least three times;

[0016] The state recognition matrix has 4 rows of content, representing the three results of camera, radar signal, level meter and sonar detection respectively;

[0017] When the radar or sound wave drives away the equipment or the level meter or sonar alarms, the camera automatically starts the video tracking function. If the camera verifies that the alarm is real, the alarm signal is actually generated on the application server.

[0018] If the camera verification alarm does not exist, the alarm signal will not be generated on the application server;

[0019] Application server command warning implementation unit.

[0020] In a specific embodiment of the present invention, the early warning implementation unit is a sonic wave driving away device.

[0021] In a specific embodiment of the present invention, it also includes a core switch;

[0022] Level meters, cameras, radars, sonars, and acoustic repellent devices are connected to the core switch via signal cables and then to the application server and data server;

[0023] The core switch is isolated from the administrative office network through a firewall, and the alarm information and video images can be viewed in the administrative office network, or they can be remotely configured;

[0024] Externally connected data collectors are connected to the core switch through the network.

[0025] In a specific embodiment of the present invention, the camera is a binocular or trinocular camera, which realizes detailed monitoring and intelligent tracking of the target object by working together with two or more cameras. It can also measure the volume of the object and realize linkage with the data processing unit.

[0026] In a specific embodiment of the present invention, the camera parameters in the state recognition matrix are collected by the camera that is closest to the target;

[0027] The camera closest to the target is selected and calculated by the application server.

[0028] In a specific embodiment of the present invention, the application unit is a mobile phone or a display; the core switch contains a wireless routing module, which is connected to the mobile phone through the operator's wireless network, and the alarm information can be sent to the mobile phone for viewing, or the core switch is connected to the wireless terminal device, and then the DTU sends the alarm information to the mobile phone text message.

[0029] In a specific embodiment of the present invention, the sonar emits sound waves in the water to obtain point cloud data of the surface shape of the discovered target, then obtains a three-dimensional model by modeling the point cloud, and finally provides information about the discovered target in the water to an application server.

[0030] The present invention provides a water area monitoring and early warning method, comprising the following steps:

[0031] Step 1: Detect dangerous objects: The radar or sonar performs a blind polling search within the detection radius on the water surface or in the water according to the set frequency, and locates the target based on the information of the reflected waves;

[0032] Step 2: Determine the tracking method and equipment to be used, including:

[0033] If there is only one detection target, the camera closest to the target is selected for key tracking and linkage;

[0034] If the camera already has an initial tracking task and multiple targets appear at the same time, the dynamic task allocation algorithm is started, and the priority of the tracking task is re-evaluated according to the characteristics of the target, and the task is inserted, replaced or deleted, and the moving speed and volume of the tracked target are displayed in turn;

[0035] Step 3: Based on the multi-dimensional parameters obtained by tracking, a state recognition matrix is ​​established for the parameter results of the camera, radar, level meter, and sonar.

[0036] When the radar or sound wave drives away the equipment or the level meter or sonar alarms, the camera automatically starts the video tracking function. If the camera verifies that the alarm is real, the alarm signal is actually generated on the application server.

[0037] If the camera verification alarm does not exist, the alarm signal will not be generated on the application server;

[0038] Step 4: Drive the target away.

[0039] In a specific embodiment of the present invention, the method for selecting the camera closest to the target is: given the motion coordinates (x m ,y θ ), according to the coordinates of the discovered target (xm ,y θ ) and the coordinates of the installed i-th camera (x i ,y i ), calculate the distance Δd between the two; select the camera with the smallest Δd for key tracking and linkage;

[0040] m, θ are integers, and i is an integer.

[0041] In a specific embodiment of the present invention, in the state recognition matrix, each state is recognized or detected at least 3 times;

[0042] The state recognition matrix has 4 rows of content, representing the three results of camera, radar signal, level meter and sonar detection respectively.

[0043] In a specific embodiment of the present invention, the wave information includes signal strength, time difference, and incident angle.

[0044] In a specific embodiment of the present invention, the target characteristics include any one or more of the target's distance, size and moving speed.

[0045] In a specific implementation of the present invention, the priority of the tracking task is calculated according to the following formula:

[0046]

[0047] k t is the priority of the tth tracking task, s t is the size or area of ​​the tth tracking task, v t is the moving speed of the t-th tracking task, d t is the distance between the tth tracking task and the water monitoring and early warning system.

[0048] In a specific embodiment of the present invention, the algorithm of dynamic task allocation satisfies the maximum fitness and the minimum dynamic disturbance degree.

[0049] The calculation formula of fitness f is:

[0050]

[0051] Among them, A is the task completion rate of camera tracking, B is the tracking accuracy of the camera, C is the resource relaxation, D is the switching rate index of various detectors, and ɑ1, ɑ2, ɑ3 and ɑ4 are the weights of each index;

[0052] Dynamic disturbance degree S formula:

[0053] Dynamic disturbance

[0054] Among them, ɑ5 is the weight of the deletion task, n5 is the total number of deletions in the new and old tasks, ɑ6 is the weight of the replacement task, n6 represents the total number of replacements in the new and old tasks, and n is the total number of new and old tasks.

[0055] In a specific embodiment of the present invention, when the camera has an initial tracking task and multiple targets appear at the same time,

[0056] According to the priority, select subtasks from the newly appeared target and directly insert them into the original task tracking plan to determine whether the subtask is successfully inserted. If not, record the new subtask that cannot be inserted and then update the new subtask; if the insertion is successful, directly update the new subtask;

[0057] Determine whether the new subtask set is empty. If it is not empty, return to determine whether the subtask is successfully inserted; if it is empty, construct a task conflict set;

[0058] Select subtasks from the conflict set and replace the atomic task;

[0059] Determine whether the task is replaced successfully. If so, continue to determine whether the task conflict set is empty. If not, perform a priority comparison and delete the task, and then continue to determine whether the conflict set is empty.

[0060] If the task conflict set is not empty, return to select a subtask from the conflict set; if it is empty, determine whether the maximum fitness and minimum dynamic disturbance are met;

[0061] If the conditions are met, an adjusted tracking plan is formed; if not, a subtask is selected from the conflict set. Compared with the prior art, the water area monitoring and early warning system and method of the present invention has the following beneficial effects:

[0062] (1) Identification is based on a multi-dimensional parameter estimation algorithm, and the detection content or number can be expanded according to actual needs. The visual recognition device performs double verification with other devices, and an alarm message is issued only when the results of the two devices are consistent. This can effectively avoid false alarms caused by detection equipment failures and make identification more accurate.

[0063] (2) Use precise detection equipment to visually track and verify the target after it is discovered, and perform three-dimensional modeling, so as to provide more accurate early warnings for targets on the water surface or in the water. If multiple targets are discovered at the same time, the application server will start the dynamic task allocation algorithm, re-evaluate the priority of the tracking task based on the distance, scale and movement speed of the target, and insert, replace or delete tasks to avoid missing more important or more dangerous objects. The dynamic task allocation algorithm needs to meet the maximum fitness and minimum dynamic disturbance. On the basis of meeting the maximum fitness, the initial tracking plan is adjusted to the minimum extent, that is, the dynamic disturbance is minimized, which can save system energy consumption to the maximum extent. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 Represents the system composition structure diagram;

[0065] Figure 2 A schematic diagram showing a matrix based on multidimensional parameter estimation;

[0066] Figure 3 Represents the overall flow chart of the system;

[0067] Figure 4 Represents the flow chart of dynamic task allocation algorithm. DETAILED DESCRIPTION

[0068] In order to further understand the present invention, the embodiments of the present invention are described below in conjunction with examples, but it should be understood that these descriptions are only for further illustrating the features and advantages of the present invention, rather than limiting the present invention.

[0069] An embodiment of the present invention discloses a water area monitoring and early warning system, comprising:

[0070] The data acquisition unit is used to monitor and collect information about dangerous objects in real time, and transmit the information to the data processing unit; the data acquisition unit includes a camera and other non-visual acquisition equipment;

[0071] Other non-visual acquisition devices of the data acquisition unit include radar, level gauge, sonar and externally connected data collector; the externally connected data may include ocean or meteorological data, which is used as a parameter item in the state recognition matrix;

[0072] A data processing unit, used to perform data algorithm calculations on dangerous object information and form a database;

[0073] The data processing unit includes an application server and a data server.

[0074] The application server performs algorithmic operations, including dynamic task allocation algorithms and state recognition matrix calculations;

[0075] The dynamic task allocation algorithm is that when the camera has an initial tracking task and multiple targets appear at the same time, the priority of the tracking task is evaluated according to the target information, and the task is inserted, replaced or deleted;

[0076] The state recognition matrix calculation is to establish a state recognition matrix for the parameters detected by the non-visual acquisition device and the camera, and when the non-visual acquisition device alarms, the camera is matched to verify the alarm and the alarm is comprehensively calculated to obtain an alarm signal; when the non-visual acquisition device alarms, if the camera does not issue an alarm, a non-alarm signal is calculated;

[0077] The parameters in the state recognition matrix include the parameters of the camera, radar, level gauge, and sonar, and may also include external access data;

[0078] Camera parameters involve the selection of cameras. The selection method is: for a target, the motion coordinates of the target are given (x m ,y θ ), according to the coordinates of the discovered target (x m ,y θ ) and the coordinates of the installed i-th camera (x i ,y i ), calculate the distance Δd between the two; select the camera with the smallest Δd for key tracking and linkage;

[0079] m, θ are integers, and i is an integer.

[0080]

[0081] The camera performs visual tracking and emits lasers to obtain point cloud data of the target's surface shape, then obtains a three-dimensional model by modeling the point cloud. Finally, it provides information such as the volume of the target to the application server, which can be displayed on the application unit.

[0082] When the camera has an initial tracking task and multiple targets appear at the same time, the dynamic task allocation algorithm is started to re-evaluate the priority of the tracking task according to the distance, scale and moving speed of the target, and insert, replace or delete the task; according to the task priority, various parameters of each task are collected; in a task, the collection of camera parameters still follows the above scheme, selecting the nearest camera, and then the camera with the smallest Δd is used for key tracking and linkage;

[0083] The application server performs identification based on multi-dimensional parameter estimation, establishes a state identification matrix for the parameter results of the camera, radar, level meter, and sonar, and identifies or detects each state at least three times; the state identification matrix can also be expanded or reduced according to the content or number of times detected by the data acquisition unit;

[0084] like Figure 2 As shown, the state recognition matrix has 4 rows of content, representing the three results of camera, radar signal, level meter, and sonar detection respectively;

[0085] When the radar or sound wave drives away the equipment or the level meter or sonar alarms, the camera automatically starts the video tracking function. If the camera verifies that the alarm is real, the alarm signal is actually generated on the application server.

[0086] If the camera verification alarm does not exist, the alarm signal will not be generated on the application server;

[0087] The alarm parameter for cameras, radars, and sonic drive devices is distance in meters; the alarm parameter for level meters is level difference in meters; the alarm parameter for sonars is the volume of the object in cubic meters; the alarm parameters can be set in advance.

[0088] The data server includes a database, which stores real-time monitoring data collected by the data collection unit, data after algorithm calculation by the application server, and system authority information;

[0089] Application unit, used to display the results of data processing and realize human-computer interaction;

[0090] The application unit is a mobile phone or a display; the location of the target, such as longitude, latitude and azimuth, running speed and scale, video images, camera tracking records, alarm information, can be displayed on the application interface of the mobile phone or the display, and a monitoring situation map can be formed. Defense zone management such as creating or deleting defense zones can be performed, and monthly, quarterly or annual health reports can be automatically output.

[0091] The early warning implementation unit is used to receive the alarm signal sent by the data processing unit to drive away the target.

[0092] The early warning implementation unit is a sound wave driving away device.

[0093] The water area monitoring and early warning system of the present invention further includes a core switch;

[0094] Level meters, cameras, radars, sonars, and acoustic repellent devices are connected to the core switch via signal cables and then to the application server and data server;

[0095] The core switch is isolated from the administrative office network through a firewall, and alarm information and video images can be viewed in the administrative office network, or they can be remotely configured; the KVM all-in-one machine is based on KVM-over-IP technology, and it has a remote management function, through which personnel can switch and view various servers and detection equipment. External data such as meteorological and oceanographic data can also be connected to the core switch through the network for remote viewing by personnel.

[0096] Externally connected data collectors are connected to the core switch through the network.

[0097] The core switch contains a wireless routing module, which is connected to the mobile phone through the operator's wireless network. The alarm information can be sent to the mobile phone for viewing, or the core switch is connected to the wireless terminal device, and then the DTU sends the alarm information to the mobile phone text message.

[0098] The camera is a binocular or trinocular camera, which realizes detailed monitoring and intelligent tracking of the target object by working together with two or more cameras. It can also measure the volume of the object and realize linkage with the data processing unit.

[0099] The sonar emits sound waves in the water to obtain point cloud data of the surface of the target, then obtains a three-dimensional model by modeling the point cloud, and finally provides information about the target found in the water to the application server.

[0100] The false alarm rate of the system of the present invention should be no greater than 5%, the software startup time should be less than 50 seconds, the average system response time should be no greater than 5 seconds, and the system has an extensible interface.

[0101] The embodiment of the present invention also provides a water area monitoring and early warning method, such as Figure 3 As shown, the following steps are included:

[0102] Step 1: Detect dangerous objects: The radar or sonar performs a blind polling search within the detection radius on the water surface or in the water according to the set frequency, and locates the target based on the information of the reflected waves;

[0103] The information of the wave includes signal strength, time difference, and incident angle.

[0104] Step 2: Determine the tracking method and equipment to be used, including:

[0105] If there is only one detection target, the camera closest to the target is selected for key tracking and linkage;

[0106] Given the motion coordinates of the target (x m ,y θ ), according to the coordinates of the discovered target (x m ,y θ ) and the coordinates of the installed i-th camera (x i ,y i ), calculate the distance Δd between the two; select the camera with the smallest Δd for key tracking and linkage;

[0107] m, θ are integers, and i is an integer.

[0108]

[0109] If the camera already has an initial tracking task and multiple targets appear at the same time, the dynamic task allocation algorithm is started, and the priority of the tracking task is re-evaluated according to the characteristics of the target, and the task is inserted, replaced or deleted, and the moving speed and volume of the tracked target are displayed in turn;

[0110] The target characteristics include any one or more of the target's distance, size, and moving speed.

[0111] The algorithm of dynamic task allocation satisfies the maximum fitness and minimum dynamic disturbance.

[0112] The calculation formula of fitness f is:

[0113]

[0114] Among them, A is the task completion rate of camera tracking, B is the tracking accuracy of the camera, C is the resource relaxation, D is the switching rate index of various detectors, and ɑ1, ɑ2, ɑ3 and ɑ4 are the weights of each index;

[0115] Dynamic disturbance degree S formula:

[0116] Dynamic disturbance

[0117] Among them, ɑ5 is the weight of the deletion task, n5 is the total number of deletions in the new and old tasks, ɑ6 is the weight of the replacement task, n6 represents the total number of replacements in the new and old tasks, and n is the total number of new and old tasks.

[0118] The priority of the tracking task is calculated according to the following formula:

[0119]

[0120] k t is the priority of the tth tracking task, s t is the size or area of ​​the tth tracking task, v t is the moving speed of the t-th tracking task, d t is the distance between the t-th tracking task and the water monitoring and early warning system. When the camera already has an initial tracking task and multiple targets appear at the same time, the application server will re-formulate the tracking priority order of the newly appeared tasks according to the distance, scale and moving speed of the target, and select subtasks to directly insert into the initial tracking plan until there are no new subtasks. At this time, a task conflict set is formed, and subtasks are selected from the conflict set for reallocation. Subtasks can be replaced or deleted until the task conflict set is empty, while satisfying the maximum fitness and minimum dynamic disturbance, and finally forming an adjusted tracking plan, which can avoid missing the timely tracking of more important or more dangerous targets.

[0121] like Figure 4 As shown, according to the priority, select subtasks from the newly appeared target and directly insert them into the original task tracking plan, and judge whether the subtask is successfully inserted. If not, record the new subtask that cannot be inserted, and then update the new subtask; if the insertion is successful, directly update the new subtask;

[0122] Determine whether the new subtask set is empty. If it is empty, return to determine whether the subtask is successfully inserted; if it is not empty, construct a task conflict set;

[0123] Select subtasks from the conflict set and replace the atomic task;

[0124] Determine whether the task is replaced successfully. If so, continue to determine whether the task conflict set is empty. If not, perform a priority comparison and delete the task, and then continue to determine whether the conflict set is empty.

[0125] If the task conflict set is empty, return to select a subtask from the conflict set; if it is not empty, determine whether the maximum fitness and minimum dynamic disturbance are met;

[0126] If it is satisfied, an adjusted tracking plan is formed; if it is not satisfied, a subtask is selected from the conflict set.

[0127] Step 3: Based on the multi-dimensional parameters obtained by tracking, a state recognition matrix is ​​established for the parameter results of the camera, radar, level meter, and sonar. The calculation results of the state recognition matrix are used to determine whether to issue an alarm signal;

[0128] When the radar or sound wave drives away the equipment or the level meter or sonar alarms, the camera automatically starts the video tracking function. If the camera verifies that the alarm is real, the alarm signal is actually generated on the application server.

[0129] If the camera verification alarm does not exist, the alarm signal will not be generated on the application server;

[0130] In the state recognition matrix, each state is recognized or detected at least 3 times;

[0131] The state recognition matrix has 4 rows of content, representing the three results of camera, radar signal, level meter and sonar detection respectively.

[0132] Step 4: Drive the target away.

[0133] The above embodiments are only used to help understand the method and core idea of ​​the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made to the present invention without departing from the principles of the present invention, and these improvements and modifications also fall within the scope of protection of the claims of the present invention.

[0134] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A water area monitoring and early warning system, characterized in that: include: A data acquisition unit is used to monitor and collect information about dangerous objects in real time, and transmit the information to a data processing unit; The data acquisition unit includes a camera and other non-visual acquisition equipment; A data processing unit, used to perform data algorithm calculations on dangerous object information and form a database; The data processing unit includes an application server and a data server. The application server performs algorithmic operations, including dynamic task allocation algorithms and state recognition matrix calculations; The dynamic task allocation algorithm is that when the camera has an initial tracking task and multiple targets appear at the same time, the priority of the tracking task is evaluated according to the target information, and the task is inserted, replaced or deleted; The state recognition matrix calculation is to establish a state recognition matrix for the parameters detected by the non-visual acquisition device and the camera, and when the non-visual acquisition device alarms, the camera is matched to verify the alarm and the alarm is comprehensively calculated to obtain an alarm signal; when the non-visual acquisition device alarms, if the camera does not issue an alarm, a non-alarm signal is calculated; The data server includes a database, which stores real-time monitoring data collected by the data collection unit, data after algorithm calculation by the application server, and system authority information; Application unit, used to display the results of data processing and realize human-computer interaction; The early warning implementation unit is used to receive the alarm signal sent by the data processing unit to drive away the target.

2. The water area monitoring and early warning system according to claim 1 is characterized in that: Other non-visual acquisition devices of the data acquisition unit include radar, liquid level meter, sonar and externally connected data acquisition device.

3. The water area monitoring and early warning system according to claim 2 is characterized in that: The application server performs identification based on multi-dimensional parameter estimation, establishes a state identification matrix for the parameter results of the camera, radar, level meter, and sonar, and identifies or detects each state at least 3 times; The state recognition matrix has 4 rows of content, representing the three results of camera, radar signal, level meter and sonar detection respectively; When the radar or sound wave drives away the equipment or the level meter or sonar alarms, the camera automatically starts the video tracking function. If the camera verifies that the alarm is real, the alarm signal is actually generated on the application server. If the camera verification alarm does not exist, the alarm signal will not be generated on the application server; Application server command warning implementation unit.

4. The water area monitoring and early warning system according to claim 3 is characterized in that: The early warning implementation unit is a sound wave driving away device.

5. The water area monitoring and early warning system according to claim 2 is characterized in that: It also includes core switches; Level meters, cameras, radars, sonars, and acoustic repellent devices are connected to the core switch via signal cables and then to the application server and data server; The core switch is isolated from the administrative office network through a firewall, and the alarm information and video images can be viewed in the administrative office network, or they can be remotely configured; Externally connected data collectors are connected to the core switch through the network.

6. The water area monitoring and early warning system according to claim 1 is characterized in that: The camera is a binocular or trinocular camera, which realizes detailed monitoring and intelligent tracking of the target object by working together with two or more cameras. It can also measure the volume of the object and realize linkage with the data processing unit.

7. The water area monitoring and early warning system according to claim 3 is characterized in that: The camera parameters in the state recognition matrix are collected by the camera that is closest to the target; The camera closest to the target is selected and calculated by the application server.

8. The water area monitoring and early warning system according to claim 5 is characterized in that: The application unit is a mobile phone or a display; the core switch contains a wireless routing module, which is connected to the mobile phone through the operator's wireless network. The alarm information can be sent to the mobile phone for viewing, or the core switch is connected to the wireless terminal device, and then the DTU sends the alarm information to the mobile phone text message.

9. The water area monitoring and early warning system according to claim 3, characterized in that: The sonar emits sound waves in the water to obtain point cloud data of the surface of the target, then obtains a three-dimensional model by modeling the point cloud, and finally provides information about the target found in the water to the application server.

10. A water area monitoring and early warning method, characterized in that: The following steps are involved: Step 1: Detect dangerous objects: The radar or sonar performs a blind polling search within the detection radius on the water surface or in the water according to the set frequency, and locates the target based on the information of the reflected waves; Step 2: Determine the tracking method and equipment to be used, including: If there is only one detection target, the camera closest to the target is selected for key tracking and linkage; If the camera already has an initial tracking task and multiple targets appear at the same time, the dynamic task allocation algorithm is started, and the priority of the tracking task is re-evaluated according to the characteristics of the target, and the task is inserted, replaced or deleted, and the moving speed and volume of the tracked target are displayed in turn; Step 3: Based on the multi-dimensional parameters obtained by tracking, a state recognition matrix is ​​established for the parameter results of the camera, radar, level meter, and sonar. When the radar or sound wave drives away the equipment or the level meter or sonar alarms, the camera automatically starts the video tracking function. If the camera verifies that the alarm is real, the alarm signal is actually generated on the application server. If the camera verification alarm does not exist, the alarm signal will not be generated on the application server; Step 4: Drive the target away.

11. The water area monitoring and early warning method according to claim 10, characterized in that: The method of selecting the camera closest to the target is: given the motion coordinates of the target (x m ,y θ ), according to the coordinates of the discovered target (x m ,y θ ) and the coordinates of the installed i-th camera (x i ,y i ), calculate the distance Δd between the two; select the camera with the smallest Δd for key tracking and linkage; m, θ are integers, and i is an integer.

12. The water area monitoring and early warning method according to claim 10, characterized in that: In the state recognition matrix, each state is recognized or detected at least 3 times; The state recognition matrix has 4 rows of content, representing the three results of camera, radar signal, level meter and sonar detection respectively.

13. The water area monitoring and early warning method according to claim 10, characterized in that: The information of the wave includes signal strength, time difference, and incident angle.

14. The water area monitoring and early warning method according to claim 10, characterized in that: The target characteristics include any one or more of the target's distance, size, and moving speed.

15. The water area monitoring and early warning method according to claim 10, characterized in that: The priority of the tracking task is calculated according to the following formula: k t is the priority of the tth tracking task, s t is the size or area of ​​the tth tracking task, v t is the moving speed of the t-th tracking task, d t is the distance between the tth tracking task and the water monitoring and early warning system.

16. The water area monitoring and early warning method according to claim 10, characterized in that: The algorithm of dynamic task allocation satisfies the maximum fitness and minimum dynamic disturbance. The calculation formula of fitness f is: Among them, A is the task completion rate of camera tracking, B is the tracking accuracy of the camera, C is the resource relaxation, D is the switching rate index of various detectors, and ɑ1, ɑ2, ɑ3 and ɑ4 are the weights of each index; Dynamic disturbance degree S formula: Among them, ɑ5 is the weight of the deletion task, n5 is the total number of deletions in the new and old tasks, ɑ6 is the weight of the replacement task, n6 represents the total number of replacements in the new and old tasks, and n is the total number of new and old tasks.

17. The water area monitoring and early warning method according to claim 16, characterized in that: When the camera has an initial tracking task and multiple targets appear at the same time, According to the priority, select subtasks from the newly appeared target and directly insert them into the original task tracking plan to determine whether the subtask is successfully inserted. If not, record the new subtask that cannot be inserted, and then update the new subtask; If the insertion is successful, the new subtask is updated directly; Determine whether the new subtask set is empty. If it is not empty, return to determine whether the subtask is inserted successfully; If it is empty, construct a task conflict set; Select subtasks from the conflict set and replace the atomic task; Determine whether the task is replaced successfully. If so, continue to determine whether the task conflict set is empty. If not, perform a priority comparison and delete the task, and then continue to determine whether the conflict set is empty. If the task conflict set is not empty, return to select a subtask from the conflict set; If it is empty, determine whether the maximum fitness and minimum dynamic disturbance are met; If it is satisfied, an adjusted tracking plan is formed; if it is not satisfied, a subtask is selected from the conflict set.

Images