A key object care system and method based on the Internet of Things
Through the collaborative work of Internet of Things technology and neural networks, effective monitoring and cross-regional response to the elderly, children and people with disabilities in complex outdoor environments is achieved, and the problem of poor outdoor monitoring of the existing technology is solved.
Patent Information
- Application Number
- CN202210784051.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-07-05
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2042-07-05
AI Technical Summary
It is difficult for existing intelligent care systems to effectively monitor the elderly, children and people with disabilities in a large range of outdoor environments, especially in complex environments.
Adopt a key object care system based on the Internet of Things, collect data through the first terminal and transmit it to the management server, use the neural network to determine whether there are abnormal situations, and generate a response plan, including the coordinated work of multiple terminals and management servers to achieve cross-regional monitoring and response.
Effectively monitor key targets in complex outdoor environments and achieve rapid response through cross-regional coordination, improving safety and timeliness for the elderly, children and people with disabilities.
Smart Images

Figure CN115150434B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of Internet of Things security technology, and in particular to a key object care system and method based on the Internet of Things. Background Art
[0002] As we all know, the elderly, children, and people with disabilities face many inconveniences in social life, and are more likely to fall into danger than ordinary people. With the improvement of social living standards, medical standards, and the decline in fertility rates, many regions around the world have entered an aging society. The elderly are prone to accidents in life due to their decreased reaction and physical functions; children under the age of 5 are relatively weak in self-care ability, discrimination, physical fitness, and reaction ability. People with disabilities face many unsolvable crises due to their own physiological deficiencies. Therefore, the above-mentioned groups should become the "key targets" of the entire society.
[0003] With the development of technology, many intelligent care systems or products have emerged. These systems / products use sensor technology, communication technology, video processing technology, biometric technology, big data technology, and Internet of Things technology to collect physiological characteristics data or video data of specific people to achieve the function of care. However, such systems are more used for indoor care or outdoor care in a small range (such as nursing homes), and the care experience is not good; while there is a lack of effective targeted monitoring for a larger range of outdoor care, but the above-mentioned key targets face even more dangers outdoors than indoors. Summary of the invention
[0004] The present invention is based on the above-mentioned problems and proposes a key object care system and method based on the Internet of Things. The first data is collected through a first terminal. When the first data belongs to a key object, a neural network on the first terminal management server determines whether there is an abnormality in the key object corresponding to the first data. When an abnormality exists, the first terminal executes a response plan. Not only can the key objects be cared for in a complex outdoor environment, but also the cross-regional search and care of key objects can be achieved by utilizing a second terminal and a second terminal management server.
[0005] In view of this, one aspect of the present invention proposes a key object care system based on the Internet of Things, comprising: a plurality of first terminals, at least one first terminal management server for managing the first terminals, a plurality of second terminals, at least one second terminal management server for managing the second terminals, and an upper-level server for managing the first terminal management server and the second terminal management server;
[0006] The first terminal includes a camera, a 5G communication module for receiving and sending data, a sound collection module, a positioning module, a light emitting module, an event processing module and a control processor;
[0007] The first terminal is configured as follows:
[0008] Acquire first data through the camera and the sound collection module, wherein the first data at least includes human behavior data, vehicle behavior data, animal behavior data and environmental data;
[0009] sending the first data to the first terminal management server;
[0010] The first terminal management server is configured as:
[0011] Using historical data to train a neural network to generate a key object neural network model, wherein the historical data at least includes human behavior historical data, vehicle behavior historical data, animal behavior historical data, and environmental historical data;
[0012] Determine whether the first data belongs to a preset key object according to a preset rule, and if not, send the first data to the upper-level server, which sends the first data to a second terminal management server managed by the upper-level server, and the second terminal management server determines whether the first data belongs to a preset key object;
[0013] If the first data belongs to a preset key object, inputting the first data into the key object neural network model;
[0014] Determine whether the key object is abnormal;
[0015] When the key object has an abnormality, generating an abnormality response plan, and sending the abnormality response plan to the first terminal;
[0016] The first terminal is configured to: receive the abnormality response plan, execute the abnormality response plan and feed back the execution result to the first terminal management server.
[0017] Optionally, the neural network includes an input layer, an initial layer, a simulated output layer, a secondary initial layer, a verification coefficient layer and an output layer;
[0018] The key object neural network model at least includes a key person neural network model, a key vehicle neural network model, a key animal neural network model and a key environment neural network model;
[0019] Wherein, in the step of using historical data to train the neural network to generate a key object neural network model, the first terminal management server is configured as follows:
[0020] Classifying the historical data to generate a plurality of different data groups;
[0021] Sending the data set as input value to the input layer;
[0022] The input layer is connected to the initial layer through matrix operation;
[0023] The initial layer obtains a first output result, activates the first output result, and sends the activated second output result to the simulation output layer;
[0024] The simulation output layer obtains simulation output values through matrix calculation, and inputs the simulation output values into the secondary initial layer;
[0025] The secondary initial layer obtains the verification output result through matrix calculation;
[0026] Performing data connection between the input value of the input layer and the secondary initial layer;
[0027] The secondary initial layer obtains a secondary output result through matrix calculation, sends the secondary output result and the verification output result to the verification coefficient layer for verification, and obtains a normalization coefficient;
[0028] Sending the normalization coefficient and the simulated output value to the output layer, the output layer normalizes the simulated output value to obtain a mimicking result;
[0029] Collect positive and / or negative feedback information;
[0030] The mimicry result is learned and modified according to the positive feedback and / or the reverse feedback information to generate a key object neural network model.
[0031] Optionally, the response plan includes:
[0032] Provide emergency medicines required by the key targets; and / or,
[0033] Playing a first aid video corresponding to the current symptoms of the key subject; and / or,
[0034] Providing supplies for managing animals; and / or, Providing supplies to cope with current environmental changes.
[0035] Optionally, the first terminal management server is configured with a key object monitoring model;
[0036] The first terminal management server is configured as:
[0037] Inputting the first data into the key subject monitoring model, and determining whether the first data matches the key subject monitoring model;
[0038] When the first data matches the key subject monitoring model, the first data is input into the key subject neural network model to determine whether there is an abnormality;
[0039] When the first data does not match the key object monitoring model, the key object monitoring model is sent to the upper-level server, and the upper-level server sends a collaborative search instruction to the second terminal management server managed by it;
[0040] The second terminal management server is configured as follows:
[0041] receiving the collaborative search instruction, and collecting second data using the second terminal;
[0042] Determining whether a key object corresponding to the key object monitoring model is found according to the second data;
[0043] The upper-level server is configured as follows:
[0044] When a key object corresponding to the key object monitoring model is found, controlling the second terminal management server to obtain current data of the key object and sending the current data to the first terminal management server;
[0045] The first terminal management server is configured to input the current data into the key object neural network model to determine whether there is an abnormality.
[0046] Optionally, the key subject care system further includes a key subject monitoring terminal, and the first terminal is configured as:
[0047] When the key object is detected, the location information of the key object is obtained through the positioning module;
[0048] Sending the location information to the first terminal management server;
[0049] The first terminal management server is configured as:
[0050] Navigation data is generated according to the location information and the location information of the key subject monitoring terminal, and is sent to the key subject monitoring terminal.
[0051] Another aspect of the present invention provides a key object care method based on the Internet of Things, the key object care system based on the Internet of Things includes multiple first terminals, at least one first terminal management server for managing the first terminals, multiple second terminals, at least one second terminal management server for managing the second terminals, and an upper-level server for managing the first terminal management server and the second terminal management server; the first terminal includes a camera, a 5G communication module for receiving and sending data, a sound collection module, a positioning module, a light-emitting module, an event processing module, and a control processor, and the key object care method includes:
[0052] The first terminal acquires first data through the camera and the sound collection module, wherein the first data at least includes human behavior data, vehicle behavior data, animal behavior data and environmental data;
[0053] sending the first data to the first terminal management server;
[0054] The first terminal management server trains a neural network using historical data to generate a key object neural network model, wherein the historical data at least includes human behavior historical data, vehicle behavior historical data, animal behavior historical data and environmental historical data;
[0055] Determine whether the first data belongs to a preset key object according to a preset rule, and if not, send the first data to the upper-level server, which sends the first data to a second terminal management server managed by the upper-level server, and the second terminal management server determines whether the first data belongs to a preset key object;
[0056] If the first data belongs to a preset key object, inputting the first data into the key object neural network model;
[0057] Determine whether the key object is abnormal;
[0058] When the key object has an abnormality, generating an abnormality response plan, and sending the abnormality response plan to the first terminal;
[0059] The first terminal receives the abnormality response plan, executes the abnormality response plan and feeds back the execution result to the first terminal management server.
[0060] Optionally, the neural network includes an input layer, an initial layer, a simulated output layer, a secondary initial layer, a verification coefficient layer and an output layer;
[0061] The key object neural network model at least includes a key person neural network model, a key vehicle neural network model, a key animal neural network model and a key environment neural network model;
[0062] The specific steps of using historical data to train the neural network and generate the neural network model of the key object are as follows:
[0063] Classifying the historical data to generate a plurality of different data groups;
[0064] Sending the data set as input value to the input layer;
[0065] The input layer is connected to the initial layer through matrix operation;
[0066] The initial layer obtains a first output result, activates the first output result, and sends the activated second output result to the simulation output layer;
[0067] The simulation output layer obtains simulation output values through matrix calculation, and inputs the simulation output values into the secondary initial layer;
[0068] The secondary initial layer obtains the verification output result through matrix calculation;
[0069] Performing data connection between the input value of the input layer and the secondary initial layer;
[0070] The secondary initial layer obtains a secondary output result through matrix calculation, sends the secondary output result and the verification output result to the verification coefficient layer for verification, and obtains a normalization coefficient;
[0071] Sending the normalization coefficient and the simulated output value to the output layer, the output layer normalizes the simulated output value to obtain a mimicking result;
[0072] Collect positive and / or negative feedback information;
[0073] The mimicry result is learned and modified according to the positive feedback and / or the reverse feedback information to generate a key object neural network model.
[0074] Optionally, the response plan includes:
[0075] Provide emergency medicines required by the key targets; and / or,
[0076] Playing a first aid video corresponding to the current symptoms of the key subject; and / or,
[0077] Providing supplies for managing animals; and / or, Providing supplies to cope with current environmental changes.
[0078] Optionally, the first terminal management server is configured with a key object monitoring model;
[0079] The key object care method also includes the steps of:
[0080] The first terminal management server inputs the first data into the key object monitoring model to determine whether the first data matches the key object monitoring model;
[0081] When the first data matches the key subject monitoring model, the first data is input into the key subject neural network model to determine whether there is an abnormality;
[0082] When the first data does not match the key object monitoring model, the key object monitoring model is sent to the upper-level server, and the upper-level server sends a collaborative search instruction to the second terminal management server managed by it;
[0083] The second terminal management server receives the collaborative search instruction and uses the second terminal to collect second data;
[0084] Determining whether a key object corresponding to the key object monitoring model is found according to the second data;
[0085] When a key object corresponding to the key object monitoring model is found, the upper server controls the second terminal management server to obtain current data of the key object, and sends the current data to the first terminal management server;
[0086] The first terminal management server inputs the current data into the key object neural network model to determine whether there is an abnormality.
[0087] Optionally, the key object care system further includes a key object monitoring terminal, and the key object care method further includes the steps of:
[0088] When the key object is detected, the first terminal obtains its own location information through the positioning module;
[0089] Sending the location information to the first terminal management server;
[0090] The first terminal management server generates navigation data according to the location information and the location information of the key object monitoring terminal, and sends the navigation data to the key object monitoring terminal.
[0091] By adopting the technical solution of the present invention, a key object care system based on the Internet of Things is provided with multiple first terminals, at least one first terminal management server for managing the first terminals, multiple second terminals, at least one second terminal management server for managing the second terminals, and a superior server for managing the first terminal management server and the second terminal management server. The first data is collected by the first terminal. When the first data belongs to a key object, the neural network on the first terminal management server determines whether there is an abnormality in the key object corresponding to the first data. When there is an abnormality, the first terminal executes the response plan; not only can key objects be cared for in a complex outdoor environment, but also cross-regional search and care of key objects can be achieved by using the second terminal and the second terminal management server. BRIEF DESCRIPTION OF THE DRAWINGS
[0092] Figure 1 is a schematic block diagram of a key object care system based on the Internet of Things provided by an embodiment of the present invention;
[0093] Figure 2 is a schematic block diagram of a first terminal provided by an embodiment of the present invention;
[0094] Figure 3 It is a flow chart of a key object care method based on the Internet of Things provided by another embodiment of the present invention. DETAILED DESCRIPTION
[0095] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.
[0096] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.
[0097] In the description of the present invention, the term "multiple" refers to two or more. Unless otherwise clearly defined, the orientation or positional relationship indicated by the terms "upper", "lower", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation of the present invention. The terms "connection", "installation", "fixation", etc. should be understood in a broad sense. For example, "connection" can be a fixed connection, a detachable connection, or an integral connection; it can be directly connected or indirectly connected through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to the specific circumstances. In addition, the terms "first", "second", etc. are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first", "second", etc. can explicitly or implicitly include one or more of the features. In the description of the present invention, unless otherwise specified, the meaning of "multiple" is two or more.
[0098] In the description of this specification, the description of the terms "one embodiment", "some implementations", "specific embodiments", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.
[0099] Refer to the following Figures 1 to 3 A key object care system and method based on the Internet of Things is described according to some embodiments of the present invention.
[0100] like Figure 1 and Figure 2 As shown, an embodiment of the present invention provides a key object care system based on the Internet of Things, including: a plurality of first terminals, at least one first terminal management server for managing the first terminals, a plurality of second terminals, at least one second terminal management server for managing the second terminals, and an upper-level server for managing the first terminal management server and the second terminal management server;
[0101] The first terminal includes a camera, a 5G communication module for receiving and sending data, a sound collection module, a positioning module, a light emitting module, an event processing module and a control processor;
[0102] The first terminal is configured as follows:
[0103] Acquire first data through the camera and the sound collection module, wherein the first data at least includes human behavior data, vehicle behavior data, animal behavior data and environmental data;
[0104] sending the first data to the first terminal management server;
[0105] The first terminal management server is configured as:
[0106] Using historical data to train a neural network to generate a key object neural network model, wherein the historical data at least includes human behavior historical data, vehicle behavior historical data, animal behavior historical data, and environmental historical data;
[0107] Determine whether the first data belongs to a preset key object according to a preset rule, and if not, send the first data to the upper-level server, which sends the first data to a second terminal management server managed by the upper-level server, and the second terminal management server determines whether the first data belongs to a preset key object;
[0108] If the first data belongs to a preset key object, inputting the first data into the key object neural network model;
[0109] It is understandable that in reality, there may be a certain object a active in area A, which is not an object that needs to be focused on in area A, but is an object that needs to be focused on in area B. For this situation, in an embodiment of the present invention, the first data is sent to the upper-level server, which is sent by the upper-level server to the second terminal management server managed by it (such as the terminal management server in area B), and the second terminal management server determines whether it belongs to the preset key object. When it is determined that this object a is a key object in area B, its data is sent to the key object neural network on the first terminal management server (i.e., the terminal management server in area A) for judgment.
[0110] Determine whether the key object is abnormal;
[0111] When the key object has an abnormality, generating an abnormality response plan, and sending the abnormality response plan to the first terminal;
[0112] The first terminal is configured to: receive the exception response plan, execute the exception response plan by the event processing module and feed back the execution result to the first terminal management server.
[0113] In an embodiment of the present invention, the first terminal and the second terminal may be smart street lamps, including at least a camera, a 5G communication module for receiving and sending data, a sound collection module, a positioning module, a lighting module, an event processing module, and a control processor, etc., and may also include an environmental monitoring module, an alarm module, etc. The first terminal management server and the second terminal management server are respectively used to manage the first terminal and the second terminal by region (such as by street or community); the upper-level server manages multiple first terminal management servers and second terminal management servers by larger region.
[0114] It can be understood that in order to provide more accurate and efficient care for key targets, the embodiments of the present invention not only focus on key personnel, but also focus on key objects, such as vehicles, animals, environments, etc. The historical data collected by the first terminal includes at least human behavior history data, vehicle behavior history data, animal behavior history data and environmental history data.
[0115] The determining according to preset rules whether the first data belongs to a preset key object may specifically be based on a preset key object label set and extracting labels from the first data to determine whether the first data belongs to a preset key object that requires special attention.
[0116] The event processing module includes a storage unit (not shown in the figure) for storing medicines and articles, a display unit (not shown in the figure) for playing videos, a fire extinguishing unit (not shown in the figure) for extinguishing fires, etc.
[0117] By adopting the technical solution of this embodiment, a key object care system based on the Internet of Things is provided with multiple first terminals, at least one first terminal management server for managing the first terminals, multiple second terminals, at least one second terminal management server for managing the second terminals, and a superior server for managing the first terminal management server and the second terminal management server. The first data is collected through the first terminal. When the first data belongs to a key object, the neural network on the first terminal management server determines whether there is an abnormality in the key object corresponding to the first data. When there is an abnormality, the first terminal executes the response plan; not only can key objects be cared for in a complex outdoor environment, but also cross-regional search and care of key objects can be achieved by using the second terminal and the second terminal management server.
[0118] It should be known that Figure 1 The block diagram of the key object care system based on the Internet of Things is shown for illustration only, and the number of modules shown therein does not limit the protection scope of the present invention.
[0119] In some possible implementations of the present invention, in order to train a neural network model of key objects with accurate and efficient recognition effect, the neural network includes an input layer, an initial layer, a simulated output layer, a secondary initial layer, a verification coefficient layer and an output layer;
[0120] The key object neural network model at least includes a key person neural network model, a key vehicle neural network model, a key animal neural network model and a key environment neural network model;
[0121] Wherein, in the step of using historical data to train the neural network to generate a key object neural network model, the first terminal management server is configured as follows:
[0122] The historical data is classified to generate a plurality of different data groups; for example, the data groups are divided into an elderly group, a child group, a truck group, a fierce dog group, an environment group, and the like.
[0123] The data set is sent as an input value to the input layer; that is, different data sets are input to obtain different training models.
[0124] The input layer is connected to the initial layer through matrix operation;
[0125] The initial layer obtains a first output result, activates the first output result, and sends the activated second output result to the simulation output layer;
[0126] In this step, the first output result is activated using an activation function. An activation function is a function added to an artificial neural network to help the network learn complex patterns in data, similar to a neuron-based model in the human brain. The activation function ultimately determines the content to be emitted to the next neuron. In an embodiment of the present invention, a Sigmoid activation function, a ReLU activation function, etc. may be used.
[0127] The simulation output layer obtains simulation output values through matrix calculation, and inputs the simulation output values into the secondary initial layer;
[0128] The secondary initial layer obtains the verification output result through matrix calculation;
[0129] Performing data connection between the input value of the input layer and the secondary initial layer;
[0130] The secondary initial layer obtains a secondary output result through matrix calculation, sends the secondary output result and the verification output result to the verification coefficient layer for verification, and obtains a normalization coefficient;
[0131] Sending the normalization coefficient and the simulated output value to the output layer, the output layer normalizes the simulated output value to obtain a mimicking result;
[0132] Collect positive and / or negative feedback information;
[0133] The mimicry result is learned and modified according to the positive feedback and / or the reverse feedback information to generate a key object neural network model.
[0134] In some possible implementations of the present invention, the coping scheme includes:
[0135] Provide emergency medicines required by the key targets; and / or,
[0136] Playing a first aid video corresponding to the current symptoms of the key subject; and / or,
[0137] Providing supplies for managing animals; and / or, Providing supplies to cope with current environmental changes.
[0138] It can be understood that in order to respond to the abnormalities occurring in the key objects in a timely manner, the control processor can control the event processing module to execute the response plan, which includes providing the emergency medicines required by the key objects, playing first aid videos corresponding to the current symptoms of the key objects, providing supplies for managing animals (pet cages, equipment for subduing animals, etc.), and providing supplies for coping with current environmental changes (such as umbrellas, masks, life jackets, etc.), one or more of the above.
[0139] For example, for an elderly person who is a focus of attention and has record information on the first terminal management server or the second terminal management server, when an abnormality is found in the elderly person, and the abnormality is an acute illness, the control processor can control the event processing module to issue a voice prompt, open the storage unit to provide the pre-stored emergency drugs to the elderly person who is ill, and also drive the display unit to play a first aid video for passers-by to provide first aid to the elderly person.
[0140] In some possible implementations of the present invention, the first terminal management server is configured with a key object monitoring model;
[0141] The first terminal management server is configured as:
[0142] Inputting the first data into the key subject monitoring model, and determining whether the first data matches the key subject monitoring model;
[0143] When the first data matches the key subject monitoring model, the first data is input into the key subject neural network model to determine whether there is an abnormality;
[0144] When the first data does not match the key object monitoring model, the key object monitoring model is sent to the upper-level server, and the upper-level server sends a collaborative search instruction to the second terminal management server managed by it;
[0145] The second terminal management server is configured as follows:
[0146] receiving the collaborative search instruction, and collecting second data using the second terminal;
[0147] Determining whether a key object corresponding to the key object monitoring model is found according to the second data;
[0148] The upper-level server is configured as follows:
[0149] When a key object corresponding to the key object monitoring model is found, controlling the second terminal management server to obtain current data of the key object and sending the current data to the first terminal management server;
[0150] The first terminal management server is configured to input the current data into the key object neural network model to determine whether there is an abnormality.
[0151] It is understandable that in practice, there may be such a situation: the key object a that has been active in area A for a long time, has not been collected by the first terminal in area A for a period of time. In order to avoid accidents, the first terminal management server in area A sends the key object module of the key object to the superior server, and the superior server searches in other areas managed by it (such as area C). The superior server sends the object model to the second terminal management server (i.e., the terminal management server in area C), controls the second terminal in area C to collect the second data, and when there is a match with the key object model in the second data, it means that the key object corresponding to the key object monitoring model is found. The superior server controls the second terminal through the second terminal management server to obtain the current data of the key object, and sends the current data to the first terminal management server to determine whether there is an abnormality. Through this embodiment, it can effectively prevent the loss of elderly people and children and the loss of valuables.
[0152] In some possible implementations of the present invention, the key subject care system further includes a key subject monitoring terminal, and the first terminal is configured as:
[0153] When the key object is detected, the location information of the key object is obtained through the positioning module;
[0154] Sending the location information to the first terminal management server;
[0155] The first terminal management server is configured as:
[0156] Navigation data is generated according to the location information and the location information of the key subject monitoring terminal, and is sent to the key subject monitoring terminal.
[0157] It can be understood that in order to better monitor key objects and arrive at the scene of the key objects in time to deal with abnormal events, the implementation mode of the present invention sets a key object monitoring terminal and determines the position of the key object through the positioning module of the first terminal, thereby realizing navigation between the key object monitoring terminal and the key object, and quickly finding the key object.
[0158] See also Figure 3 Another embodiment of the present invention provides a key object care method based on the Internet of Things, the key object care system based on the Internet of Things includes multiple first terminals, at least one first terminal management server for managing the first terminals, multiple second terminals, at least one second terminal management server for managing the second terminals, and an upper-level server for managing the first terminal management server and the second terminal management server; the first terminal includes a camera, a 5G communication module for receiving and sending data, a sound collection module, a positioning module, a light-emitting module, an event processing module, and a control processor, and the key object care method includes:
[0159] The first terminal acquires first data through the camera and the sound collection module, wherein the first data at least includes human behavior data, vehicle behavior data, animal behavior data and environmental data;
[0160] sending the first data to the first terminal management server;
[0161] The first terminal management server trains a neural network using historical data to generate a key object neural network model, wherein the historical data at least includes human behavior historical data, vehicle behavior historical data, animal behavior historical data and environmental historical data;
[0162] Determine whether the first data belongs to a preset key object according to a preset rule, and if not, send the first data to the upper-level server, which sends the first data to a second terminal management server managed by the upper-level server, and the second terminal management server determines whether the first data belongs to a preset key object;
[0163] If the first data belongs to a preset key object, inputting the first data into the key object neural network model;
[0164] It is understandable that in reality, there may be a certain object a active in area A, which is not an object that needs to be focused on in area A, but is an object that needs to be focused on in area B. For this situation, in an embodiment of the present invention, the first data is sent to the upper-level server, which is sent by the upper-level server to the second terminal management server managed by it (such as the terminal management server in area B), and the second terminal management server determines whether it belongs to the preset key object. When it is determined that this object a is a key object in area B, its data is sent to the key object neural network on the first terminal management server (i.e., the terminal management server in area A) for judgment.
[0165] Determine whether the key object is abnormal;
[0166] When the key object has an abnormality, generating an abnormality response plan, and sending the abnormality response plan to the first terminal;
[0167] The first terminal receives the abnormality response plan, executes the abnormality response plan and feeds back the execution result to the first terminal management server.
[0168] In an embodiment of the present invention, the first terminal and the second terminal may be smart street lamps, including at least a camera, a 5G communication module for receiving and sending data, a sound collection module, a positioning module, a lighting module, an event processing module, and a control processor, etc., and may also include an environmental monitoring module, an alarm module, etc. The first terminal management server and the second terminal management server are respectively used to manage the first terminal and the second terminal by region (such as by street or community); the upper-level server manages multiple first terminal management servers and second terminal management servers by larger region.
[0169] It can be understood that in order to provide more accurate and efficient care for key targets, the embodiments of the present invention not only focus on key personnel, but also focus on key objects, such as vehicles, animals, environments, etc. The historical data collected by the first terminal includes at least human behavior history data, vehicle behavior history data, animal behavior history data and environmental history data.
[0170] The determining according to preset rules whether the first data belongs to a preset key object may specifically be based on a preset key object label set and extracting labels from the first data to determine whether the first data belongs to a preset key object that requires special attention.
[0171] The event processing module includes a storage unit for storing medicines and articles, a display unit for playing videos, a fire extinguishing unit for extinguishing fires, etc.
[0172] In some possible implementations of the present invention, in order to train a neural network model of key objects with accurate and efficient recognition effect, the neural network includes an input layer, an initial layer, a simulated output layer, a secondary initial layer, a verification coefficient layer and an output layer;
[0173] The key object neural network model at least includes a key person neural network model, a key vehicle neural network model, a key animal neural network model and a key environment neural network model;
[0174] The specific steps of using historical data to train the neural network and generate the neural network model of the key object are as follows:
[0175] The historical data is classified to generate a plurality of different data groups; for example, the data groups are divided into an elderly group, a child group, a truck group, a fierce dog group, an environment group, and the like.
[0176] The data set is sent as an input value to the input layer; that is, different data sets are input to obtain different training models.
[0177] The input layer is connected to the initial layer through matrix operation;
[0178] The initial layer obtains a first output result, activates the first output result, and sends the activated second output result to the simulation output layer;
[0179] In this step, the first output result is activated using an activation function. An activation function is a function added to an artificial neural network to help the network learn complex patterns in data, similar to a neuron-based model in the human brain. The activation function ultimately determines the content to be emitted to the next neuron. In an embodiment of the present invention, a Sigmoid activation function, a ReLU activation function, etc. may be used.
[0180] The simulation output layer obtains simulation output values through matrix calculation, and inputs the simulation output values into the secondary initial layer;
[0181] The secondary initial layer obtains the verification output result through matrix calculation;
[0182] Performing data connection between the input value of the input layer and the secondary initial layer;
[0183] The secondary initial layer obtains a secondary output result through matrix calculation, sends the secondary output result and the verification output result to the verification coefficient layer for verification, and obtains a normalization coefficient;
[0184] Sending the normalization coefficient and the simulated output value to the output layer, the output layer normalizes the simulated output value to obtain a mimicking result;
[0185] Collect positive and / or negative feedback information;
[0186] The mimicry result is learned and modified according to the positive feedback and / or the reverse feedback information to generate a key object neural network model.
[0187] In some possible implementations of the present invention, the coping scheme includes:
[0188] Provide emergency medicines required by the key targets; and / or,
[0189] Playing a first aid video corresponding to the current symptoms of the key subject; and / or,
[0190] Providing supplies for managing animals; and / or, Providing supplies to cope with current environmental changes.
[0191] It can be understood that in order to respond to the abnormalities occurring in the key objects in a timely manner, the control processor can control the event processing module to execute the response plan, which includes providing the emergency medicines required by the key objects, playing first aid videos corresponding to the current symptoms of the key objects, providing supplies for managing animals (pet cages, equipment for subduing animals, etc.), and providing supplies for coping with current environmental changes (such as umbrellas, masks, life jackets, etc.), one or more of the above.
[0192] For example, for an elderly person who is a focus of attention and has record information on the first terminal management server or the second terminal management server, when an abnormality is found in the elderly person, and the abnormality is an acute illness, the control processor can control the event processing module to issue a voice prompt, open the storage unit to provide the pre-stored emergency drugs to the elderly person who is ill, and also drive the display unit to play a first aid video for passers-by to provide first aid to the elderly person.
[0193] In some possible implementations of the present invention, the first terminal management server is configured with a key object monitoring model;
[0194] The key object care method also includes the steps of:
[0195] The first terminal management server inputs the first data into the key object monitoring model to determine whether the first data matches the key object monitoring model;
[0196] When the first data matches the key subject monitoring model, the first data is input into the key subject neural network model to determine whether there is an abnormality;
[0197] When the first data does not match the key object monitoring model, the key object monitoring model is sent to the upper-level server, and the upper-level server sends a collaborative search instruction to the second terminal management server managed by it;
[0198] The second terminal management server receives the collaborative search instruction and uses the second terminal to collect second data;
[0199] Determining whether a key object corresponding to the key object monitoring model is found according to the second data;
[0200] When a key object corresponding to the key object monitoring model is found, the upper server controls the second terminal management server to obtain current data of the key object, and sends the current data to the first terminal management server;
[0201] The first terminal management server inputs the current data into the key object neural network model to determine whether there is an abnormality.
[0202] It is understandable that in practice, there may be such a situation: the key object a that has been active in area A for a long time, has not been collected by the first terminal in area A for a period of time. In order to avoid accidents, the first terminal management server in area A sends the key object module of the key object to the superior server, and the superior server searches in other areas managed by it (such as area C). The superior server sends the object model to the second terminal management server (i.e., the terminal management server in area C), controls the second terminal in area C to collect the second data, and when there is a match with the key object model in the second data, it means that the key object corresponding to the key object monitoring model is found. The superior server controls the second terminal through the second terminal management server to obtain the current data of the key object, and sends the current data to the first terminal management server to determine whether there is an abnormality. Through this embodiment, it can effectively prevent the loss of elderly people and children and the loss of valuables.
[0203] In some possible implementations of the present invention, the key subject care system further includes a key subject monitoring terminal, and the key subject care method further includes the steps of:
[0204] When the key object is detected, the first terminal obtains its own location information through the positioning module;
[0205] Sending the location information to the first terminal management server;
[0206] The first terminal management server generates navigation data according to the location information and the location information of the key object monitoring terminal, and sends the navigation data to the key object monitoring terminal.
[0207] It can be understood that in order to better monitor key objects and arrive at the scene of the key objects in time to deal with abnormal events, the implementation mode of the present invention sets a key object monitoring terminal and determines the position of the key object through the positioning module of the first terminal, thereby realizing navigation between the key object monitoring terminal and the key object, and quickly finding the key object.
[0208] By adopting the technical solution of this embodiment, a key object care system based on the Internet of Things is provided with multiple first terminals, at least one first terminal management server for managing the first terminals, multiple second terminals, at least one second terminal management server for managing the second terminals, and a superior server for managing the first terminal management server and the second terminal management server. The first data is collected through the first terminal. When the first data belongs to a key object, the neural network on the first terminal management server determines whether there is an abnormality in the key object corresponding to the first data. When there is an abnormality, the first terminal executes the response plan; not only can key objects be cared for in a complex outdoor environment, but also cross-regional search and care of key objects can be achieved by using the second terminal and the second terminal management server.
[0209] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.
[0210] According to the embodiments of the present invention as described above, these embodiments do not describe all the details in detail, nor do they limit the invention to the specific embodiments described. Obviously, many modifications and changes can be made based on the above description. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can make good use of the present invention and the modified use based on the present invention. The present invention is limited only by the claims and their full scope and equivalents.
Claims
1. A key object care system based on the Internet of Things, characterized in that: include: A plurality of first terminals, at least one first terminal management server for managing the first terminals, a plurality of second terminals, at least one second terminal management server for managing the second terminals, and an upper-level server for managing the first terminal management server and the second terminal management server; The first terminal includes a camera, a 5G communication module for receiving and sending data, a sound collection module, a positioning module, a light emitting module, an event processing module and a control processor; The first terminal is configured as follows: Acquire first data through the camera and the sound collection module, wherein the first data at least includes human behavior data, vehicle behavior data, animal behavior data and environmental data; sending the first data to the first terminal management server; The first terminal management server is configured as follows: Using historical data to train a neural network to generate a key object neural network model, wherein the historical data at least includes human behavior historical data, vehicle behavior historical data, animal behavior historical data, and environmental historical data; Determine whether the first data belongs to a preset key object according to a preset rule, and if not, send the first data to the upper-level server, which sends the first data to a second terminal management server managed by the upper-level server, and the second terminal management server determines whether the first data belongs to a preset key object; If the first data belongs to a preset key object, inputting the first data into the key object neural network model; Determine whether the key object is abnormal; When the key object has an abnormality, generating an abnormality response plan, and sending the abnormality response plan to the first terminal; The first terminal is configured to: receive the abnormality response plan, execute the abnormality response plan and feed back the execution result to the first terminal management server; The first terminal management server is configured with a key object monitoring model; the first terminal management server is also configured as: Inputting the first data into the key subject monitoring model, and determining whether the first data matches the key subject monitoring model; When the first data matches the key subject monitoring model, the first data is input into the key subject neural network model to determine whether there is an abnormality; When the first data does not match the key object monitoring model, the key object monitoring model is sent to the upper-level server, and the upper-level server sends a collaborative search instruction to the second terminal management server managed by it; The second terminal management server is configured as follows: receiving the collaborative search instruction, and collecting second data using the second terminal; Determining whether a key object corresponding to the key object monitoring model is found according to the second data; The upper-level server is configured as follows: When a key object corresponding to the key object monitoring model is found, controlling the second terminal management server to obtain current data of the key object and sending the current data to the first terminal management server; The first terminal management server is configured to input the current data into the key object neural network model to determine whether there is an abnormality.
2. The key object care system based on the Internet of Things according to claim 1 is characterized in that: The neural network includes an input layer, an initial layer, a simulated output layer, a secondary initial layer, a verification coefficient layer and an output layer; The key object neural network model at least includes a key person neural network model, a key vehicle neural network model, a key animal neural network model and a key environment neural network model; Wherein, in the step of using historical data to train the neural network to generate a key object neural network model, the first terminal management server is configured as follows: Classifying the historical data to generate a plurality of different data groups; Sending the data set as input value to the input layer; The input layer is connected to the initial layer through matrix operation; The initial layer obtains a first output result, activates the first output result, and sends the activated second output result to the simulation output layer; The simulation output layer obtains simulation output values through matrix calculation, and inputs the simulation output values into the secondary initial layer; The secondary initial layer obtains the verification output result through matrix calculation; Performing data connection between the input value of the input layer and the secondary initial layer; The secondary initial layer obtains a secondary output result through matrix calculation, sends the secondary output result and the verification output result to the verification coefficient layer for verification, and obtains a normalization coefficient; Sending the normalization coefficient and the simulated output value to the output layer, the output layer normalizes the simulated output value to obtain a mimicking result; Collect positive and / or negative feedback information; The mimicry result is learned and modified according to the positive feedback and / or the reverse feedback information to generate a key object neural network model.
3. The key object care system based on the Internet of Things according to claim 2 is characterized in that: The response options include: Provide emergency medicines required by the key targets; and / or, Playing a first aid video corresponding to the current symptoms of the key subject; and / or, Providing supplies for managing animals; and / or, Providing supplies to cope with current environmental changes.
4. The key object care system based on the Internet of Things according to any one of claims 1 to 3, characterized in that: The key object care system further includes a key object monitoring terminal, wherein the first terminal is configured as follows: When the key object is detected, the location information of the key object is obtained through the positioning module; Sending the location information to the first terminal management server; The first terminal management server is configured as follows: Navigation data is generated according to the location information and the location information of the key subject monitoring terminal, and is sent to the key subject monitoring terminal.
5. A key object care method based on the Internet of Things, characterized in that: The invention is applied to a key object care system based on the Internet of Things, the key object care system based on the Internet of Things includes a plurality of first terminals, at least one first terminal management server for managing the first terminals, a plurality of second terminals, at least one second terminal management server for managing the second terminals, and an upper-level server for managing the first terminal management server and the second terminal management server; the first terminal includes a camera, a 5G communication module for receiving and sending data, a sound collection module, a positioning module, a light-emitting module, an event processing module, and a control processor, and the key object care method includes: The first terminal acquires first data through the camera and the sound collection module, wherein the first data at least includes human behavior data, vehicle behavior data, animal behavior data and environmental data; sending the first data to the first terminal management server; The first terminal management server trains a neural network using historical data to generate a key object neural network model, wherein the historical data at least includes human behavior historical data, vehicle behavior historical data, animal behavior historical data and environmental historical data; Determine whether the first data belongs to a preset key object according to a preset rule, and if not, send the first data to the upper-level server, which sends the first data to a second terminal management server managed by the upper-level server, and the second terminal management server determines whether the first data belongs to a preset key object; If the first data belongs to a preset key object, inputting the first data into the key object neural network model; Determine whether the key object is abnormal; When the key object has an abnormality, generating an abnormality response plan, and sending the abnormality response plan to the first terminal; The first terminal receives the abnormality response plan, executes the abnormality response plan and feeds back the execution result to the first terminal management server; Wherein, the first terminal management server is configured with a key object monitoring model; the key object care method further comprises the steps of: The first terminal management server inputs the first data into the key object monitoring model to determine whether the first data matches the key object monitoring model; When the first data matches the key subject monitoring model, the first data is input into the key subject neural network model to determine whether there is an abnormality; When the first data does not match the key object monitoring model, the key object monitoring model is sent to the upper-level server, and the upper-level server sends a collaborative search instruction to the second terminal management server managed by it; The second terminal management server receives the collaborative search instruction and uses the second terminal to collect second data; Determining whether a key object corresponding to the key object monitoring model is found according to the second data; When a key object corresponding to the key object monitoring model is found, the upper server controls the second terminal management server to obtain current data of the key object, and sends the current data to the first terminal management server; The first terminal management server inputs the current data into the key object neural network model to determine whether there is an abnormality.
6. The key object care method based on the Internet of Things according to claim 5 is characterized in that: The neural network includes an input layer, an initial layer, a simulated output layer, a secondary initial layer, a verification coefficient layer and an output layer; The key object neural network model at least includes a key person neural network model, a key vehicle neural network model, a key animal neural network model and a key environment neural network model; The specific steps of using historical data to train the neural network and generate the neural network model of the key object are as follows: Classifying the historical data to generate a plurality of different data groups; Sending the data set as input value to the input layer; The input layer is connected to the initial layer through matrix operation; The initial layer obtains a first output result, activates the first output result, and sends the activated second output result to the simulation output layer; The simulation output layer obtains simulation output values through matrix calculation, and inputs the simulation output values into the secondary initial layer; The secondary initial layer obtains the verification output result through matrix calculation; Performing data connection between the input value of the input layer and the secondary initial layer; The secondary initial layer obtains a secondary output result through matrix calculation, sends the secondary output result and the verification output result to the verification coefficient layer for verification, and obtains a normalization coefficient; Sending the normalization coefficient and the simulated output value to the output layer, the output layer normalizes the simulated output value to obtain a mimicking result; Collect positive and / or negative feedback information; The mimicry result is learned and modified according to the positive feedback and / or the reverse feedback information to generate a key object neural network model.
7. The key object care method based on the Internet of Things according to claim 6 is characterized in that: The response options include: Provide emergency medicines required by the key targets; and / or, Playing a first aid video corresponding to the current symptoms of the key subject; and / or, Providing supplies for managing animals; and / or, Providing supplies to cope with current environmental changes.
8. The key object care method based on the Internet of Things according to any one of claims 5 to 7, characterized in that: The key object care system also includes a key object monitoring terminal, and the key object care method also includes the steps of: When the key object is detected, the first terminal obtains its own location information through the positioning module; Sending the location information to the first terminal management server; The first terminal management server generates navigation data according to the location information and the location information of the key object monitoring terminal, and sends the navigation data to the key object monitoring terminal.
