Operating personnel falling behavior identification and detection device

Abstract

The utility model provides an operator falling behavior identification and detection device. The operator falling behavior identification and detection device comprises a camera and a laser radar, the camera and the laser radar are installed at a preset position through the installation assembly and form a cross monitoring area so as to acquire personnel behavior data from different angles. The camera and the laser radar are respectively connected with the microcontroller through the communication interface module, the microcontroller is connected with the alarm module, and the camera, the laser radar, the microcontroller and the alarm module are respectively connected with the power supply module; and the microcontroller is embedded with local image processing logic for personnel falling identification, the microcontroller is used for carrying out real-time fusion and judgment on data transmitted by the camera and the laser radar, and when the microcontroller judges a falling event, the microcontroller immediately sends a signal to the alarm module to control the alarm module to give an alarm. According to the invention, by using the complementary characteristics of the camera and the laser radar, the posture change of the personnel in the operation process can be more comprehensively identified, and the tumble identification accuracy is significantly improved.

Description

Technical Field

[0001] This utility model relates to the field of safety management technology, and in particular to a device for recognizing and detecting worker fall behavior. Background Technology

[0002] With the increasing aging of society and rising demands for industrial safety, real-time detection and response systems for falls have gained widespread attention in fields such as medical care, industrial production, and security monitoring. Currently, common fall detection systems primarily employ the following technological approaches: single-vision recognition systems, which use cameras to capture real-time video images of the monitored area and combine this with image analysis algorithms to identify changes in posture and movement, such as standing, falling, and stillness; millimeter-wave or lidar systems, which use non-contact radar sensors to collect the target's motion trajectory and spatial point cloud data, and analyze motion trends through feature extraction; wearable sensor systems, which use sensors such as accelerometers and gyroscopes worn by the user to detect falls when rapid displacement or sudden changes in posture occur; and dual-modal fusion systems, which attempt to fuse radar or image signals with sensors to improve recognition accuracy and are commonly used in high-precision scenarios.

[0003] Among them, vision systems are the most widely used due to their ease of deployment and rich image information, while radar systems have the advantages of being all-weather and not privacy-sensitive. Both have their advantages and have demonstrated some feasibility in experimental environments. Although existing technologies have initially achieved the perception and recognition of fall behavior, they still face the following significant shortcomings and deficiencies in real-world, complex working environments:

[0004] The visual system is highly dependent on the environment: the camera is sensitive to lighting conditions, and the image recognition effect drops significantly at night, in strong backlight or dusty environments; at the same time, the image captured by the camera is easily interfered with by obstructions, leading to recognition failure.

[0005] Radar systems have limited ability to identify details: While millimeter-wave or lidar systems have the advantage of penetrating smoke and identifying spatial patterns, they are weak in judging static or minute changes in posture and have difficulty distinguishing between falling and sitting or lying down normally.

[0006] Sensor systems rely on human cooperation: Wearable devices depend on individual wearing and maintenance, and there are usage problems such as falling off, forgetting to wear, and battery life, making them unsuitable for long-term applications in complex construction sites or multi-person scenarios.

[0007] Multimodal fusion is complex to achieve: Some studies have proposed fusion algorithms, but most of them are based on software-level fusion and lack a well-structured and easy-to-install multi-sensor integrated system, making it difficult to quickly deploy in industrial scenarios. Utility Model Content

[0008] To address the aforementioned technical problems, this application provides a worker fall detection and recognition device. Utilizing the complementary characteristics of a camera and a lidar sensor, it collects image information and three-dimensional spatial contour data respectively, enabling more comprehensive identification of posture changes during work and significantly improving fall detection accuracy. To achieve the above technical features, the purpose of this utility model is as follows:

[0009] A device for recognizing and detecting worker falls includes a camera and a lidar.

[0010] The camera and the lidar are respectively mounted at preset positions via the first mounting base and the second mounting base to form an overlapping monitoring area, so as to acquire human behavior data from different angles;

[0011] The camera and the lidar are respectively connected to the microcontroller through a communication interface module. The microcontroller is connected to the alarm module. The camera, the lidar, the microcontroller, and the alarm module are respectively connected to the power module.

[0012] The microcontroller is used to perform real-time fusion and judgment of the data transmitted from the camera and the lidar. When the microcontroller detects a fall event, it immediately sends a signal to the alarm module to control the alarm module to generate an alarm.

[0013] The second mounting base includes a base fixed in a preset position; the lidar is mounted on the base via a ball joint to facilitate adjustment of the lidar's mounting angle.

[0014] The base includes a fixed seat, on which a longitudinally extending groove is provided, and a movable plate is slidably installed in the groove, and the ball hinge is installed on the movable plate.

[0015] The ball hinge includes a rotating seat fixed to the movable plate, a rotating ball movably installed inside the rotating seat, a screw hole on the rotating seat, a locking bolt screwed into the screw hole, the end of the locking bolt abutting against the rotating ball, and the rotating ball being connected to the lidar.

[0016] The fixed base is provided with a groove, a push plate is provided in the groove, and an elastic component is provided between the push plate and the fixed base.

[0017] The elastic component is a compression spring.

[0018] The alarm module is an audible and visual alarm.

[0019] The present invention has the following beneficial effects:

[0020] 1. This application utilizes the complementary characteristics of cameras and lidar to collect image information and three-dimensional spatial contour data respectively, enabling more comprehensive identification of personnel posture changes during operations and significantly improving fall detection accuracy. Especially in complex environments such as those with obstructions, low light, or smoke, the radar can still maintain effective monitoring, compensating for the weaknesses of image systems. While ensuring high detection accuracy, it provides an all-weather, non-contact, and widely deployable monitoring solution.

[0021] 2. This invention adopts a detachable modular design. Through the setting of sliding grooves and movable plates, the laser radar can be quickly installed and disassembled. At the same time, the setting of thrust plates and ball hinges makes the installation of laser radar more stable and the field of view adjustment more flexible, which can adapt to the needs of different site layouts and monitoring directions.

[0022] 3. This application has a high accuracy rate in fall detection and also has significant advantages in system structure, installation efficiency and adaptability, and has good prospects for promotion and engineering application. Attached Figure Description

[0023] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0024] Figure 1 A schematic diagram of the connection structure between the camera and the first mounting base provided in an embodiment of this utility model;

[0025] Figure 2 This is a schematic diagram of the connection structure between the lidar and the second mounting base provided in an embodiment of the present utility model;

[0026] Figure 3 A schematic diagram of the connection structure between the lidar and the ball hinge provided in an embodiment of this utility model;

[0027] Figure 4 This is a schematic diagram of the structure of the second mounting base provided in an embodiment of the present utility model;

[0028] Figure 5 This is a partial cross-sectional structural diagram of the base provided in an embodiment of the present utility model;

[0029] In the diagram: Camera 1, LiDAR 2, First mounting base 3, Second mounting base 4, Base 4a, Seat body 4a1, Slide 4a2, Movable plate 4a3, Groove 4a4, Push plate 4a5, Elastic component 4a6, Ball hinge 5, Rotating seat 5a, Rotating ball 5b, Locking bolt 5c. Detailed Implementation

[0030] The embodiments of this utility model will be further described below with reference to the accompanying drawings.

[0031] To achieve the above-mentioned technical features, the purpose of this utility model is as follows:

[0032] See appendix Figure 1-5 A fall detection and recognition device for workers includes a camera 1 and a lidar 2. The camera 1 and lidar 2 are mounted at preset positions via a first mounting base 3 and a second mounting base 4, respectively, forming an overlapping monitoring area to acquire personnel behavior data from different angles. The camera 1 and lidar 2 are connected to a microcontroller via a communication interface module. The microcontroller is connected to an alarm module, and the camera 1, lidar 2, microcontroller, and alarm module are all connected to a power supply module. The microcontroller has embedded local image processing logic for fall detection. It performs real-time fusion and judgment on the data transmitted from the camera 1 and lidar 2. When the microcontroller detects a fall event, it immediately sends a signal to the alarm module, controlling the alarm module to generate an alarm. This application utilizes the complementary characteristics of the camera and lidar to acquire image information and three-dimensional spatial contour data respectively, enabling more comprehensive identification of personnel posture changes during work and significantly improving the accuracy of fall detection. Especially in complex environments such as obstruction, low light, or smoke and dust, radar can still maintain effective monitoring, making up for the weaknesses of the imaging system. While ensuring high detection accuracy, it provides an all-weather, non-contact, and wide-area deployment monitoring solution.

[0033] In one implementation, the second mounting base 4 includes a base 4a fixed in a preset position; the lidar 2 is mounted on the base 4a via a ball joint 5. The ball joint 5 allows the lidar 2 to adjust its pitch and rotation angles by ±45° to adapt to the detection direction requirements in different scenarios.

[0034] In one embodiment, the base 4a includes a seat body 4a1, on which a slide groove 4a2 extends longitudinally. A movable plate 4a3 is slidably installed in the slide groove 4a2, and a ball hinge 5 is installed on the movable plate 4a3. The slide groove 4a2 and the movable plate 4a3 are configured to enable the quick installation and disassembly of the lidar 2.

[0035] In one embodiment, the ball hinge 5 includes a rotating seat 5a fixed on the movable plate 4a3, a rotating ball 5b movably installed inside the rotating seat 5a, a screw hole on the rotating seat 5a, a locking bolt 5c screwed into the screw hole, the end of the locking bolt 5c abutting against the rotating ball 5b, the rotating ball 5b being connected to the lidar 2, and the locking bolt 5c tightening the rotating ball 5b to prevent the lidar 2 from shifting during long-term use.

[0036] In one embodiment, the base 4a1 is provided with a groove 4a4, a push plate 4a5 is provided in the groove 4a4, and an elastic member 4a6 is provided between the push plate 4a5 and the base 4a1. When installing the movable plate 4a3, the push plate 4a5 is first pushed by an external force, which then compresses the elastic member 4a6, causing the movable plate 4a3 to retract into the groove 4a4. At this time, the elastic member 4a6 gains elastic potential energy due to compression. Then, the movable plate 4a3 is inserted into the slide groove 4a2 from top to bottom. The external force is removed, and under the action of the elastic potential energy of the elastic member 4a6, the push plate 4a5 abuts against the movable plate 4a3 to prevent the movable plate 4a3 from loosening or falling off.

[0037] In one embodiment, the elastic component 4a6 is a compression spring. By compressing the compression spring, elastic potential energy is obtained, and the push plate 4a5 abuts against the movable plate 4a3 under the elastic force of the compression spring.

[0038] In one implementation method, the alarm module in this embodiment is an audible and visual alarm. The audible and visual alarm is installed in a conspicuous position in the work area. When the terminal detects a fall event, it immediately sends a signal to the alarm module, triggering the buzzer to sound and the red LED warning light to flash. At the same time, the alarm is pushed to the remote monitoring platform or the mobile terminal of the management personnel through the 4G / 5G module.

[0039] Furthermore, the lidar is preferably a three-dimensional lidar based on the TOF (Time-of-Flight) principle, possessing a large field of view of 120°×90° and a depth detection range of 0.3m to 3.0m. The lidar housing is made of ABS plastic and aluminum alloy composite material, possessing IP66 protection capability and adaptable to complex environments such as high humidity and high dust. The camera is a 1080P resolution wide-angle industrial-grade camera with an 80°~90° horizontal field of view and a frame rate of 30fps. The camera is fixed to the wall or column surface via a universal ball joint bracket. The bracket base is fixed to the wall with expansion bolts, and the bracket head has a ball joint connection mechanism that connects to the rear shell of the camera. This structure supports angle adjustment of ±45° horizontally and ±30° vertically. The camera lens area is equipped with a transparent acrylic cover for dust and water protection, and the rear shell of the camera has heat dissipation holes to facilitate heat dissipation and improve system stability.

[0040] In practical implementation, it is recommended that the lidar be installed in the central area of ​​the ceiling at a height of 2.8m to 3.0m, looking downwards. The camera module is recommended to be installed in a corner 2.0m to 2.5m above the ground, forming an approximately 45° angle with the lidar to achieve multi-angle monitoring of human behavior in the same area. To ensure complete coverage, one device node can be deployed every 20-25 square meters, with a 10%-20% overlap between adjacent nodes to prevent blind spots. This invention can also be equipped with a remote setting platform, allowing users to visually configure detection parameters, alarm sensitivity, monitoring area boundaries, etc. All devices can be powered centrally via a DC12V power supply module or via PoE (Power over Ethernet), resulting in simple wiring and convenient installation.

Claims

1. A worker fall behavior recognition detection device, characterized in that: Including cameras and LiDAR; The camera and the lidar are respectively mounted at preset positions via the first mounting base and the second mounting base to form an overlapping monitoring area, so as to acquire human behavior data from different angles; The camera and the lidar are respectively connected to the microcontroller through a communication interface module. The microcontroller is connected to the alarm module. The camera, the lidar, the microcontroller, and the alarm module are respectively connected to the power module. The microcontroller is used to perform real-time fusion and judgment of the data transmitted from the camera and the lidar. When the microcontroller detects a fall event, it immediately sends a signal to the alarm module to control the alarm module to generate an alarm.

2. The worker fall behavior recognition detection device according to claim 1, characterized in that: The second mounting base includes a base fixed in a preset position; the lidar is mounted on the base via a ball joint to facilitate adjustment of the lidar's mounting angle.

3. The worker fall behavior recognition detection device according to claim 2, characterized in that: The base includes a fixed seat, on which a longitudinally extending groove is provided, and a movable plate is slidably installed in the groove, and the ball hinge is installed on the movable plate.

4. The worker fall behavior recognition detection device according to claim 3, characterized in that: The ball hinge includes a rotating seat fixed to the movable plate, a rotating ball movably installed inside the rotating seat, a screw hole on the rotating seat, a locking bolt screwed into the screw hole, the end of the locking bolt abutting against the rotating ball, and the rotating ball being connected to the lidar.

5. The worker fall behavior recognition detection device according to claim 3, characterized in that: The fixed base is provided with a groove, a push plate is provided in the groove, and an elastic component is provided between the push plate and the fixed base.

6. The worker fall incident recognition detection device according to claim 5, characterized by: The elastic component is a compression spring.

7. The worker fall behavior recognition detection device according to claim 1, characterized in that: The alarm module is an audible and visual alarm.

Images