A steel plant safety monitoring method based on video twinning

Abstract

The present application relates to the field of steel plant safety monitoring, and relates to a steel plant safety monitoring method based on video twinning, which comprises the following steps: constructing a three-dimensional digital twin simulation scene corresponding to different regions of a steel plant by a multi-sensor fusion technology according to a certain proportion; constructing a third-person free-roaming perspective mode in the three-dimensional digital twin simulation scene so as to freely monitor the steel plant; determining a target monitoring region of the steel plant digital twin that needs to be monitored by video and acquiring video information corresponding to the target monitoring region by using a camera device; according to a time sequence, a spatial position and a direction relationship, virtually and realistically fusing the video information of the target monitoring region with the three-dimensional digital twin simulation scene to generate a video twinning system; and according to the video twinning system, achieving real-time multi-angle observation of the target monitoring region inside the steel plant. The steel plant safety monitoring method provided by the present application brings great convenience to the safe production and efficient monitoring of a steel plant.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of steel plant safety monitoring, in particular to a steel plant safety monitoring method based on video twinning, more specifically to a method of spatiotemporal matching of real target camera information with a virtual three-dimensional scene, virtual-real fusion, for real-time safety monitoring of a steel plant. BACKGROUND

[0002] With the continuous development of science and technology and the continuous advancement of social infrastructure construction, the automation level of steel plant production equipment is continuously improving, and many safety problems in the production process of steel plants have also been exposed. Steel production is a dangerous industry involving high temperature, high pressure, harmful gases and chemicals, and other dangerous factors. Therefore, steel plant safety monitoring is crucial. Correct monitoring and management can ensure the safety of workers and reduce production accidents and equipment damage.

[0003] In recent years, with the continuous development of monitoring technology, video monitoring has been increasingly widely used in steel production. Video monitoring can monitor equipment operation status, worker safety conditions and on-site environment, etc. In addition, combined with artificial intelligence technology, monitoring data can be analyzed to find potential problems, such as detection and diagnosis of equipment failure, personnel access, temperature changes, gas leaks, etc., which helps to improve the efficiency and safety of steel plant production.

[0004] However, traditional video monitoring technology often requires a lot of manpower and resources to observe different camera pictures, which greatly increases the cost of manpower, causing waste of resources and personnel redundancy; since traditional video monitoring requires corresponding personnel to observe the pictures to avoid safety problems, it is easy to cause accidents due to personnel observation negligence; in addition, traditional video monitoring cannot match the display of the environment around the target area because it intelligently captures the pictures of the target area, and the safety monitoring effect is relatively poor; due to the many shortcomings of traditional video monitoring, designing a new steel plant safety monitoring method is not only to fill the gap, but also a major contribution to steel plant safety monitoring.

[0005] Video twinning, i.e. video + digital twinning, is a further supplement and development of digital twinning technology, which embeds real reality video pictures into a three-dimensional digital twinning system scene to realize virtual-real fusion, and can more realistically reflect reality in a three-dimensional scene. SUMMARY

[0006] In view of the above technical problems, the present application provides a steel plant safety monitoring method based on video twinning; the present application utilizes the virtual-real fusion characteristics of the video twinning system, utilizes multi-sensor fusion technology, and constructs a three-dimensional scene according to a certain proportion for the equipment environment of the steel plant site; then, video information of a target area needing safety monitoring is acquired; the video information and the three-dimensional scene are virtually and realistically fused, thereby realizing a new safety monitoring mode.

[0007] To achieve the above-mentioned purpose, the present application adopts the following technical solutions:

[0008] A steel plant safety monitoring method based on video twinning, the method comprising:

[0009] According to a certain proportion, a three-dimensional digital twinning simulation scene corresponding to different regions of the steel plant is constructed by multi-sensor fusion technology;

[0010] A third-person free-roaming perspective mode is constructed in the three-dimensional digital twinning simulation scene, so as to facilitate free monitoring of the steel plant;

[0011] A target monitoring area needing video monitoring of the steel plant digital twinning is determined, and a video information corresponding to the target monitoring area is acquired by using a camera device;

[0012] According to time series, spatial position and direction relationship, the video information of the target monitoring area is virtually and realistically fused with the three-dimensional digital twinning simulation scene, thereby generating a video twinning system;

[0013] According to the video twinning system, real-time multi-angle observation of the target monitoring area inside the steel plant is realized.

[0014] Further, the three-dimensional digital twinning simulation scene corresponding to different regions of the steel plant is constructed by multi-sensor fusion technology according to a certain proportion, and the specific method is:

[0015] 1) Point cloud data of the corresponding equipment and environment of the steel plant is acquired by using a laser radar, that is, a data set reflected back by a laser beam emitted by the laser radar when encountering an object; each point cloud data includes three-dimensional coordinates, reflection intensity, normal vector, and color information;

[0016] 2) Image data of the corresponding equipment and environment of the steel plant is acquired by using a camera device;

[0017] 3) The point cloud data acquired by the laser radar and the image data acquired by the camera device are fused, that is, the geometric information of the point cloud data and the texture information of the image data are combined to build a three-dimensional model, thereby obtaining a model corresponding to the equipment and environment of the steel plant;

[0018] 4) Import the established model corresponding to the steel plant equipment and environment into the computer software to build a three-dimensional scene of the steel plant.

[0019] Further, a third-person free-roaming perspective mode is constructed in the three-dimensional digital twin simulation scene, the direction of the scene is rotated by a mouse, and the movement of the scene is controlled by a keyboard. The specific implementation method is:

[0020] 1) Set the movement amount of the mouse in the horizontal and vertical directions as x and y, set the mouse sensitivity as m, and set the time of each frame of the video as t. Calculate the actual horizontal displacement X = x * m * t and the actual vertical displacement Y = y * m * t in the three-dimensional scene when the mouse is sliding.

[0021] 2) Apply the actual vertical displacement Y to the vertical rotation (rotation around the x-axis) of the three-dimensional scene camera, and apply the horizontal displacement X to the horizontal rotation (rotation around the y-axis) of the three-dimensional scene camera. The view angle change in the three-dimensional scene is realized by moving the mouse.

[0022] 3) Set a three-dimensional vector mov containing three dimensions (xr, yr, zr), wherein the xr component is set to the value of the horizontal input axis, the yr component is set to 0, and the zr component is set to the value of the vertical input axis. The movement direction of the three-dimensional scene is controlled by the three-dimensional vector mov.

[0023] 4) Set the movement speed of the scene as v, and set the time of each frame of the video as t. The displacement S = v * mov * t of the scene moving controlled by the keyboard is calculated, and the displacement S represents the movement in a certain component direction.

[0024] Further, the target monitoring area of the steel plant digital twin that needs to be monitored by video is determined, and the video information corresponding to the target monitoring area is obtained by using a camera device. The specific method is:

[0025] 1) Clearly define the target object and area that needs to be monitored and observe them.

[0026] 2) Use multi-camera fusion technology to fix multiple static cameras at different positions and angles to take pictures of the target object and area to obtain more comprehensive and detailed perspective information.

[0027] 3) Use dynamic cameras to move and record from different heights, angles, and positions to obtain more diverse and comprehensive information.

[0028] 4) Send request information to the target network camera in the static camera and / or the dynamic camera through a computer device to obtain the corresponding video information.

[0029] Further, for the video information obtained by the static camera and / or the dynamic camera, a DeepSORT algorithm is used to detect and track a specific target in the video information in real time.

[0030] Further, the virtual-real fusion of the video information of the target monitoring area and the three-dimensional digital twin simulation scene includes:

[0031] The video information of the camera in the target monitoring area and the object information corresponding to the three-dimensional digital twin are matched in terms of time, space and direction, so as to perform virtual-real fusion and generate a video twin system.

[0032] The video information includes video pictures captured by the static camera and the dynamic camera; and the object information corresponding to the three-dimensional digital twin refers to a three-dimensional scene modeling around the video pictures captured by the static camera and the dynamic camera.

[0033] Further, in the video twin system, for different shooting pictures of different angle cameras in the target monitoring area, the angle corresponding to the current third person free roaming perspective mode is automatically detected and the camera picture of the corresponding angle is automatically switched to.

[0034] Further, the method for automatically detecting the angle corresponding to the current third person free roaming perspective mode and automatically switching to the camera picture of the corresponding angle for different shooting pictures of different angle cameras in the target monitoring area is:

[0035] A target monitoring area includes n cameras C1, C2, C3, …, Cn. n The n cameras are uniformly distributed around the target monitoring area at an angle of (360 / n)°. Based on a certain angle, the shooting angles of the n cameras are 0°-(360 / n)°, (360 / n)°-2*(360 / n)°, …, (n-1)*(360 / n)°-360°, respectively.

[0036] In the third person free roaming perspective mode, the change of the perspective of the scene is controlled by the movement of the mouse. When the rotation angle of the mouse is 0°-(360 / n)°, the shooting picture of the camera C1 is displayed; when the rotation angle of the mouse is (360 / n)°-2*(360 / n)°, the shooting picture of the camera C2 is displayed; …, when the rotation angle of the mouse is (n-1)*(360 / n)°-360°, the shooting picture of the camera Cn is displayed. n

[0037] ​Further, for different shooting pictures of different angle cameras in the target monitoring area, the current third person free roaming perspective mode corresponding angle is automatically detected and automatically switched to the corresponding angle camera picture, and the specific method is:

[0038] 1) New n plane objects in the three-dimensional digital twin simulation scene are built, which are set as P1, P2, P3, …, P n ; the planes P1, P2, P3, …, P n are respectively used to carry the camera C1, C2, C3, …, C n ; the initial state is set as the camera C1 being activated, and C2, C3, …, C n being closed.

[0039] 2) In the third person free roaming perspective mode, the rotation angle of the scene is represented by the actual horizontal displacement X of the mouse, and the value of the actual horizontal displacement X is limited in the range of (a, b), a represents the lower limit of X, and b represents the upper limit of X; if the value of X is lower than a, it will be limited to a, and if the value of X is higher than b, it will be limited to b; the value of X is limited between 0° and 360°.

[0040] 3) When the value of X is between 0° and (360 / n)°, display the plane P1 and hide all the remaining planes, that is, only display the camera C1 shooting picture; when the value of X is between (360 / n)° and 2*(360 / n)°, display the plane P2 and hide all the remaining planes, that is, only display the camera C2 shooting picture; …; when the value of X is between (n-1)*(360 / n)° and 360°, display the plane P n and hide all the remaining planes, that is, only display the camera C n shooting picture.

[0041] Further, the number of plane objects newly built in the three-dimensional digital twin simulation scene and the number of all cameras in the three-dimensional digital twin simulation scene are consistent, and the size, shape and position of all planes should be completely consistent.

[0042] The beneficial technical effects of the present application are:

[0043] (1) The method provided by the present application greatly saves the cost of human resources, and through the video twin technology, multiple target areas that need to be monitored in the steel plant are integrated into the same video twin system, which can be observed in the three-dimensional scene, saving resources and manpower.

[0044] (2) The method provided by the application fuses the video information of the target area and the three-dimensional digital twin scene, and the system can monitor the on-site environment in two ways of automatic monitoring and manual observation, thereby avoiding on-site accidents caused by personnel negligence or system failure and greatly improving the safety production efficiency of the on-site environment.

[0045] (3) The method provided by the application generates a video twin system, and the video twin system builds a 1:1 on-site device and environment model around each target monitoring area, which helps to more realistically reflect the on-site actual situation and more vividly and realistically perceive and monitor the on-site environment. BRIEF DESCRIPTION OF DRAWINGS

[0046] Figure 1 It is a flowchart of the steel plant safety monitoring method based on video twins mentioned in the embodiments of the application.

[0047] Figure 2 It is a flowchart of the method for automatically detecting the angle corresponding to the third-person free-roaming perspective mode and automatically switching to the camera view of the corresponding angle. Wherein, θ represents the rotation angle of the scene, P1, P2 and P3 represent three planes carrying the camera view respectively. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical scheme and advantages of the application more clear and understandable, the application will be further described in detail below with examples. It should be understood that the specific examples described herein are only used to explain the application and not to limit the application. On the contrary, the application covers any alternative, modification, equivalent method and scheme defined by the claims within the essence and scope of the application. Further, in order to make the public have a better understanding of the application, some specific details are described in detail in the following detailed description of the application. The application can also be completely understood without the description of these details by those skilled in the art.

[0049] As shown in Figure 1 The application provides a steel plant safety monitoring method based on video twins, which comprises the following steps:

[0050] According to a certain ratio (preferably a ratio of 1:1), a three-dimensional digital twin simulation scene corresponding to different areas of the steel plant is constructed based on 3ds Max, Maya and Unity3D;

[0051] A third-person free-roaming perspective mode is constructed in the three-dimensional digital twin simulation scene, so as to freely monitor the steel plant;

[0052] The target video monitoring area for which video monitoring is required is determined, and video information corresponding to the target video monitoring area is obtained by using a camera device;

[0053] According to the time sequence, spatial position, and direction relationship, the video information of the target video monitoring area is fused with the three-dimensional digital twin simulation scene to generate a video twin system.

[0054] According to the video twin system, real-time multi-angle observation of the target monitoring area in the steel plant is realized.

[0055] In the video twin system, for different shooting pictures of different angle cameras in the target video monitoring area, the angle corresponding to the current third-person free-roaming perspective mode is automatically detected, and the camera picture of the corresponding angle is automatically switched to.

[0056] The three-dimensional digital twin simulation scene corresponding to different areas of the steel plant is constructed according to a certain ratio, and the specific method is as follows:

[0057] 1) Point cloud data of the corresponding equipment and environment of the steel plant is obtained by using a laser radar, that is, a data set reflected back by the laser beam emitted by the laser radar when encountering an object, each point cloud data including three-dimensional coordinates, reflection intensity, normal vector, and color information.

[0058] 2) Image data of the corresponding equipment and environment of the steel plant is obtained by using a camera device (camera or depth camera, etc.).

[0059] 3) The point cloud data obtained by the laser radar and the image data obtained by the camera device are fused, the geometric information of the point cloud data and the texture information of the image are combined, and a three-dimensional model is built to obtain a model corresponding to the equipment and environment of the steel plant.

[0060] 4) The established model corresponding to the equipment and environment of the steel plant is imported into a computer software to build a three-dimensional scene of the steel plant. The computer software used in this step can be Unity or ue4, ue5, etc. computer software, which is a conventional technology.

[0061] A third-person free-roaming perspective mode is constructed in the three-dimensional digital twin simulation scene, and the basic operation is that the left and right movements of the mouse control the rotation of the scene in the direction, and the keyboard (which can be controlled by WASD keys) controls the movement of the scene. The specific implementation method is as follows:

[0062] 1) Let the horizontal and vertical movements of the mouse be x and y respectively, let the mouse sensitivity be m, and let the time of each frame of video be t. According to this, the actual horizontal displacement X = x * m * t and the actual vertical displacement Y = y * m * t in the three-dimensional scene can be calculated when the mouse is sliding.

[0063] 2) Apply the actual vertical movement Y to the vertical rotation (rotation around the x-axis) of the three-dimensional scene camera, and apply the horizontal movement X to the horizontal rotation (rotation around the y-axis) of the three-dimensional scene camera; the change of the visual angle in the three-dimensional scene can be controlled by the movement of the mouse.

[0064] 3) Set a three-dimensional vector mov containing three dimensions (xr, yr, zr), wherein the xr component is set to the value of the horizontal input axis (keyboard A and D keys), the yr component is set to 0, and the zr component is set to the value of the vertical input axis (keyboard W and S keys). The movement direction of the three-dimensional scene is controlled by the three-dimensional vector mov.

[0065] 4) Set the movement speed of the scene as v, and the time of each frame of the video as t, then the displacement S of the scene movement controlled by the keyboard WASD is S = v * mov * t, and the displacement S represents the movement in a certain component direction.

[0066] To determine the target video monitoring area of the steel number digital twin that needs video monitoring and obtain the video information corresponding to the target video monitoring area by using a camera device; the specific method is:

[0067] 1) Clearly define the target object and area that needs to be monitored and observe them;

[0068] 2) Use multi-camera fusion technology to fix multiple static cameras at different positions and angles to take pictures of the target object and area, so as to obtain more comprehensive and detailed visual information;

[0069] 3) Use dynamic camera devices (including mobile camera devices such as mobile robots and drones) to move and record from different heights, angles, and positions, so as to obtain more diverse and comprehensive information;

[0070] 4) Use the UMP plug-in in Unity to send request information to the target network camera in the static camera device and / or the dynamic camera device, and obtain the corresponding network video stream information.

[0071] In this embodiment, for the video information obtained by the static camera device and / or the dynamic camera device, the DeepSORT algorithm is used to detect and track the specific target in the video information in real time, and the specific process is: target detection and positioning of the video frame are performed; feature extraction is performed on the target object; the target of the current frame is associated with the previous frame; for the associated target, Kalman filter is used to estimate and predict the state of the target; through the above DeepSORT algorithm, the specific target can be detected and tracked in real time.

[0072] The virtual-real fusion of the video information of the target monitoring area and the three-dimensional digital twin simulation scene includes:

[0073] The video information of the camera in the target video monitoring area and the object information corresponding to the three-dimensional digital twin are matched in terms of time, space and direction, thereby performing virtual-real fusion to generate a video twin system; the video information of the camera in the target area refers to the video pictures captured by the static camera and the dynamic camera; the object information corresponding to the three-dimensional digital twin refers to the modeling of the three-dimensional scene around the video pictures captured by the static camera and the dynamic camera.

[0074] In the video twin system, the angle corresponding to the current third-person free-roaming perspective mode is automatically detected for different camera pictures of different angles in the target video monitoring area, and the camera picture of the corresponding angle is automatically switched to. The basic method is:

[0075] Suppose there are three cameras C1, C2 and C3 uniformly distributed around the target monitoring area at an angle of 120°. With a certain angle as the reference, the camera angles of the three cameras are 0°-120°, 120°-240° and 240°-360° respectively. According to the scene roaming mode, the change of the perspective of the scene is controlled by the movement of the mouse, so when the mouse rotation angle is 0°-120°, the camera C1 picture is displayed; when the mouse rotation angle is 120°-240°, the camera C2 picture is displayed; when the mouse rotation angle is 240°-360°, the camera C3 picture is displayed.

[0076] In the video twin system, the angle corresponding to the current third-person free-roaming perspective mode is automatically detected for different camera pictures of different angles in the target video monitoring area, and the camera picture of the corresponding angle is automatically switched to. The specific implementation process is:

[0077] 1) Create three plane objects in Unity, set as P1, P2 and P3, respectively used to carry the pictures taken by cameras C1, C2 and C3. Set the initial state as C1 activated and C2 and C3 closed.

[0078] 2) According to the scene roaming mode, the rotation angle of the scene is represented by the horizontal displacement X of the mouse. The value of X can be limited to the range (a, b), where a represents the lower limit of X and b represents the upper limit of X. If the value of X is lower than a, it will be limited to a, and if the value of X is higher than b, it will be limited to b. In this way, the value of X can be limited to 0°-360°.

[0079] 3) Thus, when the value of X is in 0°-120°, the plane P1 is displayed, and all the other planes are hidden, i.e. only the picture taken by the camera C1 is displayed; when the value of X is in 120°-240°, the plane P2 is displayed, and all the other planes are hidden, i.e. only the picture taken by the camera C2 is displayed; when the value of X is in 240°-360°, the plane P3 is displayed, and all the other planes are hidden, i.e. only the picture taken by the camera C3 is displayed.

[0080] In the embodiment, the number of the plane objects newly created in the three-dimensional digital twin simulation scene is consistent with the number of all cameras in the three-dimensional digital twin simulation scene, and the size, shape and position of all the planes should be completely consistent.

[0081] The above is only a preferred embodiment of the present application, and does not limit the present application in any form. Although the present application has been disclosed as above with a preferred embodiment, it is not intended to limit the present application. Any person skilled in the art can make some changes or modifications to the above disclosed technical content without departing from the scope of the technical solution of the present application, and any equivalent embodiments with equivalent changes or modifications are still within the scope of the present application.

Claims

1. A steel plant safety monitoring method based on video twins, characterized in that, The method includes: According to a certain ratio, three-dimensional digital twin simulation scenes corresponding to different areas of the steel plant are constructed using multi-sensor fusion technology; A third-person free-roaming perspective mode is constructed in the three-dimensional digital twin simulation scene to facilitate free monitoring of the steel plant; The target monitoring area of ​​the steel plant's digital twin that needs to be monitored by video is determined, and video information corresponding to the target monitoring area is acquired using a camera device; Based on the time sequence, spatial location, and directional relationship, the video information of the target monitoring area is fused with the three-dimensional digital twin simulation scene to generate a video twin system; Based on the video twin system, real-time multi-angle observation of the target monitoring area inside the steel plant can be achieved; A third-person free-roaming perspective mode is constructed in the 3D digital twin simulation scene, where the scene's direction and rotation are controlled by the mouse, and its movement is controlled by the keyboard; the specific implementation method is as follows: Let the amount of mouse movement in the horizontal and vertical directions be x and y, respectively. Let the mouse sensitivity be m and the time of each frame of the video be t. Calculate the actual horizontal displacement X = x * m * t and the actual vertical displacement Y = y * m * t in the 3D scene when the mouse moves. The actual vertical movement Y is applied to the vertical rotation of the 3D scene camera, and the horizontal movement X is applied to the horizontal rotation of the 3D scene camera; thus, the viewpoint changes in the 3D scene can be controlled by moving the mouse. Let there be a three-dimensional vector mov containing three dimensions (xr, yr, zr), where the xr component is set to the value of the horizontal input axis, the yr component is set to 0, and the zr component is set to the value of the vertical input axis; the movement direction of the three-dimensional scene is controlled by the three-dimensional vector mov. Let the scene's movement speed be v, and the time of each frame of the video be t. Then, the displacement S controlled by the keyboard is S = v * mov * t, where the displacement S represents the movement along a certain component. The process of fusing video information of the target monitoring area with the three-dimensional digital twin simulation scene includes: The video information of the camera device within the target monitoring area and the object information corresponding to the three-dimensional digital twin are matched from three perspectives: time, space, and direction, so as to perform virtual-real fusion and generate a video twin system. The video information refers to video footage captured by both static and dynamic cameras; the object information corresponding to the three-dimensional digital twin refers to the three-dimensional scene modeling surrounding the video footage captured by both static and dynamic cameras.

2. The steel plant safety monitoring method based on video twins according to claim 1, characterized in that, The method for constructing three-dimensional digital twin simulation scenes of different areas of the steel plant according to a certain proportion using multi-sensor fusion technology is as follows: (1) Use lidar to obtain point cloud data of the corresponding equipment and environment in the steel plant, that is, the data set of the laser beam emitted by lidar reflected back from the object; each point cloud data includes three-dimensional coordinates, reflection intensity, normal vector, and color information; (2) Use camera devices to acquire image data of the corresponding equipment and environment in the steel plant; (3) The point cloud data acquired by the lidar and the image data acquired by the camera device are fused together, that is, the geometric information of the point cloud data and the texture information of the image data are combined to build a three-dimensional model and obtain a model corresponding to the steel plant equipment and environment. (4) Import the established models of the steel plant equipment and environment into computer software to build a three-dimensional scene of the steel plant.

3. The steel plant safety monitoring method based on video twins according to claim 1, characterized in that, The method involves determining the target monitoring area of ​​the steel plant's digital twin that requires video monitoring and acquiring the corresponding video information of the target monitoring area using a camera device; specifically: Identify the target objects and areas that need to be monitored and conduct observations; By using multi-camera fusion technology, multiple static cameras are fixedly placed in different positions and angles to capture images of the target object and area, thereby obtaining more comprehensive and detailed perspective information. Dynamic camera devices are used to move and record footage from different heights, angles, and positions to obtain more diverse and comprehensive information; The computer device sends a request to the target network camera in the static camera device and / or the dynamic camera device to obtain the corresponding video information.

4. The steel plant safety monitoring method based on video twins according to claim 3, characterized in that, For the video information acquired by the static camera device and / or the dynamic camera device, the DeepSORT algorithm is used to detect and track specific targets in the video information in real time.

5. The steel plant safety monitoring method based on video twins according to claim 1, characterized in that, In the video twin system, for different shooting images from cameras at different angles within the target monitoring area, the system automatically detects the angle corresponding to the current third-person free roaming view mode and automatically switches to the camera image at the corresponding angle.

6. The steel plant safety monitoring method based on video twins according to claim 5, characterized in that, The method for automatically detecting the angle corresponding to the current third-person free roaming view mode and automatically switching to the camera view at the corresponding angle for different shooting images from cameras at different angles within the target monitoring area is as follows: A target monitoring area includes n cameras C1, C2, C3, ..., C n n cameras are evenly distributed around the target monitoring area in a (360 / n)° pattern. Based on a certain angle, the shooting angles of the n cameras are 0°~(360 / n)°, (360 / n)°~2*(360 / n)°, ..., (n-1)*(360 / n)°~360°. In the third-person free roaming view mode, the scene's perspective is controlled by mouse movement. When the mouse rotation angle is between 0° and (360 / n)°, the view captured by camera C1 is displayed; when the mouse rotation angle is between (360 / n)° and 2*(360 / n)°, the view captured by camera C2 is displayed; ..., when the mouse rotation angle is between (n-1)*(360 / n)° and 360°, the view captured by camera C... n The footage was taken.

7. The steel plant safety monitoring method based on video twins according to claim 6, characterized in that, For different images captured by cameras at different angles within the target monitoring area, the system automatically detects the angle corresponding to the current third-person free roaming view mode and automatically switches to the camera image at the corresponding angle. The specific method is as follows: In the three-dimensional digital twin simulation scene, create n planar objects, denoted as P1, P2, P3, ..., P... n Planes P1, P2, P3, ..., P n These are used to carry cameras C1, C2, C3, ..., C n The captured footage; the initial state is set to camera C1 activated, C2, C3, ..., C n closure; In the third-person free roaming view mode, the rotation angle of the scene is represented by the actual horizontal displacement X of the mouse. The value of the actual horizontal displacement X is limited to the range of (a, b), where a represents the lower bound of X and b represents the upper bound of X. If the value of X is lower than a, it will be limited to a; if the value of X is higher than b, it will be limited to b. The value of X is limited to 0°~360°. When the value of X is between 0° and (360 / n)°, plane P1 is displayed, and all other planes are hidden, meaning only the image captured by camera C1 is displayed; when the value of X is between (360 / n)° and 2*(360 / n)°, plane P2 is displayed, and all other planes are hidden, meaning only the image captured by camera C2 is displayed; ...; when the value of X is between (n-1)*(360 / n)° and 360°, plane P is displayed. n Hide all other planes, that is, only show camera C. n Footage taken during filming.

8. The steel plant safety monitoring method based on video twins according to claim 7, characterized in that, The number of newly created planar objects in the three-dimensional digital twin simulation scene is consistent with the number of all cameras in the three-dimensional digital twin simulation scene.

Images