Water area safety monitoring method, system, equipment and medium based on clustering algorithm

Abstract

The present application discloses a water safety monitoring method, system, device and medium based on clustering algorithm, which mainly relates to the field of water safety monitoring technology, and is used to solve the problem that it is difficult for existing solutions to formulate appropriate camera layout strategies according to actual needs. It includes: dividing the preset waterside monitoring area into M sub-areas, and dividing the sub-areas into N*N regional point blocks; using the location data of the regional point blocks and the number of people in the preset time interval as the input data of K-mean clustering, using the preset number of cameras as the K value, clustering all the regional point blocks in the current sub-area, and obtaining K cluster centers; updating the K value; until the placement method corresponding to the updated K value cannot cover the entire sub-area; determining the camera placement position corresponding to the previous K value as the final camera placement position.

Description

Technical Field

[0001] The present application relates to the technical field of water area safety monitoring, and in particular to a water area safety monitoring method, system, equipment and medium based on a clustering algorithm. Background Art

[0002] The water area safety monitoring system can conduct 24-hour uninterrupted high-definition video intelligent monitoring of dangerous behaviors in key waters of rivers and lakes through visual and intelligent technical prevention and control means, and can automatically identify and alarm risky behaviors such as people falling into the water, effectively improving the supervision capacity and efficiency of safety monitoring work in key waters.

[0003] In the existing water safety monitoring system, the deployment of cameras is usually based on fixed positions and viewing angles, making it difficult to formulate appropriate strategies based on actual needs and flexibly adjust the monitoring scope. Especially when the flow of people is large and unevenly distributed, cameras need to be set up in every corner of the water area. The increase in cameras not only increases the cost, but also the amount of collected data that the system needs to process is large due to the large number of cameras, and the analysis and processing time is prolonged, which limits the effectiveness of the monitoring system.

[0004] Therefore, there is an urgent need for a water safety monitoring method, system, device and medium based on a clustering algorithm to solve the existing camera deployment scheme: cameras are deployed in every corner of the water area, which leads to a large number of cameras and a large amount of collected data that the system needs to process. The existing scheme is difficult to formulate a suitable camera deployment strategy according to actual needs. Summary of the invention

[0005] In view of the above-mentioned deficiencies in the prior art, the present application provides a water safety monitoring method, system, device and medium based on a clustering algorithm to solve the problem that it is difficult for existing solutions to formulate appropriate camera deployment strategies according to actual needs.

[0006] In a first aspect, the present application provides a water safety monitoring method based on a clustering algorithm, the method comprising: The preset waterfront monitoring area is divided into M sub-areas, and the sub-areas are divided into N*N regional point blocks; the location data of the regional point block is (m, x, y), m represents the mth sub-area, and the value range of x and y is [1, N]; the number of people flow in each regional point block within the preset time interval in the preset historical time period is obtained; the location data (m, x, y) of the regional point block and the number of people flow in the preset time interval are used as the input data of K-mean clustering, and the preset number of cameras is used as the K value to cluster all the regional point blocks in the current sub-area to obtain K cluster centers; according to the corresponding K cluster centers The location data (m, x, y) is used to determine the placement of the camera; the monitoring distance of the camera is determined according to the number of people corresponding to the K cluster centers; the placement position and the monitoring distance of the camera are used to determine whether the current placement method makes the camera's monitoring area cover the entire sub-area; when the entire sub-area is covered, KQ operation is performed on K to update the K value; where Q is the preset reduction amplitude; whether the placement method corresponding to the updated K value can cover the entire sub-area is determined; the KQ operation is iterated until the placement method corresponding to the updated K value cannot cover the entire sub-area; the camera placement position corresponding to the previous K value is determined to be the final camera placement position.

[0007] The water safety monitoring method provided in the embodiment of the present application can more accurately identify different crowded areas by dividing the monitoring area into sub-areas and regional blocks and obtaining the number of people in each regional block in the historical time period. The locations of different crowded areas are calculated based on K-mean clustering, and the number of cameras preset k is continuously reduced until the number of camera deployment points with the minimum number of cameras that can cover all crowded areas is calculated, thereby reducing the number of cameras and optimizing their deployment positions. This solves the problem that it is difficult to formulate a suitable camera deployment strategy according to actual needs in the existing solutions.

[0008] In one implementation of the present application, after determining that the camera placement position corresponding to the previous K value is the final camera placement position, the method further includes: Based on the historical number of people flow in each area block at each preset time interval within the preset number of days, the moving average method in the time series prediction method is used to predict the number of people flow in each area block at the current preset time interval; then the predicted total number of people flow in the area block covered by each camera at the current preset time interval is obtained; based on the range within which the total number of people flow falls, the camera focal length of the camera at the current preset time interval is adjusted; when the range falls into the preset high frequency range of people flow, the camera focal length is determined to be the a value corresponding to the monitoring distance A; when the range falls into the preset medium frequency range of people flow, the camera focal length is determined to be the b value corresponding to the monitoring distance B; when the range falls into the preset low frequency range of people flow, the camera focal length is determined to be the b value corresponding to the monitoring distance B.

[0009] In one implementation of the present application, the monitoring distance of the camera is determined according to the number of people flows corresponding to the K cluster centers, specifically including: When the number of people falls within the preset high frequency range, the monitoring distance of the camera is determined to be A; when the number of people falls within the preset medium frequency range, the monitoring distance of the camera is determined to be B; when the number of people falls within the preset low frequency range, the monitoring distance of the camera is determined to be B.

[0010] In one implementation of the present application, determining whether the current placement method enables the monitoring area of ​​the camera to cover the entire sub-area is performed based on the placement position and the monitoring distance of the camera, specifically including: Utilize the simulation function of the camera detection range provided by the preset camera monitoring system, input the placement position and the monitoring distance of the camera, and automatically generate a graphical representation of the detection range in the sub-area; determine whether the current placement method makes the camera's monitoring area cover the entire sub-area based on the graphical representation.

[0011] In a second aspect, the present application provides a water safety monitoring system based on a clustering algorithm, the system comprising: The division module is used to divide the preset waterfront monitoring area into M sub-areas, and the sub-areas into N*N regional point blocks; wherein the location data of the regional point block is (m, x, y), m represents the mth sub-area, and the value range of x and y is [1, N]; the acquisition module is used to obtain the number of people flow in each regional point block within a preset time interval within a preset historical time period; the iteration module is used to use the location data (m, x, y) of the regional point block and the number of people flow in the preset time interval as the input data of K-mean clustering, and the preset number of cameras as the K value, to cluster all the regional point blocks of the current sub-area and obtain K cluster centers; According to the position data (m, x, y) corresponding to the K cluster centers, determine the placement position of the camera; according to the number of people corresponding to the K cluster centers, determine the monitoring distance of the camera; through the placement position and the monitoring distance of the camera, determine whether the current placement method makes the camera's monitoring area cover the entire sub-area; when the entire sub-area is covered, perform KQ operation on K to update the K value; where Q is the preset reduction amplitude; determine whether the placement method corresponding to the updated K value can cover the entire sub-area; iterate the KQ operation until the placement method corresponding to the updated K value cannot cover the entire sub-area; determine the camera placement position corresponding to the previous K value as the final camera placement position.

[0012] In one implementation of the present application, the iteration module includes an adjustment unit, It is used to predict the number of people in each regional point block at the current preset time interval based on the historical number of people flow in each regional point block at each preset time interval within a preset number of days, using the moving average method in the time series prediction method; and then obtain the predicted total number of people flow in the regional point blocks in the area covered by each camera at the current preset time interval; based on the range within which the total number of people flow falls, adjust the camera focal length of the camera at the current preset time interval; wherein, when the range falls into the preset high frequency range of people flow, determine the camera focal length to be the a value corresponding to the monitoring distance A; when the range falls into the preset medium frequency range of people flow, determine the camera focal length to be the b value corresponding to the monitoring distance B; when the range falls into the preset low frequency range of people flow, determine the camera focal length to be the b value corresponding to the monitoring distance B.

[0013] In one implementation of the present application, the iteration module includes a determination unit, It is used to determine the monitoring distance of the camera as A when the number of people falls into the preset high frequency range; when the number of people falls into the preset medium frequency range, the monitoring distance of the camera is determined to be B; when the number of people falls into the preset low frequency range, the monitoring distance of the camera is determined to be B.

[0014] In one implementation of the present application, the iteration module includes a simulation unit, It is used to utilize the simulation function of the camera detection range provided by the preset camera monitoring system, input the placement position and the monitoring distance of the camera, and automatically generate a graphical representation of the detection range in the sub-area; based on the graphical representation, determine whether the current placement method makes the camera's monitoring area cover the entire sub-area.

[0015] In a third aspect, the present application provides a water safety monitoring device based on a clustering algorithm, the device comprising: processor; and a memory having executable codes stored thereon, which, when executed, causes the processor to execute a water safety monitoring method based on a clustering algorithm as described above.

[0016] In a fourth aspect, the present application provides a non-volatile computer storage medium on which computer instructions are stored. When the computer instructions are executed, they implement a water safety monitoring method based on a clustering algorithm as described above.

[0017] Those skilled in the art can understand that the present application has at least the following beneficial effects: The present application discloses a water safety monitoring method, system, device and medium based on clustering algorithm. By dividing the monitoring area into sub-areas and regional blocks, and obtaining the number of people flow in each regional block in the historical time period, it is possible to more accurately identify different crowded areas. The locations of different crowded areas are calculated according to K-mean clustering, and the number of cameras preset k is continuously reduced until the number of camera deployment points with the minimum number of cameras that can cover all crowded areas is calculated, thereby reducing the number of cameras and optimizing their deployment positions. The problem that it is difficult to formulate a suitable camera deployment strategy according to actual needs in the existing solution is solved. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Some embodiments of the present disclosure are described below with reference to the accompanying drawings, in which: Figure 1 This is a flow chart of a water area safety monitoring method based on a clustering algorithm provided in an embodiment of the present application.

[0019] Figure 2 It is a schematic diagram of the internal structure of a water area safety monitoring system based on a clustering algorithm provided in an embodiment of the present application.

[0020] Figure 3 It is a schematic diagram of the internal structure of a water safety monitoring device based on a clustering algorithm provided in an embodiment of the present application. DETAILED DESCRIPTION

[0021] It should be understood by those skilled in the art that the embodiments described below are only preferred embodiments of the present disclosure, and do not mean that the present disclosure can only be implemented through the preferred embodiments. The preferred embodiments are only used to explain the technical principles of the present disclosure, and are not used to limit the protection scope of the present disclosure. Based on the preferred embodiments provided by the present disclosure, all other embodiments obtained by ordinary technicians in this field without creative work should still fall within the protection scope of the present disclosure.

[0022] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0023] In order to clearly describe the technical solutions of the embodiments of the present application, some terms and technologies involved in the embodiments of the present application are briefly introduced below: 1. Traffic flow: The number of people passing through an area or place within a specific period of time, usually used to measure the busyness of a place or the intensity of people's activities.

[0024] 2. Camera placement point: The specific location where the camera is installed or deployed.

[0025] 3. Water area safety monitoring system: Through visualization and intelligent technical prevention and control means, it conducts 24-hour uninterrupted high-definition video intelligent monitoring of dangerous behaviors in key waters of rivers and lakes, and can realize automatic identification and alarm of risky behaviors such as people falling into the water, effectively improving the supervision capacity and efficiency of safety monitoring work in key waters.

[0026] In order to solve the existing camera deployment scheme: cameras are deployed in every corner of the water area, which leads to a large number of cameras and a large amount of collected data that the system needs to process. The existing scheme is difficult to formulate a suitable camera deployment strategy according to actual needs. This application provides a water safety monitoring method, system, device and medium based on a clustering algorithm, which intelligently analyzes the shore traffic data to determine the best camera deployment position, and realizes flexible switching of long-distance and short-distance monitoring through adjustable focus cameras, thereby reducing the number of cameras and improving the overall performance of the water safety monitoring system.

[0027] The technical solution proposed in the embodiments of the present application is described in detail below with reference to the accompanying drawings.

[0028] The embodiment provides a water area safety monitoring method based on a clustering algorithm, such as Figure 1 As shown, the method provided in the embodiment of the present application mainly includes the following steps: Step 110: Divide the preset waterfront monitoring area into M sub-areas, and divide the sub-areas into N*N area point blocks.

[0029] It should be noted that the location data of the regional point block is (m, x, y), where m represents the mth sub-region, and the value range of x and y is [1, N].

[0030] In some embodiments, the preset waterside monitoring area is divided into M sub-areas, and the sub-areas are divided into N*N area point blocks, which can be specifically: Get the bounding rectangle of the monitoring area.

[0031] The size of each sub-region is calculated according to the size of the circumscribed rectangle and the number M of sub-regions to be divided.

[0032] The entire area is divided into M sub-areas by uniform grid according to the calculated size. N can be 2m.

[0033] Step 120: Obtain the number of people flow in each area point block within a preset time interval within a preset historical time period.

[0034] It should be noted that the preset historical time period can be specifically 3 months. The preset time interval can be: taking every two hours as a time period, dividing the whole day into 12 preset time intervals, and counting the number of people flow in each monitoring area point block within the preset time interval.

[0035] Step 130: Use the location data (m, x, y) of the regional point blocks and the number of people in the preset time interval as input data for K-mean clustering, use the preset number of cameras as the K value, cluster all the regional point blocks in the current sub-region, and obtain K cluster centers.

[0036] It should be noted that the camera has a limited field of view, and the water surface within the camera field of view is considered to be a plane with a consistent elevation. The camera is installed in a fixed position and can monitor the required area. The focal length of the camera can be adjusted. If the monitoring distance is within the interval (0, A) meters, the model recognition and warning accuracy is high; if the monitoring distance is within the interval (A, B) meters, the model recognition and warning accuracy is average. The data on the number of people appearing on the shore of the water every hour can be checked.

[0037] Step 140, determine the placement position of the camera according to the position data (m, x, y) corresponding to the K cluster centers; determine the monitoring distance of the camera according to the number of people corresponding to the K cluster centers; and determine whether the current placement method makes the monitoring area of ​​the camera cover the entire sub-area through the placement position and the monitoring distance of the camera.

[0038] According to the number of people flow corresponding to the K cluster centers, the monitoring distance of the camera is determined, which can be specifically: When the number of people falls into the preset high frequency range, the monitoring distance of the camera is determined to be A; When the number of people falls within the preset frequency range, the monitoring distance of the camera is determined to be B; When the number of people falls into the preset low frequency range, the monitoring distance of the camera is determined to be B.

[0039] It should be noted that the change in monitoring distance can be achieved by adjusting the focal length of the camera.

[0040] It should be noted that according to the number of people in each monitoring area block in the 12 preset time intervals every day, the monitoring area blocks are divided into three categories, namely low-frequency people flow monitoring area blocks (the number of people falls into the preset low-frequency range), medium-frequency people flow monitoring area blocks (the number of people falls into the preset medium-frequency range), and high-frequency people flow monitoring area blocks (the number of people falls into the preset high-frequency range). The categories of monitoring area blocks in different time periods are variable. The basis for the division of monitoring area blocks is as follows: the upper quartile and lower quartile of the frequency are used as the dividing points. The frequency of the monitoring area block is defined as a low-frequency people flow monitoring area block; the frequency of the monitoring area block is defined as a medium-frequency people flow monitoring area block; the frequency of the monitoring area block is defined as a high-frequency people flow monitoring area block.

[0041] In the step, it is determined whether the current placement method enables the monitoring area of ​​the camera to cover the entire sub-area through the placement position and the monitoring distance of the camera, which can be specifically: Utilize the simulation function of the camera detection range provided by the preset camera monitoring system, input the placement position and the monitoring distance of the camera, and automatically generate a graphical representation of the detection range in the sub-area; determine whether the current placement method makes the camera's monitoring area cover the entire sub-area based on the graphical representation.

[0042] Step 150, when the entire sub-area is covered, perform KQ operation on K to update the K value; determine whether the placement method corresponding to the updated K value can cover the entire sub-area; iterate the KQ operation until the placement method corresponding to the updated K value cannot cover the entire sub-area; determine the camera placement position corresponding to the previous K value as the final camera placement position.

[0043] It should be noted that Q is the preset reduction range.

[0044] It can be understood by those skilled in the art that this step is based on the clustering algorithm K-mean clustering, and the classification center of the monitoring area point block is calculated according to the distance, and the classification center is recorded as the installation position of the camera. According to the clustering results and combined with the maximum coverage of the camera, the cameras of the high-frequency human flow monitoring area point block are ensured to be within a meters. The cameras of the medium and low-frequency human flow monitoring area point blocks are within b meters. The preset number of cameras k is continuously reduced until the minimum number of camera deployment points k that can cover all densely populated areas is calculated.

[0045] In addition, in the actual drowning prevention monitoring and early warning application, the focal length of the camera is adjustable. The closer the monitoring distance, the clearer the monitoring image and the higher the drowning warning recognition accuracy. Therefore, this application can change the focal length of the camera and adjust the camera monitoring range according to the number of people, thereby improving the drowning warning recognition accuracy. The steps are as follows: After determining that the camera placement position corresponding to the previous K value is the final camera placement position, based on the historical number of people flow in each regional point block at each preset time interval within the preset number of days, the moving average method in the time series prediction method is used to predict the number of people flow in each regional point block at the current preset time interval; then obtain the predicted total number of people flow in the regional point blocks of the area covered by each camera at the current preset time interval; based on the range within which the total number of people flow falls, adjust the camera focal length of the camera at the current preset time interval; wherein, when the range falls into the preset high frequency range of people flow, determine the camera focal length to be the a value corresponding to the monitoring distance A; when the range falls into the preset medium frequency range of people flow, determine the camera focal length to be the b value corresponding to the monitoring distance B; when the range falls into the preset low frequency range of people flow, determine the camera focal length to be the b value corresponding to the monitoring distance B.

[0046] It should be noted that the preset number of days can be 3 days. In addition, the rule for adjusting the focal length of the camera is: dynamically adjust the focal length of the camera according to the real-time changes in human flow to optimize the monitoring effect. The specific rules are as follows: prioritize ensuring that the cameras in the high-frequency monitoring area are within A meters. The cameras in the medium and low-frequency human flow monitoring area are within B meters.

[0047] It should be noted that the change in monitoring distance can be achieved by adjusting the focal length of the camera.

[0048] In addition, this application Figure 2 A water safety monitoring system based on a clustering algorithm is provided in an embodiment of the present application. Figure 2 As shown, the system provided in the embodiment of the present application mainly includes: The division module 210 is used to divide the preset waterside monitoring area into M sub-areas, and divide the sub-areas into N*N area point blocks; wherein the position data of the area point block is (m, x, y), m represents the mth sub-area, and the value range of x and y is [1, N].

[0049] The acquisition module 220 is used to acquire the number of people flow in each area point block within a preset time interval within a preset historical time period.

[0050] The iterative module 230 is used to use the position data (m, x, y) of the regional point block and the number of people in a preset time interval as input data for K-mean clustering, and use the preset number of cameras as the K value to cluster all the regional point blocks in the current sub-region to obtain K cluster centers; determine the placement position of the camera according to the position data (m, x, y) corresponding to the K cluster centers; determine the monitoring distance of the camera according to the number of people corresponding to the K cluster centers; determine whether the current placement method makes the camera's monitoring area cover the entire sub-region through the placement position and the camera's monitoring distance; when the entire sub-region is covered, perform a KQ operation on K to update the K value; wherein Q is a preset reduction amplitude; determine whether the placement method corresponding to the updated K value can cover the entire sub-region; iterate the KQ operation until the placement method corresponding to the updated K value cannot cover the entire sub-region; determine the camera placement position corresponding to the previous K value as the final camera placement position.

[0051] In addition, in the actual drowning prevention monitoring and early warning application, the focal length of the camera is adjustable. The closer the monitoring distance, the clearer the monitoring image and the higher the drowning warning recognition accuracy. Therefore, this application can change the focal length of the camera and adjust the camera monitoring range according to the number of people, thereby improving the drowning warning recognition accuracy. The steps are as follows: The adjustment unit in the iteration module 230 predicts the number of people in each regional point block at the current preset time interval based on the historical number of people flow in each regional point block at each preset time interval within the preset number of days, and adopts the moving average method in the time series prediction method to predict the number of people flow in each regional point block at the current preset time interval; then obtains the predicted total number of people flow in the regional point blocks in the area covered by each camera at the current preset time interval; based on the range within which the total number of people flow falls, adjusts the camera focal length of the camera at the current preset time interval; wherein, when the range falls into the preset high frequency range of people flow, the camera focal length is determined to be the a value corresponding to the monitoring distance A; when the range falls into the preset medium frequency range of people flow, the camera focal length is determined to be the b value corresponding to the monitoring distance B; when the range falls into the preset low frequency range of people flow, the camera focal length is determined to be the b value corresponding to the monitoring distance B.

[0052] According to the number of people flow corresponding to the K cluster centers, the monitoring distance of the camera is determined, which can be specifically: The determination unit in the iteration module 230 determines that the monitoring distance of the camera is A when the number of people falls within the preset high frequency range of people flow; when the number of people falls within the preset medium frequency range of people flow, the monitoring distance of the camera is B; when the number of people falls within the preset low frequency range of people flow, the monitoring distance of the camera is B.

[0053] The iteration module 230 determines whether the current placement mode enables the monitoring area of ​​the camera to cover the entire sub-area based on the placement position and the monitoring distance of the camera, which may be: Through the simulation unit in the iteration module 230, using the simulation function of the camera detection range provided by the preset camera monitoring system, the placement position and the monitoring distance of the camera are input, and a graphical representation of the detection range in the sub-area is automatically generated; based on the graphical representation, it is determined whether the current placement method makes the camera's monitoring area cover the entire sub-area.

[0054] The above is a method embodiment of the present application. Based on the same inventive concept, the present application embodiment also provides a water safety monitoring device based on a clustering algorithm. Figure 3 As shown, the device includes: a processor; and a memory, on which executable codes are stored. When the executable codes are executed, the processor executes a water safety monitoring method based on a clustering algorithm as in the above-mentioned embodiment.

[0055] Specifically, the server divides the preset waterfront monitoring area into M sub-areas, and divides the sub-areas into N*N regional point blocks; wherein the location data of the regional point block is (m, x, y), m represents the mth sub-area, and the value range of x and y is [1, N]; obtains the number of people flow in each regional point block within the preset time interval within the preset historical time period; uses the location data (m, x, y) of the regional point block and the number of people flow in the preset time interval as the input data of K-mean clustering, uses the preset number of cameras as the K value, clusters all the regional point blocks of the current sub-area, and obtains K cluster centers; according to the K cluster centers The camera placement position is determined based on the position data (m, x, y) corresponding to the K cluster centers; the camera monitoring distance is determined based on the number of people corresponding to the K cluster centers; the camera placement position is determined based on the placement position and the camera monitoring distance; when the entire sub-area is covered, KQ operation is performed on K to update the K value; Q is the preset reduction amplitude; whether the placement method corresponding to the updated K value can cover the entire sub-area is determined; the KQ operation is iterated until the placement method corresponding to the updated K value cannot cover the entire sub-area; the camera placement position corresponding to the previous K value is determined as the final camera placement position.

[0056] In addition, an embodiment of the present application further provides a non-volatile computer storage medium on which executable instructions are stored. When the executable instructions are executed, a water area safety monitoring method based on a clustering algorithm as described above is implemented.

[0057] So far, the technical solutions of the present disclosure have been described in combination with the above multiple embodiments, but it is easy for those skilled in the art to understand that the protection scope of the present disclosure is not limited to these specific embodiments. Without departing from the technical principles of the present disclosure, those skilled in the art can split and combine the technical solutions in the above-mentioned various embodiments, and can also make equivalent changes or replacements to the relevant technical features. Any changes, equivalent replacements, improvements, etc. made within the technical concept and / or technical principle of the present disclosure will fall within the protection scope of the present disclosure.

Claims

1. A water area safety monitoring method based on clustering algorithm, characterized in that: The method comprises: The preset waterside monitoring area is divided into M sub-areas, and the sub-areas are divided into N*N regional point blocks; wherein the location data of the regional point block is (m, x, y), m represents the mth sub-area, and the value range of x and y is [1, N]; Obtain the number of people flow in each area point block within a preset time interval within a preset historical time period; The location data (m, x, y) of the regional point blocks and the number of people in the preset time interval are used as the input data of K-mean clustering, and the preset number of cameras is used as the K value. All the regional point blocks in the current sub-region are clustered to obtain K cluster centers. According to the location data (m, x, y) corresponding to the K cluster centers, the placement of the camera is determined; according to the number of people corresponding to the K cluster centers, the monitoring distance of the camera is determined; through the placement position and the monitoring distance of the camera, it is determined whether the current placement method makes the monitoring area of ​​the camera cover the entire sub-area; When the entire sub-area is covered, perform KQ operation on K to update the K value; wherein Q is the preset reduction amplitude; determine whether the placement method corresponding to the updated K value can cover the entire sub-area; iterate the KQ operation until the placement method corresponding to the updated K value cannot cover the entire sub-area; determine the camera placement position corresponding to the previous K value as the final camera placement position.

2. The water area safety monitoring method based on clustering algorithm according to claim 1 is characterized in that: After determining that the camera placement position corresponding to the previous K value is the final camera placement position, the method further includes: Based on the historical number of people flow in each area block at each preset time interval within the preset number of days, the moving average method in the time series prediction method is used to predict the number of people flow in each area block at the current preset time interval; Then, the predicted total number of people flow in the area point blocks of the area covered by each camera in the current preset time interval is obtained; Based on the range within which the total number of people falls, adjust the focal length of the camera in the current preset time interval; when the range falls into the preset high frequency range of people flow, determine the focal length of the camera to be the a value corresponding to the monitoring distance A; when the range falls into the preset medium frequency range of people flow, determine the focal length of the camera to be the b value corresponding to the monitoring distance B; when the range falls into the preset low frequency range of people flow, determine the focal length of the camera to be the b value corresponding to the monitoring distance B.

3. The water area safety monitoring method based on clustering algorithm according to claim 1 is characterized in that: According to the number of people flow corresponding to the K cluster centers, the monitoring distance of the camera is determined, including: When the number of people falls into the preset high frequency range, the monitoring distance of the camera is determined to be A; When the number of people falls within the preset frequency range, the monitoring distance of the camera is determined to be B; When the number of people falls into the preset low frequency range, the monitoring distance of the camera is determined to be B.

4. The water area safety monitoring method based on clustering algorithm according to claim 1 is characterized in that: Based on the placement position and the monitoring distance of the camera, determine whether the current placement method enables the camera's monitoring area to cover the entire sub-area, including: Using the camera detection range simulation function provided by the preset camera monitoring system, input the placement location and monitoring distance of the camera to automatically generate a graphical representation of the detection range in the sub-area; Determine based on the graphical representation whether the current placement enables the monitoring area of ​​the camera to cover the entire sub-area.

5. A water area safety monitoring system based on clustering algorithm, characterized in that: The system comprises: A division module is used to divide the preset water bank monitoring area into M sub-areas, and divide the sub-areas into N*N area point blocks; wherein the location data of the area point block is (m, x, y), m represents the mth sub-area, and the value range of x and y is [1, N]; An acquisition module is used to obtain the number of people flow in each area point block within a preset time interval within a preset historical time period; The iterative module is used to use the location data (m, x, y) of the regional point blocks and the number of people in a preset time interval as the input data of K-mean clustering, and the preset number of cameras as the K value, cluster all the regional point blocks of the current sub-region to obtain K cluster centers; determine the placement position of the camera according to the location data (m, x, y) corresponding to the K cluster centers; determine the monitoring distance of the camera according to the number of people corresponding to the K cluster centers; determine whether the current placement method makes the camera's monitoring area cover the entire sub-region through the placement position and the camera's monitoring distance; when the entire sub-region is covered, perform KQ operation on K to update the K value; wherein Q is the preset reduction amplitude; determine whether the placement method corresponding to the updated K value can cover the entire sub-region; iterate the KQ operation until the placement method corresponding to the updated K value cannot cover the entire sub-region; determine the camera placement position corresponding to the previous K value as the final camera placement position.

6. The water area safety monitoring system based on clustering algorithm according to claim 5 is characterized in that: The iteration module includes an adjustment unit, Based on the historical number of people flow in each area block at each preset time interval within the preset number of days, the moving average method in the time series prediction method is used to predict the number of people flow in each area block at the current preset time interval; Then, the predicted total number of people flow in the area point blocks of the area covered by each camera in the current preset time interval is obtained; Based on the range within which the total number of people falls, adjust the focal length of the camera in the current preset time interval; when the range falls into the preset high frequency range of people flow, determine the focal length of the camera to be the a value corresponding to the monitoring distance A; when the range falls into the preset medium frequency range of people flow, determine the focal length of the camera to be the b value corresponding to the monitoring distance B; when the range falls into the preset low frequency range of people flow, determine the focal length of the camera to be the b value corresponding to the monitoring distance B.

7. The water area safety monitoring system based on clustering algorithm according to claim 5 is characterized in that: The iteration module includes a determination unit, When the number of people falls into the preset high frequency range, the monitoring distance of the camera is determined to be A; When the number of people falls within the preset frequency range, the monitoring distance of the camera is determined to be B; When the number of people falls into the preset low frequency range, the monitoring distance of the camera is determined to be B.

8. The water area safety monitoring system based on clustering algorithm according to claim 5 is characterized in that: The iterative module includes simulation units, for utilizing the simulation function of the camera detection range provided by the preset camera monitoring system, inputting the placement location and the monitoring distance of the camera, and automatically generating a graphical representation of the detection range in the sub-area; Determine based on the graphical representation whether the current placement enables the monitoring area of ​​the camera to cover the entire sub-area.

9. A water area safety monitoring device based on clustering algorithm, characterized in that: The device comprises: processor; and a memory storing executable codes thereon, which, when the executable codes are executed, enable the processor to execute a water area safety monitoring method based on a clustering algorithm as described in any one of claims 1 to 4.

10. A non-volatile computer storage medium, characterized in that: Computer instructions are stored thereon, and when the computer instructions are executed, the water area safety monitoring method based on clustering algorithm as described in any one of claims 1 to 4 is implemented.

Images