Method for realizing swimming pool and coast safety monitoring based on AI image recognition technology

Abstract

The embodiment of the present invention discloses a method for realizing swimming pool and coast safety monitoring based on AI image recognition technology, which relates to the field of image recognition technology. The method comprises: acquiring images of target waters in real time by at least one image acquisition device; identifying whether there is a dangerous state in the image of the target waters by AI image recognition technology; if a dangerous state is identified in the image of the target waters, an alarm is issued to the lifeguard in the target waters. The present invention can automatically and intelligently identify the dangerous state in the target waters and issue an automatic alarm through AI image recognition technology, which effectively improves the efficiency of monitoring of swimming pools and coasts, saves monitoring manpower costs, and avoids the occurrence of dangerous events.

Description

Technical Field

[0001] The present invention belongs to the field of image recognition technology, and in particular relates to a method for realizing swimming pool and coast safety monitoring based on AI image recognition technology. Background Art

[0002] In recent years, dangerous accidents have occurred frequently in waters such as swimming pools or beaches, such as drowning or injuries caused by dangerous creatures in the sea (such as sharks, poisonous jellyfish, poisonous fish, etc.). In order to reduce the occurrence of dangerous accidents and reduce the impact of dangerous accidents, existing swimming pools and coasts are mainly monitored by lifeguards on special lifeguard chairs to prevent drowning accidents and dangerous creatures in the sea. Although dangerous accidents and their impacts are reduced to a certain extent, lifeguards have to rely on manual observation to discover dangerous accidents. It is difficult to cover swimming pools with large traffic or wide coasts, resulting in poor monitoring effects, which may miss some dangerous accident phenomena, are not smart enough, have high labor costs, and have poor monitoring effects. Summary of the invention

[0003] In view of this, an embodiment of the present invention provides a method for realizing swimming pool and coast safety monitoring based on AI image recognition technology, which is used to solve the problems of poor monitoring effect, lack of intelligence and high labor cost in existing swimming pool and coast safety monitoring methods. The present invention can automatically and intelligently identify the dangerous state in the target water area image and automatically alarm through AI image recognition technology, effectively improving the efficiency of swimming pool and coast monitoring, saving monitoring labor costs, and reducing the occurrence of safety accidents.

[0004] The embodiment of the present invention provides a method for realizing swimming pool and coast safety monitoring based on AI image recognition technology, comprising the following steps:

[0005] Collecting images of the target water area in real time by at least one image acquisition device;

[0006] Use AI image recognition technology to identify whether there is a dangerous state in the image of the target water area;

[0007] If a dangerous state is identified in the image of the target water area, an alarm is issued to lifeguards in the target water area.

[0008] In an optional embodiment, the target water area includes water areas within a designated range of a swimming pool or a coast;

[0009] The method of identifying whether there is a dangerous state in the image of the target water area by using AI image recognition technology includes:

[0010] AI image recognition technology is used to identify whether there are drowning movements of people or dangerous creatures in the image of the target water area. If so, it is determined that a dangerous state exists in the image of the target water area.

[0011] In an optional embodiment, the step of identifying whether there is a person drowning or a dangerous creature in the image of the target water area by using AI image recognition technology includes:

[0012] Identify the movements of people or shapes of organisms in images of target waters through AI image recognition technology;

[0013] Based on a preset human drowning action feature library or a dangerous biological shape feature library, it is determined whether the human action or biological shape in the image of the target water area is consistent with the human drowning action feature library or the dangerous biological shape feature library.

[0014] The features in the feature library match; if so, it is determined that there are drowning actions or dangerous creatures in the image of the target water area.

[0015] In an optional embodiment, if a dangerous state is identified in the image of the target water area, an alarm is issued to a lifeguard in the target water area, including:

[0016] If a dangerous state is identified in the image of the target water area, the number of people in the surrounding area of ​​each currently identified dangerous state location and the distance value between the dangerous state location 5 and the safe zone at the edge of the target water area are identified through AI image recognition technology; wherein the surrounding area of ​​the dangerous state location is the area corresponding to a circle drawn with the dangerous state location as the center and a preset distance as the radius; the degree of danger of each dangerous state is determined according to the number of people in the surrounding area of ​​each dangerous state location and the distance value between the dangerous state location and the safe zone at the edge of the target water area;

[0017] According to the danger level of each dangerous state and the number of lifeguards in the current target waters, determine the priority rescue location among the locations of several dangerous states identified before;

[0018] Alert lifeguards at the target water area including priority rescue locations.

[0019] In an optional embodiment, the step of determining the danger level of each dangerous state according to the number of people in the surrounding area of ​​each dangerous state and the distance between the dangerous state and the safety zone at the edge of the target water area includes: 5 numbering the currently identified dangerous states;

[0020] Calculate the danger level value of each currently identified dangerous state according to the first formula;

[0021] Among them, the first formula is:

[0022]

[0023] In the first formula, G(a) represents the danger level value of the currently identified a-th dangerous state; W(a) represents the state type value of the currently identified a-th dangerous state. If the currently identified a-th dangerous state corresponds to a human drowning action, the state type value of the a-th dangerous state is 0; if the currently identified a-th dangerous state corresponds to a dangerous creature, the state type value of the a-th dangerous state is 1; D(a) represents the number of people in the surrounding area of ​​the currently identified a-th dangerous state; L(a) represents the distance value between the currently identified a-th dangerous state and the safe zone at the edge of the target water area; L0 represents a preset unit distance value; e represents a natural constant.

[0024] The step of determining the priority rescue position among the positions of the currently identified multiple dangerous states according to the degree of danger of each dangerous state and the number of lifeguards in the current target waters includes:

[0025] According to the second formula, a priority rescue position is selected from the positions of the currently identified multiple dangerous states;

[0026] The second formula is:

[0027]

[0028] In the second formula, A represents the number array of the priority rescue positions selected from the positions of the currently identified several dangerous states; b represents an integer variable, b=1, 2, ..., N; N represents the total number of positions of the currently identified dangerous states; F[] represents a judgment function, if the formula in the brackets is established, the function value is 1, if the formula in the brackets is not established, the function value is 0; N represents the total number of lifeguards in the current target waters; a is a positive integer; Indicates that the value of a is changed from 1 to N to satisfy All a values ​​of the formula are arranged in descending order to form an array; H{} represents the array integration function. If the array elements in the brackets are M, the array remains the array in the brackets unchanged. If the array elements in the brackets are greater than M, the first M element values ​​of the array in the brackets are taken to form a new array; {1,2,3,…,N} means that the element values ​​in the array are all integers from 1 to N.

[0029] In an optional embodiment, the step of sending an alarm including a priority rescue position to lifeguards in the target water area includes:

[0030] Determine whether the locations of the currently identified dangerous conditions are all priority rescue locations;

[0031] If the locations of the currently identified dangerous states are all priority rescue locations, the number of lifeguards required for each priority rescue location is calculated based on the third formula according to the danger level value of the dangerous state corresponding to the priority rescue location;

[0032] According to the order of the danger degree values ​​of the dangerous state corresponding to the priority rescue position from large to small, the number of lifeguards required for the priority rescue position is allocated to each priority rescue position in turn, and the lifeguards corresponding to each priority rescue position are obtained;

[0033] Sending an alarm including corresponding priority rescue position information to the lifeguard corresponding to each priority rescue position;

[0034] Wherein, the third formula is:

[0035]

[0036] In the third formula, m(a) represents the number of lifeguards required for the priority rescue position corresponding to the ath dangerous state currently identified; Indicates rounding down.

[0037] In an optional embodiment, after determining whether the locations of the currently identified dangerous states are all priority rescue locations, the method further includes:

[0038] If the locations of the currently identified dangerous conditions are not all priority rescue locations, then one lifeguard is assigned to each priority rescue location in order from large to small according to the danger level values ​​of the dangerous conditions corresponding to the priority rescue locations, and the lifeguard corresponding to each priority rescue location is obtained, and then the step of sending an alarm including the corresponding priority rescue location information to the lifeguard corresponding to each priority rescue location is executed.

[0039] In an optional embodiment, according to the order of the danger level values ​​of the dangerous states corresponding to the priority rescue positions from large to small, the number of lifeguards required for the priority rescue position is sequentially allocated to each priority rescue position, or one lifeguard is sequentially allocated to each priority rescue position, and the lifeguard corresponding to each priority rescue position is obtained, including:

[0040] According to the danger level values ​​of the dangerous states corresponding to the priority rescue positions, all priority rescue positions are sorted in descending order to obtain a priority rescue position sequence;

[0041] Determine the distance between each lifeguard who is not currently assigned a rescue task and the i-th priority rescue position in the priority rescue position sequence; wherein the initial value of i is 1;

[0042] Assign j lifeguards whose distance to the i-th priority rescue position is closest among the lifeguards who are not currently assigned a rescue task to the i-th priority rescue position, and obtain the lifeguard corresponding to the i-th priority rescue position; wherein the value of j is the number of lifeguards required for the i-th priority rescue position or 1;

[0043] Mark the lifeguard currently assigned to the i-th priority rescue position as the lifeguard who has been assigned a rescue mission;

[0044] Determine whether i is equal to I; wherein I is the total number of priority rescue positions in the priority rescue position sequence described in I;

[0045] If i is not equal to I, then set i=i+1, and return to execute the step of determining the distance between each lifeguard who is not currently assigned a rescue task and the i-th priority rescue position in the priority rescue position sequence.

[0046] The present invention provides a method for realizing swimming pool and coast safety monitoring based on AI image recognition technology. First, an image of a target water area is collected in real time by an image acquisition device. Then, the AI ​​image recognition technology is used to automatically and intelligently identify whether there is a dangerous state in the image. Finally, a timely alarm is issued when a dangerous state exists. The present invention can automatically, intelligently and in real time identify the dangerous state in the target water area image through AI image recognition technology and issue an automatic alarm, which effectively improves the efficiency and accuracy of swimming pool and coast monitoring, saves monitoring manpower costs, and reduces the occurrence of safety accidents. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0048] Figure 1 A flow chart of a method for implementing swimming pool and coast safety monitoring based on AI image recognition technology provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0049] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0050] It should be clear that the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0051] Figure 1 A flow chart of a method for realizing swimming pool and coast safety monitoring based on AI image recognition technology provided by an embodiment of the present invention. Figure 1 The method comprises the following steps S101-S103:

[0052] S101: Capturing images of a target water area in real time by at least one image acquisition device.

[0053] S102: Use AI image recognition technology to identify whether there is a dangerous state in the image of the target water area; if so, execute step S103.

[0054] S103: Alert lifeguards in the target waters.

[0055] The beneficial effects of the above technical solution are as follows: the embodiment of the present invention provides a method for realizing swimming pool and coast safety monitoring based on AI image recognition technology, firstly, the image of the target water area is collected in real time by an image acquisition device, and then the AI ​​image recognition technology is used to automatically and intelligently identify whether there is a dangerous state in the image, and finally, when a dangerous state exists, a timely alarm is issued. The present invention can automatically, intelligently and in real time identify the dangerous state in the target water area image through AI image recognition technology and issue an automatic alarm, which effectively improves the efficiency and accuracy of swimming pool and coast monitoring, saves monitoring labor costs, and reduces the occurrence of safety accidents.

[0056] As an optional embodiment, the target water area includes a water area within a designated range of a swimming pool or a coast;

[0057] Wherein, the step S102 may include the following steps S201-S202:

[0058] S201: Identify whether there is a person drowning or a dangerous creature in the image of the target water area by using AI image recognition technology. If so, execute S202;

[0059] S202: Determine whether a dangerous state exists in the image in which the target water area is identified.

[0060] The beneficial effects of the above technical solution are: people often drown or dangerous creatures are found in the waters near swimming pools or coasts, which will endanger the lives of people in and around the waters. Therefore, identifying these dangerous situations will effectively reduce the impact of accidents on people.

[0061] As an optional embodiment, the step S201 may include the following steps S301-S303:

[0062] S301: Identify the human action or biological shape in the image of the target water area through AI image recognition technology;

[0063] S302: Based on a preset human drowning action feature library or a dangerous creature shape feature library, determine whether the human action or creature shape in the image of the target water area matches the features in the human drowning action feature library or the dangerous creature shape feature library; if yes, execute step S303;

[0064] S303: Determine whether there is a person drowning or a dangerous creature in the image of the target water area.

[0065] The beneficial effects of the above technical solution are: a large amount of drowning action information of people and dangerous creatures can be collected in advance, and then through big data analysis technology, a feature library of drowning actions of people and a feature library of shapes of dangerous creatures can be obtained, and then based on such feature libraries, the drowning actions of people or dangerous creatures existing in the images of the target waters can be determined, thereby ensuring the accuracy of the identification of drowning actions of people and dangerous creatures.

[0066] As an optional embodiment, the step S103 may include the following steps S401-S404:

[0067] S401: Identify the number of people in the surrounding area of ​​each currently identified dangerous state location and the distance value between the dangerous state location and the safe zone at the edge of the target water area through AI image recognition technology; wherein the surrounding area of ​​the dangerous state location is the area corresponding to a circle drawn with the dangerous state location as the center and a preset distance as the radius;

[0068] S402: Determine the danger level of each dangerous state according to the number of people in the surrounding area of ​​each dangerous state location and the distance between the dangerous state location and the safety zone at the edge of the target water area;

[0069] S403: determining a priority rescue position among the positions of the currently identified multiple dangerous states according to the danger level of each dangerous state and the number of lifeguards in the current target waters;

[0070] S404: Send an alarm including a priority rescue position to lifeguards in the target water area.

[0071] As an optional embodiment, the danger level of each dangerous state is determined based on the number of people in the surrounding area of ​​the location of each dangerous state and the distance between the location of the dangerous state and the safe zone at the edge of the target water area, which further facilitates the determination of the priority rescue location in each dangerous state and the issuance of an alarm, thereby facilitating subsequent rational rescue and improving the efficiency of rescue.

[0072] As an optional embodiment, the step S402 may include the following steps S501-S502:

[0073] S501: Numbering several currently identified dangerous states;

[0074] S502: Calculating the danger level value of each currently identified dangerous state according to the first formula;

[0075] Among them, the first formula is:

[0076]

[0077] In the first formula, G(a) represents the danger level value of the currently identified a-th dangerous state; W(a) represents the state type value of the currently identified a-th dangerous state. If the currently identified a-th dangerous state corresponds to a human drowning action, the state type value of the a-th dangerous state is 0; if the currently identified a-th dangerous state corresponds to a dangerous creature, the state type value of the a-th dangerous state is 1; D(a) represents the number of people in the surrounding area of ​​the currently identified a-th dangerous state; L(a) represents the distance value between the currently identified a-th dangerous state and the safe zone at the edge of the target water area; L0 represents a preset unit distance value; e represents a natural constant.

[0078] In this embodiment, step S403 may include: selecting a priority rescue location from the locations of the currently identified multiple dangerous states according to the second formula;

[0079] Among them, the second formula is:

[0080]

[0081] In the second formula, A represents the number array of the priority rescue positions selected from the positions of the currently identified several dangerous states; b represents an integer variable, b=1, 2, ..., N; N represents the total number of positions of the currently identified dangerous states; F[] represents a judgment function, if the formula in the brackets is established, the function value is 1, if the formula in the brackets is not established, the function value is 0; M represents the total number of lifeguards in the current target waters; a is a positive integer; Indicates that the value of a is changed from 1 to N to satisfy All a values ​​of the formula are arranged in descending order to form an array; H{} represents the array integration function. If the array elements in the brackets are M, the array remains the array in the brackets unchanged. If the array elements in the brackets are greater than M, the first M element values ​​of the array in the brackets are taken to form a new array; {1,2,3,…,N} means that the element values ​​in the array are all integers from 1 to N.

[0082] The beneficial effects of the above technical solution are as follows: the first formula (1) is used to obtain the danger level value of each identified dangerous state through intelligent analysis according to the number of people around each identified dangerous state position and the distance between each identified dangerous state position and the safe zone at the edge of the target water area. Further, the corresponding rescue priority order can be obtained according to the analysis of the danger level value, so as to facilitate subsequent rational rescue; then the second formula (2) is used to screen the priority rescue position from each identified dangerous state position according to the current number of lifeguards and the danger level value of each identified dangerous state. If there are many dangerous state positions, the lifeguards with a closer distance are assigned to the positions with larger danger level values ​​for priority rescue, thereby achieving reasonable and intelligent rescue, the effect of reasonable and intelligent allocation of lifeguards, and improving the rescue efficiency of lifeguards.

[0083] As an optional embodiment, step S404 may include the following steps S601-S604:

[0084] S601: Determine whether the locations of the currently identified dangerous conditions are all priority rescue locations; if so, execute step S602.

[0085] Preferably, if the judgment result of step S601 is no, that is, the locations of the currently identified dangerous conditions are not all priority rescue locations, then S605 is executed.

[0086] S602: Calculating the number of lifeguards required for each priority rescue position based on a third formula according to the danger level value of the dangerous state corresponding to the priority rescue position;

[0087] Among them, the third formula is:

[0088]

[0089] In the third formula (3), m(a) represents the number of lifeguards required for the priority rescue position corresponding to the ath dangerous state currently identified; Indicates rounding down.

[0090] S603: Allocate the number of lifeguards required for each priority rescue position in turn according to the danger level values ​​of the dangerous states corresponding to the priority rescue positions in descending order, and obtain the lifeguards corresponding to each priority rescue position.

[0091] S604: Send an alarm including the corresponding priority rescue position information to the lifeguard corresponding to each priority rescue position.

[0092] S605: Allocate one lifeguard to each priority rescue position in order from large to small according to the danger level value of the dangerous state corresponding to the priority rescue position, obtain the lifeguard corresponding to each priority rescue position, and then return to execute step S604.

[0093] The beneficial effect of the above technical solution is: using the third formula (3) to allocate the number of personnel to each identified dangerous state position according to the danger level value of each identified dangerous state position, when the number of lifeguards is large, multiple lifeguards with correspondingly close distances are allocated for rescue according to the danger level value, thereby improving the reliability of rescue.

[0094] As an optional embodiment, the method of allocating lifeguards to priority rescue positions in S603 or S605 may include the following steps S701-S706:

[0095] S701: sorting all priority rescue positions in descending order of the danger level values ​​of the dangerous states corresponding to the priority rescue positions to obtain a priority rescue position sequence;

[0096] S702: Determine the distance between each lifeguard who is not currently assigned a rescue task and the i-th priority rescue position in the priority rescue position sequence; wherein the initial value of i is 1;

[0097] S703: assigning j lifeguards whose distance to the i-th priority rescue position is closest among the lifeguards who are not currently assigned a rescue task to the i-th priority rescue position, and obtaining the lifeguard corresponding to the i-th priority rescue position; wherein the value of j is the number of lifeguards required for the i-th priority rescue position or 1;

[0098] In this embodiment, if the positions of the identified dangerous states are all priority rescue positions, then the value of j here is the number of lifeguards required for the i-th priority rescue position, which is calculated by the third formula (3) above. If the positions of the identified dangerous states are not all (not all) priority rescue positions, then the value of j here is 1, thereby achieving the purpose of optimizing personnel allocation and effectively saving human resources.

[0099] S704: marking the lifeguard currently assigned to the i-th priority rescue position as a lifeguard who has been assigned a rescue mission;

[0100] S705: Determine whether i is equal to I; wherein i is the total number of priority rescue positions in the priority rescue position sequence described in I; if not, execute S706; if so, end the process of allocating lifeguards to the priority rescue positions;

[0101] S706: Set i=i+1, and return to execute S702.

[0102] The beneficial effect of the above technical solution is: lifeguards are assigned in order of the danger level values ​​of the priority rescue positions from large to small, and the larger the danger level value, the priority is given to the rescue of the closest lifeguard among the currently unassigned lifeguards. After rescuing the corresponding dangerous state, the nearest dangerous state position point outside the priority rescue position will be rescued until all dangerous state position points are rescued, thereby achieving the purpose of preferential allocation.

[0103] In summary, the safety monitoring method provided by the present invention has the following beneficial effects: the first formula (1) is used to obtain the danger level value of each dangerous state location according to the number of people around each identified dangerous state location and the distance value between each dangerous state location and the safe zone at the edge of the target water area, so as to obtain the danger level of each dangerous state location through intelligent analysis, and then the corresponding rescue priority order can be obtained according to the danger level value analysis, so as to facilitate subsequent rational rescue; then the second formula (2) is used to select the priority rescue position from all dangerous state locations according to the current number of lifeguards and the danger level value of each dangerous state, and if there are many dangerous state locations, the dangerous state locations with relatively large danger level values ​​are preferentially assigned to the lifeguards with a closer distance for priority rescue, so as to achieve the effect of reasonable intelligent rescue and reasonable intelligent allocation of lifeguards, and improve the rescue efficiency of lifeguards; finally, the third formula (3) is used to allocate the number of lifeguards to each dangerous state location according to the danger level value of each dangerous state, and when the number of lifeguards is large, multiple lifeguards with a relatively close distance are assigned for rescue according to the danger level value, so as to improve the reliability of rescue.

[0104] It can be seen from the contents of the above embodiments that the intelligent recognition algorithm is used to identify possible drowning situations and dangerous creatures in the water, and to send an alarm to the lifeguard, marking the location where the drowning may occur, thereby improving the efficiency of monitoring and rescue; specifically: when the possible drowning situation and dangerous creatures in the water are identified, it will be recorded as a dangerous state. If the dangerous state is identified at multiple locations in the water at the same time, the danger level value of each dangerous state position is first obtained based on the number of people around each dangerous state position and the distance of each dangerous state position from the coast, and then the dangerous state value is calculated based on the current number of lifeguards.

[0105] Personnel are assigned for each identified dangerous state, and the best allocation is made based on the distance between each lifeguard and each identified dangerous state position, which effectively improves the efficiency of rescue.

[0106] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that the flowcharts and / or block diagrams may be implemented by computer program instructions.

[0107] Each process and / or block in the flowchart and / or block diagram, and the process and / or

[0108] These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the process Figure 1 A process or multiple processes and / or boxes Figure 1 An apparatus for the method specified in one or more blocks.

[0109] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 The method specified in one or more boxes.

[0110] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to produce a computer-implemented process.

[0111] The instructions executed on a computer or other programmable device are provided to implement the process Figure 1 5 or more processes and / or boxes Figure 1 The steps of a method specified in one or more boxes.

[0112] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these changes and variations. The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for realizing swimming pool and coast safety monitoring based on AI image recognition technology, characterized in that: The following steps are involved: Collecting images of the target water area in real time by at least one image acquisition device; Use AI image recognition technology to identify whether there is a dangerous state in the image of the target water area; If a dangerous state is identified in the image of the target water area, an alarm is issued to lifeguards in the target water area; If a dangerous state is identified in the image of the target water area, an alarm is issued to a lifeguard in the target water area, including: If a dangerous state is identified in the image of the target water area, the number of people in the surrounding area of ​​each currently identified dangerous state and the distance between the dangerous state and the safe zone at the edge of the target water area are identified through AI image recognition technology; wherein the surrounding area of ​​the dangerous state is the area corresponding to a circle drawn with the dangerous state as the center and a preset distance as the radius; Determine the danger level of each dangerous state according to the number of people in the surrounding area of ​​each dangerous state and the distance between the dangerous state and the safety zone at the edge of the target water area; Determine the priority rescue position among the locations of the currently identified several dangerous states according to the danger level of each dangerous state and the number of lifeguards in the current target waters; Send an alert including the priority rescue location to lifeguards in the target water area; The step of determining the danger level of each dangerous state according to the number of people in the surrounding area of ​​each dangerous state and the distance between the dangerous state and the safety zone at the edge of the target water area includes: Numbering several currently identified dangerous conditions; Calculate the danger level value of each currently identified dangerous state according to the first formula; Among them, the first formula is: In the first formula, G(a) represents the danger level value of the ath dangerous state currently identified; W(a) represents the state type value of the ath dangerous state currently identified. If the ath dangerous state currently identified corresponds to a drowning action of a person, the state type value of the ath dangerous state is 0; if the ath dangerous state currently identified corresponds to a dangerous creature, the state type value of the ath dangerous state is 1; D(a) represents the number of people in the surrounding area of ​​the ath dangerous state currently identified; L(a) represents the distance value between the ath dangerous state currently identified and the safe zone at the edge of the target water area; L0 represents a preset unit distance value; e represents a natural constant; The step of determining the priority rescue position among the positions of the currently identified multiple dangerous states according to the degree of danger of each dangerous state and the number of lifeguards in the current target waters includes: According to the second formula, a priority rescue position is selected from the positions of the currently identified multiple dangerous states; The second formula is: In the second formula, A represents the number array of the priority rescue positions selected from the positions of the currently identified several dangerous states; b represents an integer variable, b=1, 2, ..., N; N represents the total number of positions of the currently identified dangerous states; F[] represents a judgment function, if the formula in the brackets is established, the function value is 1, if the formula in the brackets is not established, the function value is 0; M represents the total number of lifeguards in the current target waters; a is a positive integer; Indicates that the value of a is changed from 1 to N to satisfy All a values ​​of the formula are arranged in descending order to form an array; H{} represents the array integration function. If the array elements in the brackets are M, the array remains the array in the brackets unchanged. If the array elements in the brackets are greater than M, the first M element values ​​of the array in the brackets are taken to form a new array; {1,2,3,…,N} means that the element values ​​in the array are all integers from 1 to N.

2. The method for realizing swimming pool and coast safety monitoring based on AI image recognition technology as claimed in claim 1, characterized in that: The target water area includes a swimming pool or water area within a designated range near the coast; The method of identifying whether there is a dangerous state in the image of the target water area by using AI image recognition technology includes: AI image recognition technology is used to identify whether there are drowning movements of people or dangerous creatures in the image of the target water area. If so, it is determined that a dangerous state exists in the image of the target water area.

3. The method for realizing swimming pool and coast safety monitoring based on AI image recognition technology as claimed in claim 2, characterized in that: The AI ​​image recognition technology is used to identify whether there is a person drowning or dangerous creatures in the image of the target water area, including: Identify the movements of people or shapes of organisms in images of target waters through AI image recognition technology; Based on a preset human drowning action feature library or a dangerous creature shape feature library, determine whether the human action or creature shape in the image of the target water area matches the features in the human drowning action feature library or the dangerous creature shape feature library; if so, determine that there is a human drowning action or a dangerous creature in the image of the target water area.

4. The method for realizing swimming pool and coast safety monitoring based on AI image recognition technology as claimed in claim 1, wherein the step of sending an alarm including a priority rescue position to lifeguards in the target water area comprises: Determine whether the locations of the currently identified dangerous conditions are all priority rescue locations; If the locations of the currently identified dangerous states are all priority rescue locations, the number of lifeguards required for each priority rescue location is calculated based on the third formula according to the danger level value of the dangerous state corresponding to the priority rescue location; According to the order of the danger degree values ​​of the dangerous state corresponding to the priority rescue position from large to small, the number of lifeguards required for the priority rescue position is allocated to each priority rescue position in turn, and the lifeguards corresponding to each priority rescue position are obtained; Sending an alarm including corresponding priority rescue position information to the lifeguard corresponding to each priority rescue position; Wherein, the third formula is: In the third formula, m(a) represents the number of lifeguards required for the priority rescue position corresponding to the ath dangerous state currently identified; Indicates rounding down.

5. The method for realizing swimming pool and coast safety monitoring based on AI image recognition technology as claimed in claim 4, characterized in that: After determining whether the locations of the currently identified dangerous states are all priority rescue locations, the method further includes: If the locations of the currently identified dangerous conditions are not all priority rescue locations, then one lifeguard is assigned to each priority rescue location in order from large to small according to the danger level values ​​of the dangerous conditions corresponding to the priority rescue locations, and the lifeguard corresponding to each priority rescue location is obtained, and then the step of sending an alarm including the corresponding priority rescue location information to the lifeguard corresponding to each priority rescue location is executed.

6. The method for realizing swimming pool and coast safety monitoring based on AI image recognition technology as claimed in claim 5, characterized in that: According to the order of the danger degree values ​​of the dangerous states corresponding to the priority rescue positions from large to small, the number of lifeguards required for the priority rescue positions is allocated to each priority rescue position in turn, or one lifeguard is allocated to each priority rescue position in turn, and the lifeguards corresponding to each priority rescue position are obtained, including: According to the danger level values ​​of the danger states corresponding to the priority rescue positions, all priority rescue positions are sorted in descending order to obtain a priority rescue position sequence; Determine the distance between each lifeguard who is not currently assigned a rescue task and the i-th priority rescue position in the priority rescue position sequence; wherein the initial value of i is 1; Assign j lifeguards whose distance to the i-th priority rescue position is closest among the lifeguards who are not currently assigned a rescue task to the i-th priority rescue position, and obtain the lifeguard corresponding to the i-th priority rescue position; wherein the value of j is the number of lifeguards required for the i-th priority rescue position or 1; Mark the lifeguard currently assigned to the i-th priority rescue position as the lifeguard who has been assigned a rescue mission; Determine whether i is equal to I; wherein I is the total number of priority rescue positions in the priority rescue position sequence described in I; If i is not equal to I, then set i=i+1, and return to execute the step of determining the distance between each lifeguard who is not currently assigned a rescue task and the i-th priority rescue position in the priority rescue position sequence.

Images