Human body tracking-based intelligent pan-tilt head

Abstract

The application relates to an intelligent cloud platform machine based on human tracking and belongs to the technical application field of cameras, which comprises a camera unit, a cloud platform support and a cloud platform management unit, the camera unit, the cloud platform support and the cloud platform management unit are electrically connected; the camera unit is used for collecting video images in real time to extract face information and human motion track information; the cloud platform management unit is used for receiving the human motion track information to generate a rotating instruction; the cloud platform support is fixedly connected with the camera unit, the cloud platform support is used for receiving the rotating instruction and relatively rotating. Through the cooperative matching among the camera unit, the cloud platform support and the cloud platform management unit, the motion track of a construction worker in an industrial production environment is accurately monitored in real time, meanwhile, the intelligent cloud platform machine can dynamically adjust and control the rotating angle of the cloud platform, so that the camera can always be aligned with the moving construction worker.

Description

Technical Field

[0001] This application relates to the field of camera technology, and more particularly to an intelligent gimbal camera based on human body tracking. Background Technology

[0002] In industrial construction environments, safety monitoring is a crucial aspect. Construction sites typically include multiple boundary areas, such as hazardous work areas, material storage areas, and machinery operation areas. These areas require strict control to ensure the safety of construction personnel and the smooth progress of the project.

[0003] Traditional monitoring methods primarily rely on fixed cameras and manual patrols. However, relying on manual on-site inspections is not only time-consuming and labor-intensive, but also prone to failing to detect safety hazards in a timely manner due to human negligence or fatigue. Especially at large construction sites, manual patrols struggle to achieve real-time, comprehensive monitoring. Fixed cameras, on the other hand, have limited coverage, meaning they can only monitor specific areas and cannot be flexibly adjusted, resulting in blind spots in some areas, particularly when construction workers move between different areas, making comprehensive coverage difficult. Furthermore, traditional monitoring systems typically only provide video recording and lack intelligent analysis capabilities for construction worker behavior. For example, they cannot automatically identify potential safety risks (such as not wearing safety helmets or entering dangerous areas), making post-event review complex and inefficient.

[0004] Therefore, there is an urgent need for a camera device that can dynamically monitor and analyze the behavior of construction workers. Utility Model Content

[0005] This application provides an intelligent PTZ camera based on human body tracking, which aims to solve the technical problem that existing monitoring equipment cannot dynamically supervise and analyze the behavior of construction workers.

[0006] To achieve the above objectives, this application provides an intelligent gimbal camera based on human body tracking, including a camera unit, a gimbal bracket, and a gimbal management unit. The camera unit, the gimbal bracket, and the gimbal management unit are electrically connected. The camera unit is used to acquire video images in real time to extract facial information and human motion trajectory information. The gimbal management unit is used to receive human motion trajectory information to generate rotation commands. The gimbal bracket is fixedly connected to the camera unit and is used to receive the rotation commands and rotate relative to the camera unit.

[0007] Furthermore, the gimbal support includes a horizontal rotation motor and a pitch rotation motor, wherein the rotation angle of the horizontal rotation motor is 0° to 355° and the rotation angle of the pitch rotation motor is 0° to 90°.

[0008] Furthermore, the gimbal bracket is also equipped with a buffer damper made of silicone material, which is used to reduce the start-stop impact of the horizontal rotary motor and the pitch rotary motor.

[0009] Furthermore, the intelligent gimbal also includes an audio sensor connected to the gimbal management unit. The audio sensor is used to collect environmental noise data in the scene, and the environmental noise data is used to fuse with the video images collected by the camera unit.

[0010] Furthermore, the intelligent gimbal also includes a hazardous area configuration module, which is used to define the boundaries of hazardous areas in the scene through a graphical interface.

[0011] Furthermore, the intelligent gimbal also includes a trajectory analysis module, which performs real-time comparison between the boundary of the danger zone and the human movement trajectory information to determine whether a person is in the danger zone.

[0012] Furthermore, the intelligent gimbal also includes an audible and visual alarm unit, which is used to receive the judgment result of the trajectory analysis module and issue a warning message based on the judgment result.

[0013] Furthermore, the intelligent gimbal also includes a remote notification unit, which is used to receive the judgment result of the trajectory analysis module and send alarm information to the preset management personnel based on the judgment result.

[0014] Furthermore, the intelligent gimbal also includes a face comparison module, which is used to receive the face information and compare it with a preset face database, and generate a face comparison result.

[0015] Furthermore, the intelligent PTZ camera also includes a storage unit, which is connected to the camera unit and the remote notification unit. The storage unit is used to store a face database and to store the video images and alarm information.

[0016] Beneficial effects:

[0017] The intelligent PTZ camera disclosed in this application cleverly utilizes the close collaboration between the camera unit, the PTZ bracket, and the PTZ management unit to achieve accurate and real-time monitoring of the movement trajectory of construction workers in industrial production environments, thus effectively solving the problem of monitoring the movement of construction workers in complex production environments. Simultaneously, based on real-time data analysis of the movement of construction workers in the production environment, the intelligent PTZ camera can dynamically adjust and control the rotation angle of the PTZ to ensure that the camera is always pointed at the moving target. This innovative design not only significantly improves the flexibility and accuracy of monitoring but also fundamentally solves the problem of missed detection caused by the inability of fixed cameras in traditional industrial monitoring to dynamically track fast-moving targets, greatly enhancing the safety management level of industrial production environments. Attached Figure Description

[0018] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0019] Figure 1 This is a schematic diagram of the structure of an intelligent gimbal based on human body tracking, provided as an embodiment of this application.

[0020] Among them, 100 is the camera unit; 200 is the pan-tilt bracket; 300 is the pan-tilt management unit; 400 is the audio sensor; 500 is the hazardous area configuration module; 600 is the trajectory analysis module; 700 is the sound and light alarm unit; 800 is the remote notification unit; 900 is the face comparison module; and 110 is the storage unit. Detailed Implementation

[0021] The technical solutions of the present utility model will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present utility model, and not all embodiments. Based on the embodiments of the present utility model, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of the present utility model.

[0022] It should be noted that the directional terms such as left, right, up, down, front, and back in the embodiments of this utility model are only relative concepts or are based on the normal use state of the product, i.e., the direction of the product's movement, and should not be considered as limiting.

[0023] In addition, it should be noted that the dynamic terms such as "relative motion" mentioned in the embodiments of this utility model not only refer to changes in position, but also include movements such as rotation and rolling in which the position does not change relative to the position, but the state changes.

[0024] Finally, it should be noted that when a component is said to be "located on" or "set on" another component, it can be on the other component or may have an intervening component at the same time. When a component is said to be "connected to" another component, it can be directly connected to the other component or may have an intervening component at the same time.

[0025] Figure 1 This is a schematic diagram of the structure of an intelligent gimbal based on human body tracking, provided as an embodiment of this application. Figure 1 As shown, the intelligent gimbal includes a camera unit 100, a gimbal bracket 200, and a gimbal management unit 300. The camera unit 100, the gimbal bracket 200, and the gimbal management unit 300 are electrically connected. The camera unit 100 is used to acquire video images in real time to extract facial information and human motion trajectory information. The gimbal management unit 300 is used to receive human motion trajectory information to generate rotation commands. The gimbal bracket 200 is fixedly connected to the camera unit 100 and is used to receive the rotation commands and rotate relative to the camera unit.

[0026] This application, through the collaborative design of the camera unit 100, the pan-tilt bracket 200, and the pan-tilt management unit 300, solves the problem of monitoring the movement trajectory of construction workers in an industrial production environment.

[0027] Specifically, the camera unit 100 is used to acquire video images in real time, extract facial information and human movement trajectory information of construction workers. The camera unit 100 uses a high-resolution camera (such as a 200W high-definition night vision camera) to support image acquisition in complex environments (strong light, backlight, smoke). It also incorporates a human body recognition algorithm, enabling real-time analysis and tracking of human movement trajectories through a deep learning model. Furthermore, it supports a combination of binocular cameras and thermal imaging cameras to enhance detection capabilities under low light or obstructed conditions. The pan-tilt management unit 300 receives human movement trajectory information from the camera unit 100, generates rotation commands, and analyzes human contours and key point positioning (such as 68 facial features) through image sequence analysis. Combined with millimeter-wave radar and infrared sensor data, it improves the accuracy of judging the movement trajectory of construction workers. The pan-tilt bracket 200 can drive the camera unit 100 to rotate according to the commands of the pan-tilt management unit 300, achieving dynamic tracking of construction workers. The pan-tilt bracket 200 is fixedly connected to the camera unit 100, achieving stable installation through threads or rivets, and is adapted to the vibration and impact of the industrial environment.

[0028] The camera unit 100 separates the moving target from the background using a background removal algorithm (such as a Gaussian mixture model), represents the human body contour using chain code, and calculates the joint coordinates using a Shenmu algorithm to generate a two-dimensional human skeleton sequence. Combined with distance data from millimeter-wave radar, it corrects the depth information in the image to achieve three-dimensional trajectory reconstruction. The gimbal management unit 300 dynamically adjusts the gimbal rotation angle based on the human body's movement speed and direction (e.g., turning in advance after predicting the trajectory). It also communicates with the gimbal bracket 200 via a UART interface, sending PID control parameters in real time to ensure rotational synchronization.

[0029] The intelligent PTZ camera disclosed in this application works in concert with the camera unit 100, the PTZ bracket 200 and the PTZ management unit 300, and optimizes the smoothness of PTZ rotation through PID control algorithm. It combines thermal imaging and binocular camera to improve adaptability to complex environments, and solves the problem that fixed cameras in traditional industrial monitoring cannot dynamically track fast-moving targets, which leads to the risk of missed detection.

[0030] In one embodiment, the gimbal support 200 includes a horizontal rotation motor and a pitch rotation motor, wherein the horizontal rotation motor has a rotation angle of 0° to 355° and the pitch rotation motor has a rotation angle of 0° to 90°.

[0031] By incorporating horizontal and vertical rotary motors, the pan-tilt mount 200 achieves both horizontal and vertical rotation, enabling the intelligent pan-tilt unit to perform comprehensive dynamic tracking of personnel in industrial environments. Specifically, the horizontal rotary motor controls the horizontal rotation of the camera unit 100, with a rotation angle ranging from 0° to 355°. This near 360-degree horizontal rotation ensures that the camera unit 100 covers almost the entire plane, avoiding blind spots common with fixed cameras. The vertical rotary motor controls the vertical rotation of the camera unit 100, increasing its vertical adjustment capability. This allows the camera unit 100 to not only scan horizontally but also adjust its vertical viewing angle to adapt to different working environments (such as elevated platforms and underground passages). Through the coordinated operation of the horizontal and vertical rotary motors, the pan-tilt mount 200 enables the camera unit 100 to respond to personnel movement instantly, providing an efficient and accurate safety monitoring solution for industrial scenarios. This reduces the false alarm rate of intelligent pan-tilt units, improves the timeliness of early warnings, and effectively prevents safety accidents.

[0032] In one embodiment, the gimbal bracket 200 is further provided with a buffer damper, which is made of silicone material and is used to reduce the start-stop impact of the horizontal rotary motor and the pitch rotary motor.

[0033] By setting up a buffer damper to absorb and reduce the impact force generated when the motor starts and stops, the service life of the motor and its related components is extended, maintenance costs and replacement frequency are reduced, and the stability and lifespan of the intelligent gimbal machine are improved. In addition, it can also prevent mechanical vibration caused by instantaneous acceleration or deceleration from being transmitted to the camera unit 100, making the images captured by the camera unit 100 more stable, avoiding blurring caused by shaking, and improving image clarity.

[0034] In one embodiment, the intelligent gimbal also includes an audio sensor 400, which is connected to the gimbal management unit 300. The audio sensor 400 is used to collect environmental noise data in the scene, and the environmental noise data is used to fuse with the video images collected by the camera unit 100.

[0035] By integrating an audio sensor 400, the intelligent PTZ camera's ability to monitor personnel activity and abnormal situations in industrial environments is enhanced, improving the accuracy and reliability of the overall monitoring system. Specifically, the audio sensor 400 is used to collect environmental noise data in industrial scenarios. It is installed near the PTZ bracket 200 or the camera unit 100 to ensure effective capture of surrounding sound information. The audio sensor 400 can detect weak sound signals, making it suitable for complex and changing industrial environments. The PTZ management unit 300 receives and processes data from the audio sensor 400 and the camera unit 100, achieving multimodal data fusion and improving the accuracy of recognizing human movement trajectories and abnormal behaviors.

[0036] When the system starts, the audio sensor 400 begins collecting ambient noise data and removes background noise using a built-in filter. The collected sound signal is then converted into a digital signal and transmitted to the PTZ management unit 300 for further processing. Simultaneously, audio feature extraction algorithms (such as Mel-frequency cepstral coefficients, MFCC) are used to extract key features from the sound signal, such as volume and frequency distribution. The PTZ management unit 300 matches the audio features with visual features in the video image. For example, sound localization technology is used to determine the direction of the sound source, and the video image is combined to confirm the presence of human activity. Deep learning models (such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs) are used to analyze the fused data, improving the accuracy of recognizing human movement trajectories and abnormal behaviors.

[0037] By combining audio data with video images, more comprehensive and accurate scene monitoring is achieved. For example, in low-light or obstructed conditions, relying solely on video images may be insufficient for accurate judgment, while audio data can provide additional reference information. Furthermore, the integration of data from multiple sensors enables the monitoring system to better perceive changes in the surrounding environment, reducing the possibility of misjudgments from a single sensor and improving the system's reliability and stability.

[0038] In one embodiment, the intelligent PTZ camera further includes a danger zone configuration module 500, which is used to define the boundary of the danger zone in the scene through a graphical interface. The intelligent PTZ camera also includes a trajectory analysis module 600, which compares the boundary of the danger zone with the human movement trajectory information in real time to determine whether a person is in the danger zone. The intelligent PTZ camera also includes an audible and visual alarm unit 700, which receives the judgment result from the trajectory analysis module 600 and issues a warning message based on the judgment result.

[0039] The intelligent PTZ camera integrates a hazardous area configuration module 500, a trajectory analysis module 600, and an audible and visual alarm unit 700 to achieve precise monitoring and timely early warning of personnel activities in industrial environments, ensuring personnel safety.

[0040] The hazardous area configuration module 500 can define and adjust the boundaries of hazardous areas in a scene through a graphical user interface (GUI). The intelligent PTZ camera provides a visual interface that allows users to draw or adjust the boundaries of hazardous areas on the monitoring video image. The data processing unit (not shown) receives user input and converts it into coordinate data recognizable by the system, storing it in a database for later use. The communication interface communicates with the PTZ management unit 300 and other sensor modules to ensure that the defined hazardous area information can be applied to the monitoring and early warning system in real time. The trajectory analysis module 600 can compare the hazardous area boundary with human movement trajectory information in real time to determine whether a person has entered the hazardous area. The data fusion unit (not shown) receives human movement trajectory information from the camera unit 100 and boundary data from the hazardous area configuration module 500, and uses geometric matching algorithms or other machine learning methods to compare the human movement trajectory with the hazardous area boundary in real time to determine whether a person has entered the hazardous area. The decision unit (not shown) generates a corresponding judgment signal based on the comparison result and transmits it to the audible and visual alarm unit 700. The audible and visual alarm unit 700 receives the judgment result from the trajectory analysis module 600 and issues warning information (such as sound alarm and light signal) based on the result. The control circuit (not shown) can receive instructions from the trajectory analysis module 600 and control the start and stop of the audible and visual alarm unit 700.

[0041] Specifically, when the system starts, the user draws or adjusts the boundaries of the danger zone on the monitoring screen through a graphical interface (such as a mouse or touchscreen). For example, the user can select shape tools such as rectangles and polygons to mark specific areas that need to be monitored on the screen. The trajectory analysis module 600 obtains human movement trajectory information from the camera unit 100 and obtains the defined danger zone boundary data from the danger zone configuration module 500. Using geometric matching algorithms (such as point-in-polygon testing) or machine learning models (such as convolutional neural networks CNN), it compares the human movement trajectory with the danger zone boundary in real time to determine whether the person has entered the danger zone. If the comparison result shows that the person has entered the danger zone, a corresponding judgment signal is generated and transmitted to the audible and visual alarm unit 700. The audible and visual alarm unit 700 receives the judgment result from the trajectory analysis module 600, determines whether an alarm needs to be issued, and if the judgment result shows that the person has entered the danger zone, the audible and visual alarm unit 700 immediately activates the buzzer, speaker, and LED lights to emit sound and light signals.

[0042] This intelligent PTZ camera is suitable for different industrial environments. Users can adjust the dangerous area boundaries and alarm strategies of the intelligent PTZ camera in real time according to specific circumstances, so that the intelligent PTZ camera can automatically perform real-time monitoring without human intervention, reducing the possibility of human error. When it detects that someone has entered the dangerous area, the system can quickly issue voice reminders, light signal alarms or remote notifications, effectively preventing potential safety accidents.

[0043] In one embodiment, the intelligent PTZ camera further includes a remote notification unit 800, which is used to receive the judgment result of the trajectory analysis module 600 and send alarm information to a preset management personnel based on the judgment result. The intelligent PTZ camera further includes a face comparison module 900, which is used to receive the face information and compare it with a preset face database, and generate a face comparison result.

[0044] This intelligent PTZ camera, by integrating a remote notification unit 800 and a face comparison module 900, further enhances its security monitoring capabilities, enabling accurate identification and timely warning of personnel activities. It can also accurately identify the identity of personnel entering the monitored area, ensuring personnel safety in industrial environments. The remote notification unit 800 receives the judgment results from the trajectory analysis module 600 and sends alarm information to preset management personnel based on these results. The remote notification unit 800 supports multiple communication methods (such as SMS, email, and mobile application push notifications) to ensure timely delivery of alarm information to relevant personnel. Users can also view the monitoring screen and historical alarm records in real time via a web or mobile terminal. The face comparison module 900 receives facial information collected by the camera unit 100 and compares it with a preset face database to generate a face comparison result. For example, the face feature extraction unit uses deep learning algorithms (such as convolutional neural networks CNN) to extract facial feature points from video images, and the face database stores preset face information as the basis for comparison. The face comparison module 900 uses comparison algorithms (such as Euclidean distance and cosine similarity) to calculate the similarity between the currently collected facial features and the records in the face database, and generates comparison results to determine the identity of the person entering the dangerous area.

[0045] In one embodiment, the intelligent PTZ camera further includes a storage unit 110, which is connected to the camera unit 100 and the remote notification unit 800. The storage unit 110 stores a face database and also stores the video images and alarm information. By setting up the storage unit 110, the system can completely record all video images, face information, and alarm events, ensuring data integrity. It also supports long-term recording and storage of high-definition video, facilitating post-event analysis and evidence preservation, and reducing the risk of data loss. Furthermore, all alarm events and related video images are recorded by the storage unit 110, facilitating post-event analysis and responsibility allocation, thus improving system reliability.

[0046] In summary, the intelligent PTZ camera based on human body tracking described in this application, through the collaborative work of the camera unit 100, the PTZ bracket 200, and the PTZ management unit 300, achieves real-time acquisition and dynamic tracking of human movement trajectories in the monitored scene, improving the intelligence and automation level of monitoring. The PTZ bracket 200 adopts a horizontal and pitch motor structure and is equipped with a silicone buffer damper, which effectively improves rotational stability and equipment lifespan. Combined with the audio sensor 400, the hazardous area configuration module 500, the trajectory analysis module 600, the audible and visual alarm unit 700, the remote notification unit 800, the face comparison module 900, and the storage unit 110, it realizes multi-dimensional perception of personnel activities, intelligent identification of dangerous behaviors, real-time early warning, and data recording, enhancing the system's adaptability and safety protection capabilities in complex industrial environments. It has advantages in many aspects, such as reasonable structure, rapid response, comprehensive monitoring, and intelligent operation, and is suitable for various application scenarios such as industrial safety monitoring and key area protection.

[0047] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A smart gimbal camera based on human body tracking, characterized in that, include: The camera unit (100), the gimbal bracket (200), and the gimbal management unit (300) are electrically connected; The camera unit (100) is used to acquire video images in real time to extract facial information and human motion trajectory information; The gimbal management unit (300) is used to receive human motion trajectory information to generate rotation commands; The gimbal bracket (200) is fixedly connected to the camera unit (100), and the gimbal bracket (200) is used to receive the rotation command and rotate relative to the camera unit.

2. The intelligent gimbal camera based on human body tracking according to claim 1, characterized in that, The gimbal bracket (200) includes a horizontal rotation motor and a pitch rotation motor. The rotation angle of the horizontal rotation motor is 0° to 355°, and the rotation angle of the pitch rotation motor is 0° to 90°.

3. The intelligent gimbal camera based on human body tracking according to claim 2, characterized in that, The gimbal bracket (200) is also equipped with a buffer damper, which is made of silicone and is used to reduce the start-stop impact of the horizontal rotary motor and the pitch rotary motor.

4. The intelligent gimbal camera based on human body tracking according to claim 1, characterized in that, The intelligent gimbal also includes an audio sensor (400), which is connected to the gimbal management unit (300). The audio sensor (400) is used to collect environmental noise data in the scene, and the environmental noise data is used to fuse with the video images collected by the camera unit (100).

5. A smart gimbal camera based on human body tracking according to claim 1, characterized in that, The intelligent gimbal also includes a hazardous area configuration module (500), which is used to define the boundaries of hazardous areas in the scene through a graphical interface.

6. A smart gimbal camera based on human body tracking according to claim 5, characterized in that, The intelligent gimbal also includes a trajectory analysis module (600), which compares the boundary of the danger zone with the human movement trajectory information in real time to determine whether a person is in the danger zone.

7. A smart gimbal camera based on human body tracking according to claim 6, characterized in that, The intelligent PTZ camera also includes an audible and visual alarm unit (700), which is used to receive the judgment result of the trajectory analysis module (600) and issue a warning message based on the judgment result.

8. A smart gimbal camera based on human body tracking according to claim 6, characterized in that, The intelligent PTZ camera also includes a remote notification unit (800), which is used to receive the judgment result of the trajectory analysis module (600) and send alarm information to the preset management personnel based on the judgment result.

9. A smart gimbal camera based on human body tracking according to claim 8, characterized in that, The intelligent gimbal also includes a face comparison module (900), which is used to receive the face information and compare it with a preset face database, and generate a face comparison result.

10. A smart gimbal camera based on human body tracking according to claim 9, characterized in that, The intelligent PTZ camera also includes a storage unit (110), which is connected to the camera unit (100) and the remote notification unit (800). The storage unit (110) is used to store a face database and to store the video images and alarm information.

Images