False injury prevention reminding method, reminding device and smart home system

Abstract

The application provides a false injury prevention reminding method, a reminding device and a smart home system. The method comprises the following steps: acquiring a plurality of historical image data; acquiring indoor image data at a current time to obtain target indoor image data; inputting the target indoor image data into an AI model to obtain a target prediction result; in the case that the target prediction result represents that the probability of a target person contacting a sharp object is greater than a first predetermined value, controlling a signal emitting device to emit a reminding signal, the reminding signal is used to indicate the probability of the target person contacting the sharp object, and the emitting form of the reminding signal comprises at least one of the following: sound, light and electrical signal. The application solves the problem in the prior art that children are easy to contact sharp objects and cause harm due to the inability to monitor and warn children in real time in the dangerous situation of contacting sharp objects.

Description

Technical Field

[0001] This application relates to the field of smart home technology, and more specifically, to a reminder method, reminder device, computer-readable storage medium, and smart home system for preventing accidental injury. Background Technology

[0002] With social development and improved living standards, people are paying increasing attention to home safety. This is especially true for families with children, as children lack the ability to identify dangerous objects and protect themselves, making them susceptible to injury from sharp objects. Therefore, effectively preventing and mitigating potential injuries to children in indoor environments has become a pressing issue.

[0003] Traditional indoor safety measures primarily rely on physical methods, such as placing sharp objects out of children's reach or using safety boxes. However, these methods cannot completely prevent children from coming into contact with sharp objects, nor can they monitor and provide real-time warnings of potential dangers. Summary of the Invention

[0004] The main objective of this application is to provide a reminder method, reminder device, computer-readable storage medium, and smart home system to prevent accidental injury, so as to at least solve the problem in the prior art that children are prone to contact with sharp objects and suffer injuries due to the inability to monitor and warn of dangerous situations involving children coming into contact with sharp objects in real time.

[0005] To achieve the above objectives, according to one aspect of this application, a method for preventing accidental injury is provided, comprising: acquiring multiple historical image data, the historical image data including indoor historical object image data and historical target person image data, the historical object image data including image data of at least one sharp object, the target person including a child, the sharp object being an object having an end whose length in a first direction is 10 times or more of its length in a second direction, the first direction intersecting the second direction; acquiring indoor image data at the current moment to obtain target indoor image data, the indoor image data including indoor object image data and indoor target person image data; and inputting the target indoor image data into an AI (Artificial Intelligence Processing) system. In an AI (Artificial Intelligence) model, a target prediction result is obtained. The AI ​​model is a model built based on multiple historical image data. The AI ​​model is used to identify the sharp object and output a prediction result, which represents the probability that the target person will come into contact with the sharp object. If the target prediction result indicates that the probability that the target person will come into contact with the sharp object is greater than a first predetermined value, a control signal emitting device emits a reminder signal. The reminder signal is used to indicate the probability that the target person will come into contact with the sharp object. The form of the reminder signal includes at least one of the following: sound, light, and electrical signal.

[0006] Optionally, after acquiring multiple historical image data and before inputting the target indoor image data into the AI ​​model, the method further includes: establishing an initial AI model using deep learning and computer vision techniques; inputting the multiple historical image data into the initial AI model to train the initial AI model; and optimizing the parameters of the initial AI model using an optimization algorithm to obtain the AI ​​model, wherein the optimization algorithm includes at least one of stochastic gradient descent algorithm and Adam algorithm.

[0007] Optionally, after acquiring multiple historical image data and before inputting the multiple historical image data into the initial AI model, the method further includes: labeling the sharp objects and the target person in the multiple historical image data.

[0008] Optionally, if the target prediction result indicates that the probability of the target person contacting the sharp object is greater than a first predetermined value, the signal emitting device is controlled to issue a warning signal, including: if the target prediction result indicates that the probability of the target person contacting the sharp object is greater than the first predetermined value and less than or equal to a second predetermined value, controlling the signal emitting device to issue a first warning signal; if the target prediction result indicates that the probability of the target person contacting the sharp object is greater than the second predetermined value and less than or equal to a third predetermined value, controlling the signal emitting device to issue a second warning signal; and if the target prediction result indicates that the probability of the target person contacting the sharp object is greater than the third predetermined value and less than or equal to a fourth predetermined value, controlling the signal emitting device to issue a third warning signal; wherein the signal strength of the first warning signal is less than the signal strength of the second warning signal, and the signal strength of the second warning signal is less than the signal strength of the third warning signal.

[0009] Optionally, after obtaining the target prediction result, the method further includes: training the AI ​​model using online learning based on the target indoor image data and the target prediction result to obtain the trained AI model.

[0010] Optionally, after obtaining the target prediction result, the method further includes: training the AI ​​model using a fine-tuning method based on the target indoor image data and the target prediction result to obtain the trained AI model.

[0011] Optionally, obtaining indoor image data at the current moment to obtain target indoor image data includes: obtaining the indoor image data at the current moment through a smart camera to obtain the target indoor image data.

[0012] According to another aspect of this application, a reminder device for preventing accidental injury is provided, comprising: a first acquisition unit, configured to acquire multiple historical image data, the historical image data including historical object image data and historical target person image data, the historical object image data including image data of at least one sharp object, the target person including a child, the sharp object being an object having an end having a length in a first direction that is 10 times or more the length in a second direction, the first direction intersecting the second direction; and a second acquisition unit, configured to acquire indoor image data at the current moment to obtain target indoor image data, the indoor image data including indoor object image data and an image of the target person indoors. The system includes: a data input unit for inputting the target indoor image data into an AI model to obtain a target prediction result, wherein the AI ​​model is a model established based on multiple historical image data, the AI ​​model is used to identify the sharp object, and output a prediction result, wherein the prediction result represents the probability that the target person will come into contact with the sharp object; and a control unit for controlling a signal emitting device to emit a warning signal when the target prediction result indicates that the probability that the target person will come into contact with the sharp object is greater than a first predetermined value, wherein the warning signal is used to indicate the probability that the target person will come into contact with the sharp object, and the form of the warning signal includes at least one of the following: sound, light, and electrical signal.

[0013] According to another aspect of this application, a computer-readable storage medium is provided, the computer-readable storage medium including a stored program, wherein, when the program is executed, it controls the device where the computer-readable storage medium is located to perform any of the aforementioned anti-accidental injury reminder methods.

[0014] According to another aspect of this application, a smart home system is provided, comprising: a signal emitting device; a controller for the signal emitting device, comprising: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs including a reminder method for any of the aforementioned methods for preventing accidental injury.

[0015] The technical solution of this application first acquires multiple historical image data, then acquires indoor image data at the current moment to obtain target indoor image data. This target image data is then input into an AI model built based on the multiple historical image data to obtain a target prediction result representing the probability of a target person coming into contact with a sharp object. Finally, if the target prediction result indicates that the probability of the target person coming into contact with a sharp object is greater than a first predetermined value, a control signal transmitting device issues a warning signal. Compared to traditional indoor safety measures in the prior art, which mainly rely on physical means and cannot monitor and warn of the danger of children coming into contact with sharp objects in real time, leading to children easily coming into contact with sharp objects and causing injury, this application inputs the target indoor image data into an AI model used to identify sharp objects and output prediction results to obtain target prediction results. If the target prediction result indicates that the probability of the target person coming into contact with a sharp object is greater than a first predetermined value, a control signal transmitting device issues a warning signal to prevent and mitigate potential harm to children. Attached Figure Description

[0016] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0017] Figure 1 A hardware structure block diagram of a mobile terminal for performing a reminder method to prevent accidental injury, according to an embodiment of this application, is shown.

[0018] Figure 2 A flowchart illustrating a method for preventing accidental injury according to an embodiment of this application is shown.

[0019] Figure 3 A flowchart illustrating a specific method for preventing accidental injury according to an embodiment of this application is shown.

[0020] Figure 4 A structural block diagram of a reminder device for preventing accidental injury according to an embodiment of this application is shown.

[0021] The above figures include the following reference numerals:

[0022] 102. Processor; 104. Memory; 106. Transmission device; 108. Input / output device. Detailed Implementation

[0023] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0024] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0025] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of this application described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0026] As described in the background section, the existing technology cannot monitor and warn of the danger of children coming into contact with sharp objects in real time, which makes it easy for children to come into contact with sharp objects and cause injury. In order to solve the above problems, the embodiments of this application provide a reminder method, reminder device, computer-readable storage medium and smart home system to prevent accidental injury.

[0027] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0028] The methods and embodiments provided in this application can be executed on a mobile terminal, a computer terminal, or a similar computing device. Taking running on a mobile terminal as an example, Figure 1 This is a hardware structure block diagram of a mobile terminal for an embodiment of the present invention's method for preventing accidental injury. Figure 1 As shown, a mobile terminal may include one or more ( Figure 1 Only one is shown in the diagram. A processor 102 (which may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.) and a memory 104 for storing data are also shown. The mobile terminal may further include a transmission device 106 for communication functions and an input / output device 108. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the mobile terminal described above. For example, the mobile terminal may also include components that are more... Figure 1 The more or fewer components shown, or having the same Figure 1The different configurations shown.

[0029] The memory 104 can be used to store computer programs, such as application software programs and modules, like the computer program corresponding to the anti-accidental injury reminder method in this embodiment of the invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, thereby implementing the above-described method. The memory 104 may include high-speed random access memory and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the mobile terminal via a network. Examples of the aforementioned networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof. The transmission device 106 is used to receive or send data via a network. Specific examples of the aforementioned networks may include wireless networks provided by the mobile terminal's communication provider. In one example, the transmission device 106 includes a network interface controller (NIC), which can be connected to other network devices via a base station to communicate with the Internet. In one example, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0030] This embodiment provides a method for preventing accidental injury to users by running on a mobile terminal, computer terminal, or similar computing device. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0031] Figure 2 This is a flowchart of a reminder method for preventing accidental injury according to an embodiment of this application. Figure 2 As shown, the method includes the following steps:

[0032] Step S201: Acquire multiple historical image data, including indoor historical object image data and historical target person image data. The historical object image data includes image data of at least one sharp object. The target person includes a child. The sharp object is an object with an end whose length in a first direction is 10 times or more the length in a second direction. The first direction intersects with the second direction.

[0033] Specifically, the target individuals are not limited to children, but may also include blind people, people with intellectual disabilities, or other types of people, and this application does not impose specific restrictions on them; the sharp objects mentioned above include, but are not limited to, scissors, knives, and needles.

[0034] Specifically, the length of the aforementioned end of the sharp object in the first direction is much greater than its length in the second direction.

[0035] Step S202: Obtain indoor image data at the current moment to obtain target indoor image data. The indoor image data includes indoor object image data and indoor target person image data.

[0036] Step S203: Input the above-mentioned indoor image data of the target into the AI ​​model to obtain the target prediction result. The AI ​​model is a model built based on multiple historical image data. The AI ​​model is used to identify the sharp object and output the prediction result. The prediction result is a result that characterizes the probability of the target person coming into contact with the sharp object.

[0037] Step S204: If the target prediction result indicates that the probability of the target person contacting the sharp object is greater than a first predetermined value, the control signal emitting device emits a reminder signal. The reminder signal is used to indicate the probability of the target person contacting the sharp object. The form of the reminder signal includes at least one of the following: sound, light, and electrical signal.

[0038] Specifically, the aforementioned signal transmitting devices include, but are not limited to, alarms, light-emitting devices, and mobile phones. Those skilled in the art can flexibly select appropriate signal transmitting devices according to actual needs, and this application does not impose specific restrictions in this regard.

[0039] In practical applications, those skilled in the art can set the above-mentioned first predetermined value based on experience, or obtain it through multiple experiments. This application does not impose any specific restrictions on this.

[0040] Through the above embodiments, multiple historical image data are first acquired, then indoor image data at the current moment is acquired to obtain target indoor image data. This target image data is then input into an AI model built based on the multiple historical image data to obtain a target prediction result representing the probability of a target person coming into contact with a sharp object. Finally, if the target prediction result indicates that the probability of the target person coming into contact with a sharp object is greater than a first predetermined value, a control signal emitting device issues a warning signal. Compared to traditional indoor safety measures in the prior art, which mainly rely on physical means and cannot monitor and warn of the danger of children coming into contact with sharp objects in real time, leading to children easily coming into contact with sharp objects and causing injury, this application inputs the target indoor image data into an AI model used to identify sharp objects and output prediction results to obtain a target prediction result. If the target prediction result indicates that the probability of the target person coming into contact with a sharp object is greater than a first predetermined value, a control signal emitting device issues a warning signal to prevent and mitigate potential harm to children.

[0041] In one alternative approach, after acquiring multiple historical image data points and before inputting the target indoor image data into the AI ​​model, the method further includes: establishing an initial AI model using deep learning and computer vision techniques; inputting the multiple historical image data points into the initial AI model to train it; and optimizing the parameters of the initial AI model using an optimization algorithm to obtain the AI ​​model. The optimization algorithm includes at least one of stochastic gradient descent and the Adam algorithm. In this embodiment, an initial AI model is established using deep learning and computer vision techniques, and multiple historical image data points are input into the initial AI model to train it. The parameters of the initial AI model are then optimized using an optimization algorithm to obtain the AI ​​model. This prepares the system for subsequently using the AI ​​model to output target prediction results to control the signal-emitting device to issue a warning signal. Furthermore, optimizing the model parameters using an optimization algorithm ensures that the obtained AI model can accurately determine the probability of a target person contacting a sharp object based on the input image data, thus ensuring that the target prediction results obtained subsequently using the AI ​​model are relatively accurate.

[0042] Specifically, the above optimization algorithms are not limited to stochastic gradient descent and Adam algorithms, but can also be batch gradient descent, Newton's algorithm, conjugate gradient algorithm, or other algorithms. Those skilled in the art can flexibly choose appropriate optimization algorithms according to actual needs, and this application does not impose specific restrictions in this regard.

[0043] Specifically, in addition to the AI ​​models mentioned above, convolutional neural networks can be used for image classification and object detection, while recurrent neural networks can be used for sequence prediction and time series analysis.

[0044] In other embodiments, after acquiring multiple historical image data sets, and before inputting these historical image data sets into the initial AI model, the method further includes labeling the sharp objects and target individuals in the multiple historical image data sets. In this embodiment, labeling the sharp objects and target individuals in the multiple historical image data sets before inputting them into the AI ​​model further ensures better training results and accuracy for the model, preparing for the subsequent use of the obtained AI model to identify sharp objects and output target prediction results representing the probability of a target individual contacting a sharp object.

[0045] In other embodiments, after acquiring multiple historical image data sets and before inputting these historical image data sets into the initial AI model, the method further includes: performing data cleaning on the multiple historical image data sets. In this embodiment, by performing data cleaning on the historical image data sets, irrelevant information in the historical image data can be eliminated, effectively simplifying the data and preparing for the subsequent establishment of the AI ​​model.

[0046] Specifically, in addition to image annotation and data cleaning of historical image data, preprocessing operations such as grayscale transformation, image smoothing, lens distortion correction, and geometric correction can be performed on historical image data to restore useful real information, enhance the detectability of relevant information, simplify data to the maximum extent, and thus improve the reliability of feature extraction, image segmentation, matching, and recognition.

[0047] According to some exemplary embodiments of this application, when the target prediction result indicates that the probability of the target person contacting the sharp object is greater than a first predetermined value, the control signal emitting device emits a warning signal, including: when the target prediction result indicates that the probability of the target person contacting the sharp object is greater than the first predetermined value and less than or equal to a second predetermined value, controlling the signal emitting device to emit a first warning signal; when the target prediction result indicates that the probability of the target person contacting the sharp object is greater than the second predetermined value and less than or equal to a third predetermined value, controlling the signal emitting device to emit a second warning signal; and when the target prediction result indicates that the probability of the target person contacting the sharp object is greater than the third predetermined value and less than or equal to a fourth predetermined value, controlling the signal emitting device to emit a third warning signal; wherein the signal strength of the first warning signal is less than the signal strength of the second warning signal, and the signal strength of the second warning signal is less than the signal strength of the third warning signal. In this embodiment, when the target prediction result indicates that the probability of the target person contacting the sharp object falls within different ranges, the control signal emitting device emits warning signals with different signal strengths, so that the recipient of the warning signal can more clearly determine the current level of danger and take corresponding measures, further achieving the effect of preventing children from being harmed.

[0048] In practical applications, those skilled in the art can set the above-mentioned second predetermined value, third predetermined value and fourth predetermined value based on experience, or obtain them through multiple experiments. This application does not impose any specific restrictions on this.

[0049] In practical applications, those skilled in the art can set the signal strength of the first reminder signal, the second reminder signal, and the third reminder signal based on experience, or obtain them through multiple experiments. This application does not impose any specific limitations on this.

[0050] Specifically, when the reminder signal is emitted as sound, the stronger the signal strength, the louder and longer the sound lasts; when the reminder signal is emitted as light, the stronger the signal strength, the brighter the light; when the reminder signal is emitted as light, different colors of light can be used to represent different signal strengths; when the reminder signal is emitted as an electrical signal, the stronger the signal strength, the stronger the electrical signal; for example, when the signal emitting device is a mobile phone, the stronger the signal strength, the louder the vibration or ringing sound of the mobile application, and the longer the vibration or ringing sound lasts.

[0051] In another exemplary embodiment, after obtaining the target prediction result, the method further includes: training the AI ​​model using online learning based on the target indoor image data and the target prediction result to obtain the trained AI model. In this embodiment, training the AI ​​model using online learning based on the target indoor image data and the target prediction result further ensures that the AI ​​model has good generalization ability and adaptability.

[0052] According to some other exemplary embodiments of this application, after obtaining the target prediction result, the method further includes: training the AI ​​model using a fine-tuning method based on the target indoor image data and the target prediction result to obtain the trained AI model. In this embodiment, training the AI ​​model using a fine-tuning method based on the target indoor image data and the target prediction result further ensures better accuracy and reliability of the AI ​​model.

[0053] In some alternative embodiments of this application, obtaining indoor image data at the current moment to obtain target indoor image data includes: obtaining the aforementioned indoor image data at the current moment through a smart camera to obtain the aforementioned target indoor image data. In this embodiment, obtaining indoor image data at the current moment through a smart camera to obtain target indoor image data facilitates subsequent input of the target image data into an AI model to obtain target prediction results.

[0054] Of course, in addition to the smart cameras mentioned above, other devices that can collect image data in real time can also be used to collect indoor image data at the current moment. Those skilled in the art can flexibly choose appropriate acquisition devices according to actual needs, and this application does not impose specific restrictions on this.

[0055] To enable those skilled in the art to better understand the technical solution of this application, the implementation process of the anti-accidental injury reminder method of this application will be described in detail below with reference to specific embodiments.

[0056] This embodiment relates to a specific reminder method to prevent accidental injury, such as... Figure 3 As shown, it includes the following steps:

[0057] Step S1: Acquire multiple historical image data, including indoor historical object image data and historical target person image data. The historical object image data includes image data of at least one sharp object. The target person includes a child. The sharp object is an object with an end whose length in the first direction is 10 times or more the length in the second direction. The first direction intersects the second direction.

[0058] Step S2: Build an initial AI model using deep learning and computer vision technology. Input multiple historical image data into the initial AI model to train it. Optimize the parameters of the initial AI model using an optimization algorithm to obtain the AI ​​model. The optimization algorithm includes at least one of the stochastic gradient descent algorithm and the Adam algorithm.

[0059] Step S3: Obtain indoor image data at the current moment through the smart camera to obtain target indoor image data, which includes indoor object image data and indoor target person image data;

[0060] Step S4: Input the indoor image data of the target into the AI ​​model to obtain the target prediction result. The AI ​​model is used to identify sharp objects and output the prediction result, which represents the probability that the target person will come into contact with the sharp object.

[0061] Step S5: If the target prediction result indicates that the probability of the target person coming into contact with a sharp object is greater than a first predetermined value, the control signal emitting device emits a reminder signal. The reminder signal is used to indicate the probability of the target person coming into contact with a sharp object. The form of the reminder signal includes at least one of the following: sound, light, and electrical signal.

[0062] Step S6: Based on the target indoor image data and target prediction results, train the AI ​​model using online learning or fine-tuning to obtain the trained AI model.

[0063] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0064] This application also provides a reminder device to prevent accidental injury. It should be noted that the reminder device for preventing accidental injury in this application embodiment can be used to execute the reminder method for preventing accidental injury provided in this application embodiment. This device is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0065] The following describes the reminder device for preventing accidental injury provided in the embodiments of this application.

[0066] Figure 4 This is a schematic diagram of a reminder device for preventing accidental injury according to an embodiment of this application. Figure 4 As shown, the device includes:

[0067] The first acquisition unit 10 is used to acquire multiple historical image data, including indoor historical object image data and historical target person image data. The historical object image data includes image data of at least one sharp object. The target person includes a child. The sharp object is an object with an end whose length in a first direction is 10 times or more the length in a second direction. The first direction intersects with the second direction.

[0068] Specifically, the target individuals are not limited to children, but may also include blind people, people with intellectual disabilities, or other types of people, and this application does not impose specific restrictions on them; the sharp objects mentioned above include, but are not limited to, scissors, knives, and needles.

[0069] Specifically, the length of the aforementioned end of the sharp object in the first direction is much greater than its length in the second direction.

[0070] The second acquisition unit 20 is used to acquire indoor image data at the current moment and obtain target indoor image data. The indoor image data includes indoor object image data and indoor target person image data.

[0071] The input unit 30 is used to input the target indoor image data into the AI ​​model to obtain the target prediction result. The AI ​​model is a model established based on multiple historical image data. The AI ​​model is used to identify the sharp object and output the prediction result. The prediction result is a result that characterizes the probability that the target person comes into contact with the sharp object.

[0072] The control unit 40 is configured to control the signal transmitting device to issue a reminder signal when the target prediction result indicates that the probability of the target person contacting the sharp object is greater than a first predetermined value. The reminder signal is used to indicate the probability of the target person contacting the sharp object. The form of the reminder signal includes at least one of the following: sound, light, and electrical signal.

[0073] Specifically, the aforementioned signal transmitting devices include, but are not limited to, alarms, light-emitting devices, and mobile phones. Those skilled in the art can flexibly select appropriate signal transmitting devices according to actual needs, and this application does not impose specific restrictions in this regard.

[0074] In practical applications, those skilled in the art can set the above-mentioned first predetermined value based on experience, or obtain it through multiple experiments. This application does not impose any specific restrictions on this.

[0075] Through the above embodiments, a first acquisition unit acquires multiple historical image data, and a second acquisition unit acquires indoor image data at the current moment to obtain target indoor image data. The target image data is then input into an AI model built based on the multiple historical image data by an input unit to obtain a target prediction result representing the probability of a target person coming into contact with a sharp object. When the target prediction result indicates that the probability of the target person coming into contact with a sharp object is greater than a first predetermined value, the control unit controls a signal-emitting device to issue a warning signal. Compared to traditional indoor safety measures in the prior art that mainly rely on physical means and cannot monitor and warn of the danger of children coming into contact with sharp objects in real time, leading to children easily coming into contact with sharp objects and causing injury, this application inputs the target indoor image data into an AI model used to identify sharp objects and output prediction results to obtain a target prediction result. When the target prediction result indicates that the probability of the target person coming into contact with a sharp object is greater than a first predetermined value, the control signal-emitting device issues a warning signal to prevent and mitigate potential harm to children.

[0076] In one optional embodiment, the apparatus further includes: a modeling unit, configured to establish an initial AI model using deep learning and computer vision techniques after acquiring multiple historical image data and before inputting the target indoor image data into the AI ​​model; and a first training unit, configured to input the multiple historical image data into the initial AI model to train the initial AI model, and optimize the parameters of the initial AI model using an optimization algorithm to obtain the AI ​​model, wherein the optimization algorithm includes at least one of stochastic gradient descent and Adam algorithms. In this embodiment, an initial AI model is established using deep learning and computer vision techniques, and multiple historical image data are input into the initial AI model to train it. The parameters of the initial AI model are then optimized using an optimization algorithm to obtain the AI ​​model. This prepares the device for subsequently using the AI ​​model to output target prediction results to control the signal-emitting device to issue a warning signal. Furthermore, optimizing the model parameters using an optimization algorithm ensures that the obtained AI model can accurately determine the probability of a target person contacting a sharp object based on the input image data, thus ensuring that the target prediction results obtained subsequently using the AI ​​model are relatively accurate.

[0077] Specifically, the above optimization algorithms are not limited to stochastic gradient descent and Adam algorithms, but can also be batch gradient descent, Newton's algorithm, conjugate gradient algorithm, or other algorithms. Those skilled in the art can flexibly choose appropriate optimization algorithms according to actual needs, and this application does not impose specific restrictions in this regard.

[0078] Specifically, in addition to the AI ​​models mentioned above, convolutional neural networks can be used for image classification and object detection, while recurrent neural networks can be used for sequence prediction and time series analysis.

[0079] In other embodiments, the above-described apparatus further includes an annotation unit, configured to annotate the sharp objects and target individuals in the multiple historical image data sets after acquiring them, and before inputting the multiple historical image data sets into the initial AI model. In this embodiment, annotating the sharp objects and target individuals in the multiple historical image data sets before inputting them into the AI ​​model further ensures better training results and accuracy of the model, preparing for the subsequent use of the obtained AI model to identify sharp objects and output target prediction results representing the probability of a target individual contacting a sharp object.

[0080] In other embodiments, the apparatus further includes a cleaning unit, configured to clean the historical image data after acquiring it and before inputting it into the initial AI model. In this embodiment, by cleaning the historical image data, irrelevant information can be eliminated, effectively simplifying the data and preparing it for subsequent AI model building.

[0081] Specifically, in addition to image annotation and data cleaning of historical image data, preprocessing operations such as grayscale transformation, image smoothing, lens distortion correction, and geometric correction can be performed on historical image data to restore useful real information, enhance the detectability of relevant information, simplify data to the maximum extent, and thus improve the reliability of feature extraction, image segmentation, matching, and recognition.

[0082] According to some exemplary embodiments of this application, the control unit includes: a first control module, configured to control the signal emitting device to emit a first warning signal when the target prediction result indicates that the probability of the target person contacting the sharp object is greater than the first predetermined value and less than or equal to a second predetermined value; a second control module, configured to control the signal emitting device to emit a second warning signal when the target prediction result indicates that the probability of the target person contacting the sharp object is greater than the second predetermined value and less than or equal to a third predetermined value; and a third control module, configured to control the signal emitting device to emit a third warning signal when the target prediction result indicates that the probability of the target person contacting the sharp object is greater than the third predetermined value and less than or equal to a fourth predetermined value; wherein the signal strength of the first warning signal is less than the signal strength of the second warning signal, and the signal strength of the second warning signal is less than the signal strength of the third warning signal. In this embodiment, when the target prediction result indicates that the probability of the target person contacting the sharp object falls within different ranges, the signal emitting device emits warning signals with different signal strengths, enabling the recipient of the warning signal to clearly determine the current level of danger and take corresponding measures, further preventing potential harm to children.

[0083] In practical applications, those skilled in the art can set the above-mentioned second predetermined value, third predetermined value and fourth predetermined value based on experience, or obtain them through multiple experiments. This application does not impose any specific restrictions on this.

[0084] In practical applications, those skilled in the art can set the signal strength of the first reminder signal, the second reminder signal, and the third reminder signal based on experience, or obtain them through multiple experiments. This application does not impose any specific limitations on this.

[0085] Specifically, when the reminder signal is emitted as sound, the stronger the signal strength, the louder and longer the sound lasts; when the reminder signal is emitted as light, the stronger the signal strength, the brighter the light; when the reminder signal is emitted as light, different colors of light can be used to represent different signal strengths; when the reminder signal is emitted as an electrical signal, the stronger the signal strength, the stronger the electrical signal; for example, when the signal emitting device is a mobile phone, the stronger the signal strength, the louder the vibration or ringing sound of the mobile application, and the longer the vibration or ringing sound lasts.

[0086] In another exemplary embodiment, the apparatus further includes a second training unit, configured to, after obtaining the target prediction result, train the AI ​​model using online learning based on the target indoor image data and the target prediction result, to obtain the trained AI model. In this embodiment, training the AI ​​model using online learning based on the target indoor image data and the target prediction result further ensures good generalization ability and adaptability of the AI ​​model.

[0087] According to some other exemplary embodiments of this application, the above-described apparatus further includes: a third training unit, configured to, after obtaining the target prediction result, train the AI ​​model using a fine-tuning method based on the target indoor image data and the target prediction result, to obtain the trained AI model. In this embodiment, training the AI ​​model using a fine-tuning method based on the target indoor image data and the target prediction result further ensures better accuracy and reliability of the AI ​​model.

[0088] In some alternative embodiments of this application, the second acquisition unit includes an acquisition module, configured to acquire the indoor image data at the current moment through a smart camera to obtain the target indoor image data. In this embodiment, the target indoor image data is obtained by acquiring the indoor image data at the current moment through a smart camera, so that the target image data can be subsequently input into an AI model to obtain a target prediction result.

[0089] Of course, in addition to the smart cameras mentioned above, other devices that can collect image data in real time can also be used to collect indoor image data at the current moment. Those skilled in the art can flexibly choose appropriate acquisition devices according to actual needs, and this application does not impose specific restrictions on this.

[0090] The aforementioned anti-accidental injury reminder device includes a processor and a memory. The first acquisition unit, the second acquisition unit, the input unit, and the control unit are all stored as program units in the memory. The processor executes the program units stored in the memory to achieve the corresponding functions. All of the above modules are located in the same processor; or, the above modules are located in different processors in any combination.

[0091] This invention provides a computer-readable storage medium including a stored program, wherein, when the program is executed, it controls the device containing the computer-readable storage medium to perform the aforementioned method for preventing accidental injury.

[0092] Specifically, methods for preventing accidental injury include:

[0093] Step S201: Acquire multiple historical image data, including indoor historical object image data and historical target person image data. The historical object image data includes image data of at least one sharp object. The target person includes a child. The sharp object is an object with an end whose length in a first direction is 10 times or more the length in a second direction. The first direction intersects with the second direction.

[0094] Specifically, the target individuals are not limited to children, but may also include blind people, people with intellectual disabilities, or other types of people, and this application does not impose specific restrictions on them; the sharp objects mentioned above include, but are not limited to, scissors, knives, and needles.

[0095] Specifically, the length of the aforementioned end of the sharp object in the first direction is much greater than its length in the second direction.

[0096] Step S202: Obtain indoor image data at the current moment to obtain target indoor image data. The indoor image data includes indoor object image data and indoor target person image data.

[0097] Step S203: Input the above-mentioned indoor image data of the target into the AI ​​model to obtain the target prediction result. The AI ​​model is a model built based on multiple historical image data. The AI ​​model is used to identify the sharp object and output the prediction result. The prediction result is a result that characterizes the probability of the target person coming into contact with the sharp object.

[0098] Step S204: If the target prediction result indicates that the probability of the target person contacting the sharp object is greater than a first predetermined value, the control signal emitting device emits a reminder signal. The reminder signal is used to indicate the probability of the target person contacting the sharp object. The form of the reminder signal includes at least one of the following: sound, light, and electrical signal.

[0099] Specifically, the aforementioned signal transmitting devices include, but are not limited to, alarms, light-emitting devices, and mobile phones. Those skilled in the art can flexibly select appropriate signal transmitting devices according to actual needs, and this application does not impose specific restrictions in this regard.

[0100] In practical applications, those skilled in the art can set the above-mentioned first predetermined value based on experience, or obtain it through multiple experiments. This application does not impose any specific restrictions on this.

[0101] Optionally, after acquiring multiple historical image data and before inputting the target indoor image data into the AI ​​model, the method further includes: establishing an initial AI model using deep learning and computer vision techniques; inputting multiple historical image data into the initial AI model to train the initial AI model; and optimizing the parameters of the initial AI model using an optimization algorithm to obtain the AI ​​model, wherein the optimization algorithm includes at least one of stochastic gradient descent algorithm and Adam algorithm.

[0102] Optionally, after acquiring multiple historical image data, and before inputting the multiple historical image data into the initial AI model, the method further includes: labeling the sharp objects and target persons in the multiple historical image data.

[0103] Optionally, if the target prediction result indicates that the probability of the target person contacting the sharp object is greater than a first predetermined value, the control signal emitting device issues a warning signal, including: if the target prediction result indicates that the probability of the target person contacting the sharp object is greater than the first predetermined value and less than or equal to a second predetermined value, controlling the signal emitting device to issue a first warning signal; if the target prediction result indicates that the probability of the target person contacting the sharp object is greater than the second predetermined value and less than or equal to a third predetermined value, controlling the signal emitting device to issue a second warning signal; and if the target prediction result indicates that the probability of the target person contacting the sharp object is greater than the third predetermined value and less than or equal to a fourth predetermined value, controlling the signal emitting device to issue a third warning signal; wherein the signal strength of the first warning signal is less than the signal strength of the second warning signal, and the signal strength of the second warning signal is less than the signal strength of the third warning signal.

[0104] Optionally, after obtaining the target prediction result, the above method further includes: training the above AI model using online learning based on the above indoor image data of the target and the above target prediction result, to obtain the trained AI model.

[0105] Optionally, after obtaining the target prediction result, the above method further includes: training the above AI model using a fine-tuning method based on the above indoor image data of the target and the above target prediction result, to obtain the trained AI model.

[0106] Optionally, obtaining indoor image data at the current moment to obtain target indoor image data includes: obtaining the aforementioned indoor image data at the current moment through a smart camera to obtain the aforementioned target indoor image data.

[0107] This invention provides a smart home system, including: a signal emitting device; a controller for the signal emitting device, including: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include a reminder method for any of the above-described methods for preventing accidental injury.

[0108] It is obvious to those skilled in the art that the modules or steps of the present invention described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. They can be implemented using computer-executable program code, and thus can be stored in a storage device for execution by a computing device. In some cases, the steps shown or described can be performed in a different order than those described herein, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, the present invention is not limited to any particular combination of hardware and software.

[0109] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0110] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0111] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0112] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0113] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0114] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0115] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0116] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0117] As can be seen from the above description, the embodiments of this application achieve the following technical effects:

[0118] 1) In the anti-accidental injury reminder method of this application, multiple historical image data are first acquired, then indoor image data at the current moment is acquired to obtain target indoor image data. The target image data is then input into an AI model established based on the multiple historical image data to obtain a target prediction result representing the probability of a target person contacting a sharp object. Finally, if the target prediction result indicates that the probability of the target person contacting a sharp object is greater than a first predetermined value, a control signal emitting device issues a reminder signal. Compared to traditional indoor protection measures in the prior art that mainly rely on physical means and cannot monitor and warn of the danger of children contacting sharp objects in real time, leading to children easily coming into contact with sharp objects and causing injury, this application inputs the target indoor image data into an AI model used to identify sharp objects and output prediction results to obtain target prediction results. If the target prediction result indicates that the probability of the target person contacting a sharp object is greater than a first predetermined value, a control signal emitting device issues a reminder signal to prevent and mitigate potential injuries to children.

[0119] 2) In the anti-accidental injury reminder device of this application, multiple historical image data are acquired through a first acquisition unit, and indoor image data at the current moment is acquired through a second acquisition unit to obtain target indoor image data. The target image data is then input into an AI model established based on multiple historical image data through an input unit to obtain a target prediction result representing the probability of a target person coming into contact with a sharp object. When the target prediction result indicates that the probability of a target person coming into contact with a sharp object is greater than a first predetermined value, the control unit controls the signal emitting device to issue a reminder signal. Compared with the traditional indoor protection measures in the prior art, which mainly rely on physical means and cannot monitor and warn of the danger of children coming into contact with sharp objects in real time, thus making it easy for children to come into contact with sharp objects and cause injury, this application inputs the target indoor image data into an AI model used to identify sharp objects and output prediction results to obtain target prediction results. When the target prediction result indicates that the probability of a target person coming into contact with a sharp object is greater than a first predetermined value, the control signal emitting device issues a reminder signal to prevent and mitigate potential injuries to children.

[0120] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A reminder method to prevent accidental injury, characterized in that, include: Acquire multiple historical image data, including indoor historical object image data and historical target person image data. The historical object image data includes image data of at least one sharp object. The target person includes a child. The sharp object is an object with an end whose length in a first direction is 10 times or more the length in a second direction. The first direction intersects the second direction. Acquire indoor image data at the current moment to obtain target indoor image data, wherein the indoor image data includes indoor object image data and indoor target person image data; The target indoor image data is input into the AI ​​model to obtain the target prediction result. The AI ​​model is a model built based on multiple historical image data. The AI ​​model is used to identify the sharp object and output the prediction result, which is a result representing the probability that the target person will come into contact with the sharp object. If the target prediction result indicates that the probability of the target person contacting the sharp object is greater than a first predetermined value, the control signal emitting device emits a reminder signal. The reminder signal is used to indicate the probability of the target person contacting the sharp object. The form of the reminder signal includes at least one of the following: sound, light, and electrical signal.

2. The reminder method for preventing accidental injury according to claim 1, characterized in that, After acquiring multiple historical image data sets, and before inputting the target indoor image data into the AI ​​model, the method further includes: Utilize deep learning and computer vision technologies to build an initial AI model; Multiple historical image data are input into the initial AI model to train the initial AI model, and the parameters of the initial AI model are optimized using an optimization algorithm to obtain the AI ​​model. The optimization algorithm includes at least one of the stochastic gradient descent algorithm and the Adam algorithm.

3. The reminder method for preventing accidental injury according to claim 2, characterized in that, After acquiring multiple historical image data sets, and before inputting these multiple historical image data sets into the initial AI model, the method further includes: The sharp objects and the target individuals in the multiple historical image data are labeled.

4. The reminder method for preventing accidental injury according to any one of claims 1 to 3, characterized in that, If the target prediction result indicates that the probability of the target person coming into contact with the sharp object is greater than a first predetermined value, the control signal transmitting device issues a warning signal, including: If the target prediction result indicates that the probability of the target person coming into contact with the sharp object is greater than the first predetermined value and less than or equal to the second predetermined value, the signal transmitting device is controlled to issue a first warning signal. If the target prediction result indicates that the probability of the target person coming into contact with the sharp object is greater than the second predetermined value and less than or equal to the third predetermined value, the signal transmitting device is controlled to issue a second reminder signal. If the target prediction result indicates that the probability of the target person coming into contact with the sharp object is greater than the third predetermined value and less than or equal to the fourth predetermined value, the signal transmitting device is controlled to issue a third reminder signal. Wherein, the signal strength of the first reminder signal is less than the signal strength of the second reminder signal, and the signal strength of the second reminder signal is less than the signal strength of the third reminder signal.

5. The reminder method for preventing accidental injury according to any one of claims 1 to 3, characterized in that, After obtaining the target prediction result, the method further includes: Based on the target indoor image data and the target prediction results, the AI ​​model is trained using online learning to obtain the trained AI model.

6. The reminder method for preventing accidental injury according to any one of claims 1 to 3, characterized in that, After obtaining the target prediction result, the method further includes: Based on the target indoor image data and the target prediction results, the AI ​​model is trained using a fine-tuning method to obtain the trained AI model.

7. The reminder method for preventing accidental injury according to any one of claims 1 to 3, characterized in that, Acquire indoor image data at the current moment to obtain target indoor image data, including: The target indoor image data is obtained by acquiring the indoor image data at the current moment through a smart camera.

8. A reminder device to prevent accidental injury, characterized in that, include: The first acquisition unit is used to acquire multiple historical image data, including indoor historical object image data and historical target person image data. The historical object image data includes image data of at least one sharp object, and the target person includes a child. The sharp object is an object with an end whose length in a first direction is 10 times or more the length in a second direction, and the first direction intersects the second direction. The second acquisition unit is used to acquire indoor image data at the current moment and obtain target indoor image data, wherein the indoor image data includes indoor object image data and indoor target person image data; The input unit is used to input the target indoor image data into the AI ​​model to obtain the target prediction result. The AI ​​model is a model built based on multiple historical image data. The AI ​​model is used to identify the sharp object and output the prediction result, which is a result representing the probability that the target person will come into contact with the sharp object. The control unit is configured to, when the target prediction result indicates that the probability of the target person contacting the sharp object is greater than a first predetermined value, control the signal emitting device to emit a reminder signal, the reminder signal being used to indicate the probability of the target person contacting the sharp object, and the form of the reminder signal emitting includes at least one of the following: sound, light, and electrical signal.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein, when the program is executed, it controls the device on which the computer-readable storage medium is located to perform the reminder method for preventing accidental injury as described in any one of claims 1 to 7.

10. A smart home system, characterized in that, include: Signal transmitting device; The controller of the signal emitting device includes: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include a reminder method for preventing accidental injury as described in any one of claims 1 to 7.

Images