A swimming drowning monitoring method, system and computer storage medium

Abstract

The present invention provides a method, system and computer storage medium for monitoring drowning in swimming. The intelligent wearable device is used to collect physiological characteristic data and motion characteristic data of each swimmer, including: marking a swimmer whose physiological characteristic data is in an abnormal state as a target drowning person; determining the supportable object closest to the target drowning person; determining the first moving direction of the target drowning person according to the motion characteristic data of the target drowning person; when the first moving direction of the target drowning person is toward the supportable object, and the physiological characteristic data of the target drowning person returns to a normal state and then changes to an abnormal state, it is determined that the target drowning person is in a drowning state. Whether the swimmer is in an abnormal state can be directly determined by the physiological characteristic data and the motion characteristic data, and combined with the changes in the physiological characteristic data of the target drowning person asking for help from the nearby supportable object, it is further determined that the target drowning person is truly in a drowning state.

Description

Technical Field

[0001] The present invention relates to the field of drowning prevention, and in particular to a swimming drowning monitoring method, system and computer storage medium. Background Art

[0002] Swimming is a very popular form of exercise and leisure, but as a water sport, it carries a high risk of drowning. Especially for beginners and swimmers with poor physical fitness, swimmers are prone to drowning due to incorrect swimming posture, fatigue, illness or psychological fear.

[0003] At present, most swimming pools are equipped with lifeguards, but due to the limited number of lifeguards, it is difficult to monitor the swimming status of each swimmer at all times, and lifeguards are also prone to fatigue after working for a long time and cannot rescue drowning swimmers in time. In addition, some swimming pools are equipped with anti-drowning systems, which install underwater cameras around and at the bottom of the pool to capture and determine whether the swimmer's movements are in a state of drowning struggle, and then provide early warnings. However, when swimmers are swimming, due to the large amount of water splashes in the pool or the reflection of the water, the captured images are unclear or distorted, and timely and effective supervision cannot be provided. Summary of the invention

[0004] In view of this, an object of the present invention is to provide a swimming drowning monitoring method, system and computer storage medium, which solve the problem in the prior art that timely and effective supervision effects cannot be provided to drowning swimmers.

[0005] In a first aspect, the present application provides a method for monitoring drowning in swimming, wherein each swimmer wears a smart wearable device, and the smart wearable device is used to collect physiological characteristic data and motion characteristic data of each swimmer, and the method comprises:

[0006] Marking the swimmer whose physiological characteristic data is in an abnormal state as a target drowning person;

[0007] Taking the target drowning person as the center, determining the supportable object closest to the target drowning person; the supportable object includes other swimmers or auxiliary rescue facilities;

[0008] Determining a first moving direction of the target drowning person according to the motion characteristic data of the target drowning person;

[0009] When the first moving direction of the target drowning person is toward the supportable object, and the physiological characteristic data of the target drowning person returns to a normal state and then changes to an abnormal state, it is determined that the target drowning person is in a drowning state.

[0010] In one embodiment, the physiological characteristic data includes heart rate data, and the heart rate data includes dynamic heart rate and static heart rate; marking a swimmer whose physiological characteristic data is in an abnormal state as a target drowning person specifically includes:

[0011] When the dynamic heart rate of a swimmer exceeds the maximum heart rate range, the swimmer is marked as a target drowning person.

[0012] In one embodiment, before marking the swimmer as a target drowning person, the method further includes:

[0013] Obtaining a static heart rate of each of the swimmers;

[0014] A maximum heart rate range for each of the swimmers is determined based on the static heart rate.

[0015] In one embodiment, before marking the swimmer as a target drowning person, the method further includes:

[0016] determining the movement pattern and movement intensity of each of said swimmers;

[0017] Obtaining historical dynamic heart rate of each swimmer's historical swimming data corresponding to the exercise mode and the exercise intensity;

[0018] A maximum heart rate range of the swimmer is determined based on the historical dynamic heart rate.

[0019] In one embodiment, the physiological characteristic data further includes blood pressure data and blood oxygen data, the blood pressure data includes dynamic blood pressure, and the blood oxygen data includes dynamic blood oxygen; after marking the swimmer as a target drowning person, the following is further included:

[0020] Obtaining the dynamic blood pressure, dynamic blood oxygen and dynamic heart rate of the target drowning person within a preset time range before the time point of the abnormal state;

[0021] When the dynamic heart rate at a certain time node is lower than the preset minimum heart rate, the time node is determined as the initial diving time node;

[0022] Between the initial diving time node and the time node of the abnormal state, if the dynamic blood pressure does not show a trend of first increasing and then stabilizing or decreasing, and the dynamic blood oxygen does not show a trend of first increasing or decreasing, the swimmer is marked from a target drowning person to be determined.

[0023] In one embodiment, after marking the swimmer from a target drowning person to be determined, the method further includes:

[0024] Acquire image information of a plurality of drowning persons to be determined as targets;

[0025] By combining upsampling with downsampling, the plurality of image information are fused to obtain a fused image;

[0026] Inputting the fused image into a preset depth recognition model to identify whether the target drowning person to be determined is in drowning behavior;

[0027] When the target drowning person to be determined does not have any drowning behavior, the mark of the target drowning person to be determined is cancelled.

[0028] In one embodiment, the fusing of the plurality of image information to obtain a fused image by combining upsampling with downsampling specifically includes:

[0029] Performing a convolution operation on each of the image information to reduce the size of the image information to obtain a scaled image;

[0030] The zoomed image is placed at the center of a preset enlarged template image, and adjacent positions of corresponding pixel values ​​in the zoomed image are assigned values ​​according to pixels of the image information to obtain restored image information;

[0031] Each of the restored image information is superimposed according to the number of channels and the spatial dimension, so as to obtain the fused image.

[0032] In one embodiment, after the first moving direction of the target drowning person is toward the supportable object, the method further includes:

[0033] When the target drowning person is located within the preset rescueable range of the supportable object, obtaining the second moving direction of the target drowning person again;

[0034] If the second moving direction is away from the supportable object, it is determined that the target drowning person is not in a drowning state.

[0035] In a second aspect, the present application provides a swimming drowning monitoring system, comprising a processor and a memory; wherein the memory stores a computer program, and the computer program is used to be loaded by the processor and execute the swimming drowning monitoring method as described in any one of the first aspects.

[0036] In a third aspect, the present application provides a computer-readable storage medium, characterized in that the computer-readable storage medium stores instructions, and the instructions are used by a processor to load and execute the swimming drowning monitoring method as described in any one of the first aspects.

[0037] In the swimming drowning monitoring method, system and computer storage medium of the present application, the physiological characteristic data and motion characteristic data of the swimmer are collected by the smart wearable device, and the physiological characteristic data and motion characteristic data can be used to directly determine whether the swimmer is in an abnormal state, and the swimmer in the siege state is determined to be the target drowning person. Combined with the changes in the physiological characteristic data of the target drowning person asking for help from nearby supportive objects, it is further determined that the target drowning person is actually in a drowning state, so as to avoid the situation where some swimmers whose physiological characteristics change greatly during diving or swimming are not in a drowning state, thereby improving the accuracy of the swimming drowning monitoring method. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0039] Figure 1 Schematic diagram of the process of the swimming drowning monitoring method in the embodiment of the present application. DETAILED DESCRIPTION

[0040] The specific embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the description 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.

[0041] In the description of the present invention, unless otherwise clearly specified and limited, the terms "set", "install", "connection" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to specific circumstances.

[0042] The directions or positional relationships indicated by terms such as “upper”, “lower”, “left”, “right”, “front”, “back”, “top”, “bottom”, “inside” and “outside” are based on the directions or positional relationships shown in the accompanying drawings, or are the directions or positional relationships in which the inventive product is usually placed when used. They are only for the convenience of description and simplified description, and do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore should not be understood as a limitation on the present invention.

[0043] The terms "first", "second", "third", etc. are merely used to distinguish elements of similar nature, and do not indicate or imply relative importance or a particular order.

[0044] The terms "comprises," "comprising," or any other variations thereof, are intended to cover a non-exclusive inclusion of the elements listed and may also include additional elements not expressly listed.

[0045] In the system architecture of the method for drowning in swimming in this embodiment, multiple smart wearable devices, multiple cameras, a cloud server and a PC are included. The smart wearable device is provided with a heart rate sensor, a blood oxygen sensor, a blood pressure sensor, a temperature sensor, a speed sensor and other sensors, which can collect the physiological characteristic data and motion characteristic data of the user's swimmer. The camera includes an underwater camera installed at the bottom of the swimming pool, and also includes a high-speed camera installed above the swimming pool. The underwater camera can take the swimming action of the swimmer underwater, and the high-speed camera is used to determine the position of the swimmer's head, whether it is above the water surface or below the water surface. A communication module that is wirelessly connected to the cloud is provided in the smart wearable device and the camera, and the physiological characteristic data and motion characteristic data collected by the smart wearable device, as well as the image information collected by the camera can be wirelessly transmitted to the cloud server. The PC end is used to receive the physiological characteristic data, motion characteristic data and image information transmitted by the cloud server, and process the relevant data, so as to identify whether the swimmer is in a drowning state, and after determining that the swimmer is in a drowning state, an alarm is issued by sending an alarm command, so that the lifeguard at the poolside can quickly rescue the drowning swimmer.

[0046] Among them, the smart wearable device can be one of a smart bracelet, a smart watch, and a smart phone. The smart bracelet is taken as an example for description; the smart bracelet collects the swimmer's physiological characteristic data and motion characteristic data through sensors according to a preset time period, and sends them to the cloud server.

[0047] like Figure 1 As shown, this embodiment provides a method for monitoring drowning in swimming, where each swimmer wears a smart wearable device, and the smart wearable device is used to collect physiological characteristic data and motion characteristic data of each swimmer. The method includes:

[0048] Step S100: marking the swimmer whose physiological characteristic data is in an abnormal state as a target drowning person;

[0049] Step S200: Taking the target drowning person as the center, determining the supportable object closest to the target drowning person; the supportable object includes other swimmers or auxiliary rescue facilities;

[0050] Step S300: determining a first moving direction of the target drowning person according to the motion characteristic data of the target drowning person;

[0051] Step S400: When the first moving direction of the target drowning person is toward the supportable object, and the physiological characteristic data of the target drowning person returns to a normal state and then changes to an abnormal state, it is determined that the target drowning person is in a drowning state.

[0052] When a swimmer drowns, he usually enters the "drowning" state. Only a small number of swimmers who enter the "drowning" state have excessive nervous reactions due to panic or fear, and the throat produces coughing and swallowing reflexes. After the laryngeal spasm and glottis closure, the airway is completely obstructed and suffocated, which further causes death from brain hypoxia and myocardial hypoxia. Or due to the effect of nerve reflexes, when the laryngeal spasm, the heart can reflexively stop beating through the vagal reflex; and reflex malignant arrhythmia occurs after suffocation, and finally the heart stops beating. However, most swimmers who enter the "drowning" state usually do not die immediately. Instead, after a short panic, they seek nearby support objects and move closer to them in the hope of getting help from them. In this process, the swimmer will first go through a period of rational thinking, and slow down the heartbeat and oxygen consumption until the body is hypoxic due to holding the breath for a long time.

[0053] In the swimming drowning monitoring method of the present embodiment, the physiological characteristic data and motion characteristic data of the swimmer are collected by the smart wearable device, and the physiological characteristic data and motion characteristic data can be used to directly determine whether the swimmer is in an abnormal state, and the swimmer in the siege state is determined to be the target drowning person. Combined with the changes in the physiological characteristic data of the target drowning person asking for help from nearby support objects, it is further determined that the target drowning person is actually in a drowning state, so as to avoid the situation where some swimmers whose physiological characteristics change greatly during diving or swimming are not in a drowning state, thereby improving the accuracy of the swimming drowning monitoring method.

[0054] In step S100, within a few seconds after the swimmer enters the drowning state, the swimmer will be affected by emotions such as fear. The swimmer's heart rate changes the most during this process. Therefore, it can be determined whether the swimmer is in an abnormal state by judging the change in the swimmer's heart rate.

[0055] In one embodiment, the physiological characteristic data includes heart rate data, and the heart rate data includes dynamic heart rate and static heart rate; a swimmer whose physiological characteristic data is in an abnormal state is marked as a target drowning person, specifically including: when the dynamic heart rate of a swimmer exceeds the maximum heart rate range, the swimmer is marked as a target drowning person.

[0056] Static heart rate refers to the number of heartbeats per minute when the human body is at rest. For swimmers, static heart rate is usually measured when the body is fully rested and in a relaxed state. It is usually measured within a period of time before swimming when the wearable device is worn. At this time, the body has not been disturbed by external activities and other factors, and can more accurately reflect the body's basic physiological state. The static heart rate of a normal adult is generally 60-100 beats per minute. Some swimmers or professional swimmers who are in good physical condition have enhanced cardiopulmonary function due to long-term swimming exercises, and their static heart rate is often lower than that of the general population, which may be around 40-60 beats per minute.

[0057] Dynamic heart rate refers to the number of times the heart beats per minute during swimming. It varies with factors such as swimming intensity, time, and swimming style. For example, when a swimmer is doing low-intensity breaststroke or backstroke, the heart rate is about 50%-60% of the maximum heart rate; when a swimmer is doing moderate-intensity freestyle, the heart rate is about 60%-75% of the maximum heart rate; and when a swimmer is doing high-intensity competitive swimming, the heart rate is about 80%-95% of the maximum heart rate or above.

[0058] Maximum heart rate is the highest heart rate a person can reach during exercise (HRmax). Usually, a person's maximum heart rate can be roughly estimated by age, that is, Hrmax=220-0.7*age. However, due to the large individual differences, it is more accurate to determine the swimmer's maximum heart rate range by static heart rate and dynamic heart rate.

[0059] In one embodiment, before marking the swimmer as a target drowning person, the method further includes:

[0060] Obtaining a static heart rate of each of the swimmers;

[0061] A maximum heart rate range for each of the swimmers is determined based on the static heart rate.

[0062] In this embodiment, a scale of the swimmer's maximum heart rate and static heart rate is established based on the swimmer's physical attribute information such as height, weight, and vital capacity. The maximum heart rate corresponding to the physical attribute information is recorded in the scale as a multiple of the static heart rate. The maximum heart rate range can be directly obtained by retrieving the relevant multiples in the scale and multiplying it by the static heart rate.

[0063] In addition, for customers who swim in the pool more than once, the cloud server also records the physiological and sports characteristics data of each swimmer during their historical swimming process, and uses the camera to determine the sports mode (for example, breaststroke, backstroke or freestyle, sprint swimming, etc.) corresponding to the different physiological characteristics of the swimmer. Swimmers can bind their accounts by logging into the APP on their smart wearable devices. During the swimming process, the swimmer's swimming data can be wirelessly transmitted to the cloud server through the smart wearable device for storage.

[0064] In one embodiment, before marking the swimmer as a target drowning person, the method further includes:

[0065] determining the movement pattern and movement intensity of each of said swimmers;

[0066] Obtaining historical dynamic heart rate of each swimmer's historical swimming data corresponding to the exercise mode and the exercise intensity;

[0067] A maximum heart rate range of the swimmer is determined based on the historical dynamic heart rate.

[0068] In this embodiment, the swimmer's motion pattern can be determined by the swimmer's moving speed and the image information captured by the high-speed camera, and multiple high-speed cameras are used to shoot the swimmer from different angles (such as the side and top of the pool). These cameras should be able to clearly capture the swimmer's body contour, limb movements, and position changes in the water. For example, installing cameras on both sides of the pool can well observe the details of the swimmer's arm strokes and leg kicks; cameras installed from the top are helpful for observing the swimmer's swimming trajectory in the pool.

[0069] Computer vision technology is used to identify key points of the human body in the captured images. These key points include but are not limited to the head, shoulders, elbows, wrists, hips, knees and ankles. By identifying the positions of these key points in each frame of the image, a body posture model of the swimmer can be constructed. For example, by analyzing the position changes of the wrists and elbows, the amplitude and frequency of the arm stroke can be understood; by observing the movements of the ankles and knees, the leg kick can be analyzed. According to the position changes of the key points or the overall outline of the swimmer's body in the image sequence, the movement trajectory of the swimmer in the pool is tracked. This can be achieved by using the optical flow method or the target tracking algorithm based on deep learning. For example, by calculating the changes in the position of the swimmer's head between adjacent frames, the movement trajectory of the swimmer in the horizontal direction (forward or backward) and the vertical direction (floating up and down) can be obtained. This trajectory information can intuitively reflect whether the swimmer's swimming path is stable and how linear it is.

[0070] The action cycle is divided according to the repeatability of the swimmer's limb movements. Taking freestyle as an example, a complete arm stroke cycle can start from the arm entering the water, through paddling and pushing, to the arm leaving the water and moving the arm in the air, and then returning to the water to end. By analyzing the arm movements in the image, the starting and ending points of each cycle are determined, and then the action details in each cycle, such as stroke speed, stroke angle, etc., are analyzed.

[0071] The swimming style is determined by analyzing the swimmer's limb movements and body posture. For example, if the swimmer is observed lying prone on the water, moving forward by alternating arm strokes and legs up and down, and the arm strokes are in an S shape, it is likely freestyle; if the swimmer is seen paddling with both arms at the same time, pushing the water outward from the chest, and the leg movements are a continuous movement of pulling in the legs, turning the feet, kicking the legs, and clamping the legs, it is most likely breaststroke.

[0072] At the same time, the swimmer's exercise intensity can be determined in combination with the swimmer's movement speed, and the historical dynamic heart rate of the same intensity in the corresponding exercise mode can be determined based on the exercise intensity. The maximum heart rate range can be obtained by multiplying the historical dynamic heart rate by the corresponding coefficient.

[0073] Because some swimmers' dynamic heart rate will exceed the maximum heart rate range when they are diving, but the emotions of the diver are relatively stable, so their blood pressure and blood oxygen psychological characteristic parameters are different from those of the drowning patients. Therefore, in order to avoid the interference caused by the diver, the two parameters of blood pressure and blood oxygen can be combined to further identify whether the target drowning person is in a drowning state.

[0074] In one embodiment, the physiological characteristic data further includes blood pressure data and blood oxygen data, the blood pressure data includes dynamic blood pressure, and the blood oxygen data includes dynamic blood oxygen; after marking the swimmer as a target drowning person, the following is further included:

[0075] Obtaining the dynamic blood pressure, dynamic blood oxygen and dynamic heart rate of the target drowning person within a preset time range before the time point of the abnormal state;

[0076] When the dynamic heart rate at a certain time node is lower than the preset minimum heart rate, the time node is determined as the initial diving time node;

[0077] Between the initial diving time node and the time node of the abnormal state, if the dynamic blood pressure does not show a trend of first increasing and then stabilizing or decreasing, and the dynamic blood oxygen does not show a trend of first increasing or decreasing, the swimmer is marked from a target drowning person to be determined.

[0078] In this embodiment, when a swimmer enters the water from the surface and starts diving, the human body instinctively holds its breath, which will cause a strong diving reflex, and symptoms such as apnea, bradycardia, and severe peripheral vascular contraction will occur. In a short period of time, the swimmer's heart rate will drop rapidly and fall below the preset minimum heart rate. Therefore, the time point below the preset minimum heart rate is used as the initial diving time point for entering the diving reaction. For drowning patients, due to the psychological state of fear, their heart rate is mostly in an upward state in the early stage, and remains stable after rising to a certain extent.

[0079] The blood pressure sensor can continuously obtain the swimmer's dynamic blood pressure according to the sampling cycle, generally automatically measuring once every 15-20 seconds. When the diving reflex occurs, the blood pressure will rise in its initial stage. This is because the peripheral blood vessels (especially the blood vessels in the skin and limbs) will strongly contract under triggering factors such as cold water stimulating the face. This vasoconstriction causes more blood to concentrate around the core organs, such as the heart, brain and other important organs, resulting in increased blood pressure. If the diving time is prolonged, with the combined influence of various factors such as body metabolism, the blood pressure may stabilize or have a downward trend. For example, during a long dive, the body's oxygen reserves gradually decrease, and the heart function may be affected to a certain extent, so that the blood pressure cannot be maintained at the increased level.

[0080] The blood oxygen sensor can continuously emit red light and infrared light of specific wavelengths to the human body. After the light penetrates the skin, tissues and blood, it is received by the photodetector. Since oxygenated hemoglobin and deoxygenated hemoglobin have different absorption characteristics for red light and infrared light, the dynamic blood oxygen is calculated by measuring the absorption degree of the two lights. For a period of time after the start of the diving reflex, the blood oxygen content can rise slightly from the normal level. This is mainly due to the redistribution of blood, which allows more oxygen-rich blood to supply important organs first. For example, after the redistribution of blood, the oxygen carried by hemoglobin can be more effectively used by organs such as the heart and brain. In addition, in the early stage of the diving reflex, the human body's metabolic rate may decrease, and the body's consumption of oxygen slows down, which also helps maintain the blood oxygen content. As the diving time increases, the body's oxygen reserves will gradually be consumed due to the inability to obtain fresh oxygen from the outside world. In particular, when the diver reaches the limit of breath holding, the blood oxygen content will gradually decrease.

[0081] Since some drowning swimmers will also enter the diving reflex, and after entering the diving reflex, only the changes in dynamic blood pressure and dynamic blood oxygen cannot fully determine whether the swimmer is in a drowning state or diving. Therefore, it is necessary to further combine the image information of the swimmer's body shape to determine whether the swimmer is drowning.

[0082] In one embodiment, after marking the swimmer from a target drowning person to be determined, the method further includes:

[0083] Acquire image information of a plurality of drowning persons to be determined as targets;

[0084] By combining upsampling with downsampling, the plurality of image information are fused to obtain a fused image;

[0085] Inputting the fused image into a preset depth recognition model to identify whether the target drowning person to be determined is in drowning behavior;

[0086] When the target drowning person to be determined does not have any drowning behavior, the mark of the target drowning person to be determined is cancelled.

[0087] In this embodiment, image information of the target drowning person underwater can be obtained through multiple underwater cameras. By fusing multiple image information, the influence of splashes and reflections on the posture recognition of the target drowning person to be determined can be effectively filtered out, which facilitates the deep recognition model to extract posture features with higher accuracy and determine whether the target drowning person to be determined is in drowning behavior.

[0088] The dataset used to train the deep learning model is the swimming and drowning detection dataset, which includes three labeled categories (drowning people, swimming people, and people leaving the water). The scenes are mainly indoor swimming pools and outdoor swimming pools. The dataset is divided into training set, validation set, and test set. Among them, the deep learning model learns from the data in the training set and adjusts its own parameters to build an algorithm that can effectively predict or classify the input data. The validation set is used to evaluate the performance of different model architectures or hyperparameter settings during the model training process. For example, when training a neural network, different combinations of hyperparameters such as the number of layers, the number of neurons in each layer, and the learning rate may be tried. By evaluating the model on the validation set, the performance indicators such as accuracy, recall rate, F1 value, etc. of the model under different combinations can be compared. The test set is used to evaluate the performance and generalization ability of the model in actual application scenarios.

[0089] In one embodiment, the fusing of the plurality of image information to obtain a fused image by combining upsampling with downsampling specifically includes:

[0090] Performing a convolution operation on each of the image information to reduce the size of the image information to obtain a scaled image;

[0091] The zoomed image is placed at the center of a preset enlarged template image, and adjacent positions of corresponding pixel values ​​in the zoomed image are assigned values ​​according to pixels of the image information to obtain restored image information;

[0092] Each of the restored image information is superimposed according to the number of channels and the spatial dimension, so as to obtain the fused image.

[0093] In this embodiment, the image information is first downsampled. For example, by setting the convolution operation with a basic step size of 1, if the size of the image information is H*W*C (H is the height, W is the width, and C is the number of channels), a convolution kernel with a step size of 2 is used for convolution, and the size can be H / 2*W / 2*C. / (C / The scaled image is then upsampled. For example, the 2*2 scaled image is placed at the center of the 4*4 enlarged template image. Through the nearest neighbor upsampling method, each pixel value in the unprocessed image data is assigned in a copy manner in the enlarged template image, thereby restoring the size of the scaled image to its original size and obtaining the restored image information. When performing image fusion, the similar features of the restored image information of different scales are fused by element-by-element addition, thereby reflecting the swimmer's posture.

[0094] Downsampling can extract high-level semantic information from images because it reduces the image size. In this process, some local details can be ignored and the focus can be placed on more abstract and representative features. For example, the texture details of the palm and skin in the image information will be gradually ignored, but the overall outline of the swimmer and the spatial relationship with the surrounding environment will be more clearly presented. Upsampling can restore the size of the image and reintroduce the detail information. By fusing the semantic information obtained by downsampling with the detail information obtained by upsampling, the generated fused image can contain rich details and high-level semantic information at the same time, construct a multi-level feature pyramid, and facilitate the deep learning model to accurately obtain the swimmer's body information.

[0095] Step S200: Taking the target drowning person as the center, determine the supportable object closest to the target drowning person; the supportable object includes other swimmers or auxiliary rescue facilities.

[0096] The camera is also equipped with an infrared sensor device, which sends infrared light to the target drowning person and multiple nearby supportable objects. By emitting infrared light and measuring its reflection time, other swimmers or auxiliary rescue facilities whose reflection time is closest to the target drowning person are considered supportable objects.

[0097] Step S300: determining a first moving direction of the target drowning person according to the motion characteristic data of the target drowning person.

[0098] The smart wearable device is also equipped with a gyroscope, which can sense the forward direction and azimuth of the target drowning person, thereby determining the first moving direction of the target drowning person.

[0099] Step S400: When the first moving direction of the target drowning person is toward the supportable object, if the physiological characteristic data of the target drowning person returns to a normal state and then becomes an abnormal state, it is determined that the target drowning person is in a drowning state.

[0100] When the target drowning person enters the rescue range of the supportable object, due to the instinct of survival, the heart rate of the target drowning person will speed up again in order to seek help from the supportable object as soon as possible. During this process, the dynamic heart rate exceeds the maximum heart rate range, and due to prolonged hypoxia, the dynamic blood oxygen will also decrease.

[0101] In one embodiment, after the first moving direction of the target drowning person is toward the supportable object, the method further includes:

[0102] When the target drowning person is located within the preset rescueable range of the supportable object, obtaining the second moving direction of the target drowning person again;

[0103] If the second moving direction is away from the supportable object, it is determined that the target drowning person is not in a drowning state.

[0104] In this embodiment, if the target drowning person is not a real drowning person, but a diver with high training intensity, there may be other people or auxiliary rescue facilities on the road ahead that are not observed during the diving path, that is, the supportable object becomes an obstacle on its forward path, so it chooses a second moving direction different from the first moving direction to move away from the supportable object. Such diver with high training intensity can also be excluded. Among them, when the angle between the first moving direction and the second moving direction is greater than 45°, it can be determined that the second moving direction is away from the supportable object.

[0105] To summarize, in the swimming drowning monitoring method of the embodiment of the present application, the physiological characteristic data and motion characteristic data of the swimmer are collected by the smart wearable device, and the physiological characteristic data and motion characteristic data can be used to directly determine whether the swimmer is in an abnormal state, and the swimmer in the siege state is determined to be the target drowning person. Combined with the changes in the physiological characteristic data of the target drowning person asking for help from nearby supportive objects, it is further determined that the target drowning person is actually in a drowning state, so as to avoid the possibility that some swimmers whose physiological characteristics change greatly during diving or swimming are not in a drowning state, thereby improving the accuracy of the swimming drowning monitoring method.

[0106] Based on the same inventive concept as the above-mentioned embodiment, the present application further provides a swimming drowning monitoring device, in which each swimmer wears a smart wearable device, and the smart wearable device is used to collect physiological characteristic data and motion characteristic data of each swimmer, including:

[0107] A target determination module, used for marking a swimmer whose physiological characteristic data is in an abnormal state as a target drowning person;

[0108] A support determination module, used to determine the closest supportable object to the target drowning person, with the target drowning person as the center; the supportable object includes other swimmers or auxiliary rescue facilities;

[0109] a direction determination module, used to determine a first moving direction of the target drowning person according to the motion characteristic data of the target drowning person;

[0110] The state determination module is used to determine that the target drowning person is in a drowning state when the first moving direction of the target drowning person is toward the supportable object and the physiological characteristic data of the target drowning person returns to a normal state and then becomes an abnormal state.

[0111] Based on the same inventive concept as the above embodiments, this embodiment also provides a swimming drowning monitoring system, including: a processor and a memory; wherein the memory stores a computer program, and the computer program is used to be loaded by the processor and execute the above-mentioned swimming drowning monitoring method.

[0112] Based on the same inventive concept as the above embodiments, this embodiment further provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, and the instructions are used for a processor to load and execute the above-mentioned swimming drowning monitoring method.

[0113] In the embodiments of the mobile terminal and computer-readable storage medium provided in the present application, all technical features of the above-mentioned control method embodiments are included, and the expansion and explanation content of the specification are basically the same as those of the above-mentioned method embodiments, which will not be repeated here.

[0114] The embodiment of the present application further provides a computer program product, which includes a computer program code. When the computer program code runs on a computer, the computer executes the methods in the above various possible implementation modes.

[0115] An embodiment of the present application also provides a chip, including a memory and a processor, the memory is used to store a computer program, and the processor is used to call and run the computer program from the memory, so that a device equipped with the chip executes the methods in various possible implementation modes as described above.

[0116] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0117] In the present application, the same or similar terminology concepts, technical solutions and / or application scenario descriptions are generally described in detail only the first time they appear. When they appear again later, they are generally not repeated for the sake of brevity. When understanding the technical solutions and other contents of the present application, for the same or similar terminology concepts, technical solutions and / or application scenario descriptions that are not described in detail later, reference can be made to the previous related detailed descriptions.

[0118] In the present application, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0119] The various technical features of the technical solution of the present application can be arbitrarily combined. In order to make the description concise, not all possible combinations of the various technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of the present application.

[0120] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product is stored in a storage medium as above, including a number of instructions for a terminal device to execute the method of each embodiment of the present application. The above is only a preferred embodiment of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made by using the content of the specification and drawings of this application, or directly or indirectly used in other related technical fields, is similarly included in the patent protection scope of the present application.

[0121] It should be noted that the various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0122] 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 a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the attached claims.

Claims

1. A method for monitoring drowning while swimming, wherein each swimmer wears a smart wearable device, and the smart wearable device is used to collect physiological characteristic data and motion characteristic data of each swimmer, characterized in that: The method comprises: Marking the swimmer whose physiological characteristic data is in an abnormal state as a target drowning person; Taking the target drowning person as the center, determining the supportable object closest to the target drowning person; the supportable object includes other swimmers or auxiliary rescue facilities; Determining a first moving direction of the target drowning person according to the motion characteristic data of the target drowning person; When the first moving direction of the target drowning person is toward the supportable object, and the physiological characteristic data of the target drowning person returns to a normal state and then changes to an abnormal state, it is determined that the target drowning person is in a drowning state; The physiological characteristic data includes heart rate data, and the heart rate data includes dynamic heart rate and static heart rate; marking a swimmer whose physiological characteristic data is in an abnormal state as a target drowning person specifically includes: When the dynamic heart rate of a swimmer exceeds the maximum heart rate range, the swimmer is marked as a target drowning person; The physiological characteristic data also includes blood pressure data and blood oxygen data, wherein the blood pressure data includes dynamic blood pressure, and the blood oxygen data includes dynamic blood oxygen; after marking the swimmer as a target drowning person, the following further includes: Obtaining the dynamic blood pressure, dynamic blood oxygen and dynamic heart rate of the target drowning person within a preset time range before the time point of the abnormal state; When the dynamic heart rate at a certain time node is lower than the preset minimum heart rate, the time node is determined as the initial diving time node; Between the initial diving time node and the time node of the abnormal state, if the dynamic blood pressure does not show a trend of first increasing and then stabilizing or decreasing, and the dynamic blood oxygen does not show a trend of first increasing or decreasing, the swimmer is marked from a target drowning person to be determined.

2. The method for monitoring drowning in swimming according to claim 1, characterized in that: Prior to marking the swimmer as a target drowning person, also include: Obtaining a static heart rate of each of the swimmers; A maximum heart rate range for each of the swimmers is determined based on the static heart rate.

3. The method for monitoring drowning in swimming according to claim 1, characterized in that: Prior to marking the swimmer as a target drowning person, also include: determining the movement pattern and movement intensity of each of said swimmers; Obtaining historical dynamic heart rate of each swimmer's historical swimming data corresponding to the exercise mode and the exercise intensity; A maximum heart rate range of the swimmer is determined based on the historical dynamic heart rate.

4. The method for monitoring drowning in swimming according to claim 1, characterized in that: After marking the swimmer from a target drowning person to be determined as a target drowning person to be determined, the method further includes: Acquire image information of a plurality of drowning persons to be determined as targets; By combining upsampling with downsampling, the plurality of image information are fused to obtain a fused image; Inputting the fused image into a preset depth recognition model to identify whether the target drowning person to be determined is in drowning behavior; When the target drowning person to be determined does not have any drowning behavior, the mark of the target drowning person to be determined is cancelled.

5. The method for monitoring drowning in swimming according to claim 4, characterized in that: The fusing of the plurality of image information to obtain a fused image by combining upsampling with downsampling specifically includes: Performing a convolution operation on each of the image information to reduce the size of the image information to obtain a scaled image; The zoomed image is placed at the center of a preset enlarged template image, and adjacent positions of corresponding pixel values ​​in the zoomed image are assigned values ​​according to pixels of the image information to obtain restored image information; Each of the restored image information is superimposed according to the number of channels and the spatial dimension, so as to obtain the fused image.

6. The method for monitoring drowning in swimming according to any one of claims 1 to 5, characterized in that: After the first moving direction of the target drowning person is toward the supportable object, the method further includes: When the target drowning person is located within the preset rescueable range of the supportable object, obtaining the second moving direction of the target drowning person again; If the second moving direction is away from the supportable object, it is determined that the target drowning person is not in a drowning state.

7. A swimming drowning monitoring system, characterized in that: It comprises a processor and a memory; wherein the memory stores a computer program, and the computer program is used for the processor to load and execute the swimming drowning monitoring method as described in any one of claims 1-6.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores instructions, and the instructions are used by a processor to load and execute the swimming drowning monitoring method as described in any one of claims 1-6.

Images