Multi-mode indoor monitoring unmanned aerial vehicle system and control method thereof

Abstract

The invention discloses an indoor health and safety detection unmanned aerial vehicle system and method, and belongs to the technical field of intelligent security and health monitoring. The unmanned aerial vehicle integrates a multi-mode sensor and an edge calculation unit, and real-time monitoring of human body vital signs, pet safety and environmental risks is achieved through a millimeter wave radar, a dual-light camera (visible light and thermal imaging), a non-contact infrared body temperature sensor and a multi-gas detector. According to the system, a multi-source data fusion algorithm is adopted, a space-time alignment technology and a dynamic weight distribution model are combined, respiratory frequency, body temperature distribution and abnormal behavior characteristics are accurately analyzed, and graded response is triggered based on a three-level early warning mechanism (primary / intermediate / emergency). According to the invention, the maneuverability of the unmanned aerial vehicle is creatively combined with a multi-mode sensing technology, and the system is suitable for scenes such as family nursing, smart pension, pet monitoring and environmental safety, and has the advantages of high-precision monitoring and high cost performance.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent security and health monitoring, and specifically relates to a multi-modal indoor monitoring drone system and its control method, which realizes intelligent identification and multi-level response to human health status, pet safety, and environmental risks. Background Art

[0002] The prior art has the following defects: the health monitoring devices applied in residential homes have single functions and cannot achieve multi-parameter fusion analysis; it is difficult for drones to coordinate environmental perception and biological detection; the data of the home security system and the health management platform are fragmented; there is a lack of effective technical means for detecting abnormal pet behaviors. Summary of the Invention

[0003] The system architecture of the present invention mainly includes a drone body, a sensing matrix, and a core processor.

[0004] The drone body adopts a four-axis folding fuselage (expanded size ≤ 500 mm), a three-axis mechanical stabilization gimbal, and a replaceable battery module (endurance ≥ 45 minutes).

[0005] The sensing matrix includes a visible light-infrared dual-mode camera (1080P@30fps + thermal imaging 384×288), a 60GHz millimeter-wave radar module (detection range 0.2 - 5m), a non-contact infrared body temperature sensor (accuracy ±0.3°C), and a multi-gas detector (CO / smoke / VOC).

[0006] The core processor adopts an edge computing unit (NPU 4TOPS), a multi-source data fusion processing algorithm, and is equipped with an adaptive flight control module.

[0007] The technical innovation points of the present invention include a vital sign fusion detection algorithm, a three-level early warning mechanism, and a pet feature recognition system.

[0008] Vital sign fusion detection algorithm: extracting the breathing frequency from the phase change of the millimeter-wave radar (accuracy 0.1Hz), dynamically tracking the body surface temperature distribution with the infrared image, and judging the fall / syncope state by visual gesture recognition.

[0009] Three-level early warning mechanism: primary warning: the vital signs deviate from the normal value by 10%, intermediate warning: continuous abnormality exceeds 15 minutes, emergency response: detecting a fall / asphyxiation / fire.

[0010] Pet feature recognition system: constructing a three-dimensional bone model library for dogs and cats, analyzing the abnormal coefficient of the movement trajectory, and jointly monitoring the body temperature and respiration of pets.

[0011] This solution achieves high cost performance through the following designs: The hardware adopts an industrial-grade mature solution (such as TI millimeter-wave radar module), the algorithm is deployed in the edge computing unit to reduce the dependence on the cloud, the existing network infrastructure of smart home is reused, and the modular design supports function expansion (such as adding an ultraviolet disinfection module). Description of the Drawings

[0012] Figure 1 : Schematic diagram of the system architecture and data flow

[0013] Figure 2 : Profile diagram of the spatial distribution of sensors

[0014] Figure 3 : Flow chart of the multi-modal data fusion algorithm

[0015] Figure 4 : State machine diagram of the emergency response protocol

[0016] Figure 5 and schematic diagram of the pet health monitoring interface Specific Embodiments

[0017] Elderly care scenario: The drone detects that the breathing rate in the bedroom is lower than 8 times per minute. The drone automatically switches to the emergency scanning mode: The millimeter-wave radar confirms the amplitude of chest movement, the thermal imaging locates the human body orientation, and the visible light camera captures facial features.

[0018] The data fusion determines the risk level of respiratory failure and triggers a level 3 response: Locally activate the sound and light alarm, call the associated service robot to deliver first aid drugs, and push the alarm information to the elderly care service platform.

[0019] Fire warning application: The gas sensor detects that the CO concentration exceeds 100 ppm, and the drone performs the following action sequence: Start the thermal imaging scan to locate the heat source; The visible light camera confirms the burning substance; Autonomously fly to the nearest smart socket to cut off the power supply; Activate all alarm devices through the mesh network.

[0020] Pet health warning: The home surveillance camera captures that the pet dog is acting abnormally and restlessly in the dog cage. The drone flies in front of the dog cage and conducts joint monitoring of the pet's body temperature and respiration. It senses that the body temperature is 39.6 degrees and the respiration is 27 times per minute, which is higher than the normal level. It is initially judged that the dog has a fever. The system immediately notifies the owner, and the owner makes an appointment with a pet doctor to diagnose and treat the dog ( Figure 5 )

Claims

1. An indoor health and safety monitoring drone system, characterized in that include: The drone platform is equipped with multimodal sensors, positioning reference points deployed in the flight area, a multi-source data fusion processing unit, a communication module supporting the MQTT / HTTP protocol, a cloud-based health data analysis platform, a vital signs fusion detection algorithm, a three-level early warning mechanism, and a pet feature recognition system.

2. The system according to claim 1, characterized in that The sensor array includes: a dual-band life detection module of millimeter-wave radar and infrared thermal imaging, an environmental perception module of visible light camera and gas detector, and a double-layer sensor layout distributed with the drone body and gimbal. The system as claimed in claim 1, wherein the dynamic scanning method includes: a search strategy combining a preset cruise path with a random path, adaptive flight altitude adjustment (1.5-3m) based on personnel density, and precise hovering monitoring guided by heat source positioning.

3. The system as claimed in claim 1, wherein the data processing method thereof comprises: wavelet denoising of radar signals, feature registration of thermal imaging and visible light images, and sliding window comparison and analysis of time series data (window length 5-30 minutes).

4. The system as claimed in claim 1, wherein the emergency response protocol comprises: Fall detection → automatically lower the height and start voice inquiry, apnea → trigger nearby equipment light warning, fire confirmation → autonomously start emergency escape route guidance.

5. The system according to claim 1, characterized in that Support: wireless communication with smart home hub, push HL7 standard data packets to medical service platform, and establish UWB precise positioning linkage with mobile robots.

6. The system as described in claim 1, wherein its privacy protection mechanism includes: local encrypted storage of sensitive biometric data, real-time face blurring processing in video streams, and feature desensitization processing before data upload.

7. The system as claimed in claim 1, wherein the pet monitoring function realizes: animal identification and classification based on YOLOv5, generation of pet activity heat map, and correlation analysis between abnormal barking and movement status.

8. The system according to claim 1, characterized in that Hardware configuration: ESP32-C6 wireless communication module, TIAWR6843 millimeter-wave radar chip, and integrated MLX90640 infrared sensor array.

Images