An intelligent home system based on security technology

Abstract

This application discloses a smart home system based on security technology, relating to the fields of smart home and smart elderly care technology, including: a terminal acquisition layer, a network edge computing layer, a cloud platform layer, and an interactive application layer; the terminal acquisition layer integrates elderly care wristbands, environmental sensors, fall radar, smart cameras, door magnets, positioning tags, and companion robots; the cloud platform layer deploys a data fusion engine, an AI big data model, a video analysis module, and a positioning service engine. This invention achieves functions such as health anomaly monitoring and medication reminders, environmental safety early warning and automatic control, anti-wandering voice reminders, dangerous area protection, real-time positioning of people and objects, multimodal fall detection, voice control of home appliances and monitoring of abnormal electricity use, intelligent object search, AI big data model video retrieval, and community-linked missing person search. Through multimodal data fusion and AI big data model-driven interaction, it achieves proactive smart monitoring of the elderly's lives, improving the level of intelligence in elderly care monitoring and the user experience.

Description

Technical Field

[0001] This application relates to the fields of smart home and smart elderly care technology, specifically to a smart home system based on security technology. Background Technology

[0002] With the increasing trend of population aging, the home safety and health care of elderly people living alone and those in empty nests have become a growing social concern. Currently, while various smart home systems and elderly care monitoring devices exist on the market, most are functionally limited and fragmented. For example, traditional security systems focus on intrusion alarms, while health monitoring devices operate independently of the home environment, preventing data linkage and comprehensive analysis. This results in a lack of proactive warnings and interventions for potential risks such as falls, getting lost, missed medication, or forgetting to turn off the stove in the kitchen. Furthermore, existing human-computer interaction methods are not user-friendly for the elderly; complex interfaces and processes increase the learning curve, and there is a lack of proactive companionship and convenient interaction methods. In emergencies, existing systems often only notify family members, lacking an emergency response mechanism that coordinates with the community and medical institutions, leading to low rescue efficiency and missed opportunities for optimal assistance. Therefore, there is an urgent need for a comprehensive and intelligent home-based elderly care monitoring system that integrates safety protection, health monitoring, intelligent interaction, and community collaboration. Summary of the Invention

[0003] In view of this, this application provides a smart home system based on security technology, which solves the technical problems existing in the prior art, such as incomplete home monitoring of the elderly, untimely risk warning, lack of coordination in emergency response, and inconvenient human-computer interaction.

[0004] To achieve the above objectives, the present invention provides the following technical solution: A smart home system based on security technology includes: a terminal acquisition layer, a network edge computing layer, a cloud platform layer, and an interactive application layer; The terminal acquisition layer includes wearable health monitoring devices, scene detection modules, positioning tags, companion robots, and intelligent control terminals. The terminal acquisition layer is used to collect home scene data. The network edge computing layer deploys edge gateways for data preprocessing, anomaly identification, and rapid linkage control, and enables data communication between the terminal acquisition layer and the cloud platform layer, supporting Wi-Fi, Zigbee, Bluetooth, LoRa, and 4G / 5G hybrid networking; The cloud platform layer is deployed on the home smart gateway or cloud server, including a data fusion engine, AI big model, video analysis module, location service engine and rule engine. The cloud platform layer is used for data storage, AI model operation, permission management and data synchronization. The interactive application layer includes user mobile terminals, community management terminals, smartwatches, and temporary visitor web terminals. The interactive application layer is used for early warning push notifications, receiving control commands, and two-way audio and video communication.

[0005] As a further aspect of the present invention: the wearable health monitoring device includes an elderly care bracelet, which is used to collect the elderly's physiological parameters and exercise data in real time and transmit them to the cloud platform layer. The physiological parameters include heart rate, blood pressure, and blood oxygen data. The cloud platform layer runs a dynamic threshold algorithm. When the collected monitoring data exceeds the preset dynamic threshold, an alarm message is sent through the interactive application layer. The cloud platform layer also sends reminder messages to the wearable health monitoring device or the interactive application layer according to the preset medication schedule. Medication reminders are provided through the wearable health monitoring device or the interactive application layer, and the medication action is confirmed using a posture recognition algorithm based on the collected image data.

[0006] As a further aspect of the present invention: the scene detection module includes an environmental sensor, a door magnetic detector, a human presence sensor, and a passive infrared detector; Environmental sensors include gas leak sensors, temperature and humidity sensors, and smoke detectors. The environmental sensors upload the detected environmental data to the cloud platform layer and are linked with gas valves and power switches through smart sockets. When the environmental data is abnormal, the cloud platform layer triggers a local alarm and sends alarm information to the user's mobile terminal in the interactive application layer. At the same time, it automatically performs the operation of cutting off the gas or power. The user can remotely and manually control the gas or power switch through the user's mobile terminal in the interactive application layer. The door magnetic detector is deployed at the entrance door. When the system is armed, if the door magnetic detector detects that the entrance door is open and the human presence sensor detects that an elderly person is near the door, the detection data is sent to the cloud platform. The cloud platform then triggers the voice module to play a preset reminder voice to prevent the elderly person from getting lost. The passive infrared detector is deployed in the kitchen. When the system is armed, if the passive infrared detector detects that someone has entered the kitchen and the system determines that it is outside of normal cooking hours, the detection data is sent to the cloud platform. The cloud platform then cuts off the kitchen power supply through the smart socket of the smart control terminal and reminds the elderly person through the voice module. The smart control terminal includes a smart socket and a smart gas valve.

[0007] As a further embodiment of the present invention: the scene detection module also includes a fall detection radar and an intelligent camera; The fall detection radar uses millimeter-wave radar to collect point cloud data and identifies human posture through point cloud imaging technology. The smart camera runs a lightweight posture estimation network to collect human image information and identify human posture based on the posture estimation network. The wearable health monitoring device has a built-in accelerometer sensor. The data fusion engine at the cloud platform layer uses DS evidence theory to fuse the fall detection data uploaded by the fall detection radar, smart camera, and wearable health monitoring device, and outputs the fall probability. When the fusion confidence exceeds a preset threshold, an alarm is triggered. When the alarm is triggered, the cloud platform layer controls the smart camera to capture video of a certain period before and after the alarm and push it to the user's mobile terminal at the interactive application layer. The user can call the community management center and ambulance with one click through the user's mobile terminal.

[0008] As a further aspect of the present invention: the positioning tag is attached to the elderly or valuable items, and the positioning service engine uses UWB or Bluetooth positioning technology to perform real-time positioning and tracking of the positioning tag, and presets an electronic fence; when the located target leaves the electronic fence, the cloud platform layer sends an alarm message to the user's mobile terminal in the interactive application layer.

[0009] As a further aspect of the present invention: when the located target, such as an elderly person, leaves the electronic fence and triggers an alarm, the user generates a shared link with temporary permissions through the user's mobile terminal in the interactive application layer and sends it to the community management personnel. The community management personnel log in to the temporary visitor web terminal through the shared link to view the geographical location information of the located elderly person in real time and assist the user in finding the elderly person.

[0010] As a further aspect of the present invention: the companion robot integrates a voice interaction module, which is used to play programs that the elderly prefer according to a preset schedule, and guide the elderly to the restaurant by voice before mealtime. The companion robot uses a camera to identify whether the elderly are seated for a meal. If they are not seated, it will repeatedly remind them and push the information to the user's mobile terminal in the interactive application layer.

[0011] As a further aspect of the present invention: the AI ​​large model is deployed on a cloud server to receive the user's voice input commands, perform semantic parsing on the voice commands, extract time, location and event features, and retrieve the corresponding camera footage from the video storage server based on the features.

[0012] As a further aspect of the present invention: the intelligent camera is also used to detect and identify preset items. The cloud platform layer controls one or more intelligent cameras to rotate and scan according to the user's voice query command, and feeds back the image and location information of the detected target items to the interactive application layer.

[0013] As a further aspect of this invention, the interactive application layer supports two-way audio and video communication with indoor smart screens or smartwatches, allowing seniors to initiate audio and video calls with family members via voice commands or one-click operation. This system integrates multimodal sensing devices, IoT communication technology, and artificial intelligence algorithms to provide comprehensive safety protection, health monitoring, behavioral intervention, and daily living assistance for seniors living at home.

[0014] As can be seen from the above technical solution, the advantages of the present invention are: 1. This application improves the accuracy of fall detection and reduces false alarms and missed alarms by combining radar, vision, and wearable data with DS evidence theory-based fusion decision-making. Furthermore, it employs proactive safety protection, not only providing alarms for abnormal situations but also intervening in advance when elderly individuals accidentally enter dangerous areas or attempt to go out alone, cutting off power or issuing voice reminders to reduce safety hazards through behavior prediction and area-based defense. Additionally, through a rotatable camera and target detection model, it automatically identifies and announces the location of frequently sought items, demonstrating high practicality. The robot proactively reminds users to eat and play programs according to their schedule, and visually confirms the execution, providing both daily living assistance and emotional companionship.

[0015] 2. Users can use voice commands to view surveillance footage and control devices without needing to learn a complex user interface. Furthermore, when an elderly person goes missing, family members can grant temporary access to community management personnel, enabling a rapid social response and assistance in finding them. 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 of this application and their descriptions are used to explain this application and do not constitute an undue limitation of this application.

[0017] Figure 1 This is a schematic diagram of the composition structure of a smart home system based on security technology according to this application.

[0018] Figure 2 This is yet another structural schematic diagram of this application.

[0019] Figure 3 This is a schematic diagram of the fall detection and rescue process in this embodiment.

[0020] Figure 4 This is a flowchart illustrating the AI ​​large-scale intelligent video retrieval steps in this embodiment.

[0021] Figure 5 This is a schematic diagram illustrating the steps of a smart elderly care monitoring method based on multimodal perception and AI large model in this embodiment. Detailed Implementation

[0022] To make the purpose, technical solutions, and advantages of this application more clear and understandable, the following further elaborates on this application in combination with the embodiments and the accompanying drawings. Here, the illustrative embodiments of this application and their descriptions are used to explain this application, but do not limit this application.

[0023] Reference Figures 1 to 5 , such as Figures 1 to 2 shown, this embodiment provides a smart home system based on security technology. The system architecture includes: a terminal acquisition layer, a network edge computing layer, a cloud platform layer, and an interaction application layer. The terminal acquisition layer deploys a variety of intelligent devices for collecting home scenario data, including wearable health monitoring devices, scenario detection modules, positioning tags, companion robots, and intelligent control terminals. The network edge computing layer deploys edge gateways for data preprocessing, anomaly recognition, and fast linkage control, and realizes data communication between the terminal acquisition layer and the cloud platform layer, supporting Wi-Fi, Zigbee, Bluetooth, LoRa, and 4G / 5G hybrid networking.

[0024] Specifically, the wearable health monitoring device includes an elderly care bracelet, which integrates a heart rate, blood pressure, blood oxygen sensor, and a six-axis accelerometer and communicates with the home gateway via Bluetooth. The elderly care bracelet in this embodiment is used to collect the physiological parameters and exercise data of the elderly in real time and transmit them to the cloud platform layer. The physiological parameters include heart rate, blood pressure, and blood oxygen data. The cloud platform layer runs a dynamic threshold algorithm. When the collected monitoring data exceeds the preset dynamic threshold, it sends an alarm message through the interaction application layer. The cloud platform layer also sends reminder messages to the wearable health monitoring device or the interaction application layer according to the preset medication schedule, and performs medication reminders through the wearable health monitoring device or the interaction application layer, and uses the gesture recognition algorithm to confirm the medication action based on the collected image data.

[0025] The scenario detection module includes environmental sensors, door magnetic detectors, human presence sensors, and passive infrared detectors. Specifically, the environmental sensors include gas leakage sensors, temperature and humidity sensors, smoke detectors, and water immersion sensors. The gas leakage sensor uses the MQ-5 model, and the smoke detector uses the MQ-5 model. The temperature and humidity sensor uses the DHT22 model. The environmental sensors are all connected to the gateway through Zigbee modules. The environmental sensors upload the detected environmental data to the cloud platform layer and are linked with the gas valve and power switch through smart sockets. When the environmental data is abnormal, the cloud platform layer triggers a local alarm and sends an alarm message to the user's mobile terminal in the interaction application layer, and at the same time automatically executes the operation of cutting off the gas or power supply. The user remotely manually controls the gas or power switch through the user's mobile terminal in the interaction application layer.

[0026] A door magnetic detector is deployed on the entrance door, using a reed switch to detect whether the door is open or closed. When the system is armed, if the door magnetic detector detects the entrance door is open and the human presence sensor detects an elderly person near the door, the detection data is sent to the cloud platform. The cloud platform then triggers a voice module to play a preset reminder to prevent the elderly person from getting lost. The human presence sensor uses a 24GHz millimeter-wave radar, installed above the door frame, to detect if someone is approaching. A passive infrared detector is installed on the kitchen ceiling to detect human movement. When the system is armed, if the passive infrared detector detects someone entering the kitchen and the system determines it is outside of normal cooking hours, the detection data is sent to the cloud platform. The cloud platform then cuts off the kitchen power through a smart socket on the smart control terminal and reminds the elderly person via a voice module. The smart control terminal includes a smart socket and a smart gas valve. In this embodiment, the smart socket integrates an energy metering chip and a relay, allowing for remote switching and real-time collection of electricity consumption data. The smart gas valve is driven by a motor and controlled by a gateway.

[0027] The scene detection module also includes a fall detection radar and a smart camera. The fall detection radar uses millimeter-wave radar to collect point cloud data and identifies human posture through point cloud imaging technology. The smart camera runs a lightweight posture estimation network to collect human image information and identify human posture based on the posture estimation network. The wearable health monitoring device has a built-in accelerometer sensor. The data fusion engine at the cloud platform layer uses DS evidence theory to fuse the fall detection data uploaded by the fall detection radar, smart camera, and wearable health monitoring device, and outputs the fall probability. When the fusion confidence exceeds a preset threshold, an alarm is triggered. When an alarm is triggered, the cloud platform layer controls the smart camera to capture video of a certain period before and after the alarm and pushes it to the user's mobile terminal at the interactive application layer. The user can then use the mobile terminal to call the community management center and an ambulance with one click. The smart camera is also used for target detection and recognition of preset objects. Based on the user's voice query command, the cloud platform layer controls one or more smart cameras to rotate and scan, and feeds back the image and location information of the detected target object to the interactive application layer.

[0028] The system's location tags are attached to the elderly person or valuables. The location service engine uses UWB or Bluetooth positioning technology to track the tags in real time and presets an electronic fence. In this embodiment, a UWB tag is embedded in the elderly person's wristband, working with indoor UWB base stations to achieve centimeter-level positioning. When the located target leaves the electronic fence, the cloud platform layer sends an alarm message to the user's mobile terminal in the interactive application layer. When the located target, such as an elderly person, leaves the electronic fence and triggers an alarm, the user generates a shared link with temporary permissions through the user's mobile terminal in the interactive application layer and sends it to the community management personnel. The community management personnel log in to the temporary visitor web terminal through the shared link to view the location information of the located elderly person in real time and assist the user in finding the elderly person.

[0029] The companion robot is a mobile robot with a screen and a LiDAR navigation module. It integrates a voice interaction module to play programs that the elderly prefer according to a preset schedule. Before mealtime, it guides the elderly to the restaurant by voice. The companion robot uses a camera to identify whether the elderly are seated for a meal. If they are not seated, it will repeatedly remind them and push the information to the user's mobile terminal in the interactive application layer.

[0030] In this embodiment, the network edge computing layer adopts a hybrid networking scheme. A low-power sensor network is built within the home via Zigbee, the elderly care wristband is connected via BLE Bluetooth, high-bandwidth devices such as cameras and radars are connected via Wi-Fi, and the gateway is connected to the cloud via Ethernet or 4G / 5G.

[0031] In this embodiment, the cloud platform layer is deployed on a home smart gateway or cloud server, including a data fusion engine, an AI big data model, a video analytics module, a location service engine, and a rules engine. The data fusion engine is responsible for parsing, time-aligning, and initially cleaning data from different protocols. The rules engine runs preset condition-action scripts, such as "IF gas concentration > threshold THEN close valve AND alarm". The video analytics module uses YOLOv8 and OpenPose algorithms to perform object detection and human pose estimation. The location service engine uses the TDOA algorithm to calculate UWB tag locations and manages electronic fences. The AI ​​big data model, deployed on the cloud server, receives user voice input commands, performs semantic parsing of the voice commands, extracts time, location, and event features, and retrieves and retrieves corresponding camera footage from the video storage server based on these features. In this embodiment, the AI ​​big data model is deployed in the cloud as a multimodal big data model based on the Transformer architecture, enabling joint understanding of text, voice, and images.

[0032] The interactive application layer is used for early warning push notifications, control command reception, and two-way audio and video communication, providing a multi-terminal interactive interface. It includes user mobile terminals, community management terminals, smartwatches, and a web interface for temporary visitors. User mobile terminals support real-time status viewing, alarm reception, remote control, video intercom, and item search. Smartwatches, worn by the elderly, support one-click calling and voice interaction. The web interface of the community management terminal allows viewing the real-time status of elderly residents within the jurisdiction, but authorization is required. The web interface for temporary visitors is accessible via a temporary link and is limited to viewing the real-time location of lost elderly residents.

[0033] like Figure 5 As shown, this application also discloses a smart elderly care monitoring method based on multimodal perception and AI big data models, which uses a smart home system based on security technology as described above for elderly care monitoring, including the following steps: Step S01: Deploy the perception layer devices and connect them to the network edge computing layer; Step S02: The data fusion engine of the cloud platform layer collects data from the terminal acquisition layer in real time, and performs preprocessing and feature extraction; Step S03: The rule engine performs logical judgment on the data according to preset rules, triggering local or remote alarms and control commands; Step S04: The video analysis module analyzes the video stream captured by the smart camera to identify falls, object locations, and elderly behavior; Step S05: The location service engine calculates the real-time location of the target and compares it with the electronic fence; Step S06: The AI ​​big data model receives the user's natural language query, parses it, and retrieves the corresponding historical video clips or real-time footage; Step S07: The interactive application layer presents alarm information, real-time status, and interactive interface to the user.

[0034] like Figure 3 As shown, in the second embodiment, the steps for fall detection and rescue are as follows: Step 1: Data Acquisition. The millimeter-wave radar outputs point cloud data at 20fps. The PointNet++ model extracts the human skeleton and posture, outputting the fall probability Pr. The smart camera acquires RGB images at 30fps, extracts keypoint coordinates using OpenPose, and uses an LSTM temporal network to determine if a fall has occurred, outputting the probability Pc. The elderly care wristband's accelerometer samples at 50Hz. If it detects an impact peak greater than 2.5g followed by a period of stillness exceeding 30 seconds, it triggers a fall candidate, outputting the probability Pw, which takes the value 0 or 1.

[0035] Step 2: Data Fusion. The cloud platform layer uses Dempster's evidence theory for fusion, defining the recognition framework G = {fallen, not fallen}. Pr, Pc, and Pw are converted into basic probability assignments mr, mc, and mw, respectively. For example, mr({fallen}) = Pr, mr({not fallen}) = 1 - Pr, mr(G) = 0. Then, using Dempster's composition rule, orthogonal sums are calculated on mr, mc, and mw to obtain the fused probability assignment m.

[0036] Step 3: Decision and Alarm. Set the fusion confidence threshold T=0.85. If m({fall})>T, it is determined to be a fall event. The system immediately performs the following operations: 1) Capture a short video of 10 seconds before and after the fall using a camera and upload it to the cloud; 2) Push alarm information and short video to all family members via the user's mobile terminal APP; 3) Ask the elderly person through voice synthesis: "Did you fall? Please answer if you need help." If the elderly person does not respond or answers "yes," the system automatically proceeds to the next step.

[0037] Step 4: One-click emergency assistance. After a family member opens the user's mobile terminal (phone) APP, a red alarm card will pop up on the screen, containing two buttons: "Call the Community Management Center" and "Call 120". Clicking either button will automatically dial the preset number and send the elderly person's location information and a link to the live video feed to the recipient via SMS.

[0038] like Figure 4 As shown, in the third embodiment, the AI ​​large-scale model intelligent video retrieval steps are as follows: the process of retrieving surveillance footage through natural language: Step 1: Voice Input. The elderly person or their family member can speak commands via a mobile app or smart screen in the room, such as: "Can you check what Grandpa did in the living room last night?" Step 2: Speech Recognition and Semantic Analysis. The front-end ASR module converts speech into text and transmits it to the cloud-based AI model. The AI ​​model performs intent recognition and entity extraction, extracting key information: Time = "last night", which the system automatically parses into a specific time period, such as 2024-05-15 19:00 to 24:00; Location = "living room"; Target Person = "Grandpa"; Action = "what to do", i.e., retrieving all activities.

[0039] Step 3: Video Retrieval. The AI ​​model sends the parsed structured query to the video storage server. The video storage server quickly locates the corresponding video file based on the timestamp and camera ID, such as a living room camera. Then, using video summarization technology, through motion detection and keyframe extraction, it generates a list of thumbnails containing images of people's activities.

[0040] Step 4: Result Return and Interaction. The AI ​​big data model returns the list of thumbnails and their corresponding time points to the user's mobile device, along with a voice announcement: "Last night at 7:20 PM, Grandpa was watching TV in the living room; at 9:10 PM, Grandpa got up and went to his bedroom." The user can then ask further questions such as, "What were his facial expressions like while watching TV?" The big data model then retrieves video frames from the corresponding time period for facial expression analysis and returns the results.

[0041] In the fourth embodiment, the intelligent item search process is as follows, illustrating the process of an elderly person searching for their keys: (a) An elderly man said to the companion robot, “I can’t find my keys.” (b) After the voice command is converted into text by the front-end ASR module, the item search service is triggered. The system controls the smart cameras in the living room and bedroom to start rotating and scanning, and at the same time calls the companion robot to go to the area where keys are usually placed, such as the entryway or coffee table, to search.

[0042] (c) The video stream captured by the camera is input into the YOLOv5s object detection model in real time. The YOLOv5s object detection model is pre-trained with detection weights for more than ten items such as keys, remote controls, glasses, and mobile phones.

[0043] (d) When any camera detects the key target in the frame, record the target's position and confidence level in the image. The system estimates the actual position of the key in the room, such as the coordinates of the key on the coffee table, through multi-camera collaborative localization.

[0044] (e) The robot moves to the target location, illuminates the target with a laser pointer, and announces via voice: "The key is next to the remote control on the coffee table." At the same time, a live screen pops up on the user's mobile app, highlighting the key with a red frame.

[0045] In the fifth embodiment, the procedure for handling the elderly person after they leave the electronic fence is as follows: (a) The location service engine detects that the UWB coordinates of the elderly person's elderly care wristband exceed the boundary of the community, i.e. the set electronic fence, triggers the missing person alarm, and pushes the alarm information to the family's mobile phone, i.e. the user's mobile terminal.

[0046] (b) After family members confirm, they click the "Seek Community Assistance" button in the user's mobile app interface. The system generates a temporary shared link containing a time-sensitive token, such as valid for 2 hours, with access restrictions, allowing only location viewing and no device control.

[0047] (c) Family members send the link to the community property management personnel via WeChat or SMS.

[0048] (d) Property management personnel click the link and open a temporary web-based map interface in their browser to view the elderly person's movement trajectory and current location in real time.

[0049] (e) The system records this shared log, including the generation time, user IP, number of accesses, etc., to ensure security and traceability.

[0050] In the sixth embodiment, the monitoring process for the electrical safety of the elderly is as follows: (a) The smart socket collects the power consumption of appliances such as rice cookers and kettles in real time and uploads the data to the gateway every 15 seconds.

[0051] (b) The cloud platform layer runs online learning algorithms to establish a 24-hour power consumption pattern baseline for each appliance. For example, rice cookers typically operate from 7:00-7:30 in the morning and from 17:30-18:00 in the evening, and their power curves exhibit a specific shape.

[0052] (c) The real-time power data is compared with the baseline. If the rice cooker is detected to be working continuously for more than 1 hour at an unexpected time (such as 23:00 at night), or the power curve is abnormal (such as continuous high power), it is judged as abnormal.

[0053] (d) The system pushes a reminder to family members' mobile phones: "The rice cooker may have been left running for 1 hour." If there is no manual intervention within 30 minutes, the system can automatically cut off the power supply through the smart socket and issue a voice reminder to the elderly.

[0054] In summary, this invention, through the deep integration of IoT, multimodal sensing, and artificial intelligence technologies, constructs a comprehensive, proactive, and intelligent elderly care monitoring environment, significantly improving the quality of life and safety of the elderly.

[0055] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. In some cases, the actions or steps recorded in the specification and claims can be performed in a different order than that shown in the embodiments, and the desired result can still be achieved. In addition, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result; in some embodiments, multitasking and parallel processing are also feasible or advantageous.

[0056] Those skilled in the art will understand that all or part of the steps of the above embodiments can be implemented by hardware or by a program instructing the relevant hardware to implement them. The program can be stored in a computer-readable storage medium, such as a read-only memory, a disk, or an optical disk.

[0057] 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 the embodiments of 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 smart home system based on security technology, characterized in that, include: Terminal acquisition layer, network edge computing layer, cloud platform layer, and interactive application layer; The terminal acquisition layer includes wearable health monitoring devices, scene detection modules, positioning tags, companion robots, and intelligent control terminals. The terminal acquisition layer is used to collect home scene data. The network edge computing layer deploys an edge gateway for data preprocessing and to enable data communication between the terminal acquisition layer and the cloud platform layer. The cloud platform layer is deployed on a home smart gateway or cloud server and includes a data fusion engine, an AI big model, a video analysis module, a location service engine, and a rule engine. The cloud platform layer is used for data storage, AI model operation, access management, and data synchronization. The interactive application layer includes user mobile terminals, community management terminals, smartwatches, and temporary visitor web terminals. The interactive application layer is used for early warning push, control command reception, and two-way audio and video communication.

2. The smart home system based on security technology according to claim 1, characterized in that, The wearable health monitoring device includes an elderly care wristband, which is used to collect the elderly's physiological parameters and movement data in real time and transmit them to the cloud platform layer. The cloud platform layer runs a dynamic threshold algorithm. When the collected monitoring data exceeds a preset dynamic threshold, an alarm message is sent through the interactive application layer. The cloud platform layer also sends reminder messages to the wearable health monitoring device or the interactive application layer according to a preset medication schedule. Medication reminders are provided through the wearable health monitoring device or the interactive application layer, and the medication taking action is confirmed using a posture recognition algorithm based on the collected image data.

3. The smart home system based on security technology according to claim 1, characterized in that, The scene detection module includes an environmental sensor, a door magnetic detector, a human presence sensor, and a passive infrared detector. The environmental sensors include a gas leak sensor, a temperature and humidity sensor, and a smoke detector. The environmental sensors upload the detected environmental data to the cloud platform layer. When the environmental data is abnormal, the cloud platform layer triggers a local alarm and sends an alarm message to the user's mobile terminal in the interactive application layer, automatically cutting off the gas or power supply. The user can remotely and manually control the gas or power switch through the user's mobile terminal in the interactive application layer. When the system is armed, if the door magnetic detector detects that the entrance door is opened and the human presence sensor detects that an elderly person is near the door, the detection data is sent to the cloud platform layer. The cloud platform layer then triggers the voice module to play a preset reminder voice. The passive infrared detector is deployed in the kitchen. When the passive infrared detector detects that someone has entered the kitchen, it sends the detection data to the cloud platform layer. The cloud platform layer then cuts off the kitchen power supply through the smart control terminal and reminds the elderly person through the voice module. The smart control terminal includes a smart socket and a smart gas valve.

4. The smart home system based on security technology according to claim 3, characterized in that, The scene detection module also includes a fall detection radar and an intelligent camera; The fall detection radar is used to collect point cloud data and identify human posture; the intelligent camera is used to collect human image information and identify human posture based on a posture estimation network. The wearable health monitoring device has a built-in accelerometer sensor. The data fusion engine of the cloud platform layer uses DS evidence theory to fuse the fall detection data uploaded by the fall detection radar, the smart camera, and the wearable health monitoring device, and outputs the fall probability. When the fusion confidence exceeds a preset threshold, an alarm is triggered. When the alarm is triggered, the cloud platform layer controls the smart camera to capture video of a certain period before and after the alarm and push it to the user's mobile terminal of the interactive application layer. The user can call the community management center and ambulance with one click through the user's mobile terminal.

5. The smart home system based on security technology according to claim 1, characterized in that, The location tag is attached to the elderly or valuables. The location service engine uses UWB or Bluetooth positioning technology to track the location tag in real time and presets an electronic fence. When the located target leaves the electronic fence, the cloud platform layer sends an alarm message to the user's mobile terminal in the interactive application layer.

6. The smart home system based on security technology according to claim 5, characterized in that, When the located target leaves the electronic fence and triggers an alarm, the user generates a shared link with temporary permissions through the user's mobile terminal in the interactive application layer and sends it to the community management terminal. The community management personnel log in to the temporary visitor web terminal through the shared link to view the geographical location information of the located target in real time.

7. The smart home system based on security technology according to claim 1, characterized in that, The companion robot integrates a voice interaction module, which is used to play programs at set times according to a preset schedule and guide the elderly to the restaurant by voice before mealtime. The companion robot uses a camera to identify whether the elderly are seated for a meal. If they are not seated, it will repeatedly remind them and push the information to the user's mobile terminal.

8. The smart home system based on security technology according to claim 4, characterized in that, The AI ​​model is deployed on a cloud server to receive users' voice input commands, perform semantic parsing on the voice commands, extract time, location and event features, and retrieve the corresponding camera footage from the video storage server based on the features.

9. The smart home system based on security technology according to claim 8, characterized in that, The intelligent camera is also used to detect and identify preset items. The cloud platform layer controls one or more intelligent cameras to rotate and scan according to the user's voice query command, and feeds back the image and location information of the detected target items to the interactive application layer.

10. The smart home system based on security technology according to claim 8, characterized in that, The interactive application layer supports two-way audio and video communication with indoor smart screens or smartwatches, allowing seniors to initiate audio and video calls with family members via voice commands or one-click operation.

Images