An artificial intelligence-based drowning prevention and early warning rescue method and system

Abstract

The present application relates to a kind of artificial intelligence-based anti-drowning early warning rescue method and system, the method includes: water area is divided into multiple sub-regions, automatically upgrade column, sound wave sensor and underwater shooting device are fixedly installed in each sub-region, and net-shaped rescue device located above automatically upgrade column and equipped with water level detector is laid in water area;Real-time receive the sound wave data collected by each sound wave sensor and the human face image collected by underwater shooting device, and then judge whether there is a target in the corresponding sub-region in fixed position stays time exceeds time threshold, if so, the early warning signal of corresponding sub-region is generated;Based on the sub-region included in early warning signal, the automatically upgrade column corresponding to the sub-region is controlled to rise, so that the net-shaped rescue device laid above automatically upgrade column rises.The present application greatly improves the accuracy and timeliness of early warning, and can automatically take rescue measures, greatly improves the safety of drowning personnel.

Description

Technical Field

[0001] The present invention relates to the field of safety monitoring, and in particular to an artificial intelligence-based drowning prevention, early warning and rescue method and system. Background Art

[0002] With the increasing frequency of drowning incidents, water safety is a growing concern. Current water safety systems rely heavily on manual patrols, which can lead to accidents when patrols are inadequate. Furthermore, rescue efforts are slow when drowning in the deep sea, making it easy to miss critical rescue opportunities. Summary of the Invention

[0003] In order to solve the above problems, the present invention proposes an artificial intelligence-based drowning prevention, early warning and rescue method and system.

[0004] The specific plan is as follows:

[0005] An artificial intelligence-based drowning prevention and early warning rescue method comprises the following steps:

[0006] The water area is divided into multiple sub-areas. Automatic upgrade columns are fixedly installed in each sub-area. A net-shaped rescue device equipped with a water level detector is deployed above the automatic upgrade columns in the water area. At the same time, an acoustic sensor for detecting underwater targets and an underwater camera for capturing underwater facial images are deployed in each sub-area.

[0007] Receive water level data collected by the water level detector in real time, and control the net rescue device to move toward the deep sea when the water level data is less than the low water level threshold; control the net rescue device to move toward the beach when the water level data is greater than the high water level threshold;

[0008] Receive the acoustic wave data collected by each acoustic wave sensor in real time, and determine whether there is a target in the corresponding sub-area that has stayed in a fixed position for longer than a time threshold based on the received acoustic wave data. If so, generate an early warning signal for the corresponding sub-area;

[0009] Receive facial images captured by underwater cameras in real time and track their trajectory. When a trajectory stays at a fixed location for longer than a threshold, an early warning signal is generated for the corresponding sub-area.

[0010] Based on the sub-area included in the early warning signal, the automatic upgrading column corresponding to the sub-area is controlled to rise, so that the net rescue device arranged above the automatic upgrading column rises.

[0011] Furthermore, the conditions for controlling the movement of the net rescue device toward the deep sea include not only the water level data being less than the low water level threshold, but also the current tide state being low tide; the conditions for controlling the movement of the net rescue device toward the beach include not only the water level data being greater than the high water level threshold, but also the current tide state being high tide.

[0012] Furthermore, the net rescue device adopts a mesh topology nylon rope; the acoustic wave sensor adopts a sonic sodium detector; and the underwater shooting device adopts a thermal imaging scanner.

[0013] Furthermore, face capture cameras are installed at the entrance of the seaside scenic area, and positioning devices are distributed to each tourist; the face capture cameras installed at the entrance of the scenic area collect facial images of each tourist entering the scenic area, and a correspondence between the facial images and the positioning devices is established; the facial emotion attributes corresponding to each facial image are analyzed based on the collected facial images, and facial images with negative facial emotion attributes are marked as key focus objects; the real-time location information of the key focus objects is obtained based on the correspondence between the facial images and the positioning devices, and when the time it stays at a fixed position in a sub-area of ​​the water area exceeds a time threshold, an early warning signal for the corresponding sub-area is generated.

[0014] Furthermore, in the determination of the focus object, in addition to the facial emotion attribute being of the negative type, it also includes: determining whether there is a facial image whose distance from the negative type facial image is less than a distance threshold for a period of time. If so, the negative type facial image is determined as a non-focus observation object.

[0015] Furthermore, the positioning device adopts a positioning bracelet.

[0016] Furthermore, the positioning bracelet is configured to issue an early warning when the current position obtained by the positioning bracelet exceeds a safe range.

[0017] Furthermore, the positioning bracelet uses voice warning when issuing warnings.

[0018] An artificial intelligence-based drowning prevention and warning rescue system includes automatically upgraded columns, acoustic sensors, and underwater photography devices fixedly installed in each sub-area of ​​a water area, a mesh rescue device and a control terminal arranged above the automatically upgraded columns and equipped with a water level detector; the system implements the above-mentioned method of the embodiment of the present invention.

[0019] Furthermore, the system also includes automatic upgrading columns, acoustic wave sensors and underwater shooting devices fixedly installed in each sub-area of ​​the water area, and a mesh rescue device and a control terminal arranged above the automatic upgrading columns and equipped with a water level detector; the system implements the above-mentioned method of the embodiment of the present invention.

[0020] The present invention adopts the above technical solution, which can provide timely warning and fast rescue processing, greatly improving the safety of drowning people. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 Shown is a flow chart of a method according to a first embodiment of the present invention. DETAILED DESCRIPTION

[0022] To further illustrate various embodiments, the present invention provides accompanying drawings. These drawings form part of the present disclosure and are primarily used to illustrate the embodiments and, in conjunction with the relevant description in the specification, to explain the operating principles of the embodiments. By referring to these drawings, those skilled in the art will be able to understand other possible implementations and the advantages of the present invention.

[0023] The present invention will now be further described with reference to the accompanying drawings and specific embodiments.

[0024] Example 1:

[0025] The embodiment of the present invention provides an artificial intelligence-based drowning prevention and early warning rescue method, such as Figure 1 As shown, the method includes the following steps:

[0026] S101: Divide the water area of ​​the seaside scenic area into multiple sub-areas, install an automatic upgrading column in each sub-area, and deploy a net rescue device in the water area, which is located above the automatic upgrading column and equipped with a water level detector; at the same time, deploy an acoustic wave sensor for detecting underwater targets and an underwater shooting device for capturing underwater facial images in each sub-area.

[0027] In this embodiment, the water area is divided into a grid pattern, with at least four automatic upgrade columns installed in each sub-area, forming a rectangular shape. Normally, the automatic upgrade columns are in a non-elevated position. When an early warning signal is generated, the columns are controlled to rise, supporting the net-shaped rescue device above them.

[0028] In this embodiment, the net-shaped rescue device adopts a net-shaped topology nylon rope to facilitate rescue; the acoustic wave sensor adopts a sonic sodium detector; and the underwater shooting device adopts a thermal imaging scanner.

[0029] S102: Receive water level data collected by the water level detector in real time, and control the net rescue device to move toward the deep sea when the water level data is less than the low water level threshold; and control the net rescue device to move toward the beach when the water level data is greater than the high water level threshold.

[0030] Furthermore, while water level fluctuations are typically caused by tidal fluctuations, in special cases, they may be caused by human intervention. To avoid this impact, in this embodiment, the conditions for controlling the movement of the net rescue device toward the deep sea include not only the water level data being less than a low water level threshold but also the current tide being low. The conditions for controlling the movement of the net rescue device toward the beach include not only the water level data being greater than a high water level threshold but also the current tide being high. Those skilled in the art can freely determine the low and high water level thresholds, as well as the movement distance, based on actual applications, and are not limited here.

[0031] S103: Receive acoustic wave data collected by each acoustic wave sensor in real time and, based on the received acoustic wave data, determine whether a target within the corresponding sub-area has remained at a fixed location for a period exceeding a time threshold. If so, generate a warning signal for the corresponding sub-area. Receive facial images collected by the underwater camera in real time and track their trajectory. If a trajectory remains at a fixed location for a period exceeding a time threshold, generate a warning signal for the corresponding sub-area.

[0032] When a target remains at a fixed location for longer than a time threshold, there's a high probability of drowning. Therefore, in this embodiment, a warning message is generated based on this information. By simultaneously analyzing both acoustic and trajectory data, we can avoid missing either one and improve detection accuracy. Those skilled in the art can determine the time threshold based on their own experience or experimental data, and this is not a limitation here.

[0033] S104: Based on the sub-area included in the early warning signal, the automatic upgrade column corresponding to the sub-area is controlled to rise, so that the net rescue device arranged above the automatic upgrade column rises.

[0034] In this embodiment, the water area is divided so that the rescue position can be precisely controlled, which greatly improves the efficiency of the rescue.

[0035] Furthermore, in this embodiment, the following warning methods are added on the basis of the above method:

[0036] S201: Deploy facial capture cameras at the entrance of the scenic area and distribute positioning devices to each tourist.

[0037] S202: The facial images of tourists entering the scenic area are collected by facial capture cameras installed at the entrance of the scenic area, and a correspondence between the facial images and the positioning devices is established.

[0038] S203: Analyze the facial emotion attributes corresponding to each facial image based on the collected facial images, and mark facial images with negative facial emotion attributes as key attention objects.

[0039] S204: Based on the correspondence between the facial image and the positioning device, the real-time location information of the focus object is obtained, and when the time it stays at a fixed position in a sub-area of ​​the water area exceeds a time threshold, an early warning signal for the corresponding sub-area is generated.

[0040] In this embodiment, the positioning device uses a positioning wristband. To facilitate alerting visitors, this embodiment also includes: when the positioning wristband determines the visitor's current location exceeds the safe range, the positioning wristband issues an early warning. The warning can be voice-activated, allowing visitors to receive timely warning information. The safe range needs to be pre-configured in the positioning wristband.

[0041] The analysis of facial emotion attributes can be obtained through a pre-trained existing machine learning model. The machine learning model outputs the facial emotion attributes or the type of facial emotion attributes (negative, positive, and normal) to achieve the recognition of the focus object. In this embodiment, the negative type of facial emotion attributes are set to include sadness, sorrow, etc.

[0042] Through the above-mentioned emotional judgment, we can focus on key personnel and improve the timeliness of rescue.

[0043] Furthermore, in order to improve the accuracy of the above-mentioned key personnel determination, this embodiment also includes eliminating the influence of accompanying persons in the following manner, specifically: determining whether there is a facial image whose distance from the negative type facial image is less than a distance threshold for a period of time. If so, the negative type facial image is determined as a non-key observation object.

[0044] The embodiments of the present invention greatly improve the accuracy and timeliness of early warnings, and can automatically take rescue measures, thereby greatly improving the safety of drowning people.

[0045] Example 2:

[0046] The present invention also provides an artificial intelligence-based drowning prevention and early warning rescue system, comprising: automatically upgraded columns, acoustic sensors, and underwater camera devices fixedly installed in each sub-area of ​​the water; a network rescue device equipped with a water level detector and positioned above the automatically upgraded columns; a control terminal; facial capture cameras deployed at the entrance to a seaside scenic area; and a positioning device distributed to each visitor. The control terminal, which can be a PC, implements steps S102-S104 and S202-S204 of the first embodiment of the present invention.

[0047] Although the present invention has been particularly shown and described in conjunction with preferred embodiments, it will be understood by those skilled in the art that various changes in form and details may be made to the present invention without departing from the spirit and scope of the invention as defined in the appended claims, and all such changes are within the scope of protection of the present invention.

Claims

1. An artificial intelligence-based drowning prevention and early warning rescue method, characterized in that: The following steps are involved: The water area is divided into multiple sub-areas. Automatic upgrade columns are fixedly installed in each sub-area. A net-shaped rescue device equipped with a water level detector is deployed above the automatic upgrade columns in the water area. At the same time, an acoustic sensor for detecting underwater targets and an underwater camera for capturing underwater facial images are deployed in each sub-area. Receive water level data collected by the water level detector in real time, and control the net rescue device to move toward the deep sea when the water level data is less than the low water level threshold; control the net rescue device to move toward the beach when the water level data is greater than the high water level threshold; Receive the acoustic wave data collected by each acoustic wave sensor in real time, and determine whether there is a target in the corresponding sub-area that has stayed in a fixed position for longer than a time threshold based on the received acoustic wave data. If so, generate an early warning signal for the corresponding sub-area; Receive facial images captured by underwater cameras in real time and track their trajectory. When a trajectory stays at a fixed location for longer than a threshold, an early warning signal is generated for the corresponding sub-area. Based on the sub-area included in the early warning signal, the automatic upgrade column corresponding to the sub-area is controlled to rise, so that the net rescue device arranged above the automatic upgrade column rises; The method further comprises: A facial capture camera is set up at the entrance of a seaside scenic area, and a positioning device is distributed to each tourist. The facial capture camera set up at the entrance of the scenic area collects facial images of each tourist entering the scenic area, and a correspondence between the facial images and the positioning device is established. The facial emotion attributes corresponding to each facial image are analyzed based on the collected facial images, and facial images with negative facial emotion attributes are marked as key focus objects. The real-time location information of the key focus objects is obtained based on the correspondence between the facial images and the positioning device. When the time it stays at a fixed position in a sub-area of ​​the water area exceeds a time threshold, an early warning signal for the corresponding sub-area is generated.

2. The artificial intelligence-based drowning prevention and early warning rescue method according to claim 1, characterized in that: The conditions for controlling the movement of the net rescue device toward the deep sea include not only the water level data being less than the low water level threshold, but also the current tide state being low tide; the conditions for controlling the movement of the net rescue device toward the beach include not only the water level data being greater than the high water level threshold, but also the current tide state being high tide.

3. The artificial intelligence-based drowning prevention and early warning rescue method according to claim 1, characterized in that: The net rescue device uses a mesh topology nylon rope; the acoustic wave sensor uses a sonic sodium detector; and the underwater shooting device uses a thermal imaging scanner.

4. The artificial intelligence-based drowning prevention and early warning rescue method according to claim 1, characterized in that: In the determination of the focus object, in addition to the facial emotion attribute being of the negative type, it also includes: judging whether there is a face image whose distance from the negative type face image is less than a distance threshold for a period of time. If so, the negative type face image is determined as a non-focus observation object.

5. The artificial intelligence-based drowning prevention and early warning rescue method according to claim 1, characterized in that: The positioning device uses a positioning bracelet.

6. The artificial intelligence-based drowning prevention and early warning rescue method according to claim 5, characterized in that: The positioning bracelet is configured to issue an early warning when the current position obtained by the positioning bracelet exceeds a safe range.

7. The artificial intelligence-based drowning prevention and early warning rescue method according to claim 6, characterized in that: The positioning bracelet uses voice warning when issuing warnings.

8. An artificial intelligence-based drowning prevention and early warning rescue system, characterized by: The system comprises automatic upgrading columns, acoustic wave sensors and underwater shooting devices fixedly installed in each sub-area of ​​the water area, a mesh rescue device and a control terminal arranged above the automatic upgrading columns and equipped with a water level detector; a face capture camera arranged at the entrance of the seaside scenic area and a positioning device distributed to each tourist; the system implements the method as described in any one of claims 1 to 7.

Images