A home monitoring system and method

Abstract

The application discloses a kind of family monitoring system and method, system includes the monitoring host of integrated main control, audio and video acquisition, audio playback and wireless communication module and is realized data interaction by SDIO, I2S interface, supports the hand-held terminal of local wireless dual-mode communication with Internet.Adaptive optimization is reduced by interface to reduce terminal computing power consumption, realize audio intelligent noise reduction and abnormal audio priority retention;Desensitization fragment cloud relay storage architecture is constructed, and original data privacy leakage is prevented;Adaptive switching video encoding mode, rely on adjustable antenna beam control and radio frequency adaptation capability, significantly improve the transmission stability of long-distance, shielding scene.Simultaneously integrated sleep disturbance-free, offline emergency control, multi-stage power outage emergency power supply and two-way voice intercom function, solve the problem that traditional equipment network is strongly dependent, long-distance monitoring effect is poor, security is low, is easily disturbed, emergency support is missing, realize all-weather high reliability, humanized intelligent family monitoring.

Description

Technical Field

[0001] This invention relates to the field of smart home monitoring technology, specifically to a home monitoring system and method suitable for scenarios involving wall obstruction and long-distance monitoring, featuring dual-network adaptive communication, optimized hardware interface transmission, intelligent audio recognition, radio frequency adaptive control, sleep do-not-disturb mode, and offline emergency control functions. Background Technology

[0002] Smart home monitoring devices are widely used in scenarios such as infant and toddler care, elderly home monitoring, and early warning of abnormal home conditions. Existing traditional home monitoring systems mostly adopt a single cloud network transmission architecture, and the data transmission and command interaction of devices are highly dependent on the Internet link. When abnormal conditions such as network interruption or network fluctuation occur, the monitoring devices cannot operate normally, posing security risks of monitoring interruption or failure, and the overall reliability of the devices is poor.

[0003] Meanwhile, existing monitoring devices mostly use general transmission interfaces and general data processing logic for audio and video data transmission. They lack hardware-level customization and optimization for the low latency, low computational overhead, high-fidelity audio, and retention of critical abnormal data required in home monitoring scenarios. This results in problems such as high terminal computational load, poor bandwidth adaptability, video stuttering, low audio signal-to-noise ratio, and severe environmental noise interference. Crucially, traditional monitoring devices cannot intelligently identify high-risk audio signals such as infant cries or abnormal calls for help, and cannot prioritize the retention and transmission of abnormal audio data, easily leading to the loss of core monitoring data. This results in insufficient device intelligence and monitoring reliability.

[0004] In terms of privacy and security, most existing monitoring devices upload the complete original audio and video data to cloud storage. The cloud has decryption and viewing permissions for the original data, posing a significant risk of user privacy leaks and making it difficult to meet the security needs of private home scenarios. At the same time, traditional devices use fixed video encoding and compression strategies, which cannot adaptively adjust encoding parameters according to real-time network conditions. In scenarios with limited local area network bandwidth, problems such as large data transmission volumes, network congestion, data packet loss, and video stuttering are likely to occur.

[0005] Furthermore, existing monitoring devices mostly use fixed-parameter omnidirectional antennas, which cannot adaptively adjust the antenna beam, modulation method, and transmission power according to the location of the handheld terminal, real-time signal quality, and regional radio frequency specifications. In scenarios involving wall obstruction in homes or long-distance transmission, they suffer from severe signal degradation, weak anti-interference capabilities, and poor transmission stability. Additionally, traditional monitoring devices lack user-friendly intelligent monitoring logic, cannot identify the sleep state of the monitored individual, and continuous voice prompts and calls at night can easily disturb neighbors, resulting in a poor user experience.

[0006] The command parsing and execution of existing monitoring equipment heavily rely on cloud servers. When the network is interrupted, independent local emergency control is impossible, creating a security vulnerability where emergency operations fail. Furthermore, the equipment only supports a single mains power supply mode and lacks emergency power supply and alarm mechanisms for power outages. Monitoring functions immediately terminate when there is an abnormal mains power failure, making 24 / 7 uninterrupted monitoring impossible.

[0007] In summary, the reliability, intelligence level, and security of existing technologies are insufficient to meet the actual needs of high-quality home monitoring. Summary of the Invention

[0008] The purpose of this invention is to overcome the technical shortcomings of existing home monitoring devices, such as strong network dependence, poor audio and video transmission performance, weak abnormal audio recognition capability, low privacy and security, poor adaptability to radio frequency environments, disturbance to neighbors during nighttime monitoring, lack of offline emergency capabilities, and lack of power outage protection. This invention provides a home monitoring system and method. By constructing a dual-end collaborative hardware architecture, combined with dual-network adaptive switching, optimized hardware interface transmission, intelligent audio recognition and retention, privacy-desensitized transmission, adaptive radio frequency parameter control, sleep mode with non-disturbing linkage, local offline emergency control, and multi-level emergency power supply design, this invention achieves intelligent home monitoring with full-scene capabilities, high reliability, high security, and low interference.

[0009] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides a home monitoring system, including a monitoring host and a handheld terminal; the monitoring host is configured with a host control unit, a video acquisition unit, an audio acquisition unit, and an audio playback unit, all of which are electrically connected to the host control unit, and the host control unit has a built-in wireless communication module; the handheld terminal is configured with a terminal communication unit, an audio optimization processing unit, a terminal control unit, and a display interaction unit such as a touch screen, the terminal communication unit being connected to the terminal control unit via an SDIO interface and the audio optimization processing unit via an I2S interface, respectively, and the handheld terminal supports Internet access and local wireless communication; the host control unit is configured to monitor the network status of the handheld terminal and adaptively select a cross-network relay transmission mode or a WPA3 local encrypted wireless transmission mode according to the network status, and the host control unit is also used to identify the deep sleep state of the monitored object based on the audio and video acquisition data.

[0010] Furthermore, the SDIO interface features adaptive timing adjustment, dual-buffered queue management, and interrupt merging mechanisms. Through its own mechanisms, the SDIO interface works with the host control unit to dynamically adjust the compression rate and transmission rate of audio and video data, thereby reducing the computational overhead of the terminal control unit and adapting to real-time network bandwidth changes.

[0011] Furthermore, the I2S interface is used to transmit audio data. The audio optimization processing unit works with the I2S interface to perform dynamic filtering and adaptive gain control on the audio data to optimize the audio signal-to-noise ratio. The audio optimization processing unit can identify abnormal audio data such as baby crying or abnormal calls for help, and prioritize the retention, transmission and storage of abnormal audio data.

[0012] Furthermore, the system supports dual communication modes: remote internet communication and local wireless communication. When the handheld terminal is connected to the internet, the system prioritizes the use of the local wireless LAN to transmit real-time audio and video data and control commands. At the same time, identity verification is completed through the cloud. Only key event segments that have been anonymized with user authorization are uploaded to the private cloud storage. The cloud only serves as a remote access relay channel and cannot decrypt the original audio and video data.

[0013] Furthermore, the host control unit can adaptively switch the video encoding mode according to the network status of the handheld terminal; it uses low compression encoding when connected to the Internet, and high compression encoding (higher than the low compression encoding) when communicating on a local wireless LAN, and adds anomaly markers to local communication data packets.

[0014] Furthermore, the monitoring host is equipped with an adjustable antenna array that is electrically connected to the host control unit; the host control unit can adjust the antenna beam according to the position of the handheld terminal, enable robust modulation and forward error correction mechanisms when the signal deteriorates, and adaptively adjust the transmission power and communication channel to adapt to regional radio frequency specifications.

[0015] Furthermore, the voice commands collected by the display interaction unit can be sent to the audio playback unit via a wireless communication link to achieve two-way voice communication; the host control unit can identify the deep sleep state of the monitored object and send a sleep recognition signal. After receiving the sleep recognition signal, the terminal control unit sends a mute instruction to the host control unit, causing the monitoring host to switch to non-voice reminder mode and stop voice call requests.

[0016] Furthermore, when the handheld terminal is in local wireless LAN communication mode, the emergency control commands generated by the display interaction unit can be parsed and executed independently by the terminal main control unit locally, without the need for cloud intervention.

[0017] Furthermore, the monitoring host is equipped with an emergency power supply unit, which includes a supercapacitor, an intelligent power module, and a backup battery interface. When the mains power fails, the supercapacitor provides instantaneous power and reports an alarm. After the backup battery is connected, the system enters a low-power operation mode. After the mains power is restored, the system automatically switches to the normal operating power supply state.

[0018] Secondly, the present invention provides a home monitoring method, applied to the home monitoring system described in the first aspect, comprising the following steps: S1: The monitoring host collects audio and video data through the video acquisition unit and the audio acquisition unit. The host main control unit completes audio and video synchronization and privacy desensitization processing. The emergency power supply unit monitors the power supply status of the equipment in real time. S2: The main control unit monitors the internet connection status of the handheld terminal in real time and determines the system communication mode. When the internet connection is detected to be normal, the system prioritizes transmitting real-time audio and video data and control commands through the local wireless LAN. After the handheld terminal is verified by the cloud, it uploads de-identified key event fragments to the private cloud storage according to user authorization. The cloud only acts as a remote relay and cannot decrypt the original data. When the internet connection is detected to be interrupted, the system automatically switches to the local communication mode. The monitoring host or handheld terminal initiates and establishes a WPA3 encrypted wireless LAN to complete the identity verification of the network devices and realize direct communication without a network. S3: Audio and video data are transmitted to the terminal main control unit via wireless communication link and SDIO interface. Video data is output to display interaction unit to complete screen display. Audio data is transmitted to audio optimization processing unit via I2S interface for noise reduction and gain optimization before being sent back to realize audio playback, storage or abnormal audio retention and forwarding. S4: User control commands are transmitted back to the monitoring host via a wireless link to enable remote control of the equipment; when the system is in local communication mode, emergency control commands are parsed and executed independently by the terminal main control unit to ensure the effectiveness of offline emergency control. S5: The main control unit regulates the adjustable antenna array to optimize the signal beam, and adjusts the signal modulation method and transmission power in accordance with regional radio frequency specifications to ensure stable data transmission over long distances with low interference; S6: The host control unit intelligently identifies the deep sleep state of the monitored object based on the collected audio and video data, generates and sends out a sleep recognition signal; after receiving the signal, the terminal control unit sends out a mute command, controls the monitoring host to switch to non-voice reminder mode, terminates voice calls, and realizes sleep monitoring without disturbance.

[0019] The beneficial effects of this invention include: Compared with existing technologies, it has the following technical advantages: 1. This invention adopts a dual-end collaborative hardware architecture of a monitoring host and a handheld terminal, configured with a dual network communication system, which can adaptively switch communication modes under normal internet conditions and internet interruption conditions. When the internet is normal, local area network is used first to transmit data to ensure real-time monitoring; when the internet is interrupted, a WPA3 encrypted local area network is automatically established to achieve direct monitoring without a network connection. This overcomes the shortcomings of traditional devices that are highly dependent on the network and fail when the network is disconnected, and significantly improves the stability of device operation and the adaptability to different scenarios.

[0020] 2. This invention integrates adaptive timing adjustment, dual-buffered queue management, and interrupt merging mechanisms through the SDIO interface. It can dynamically adjust the compression and transmission rates of audio and video data based on real-time network bandwidth fluctuations, effectively reducing the computational load on the terminal's main control unit and solving the problems of high computing power consumption and poor transmission stability in traditional devices. Simultaneously, through the I2S interface and in conjunction with the audio optimization processing unit, it achieves hardware-level dynamic filtering and adaptive gain control, improving the audio signal-to-noise ratio. Furthermore, it prioritizes the retention, transmission, and storage of high-risk audio data such as infant cries and abnormal distress calls, effectively ensuring the traceability of critical monitoring data and enhancing monitoring security.

[0021] 3. This invention constructs a hierarchical privacy and security protection system. Only key abnormal event fragments authorized by the user are anonymized and uploaded to private cloud storage. The cloud only serves as a remote access relay and does not have the decryption authority for the original audio and video data. This avoids the risk of original privacy data being leaked to the cloud from the source of the data, and balances the convenience of remote monitoring with the security of family privacy.

[0022] 4. This invention can adaptively switch video encoding modes according to the real-time network conditions of the handheld terminal. Low-compression encoding is used when connected to the internet, while high-compression encoding is used when communicating via a local wireless LAN. Anomaly markers are added to local communication data packets, adapting to different bandwidth environments of remote internet and local LAN, facilitating subsequent tracing and retrieval of abnormal events. Simultaneously, this invention features an adjustable antenna array, which can dynamically adjust the antenna beam. Combined with forward error correction and robust modulation mechanisms, it adaptively adapts to regional RF specifications and adjusts the transmit power and communication channel, effectively improving signal degradation problems in scenarios such as home wall obstruction and long-distance monitoring, significantly enhancing the anti-interference capability and operational stability of the device for long-distance transmission.

[0023] 5. This invention features a smart sleep state recognition and do-not-disturb linkage mechanism, which can accurately identify the deep sleep state of the monitored object, automatically trigger a mute command, turn off the device's active voice calls and switch to a non-voice reminder mode, effectively solving the problem of disturbing the peace with nighttime voice prompts of traditional monitoring devices, achieving all-weather humanized smart monitoring, and greatly optimizing the user experience.

[0024] 6. This invention has local offline emergency control capability. Under local wireless communication conditions, the emergency control commands generated by the handheld terminal can be parsed and executed locally by the terminal main control unit without the participation of the cloud server. This effectively makes up for the safety shortcomings of traditional equipment after network outage and improves the operational reliability of the equipment under extreme conditions.

[0025] 7. This invention is equipped with a multi-level emergency power supply unit, which realizes instantaneous power supply and abnormal alarm through supercapacitors when the mains power fails. It also achieves low-power operation with the backup battery interface. After the mains power is restored, it can automatically switch to the normal power supply mode, effectively avoiding monitoring interruption caused by abnormal mains power failure, realizing uninterrupted monitoring operation around the clock, and improving the fault tolerance and operational reliability of the equipment. Attached Figure Description

[0026] Figure 1 This is a structural block diagram of a home monitoring system provided in an embodiment of the present invention.

[0027] Figure 2 This is a flowchart of a home monitoring method provided in an embodiment of the present invention. Detailed Implementation

[0028] The present invention will be further described in detail below with reference to specific embodiments. The specific embodiments described are only used to explain the present invention and are not intended to limit the scope of protection of the present invention.

[0029] Please refer to Figure 1 This invention discloses a home monitoring system, which mainly includes a monitoring host and a handheld terminal. The monitoring host and the handheld terminal complete data and command interaction through a wireless communication link. This system is suitable for scenarios such as infant and toddler care, home-based elderly monitoring, and home anomaly early warning. It can operate stably under complex conditions such as network outages, network interruptions, long-distance transmission obstructions, power supply abnormalities, and the monitored person sleeping.

[0030] Specifically, the monitoring host includes a main control unit, a video acquisition unit, an audio acquisition unit, and an audio playback unit. The video acquisition unit is responsible for real-time acquisition of dynamic video data of the monitored scene; the audio acquisition unit picks up ambient audio; and the audio playback unit is used for remote voice playback. All three are electrically connected to the main control unit, which coordinates the start / stop of acquisition and audio output. The wireless communication module integrated within the main control unit acts as the wireless transceiver hub for the entire device, responsible for establishing a two-way wireless data link between the host and the handheld terminal.

[0031] The handheld terminal is equipped with a terminal communication unit, an audio optimization processing unit, a terminal main control unit, and a display interaction unit. The terminal communication unit establishes a data connection with the terminal main control unit via an SDIO interface, while the audio optimization processing unit establishes a data connection with the terminal main control unit via an I2S interface. The terminal communication unit handles wireless data transmission and reception and converts audio and video at high speed via SDIO. The audio optimization processing unit specifically processes audio data streams and interacts with the main control unit via an I2S serial bus. The handheld terminal has two communication paths: internet access for cross-network segment remote cloud access and local wireless communication for short-range LAN direct connection interaction. The main control unit monitors the network connection status of the handheld terminal in real time to determine the system communication mode. When a normal internet connection is detected, the system prioritizes transmitting real-time audio and video data and control commands via the local wireless LAN. When an internet connection is detected to be interrupted, the main control unit automatically controls the monitoring host or handheld terminal to initiate and establish a WPA3 encrypted wireless LAN, completes network device identity verification, and achieves direct communication without a network connection.

[0032] The SDIO interface integrates adaptive timing adjustment, dual-buffered queue management, and interrupt merging mechanisms. The relevant control logic is executed by the SDIO controller within the terminal's main control unit. The terminal's main control unit monitors the load status of the SDIO bus and network jitter parameters in real time, dynamically adjusting the clock frequency and data transmission window of the SDIO interface. It divides the memory into two independent buffers; when data is written to one buffer, the other is transferred out via DMA. The SDIO controller merges scattered interrupts and reports them to the CPU, reducing terminal computational overhead. The main control unit dynamically adjusts the audio and video compression rate and transmission rate based on changes in network bandwidth.

[0033] The I2S interface handles serial audio transmission. The audio optimization processing unit has a built-in DSP and uses the MFCC algorithm for audio processing: audio is framed in 20ms increments with 10ms frame overlap, and 12-dimensional MFCC features are extracted after Hanning window preprocessing. The device locally stores standard feature templates for crying and distress calls. When the real-time feature has a cosine similarity ≥ 0.75 with the template, it is identified as abnormal audio. Abnormal audio is cached, transmitted, and stored with higher priority than regular audio. This recognition scheme does not require cloud computing power or complex deep learning models and can run locally on the embedded system. After identifying anomalies, a priority marker is added to the data packet, and it is stored in a high-priority storage queue to avoid overwriting critical data.

[0034] The system supports dual-mode communication: internet and local wireless. In network mode, real-time audio / video and control commands prioritize the local area network (LAN), with the cloud only performing identity verification. The host encrypts and de-identifies key event segments before uploading them to the private cloud; the cloud does not possess the decryption key and only acts as a remote relay. To prevent the cloud from decrypting the original data, the host generates an encryption key locally to encrypt the segments to be uploaded; the decryption key is stored only on the handheld terminal, while the cloud only stores the encrypted data. When offline, the host or handheld terminal actively establishes a WPA3 encrypted LAN, and the device verifies the identity key to achieve direct offline connection.

[0035] The host switches encoding based on the handheld terminal's network connection status: low compression encoding is used for external networks, high compression encoding is used for local area networks, and anomaly markers are added to local data packets.

[0036] The directional antenna array is controlled by the host machine, which adjusts the antenna beam according to the terminal location. When the signal deteriorates, robust modulation and forward error correction are activated, and the transmit power and communication channel are adjusted in accordance with the regional radio frequency specifications to improve the quality of communication in obstructed and long-distance communication.

[0037] Users record voice input through interactive display units such as touch screens, which are then transmitted wirelessly to the host audio playback unit to achieve two-way communication. The host main control unit intelligently identifies the deep sleep state of the monitored object based on the collected audio and video data and sends a sleep recognition signal. After receiving the signal, the terminal main control unit sends back a mute command, and the host disables active voice calls and switches to non-voice reminders.

[0038] When communicating offline locally, emergency control commands are parsed and executed locally by the terminal master controller, without the need for cloud forwarding and verification.

[0039] The monitoring host is equipped with a dedicated emergency power supply unit, which consists of a supercapacitor, an intelligent power module, and a backup battery interface. The intelligent power module manages the overall power supply. When the mains power is normal, it supplies power to the host while simultaneously storing energy in the supercapacitor. In the event of a power outage, the supercapacitor momentarily discharges to maintain short-term equipment operation and reports alarm information to the upper-level main control unit. When the backup battery is connected to the dedicated interface, the intelligent power module automatically reduces its frequency and switches to a low-power operating mode. When the mains power is restored, the system automatically switches back to the regular mains power supply circuit. When the regular mains power supply is normal, the system is powered by the mains power, while the intelligent power module stores energy in the supercapacitor. When a mains power outage is detected, the supercapacitor momentarily releases energy to temporarily power the equipment and actively reports an alarm to alert the user of the power supply anomaly. When an external backup battery is connected to the interface, the system automatically switches to a low-power operating mode, extending the power outage monitoring duration. When the mains power supply is restored, the system automatically switches back to the regular mains power supply state, achieving uninterrupted monitoring during power outages and effectively improving the reliability and fault tolerance of the equipment.

[0040] Please refer to Figure 2 The workflow of the home monitoring method of the present invention is as follows: S1: The monitoring host synchronously collects on-site audio and video data through the video acquisition unit and audio acquisition unit. The host main control unit completes the audio and video time axis synchronization alignment and privacy desensitization preprocessing. The emergency power supply unit collects voltage parameters in real time and monitors abnormal power supply conditions of the equipment throughout the process.

[0041] S2: The main control unit monitors the internet connection status of the handheld terminal in real time and adaptively determines the system communication mode. When the internet connection is normal, the system prioritizes transmitting real-time audio and video data and control commands through the local wireless LAN. After the handheld terminal completes cloud identity verification, it only uploads user-authorized de-identified key event fragments to the private cloud storage. The cloud only acts as a remote relay and cannot decrypt the original data. When the internet connection is interrupted, the system automatically switches to the local communication mode. The monitoring host or handheld terminal actively initiates and establishes a WPA3 encrypted wireless LAN. After completing device identity key verification, the device achieves direct communication without a network connection.

[0042] S3: The collected audio and video data are transmitted via a wireless communication link. The video data is transmitted at high speed to the terminal main control unit through the SDIO interface and finally output to the display and interaction unit to complete the real-time screen display. The audio data is transmitted to the audio optimization and processing unit through the I2S interface. Abnormal audio is identified by MFCC feature extraction + 0.75 similarity threshold matching. Dynamic filtering, adaptive gain optimization, and abnormal audio priority retention processing are implemented simultaneously. The optimized audio data is sent back to enable playback, storage, or abnormal audio forwarding.

[0043] S4: User's regular control commands are transmitted back to the monitoring host via a wireless link to enable remote control of the device; when the system is in local offline communication mode, all emergency control commands are parsed and executed independently by the terminal main control unit without cloud intervention, ensuring the effectiveness and timeliness of offline emergency control.

[0044] S5: The main control unit adjusts the directional antenna array in real time to optimize the signal beam direction, and adaptively adjusts the signal modulation method, transmission power and communication channel in combination with regional radio frequency specifications to effectively avoid interference and ensure low interference and high stability data transmission in long-distance and multi-obstruction scenarios.

[0045] S6: The main control unit of the host intelligently identifies the deep sleep state of the monitored object based on audio and video data, generates and sends out sleep recognition signals; after receiving the signals, the terminal main control unit sends out a mute command, controls the monitoring host to switch to non-voice reminder mode, terminates the active voice call request, and finally realizes stable, humane, and undisturbed intelligent sleep monitoring around the clock.

[0046] For those skilled in the art, various other corresponding changes and modifications can be made based on the technical solutions and concepts described above, and all such changes and modifications should fall within the protection scope of the claims of this invention.

[0047] It should be noted that the embodiments described above are only some embodiments of the present invention, and not all embodiments. The singular forms "a," "described," and "itself" used in the embodiments and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise.

Claims

1. A home monitoring system, characterized in that, Includes monitoring host and handheld terminal; The monitoring host is equipped with a host main control unit, a video acquisition unit, an audio acquisition unit, and an audio playback unit. The video acquisition unit, audio acquisition unit, and audio playback unit are all electrically connected to the host main control unit. The host main control unit has a built-in wireless communication module. The handheld terminal is equipped with a terminal communication unit, an audio optimization processing unit, a terminal main control unit, and a display interaction unit. The terminal communication unit is connected to the terminal main control unit through an SDIO interface, and the audio optimization processing unit is connected to the terminal main control unit through an I2S interface. The handheld terminal supports Internet access and local wireless communication. The SDIO interface features adaptive timing adjustment, dual-buffered queue management, and interrupt merging mechanisms. The terminal control unit monitors the SDIO bus load status and network jitter parameters in real time, dynamically adjusting the clock frequency and data transmission window of the SDIO interface. Two independent buffers are partitioned in memory, with one buffer writing data while the other is transmitted via DMA. The SDIO controller merges and reports scattered interrupts to the terminal control unit. The host control unit dynamically adjusts the compression rate and transmission rate of audio and video data according to changes in network bandwidth to reduce the computational overhead of the terminal control unit. The audio optimization processing unit has a built-in DSP and receives audio data through the I2S interface. It uses MFCC feature extraction combined with local pre-stored abnormal sound feature template matching to identify abnormal audio. After adding a priority mark to the data packet, the identified abnormal audio is preferentially stored in a high-priority storage queue so that the caching, transmission and storage priority of abnormal audio is higher than that of regular audio. The audio optimization processing unit does not require cloud computing power and complex deep learning models and can run locally in the embedded system. The host control unit monitors the internet connection status of the handheld terminal in real time; When the internet connection is interrupted, the system automatically switches to local communication mode, whereby the monitoring host or the handheld terminal initiates and establishes a WPA3 encrypted wireless local area network, completes the identity key verification of the network devices, and realizes direct communication without a network.

2. The home monitoring system according to claim 1, characterized in that, The monitoring host is equipped with an adjustable antenna array electrically connected to the host main control unit. The host main control unit can adjust the antenna beam according to the position of the handheld terminal, enable robust modulation and forward error correction mechanism when the signal deteriorates, and adaptively adjust the transmission power and communication channel to adapt to regional radio frequency specifications.

3. The home monitoring system according to claim 1, characterized in that, The monitoring host is equipped with an emergency power supply unit, which includes a supercapacitor, an intelligent power module, and a backup battery interface. When the mains power is normal, the intelligent power module supplies power to the monitoring host and stores energy in the supercapacitor. When the mains power fails, the supercapacitor discharges instantaneously to maintain short-term operation and reports an alarm. After the backup battery is connected, the intelligent power module automatically reduces its frequency and switches to a low-power operation mode to achieve uninterrupted monitoring during power outages. After the mains power is restored, the system automatically switches back to the normal operating power supply circuit.

4. The home monitoring system according to claim 1, characterized in that, The host control unit can adaptively switch the video encoding mode according to the network status of the handheld terminal; it uses low compression encoding when connected to the Internet, and uses high compression encoding with a higher compression rate than the low compression encoding when communicating on a local wireless LAN, and adds anomaly markers to local communication data packets.

5. The home monitoring system according to claim 1, characterized in that, When the Internet is connected normally, the system prioritizes the use of the local wireless LAN to transmit real-time audio and video data and control commands, while using the cloud to complete identity verification. The host main control unit generates encryption keys locally, encrypts and desensitizes the key event segments to be uploaded, and then uploads them to the private cloud storage. The decryption key is only stored locally on the handheld terminal. The cloud only stores the ciphertext and serves as a remote access relay channel, and cannot decrypt the original audio and video data.

6. The home monitoring system according to claim 1, characterized in that, The host control unit can identify the deep sleep state of the monitored object and send a sleep recognition signal. After receiving the sleep recognition signal, the terminal control unit sends a mute command to the host control unit, causing the monitoring host to switch to non-voice reminder mode and stop voice call requests.

7. The home monitoring system according to claim 1, characterized in that, When the handheld terminal is in local wireless LAN communication mode, the emergency control commands generated by the display interaction unit are parsed and executed independently by the terminal main control unit locally, without the need for cloud intervention.

8. The home monitoring system according to claim 1, characterized in that, When the audio optimization processing unit performs audio processing, it performs 20ms frame segmentation, 10ms frame overlap, and Hanning window preprocessing on the audio to extract 12-dimensional MFCC features. These features are then matched with locally stored standard feature templates for crying and calling for help using cosine similarity. When the similarity is ≥0.75, it is determined that there is abnormal audio.

9. A method for home monitoring implemented using the home monitoring system of claim 2, characterized in that, Includes the following steps: S1: The monitoring host collects audio and video data through the video acquisition unit and the audio acquisition unit, and the host main control unit completes audio and video synchronization and privacy desensitization processing; S2: The host control unit monitors the Internet connectivity status of the handheld terminal in real time and determines the system communication mode; When the Internet connection is detected to be disconnected, the system automatically switches to local communication mode. The monitoring host or handheld terminal initiates and establishes a WPA3 encrypted wireless LAN, completes device identity key verification, and realizes direct communication without Internet access. S3: Audio and video data are transmitted to the terminal main control unit via wireless communication link and SDIO interface, and video data is output to the display interaction unit for display; audio data is transmitted to the audio optimization processing unit via I2S interface, and abnormal audio is identified by the built-in DSP using MFCC feature extraction combined with local pre-stored abnormal sound template matching, and a priority mark is added to the abnormal audio data packet, which is preferentially stored in the high-priority storage queue, and then noise reduction and gain optimization processing is performed before being sent back. S4: User control commands are transmitted back to the monitoring host via a wireless link to enable remote control of the device; when the system is in local communication mode, emergency control commands generated by the display interaction unit are parsed and executed independently by the terminal main control unit locally to ensure the effectiveness of offline emergency control. S5: The host main control unit regulates the adjustable antenna array to optimize the antenna beam, and adjusts the signal modulation method and transmission power in combination with regional radio frequency specifications to ensure stable data transmission over long distances with low interference; S6: The host main control unit identifies the deep sleep state of the monitored object based on the collected audio and video data, generates and sends a sleep recognition signal; after receiving the signal, the terminal main control unit sends a mute command to control the monitoring host to switch to non-voice reminder mode, terminate voice calls, and realize sleep monitoring without disturbance.

Images