Visual real-time imaging method and system based on smart city CIM

Abstract

The invention discloses a visual real-time imaging method and system based on a smart city CIM, and belongs to the technical field of image processing, and the method comprises the steps: obtaining N initial images collected by N cameras in real time; dividing the N initial images into a reference image set and a background image set according to whether the fire fighting equipment is recorded in the initial images, and extracting equipment images of the fire fighting equipment at different angles from the reference image set to obtain an equipment image set; generating a complete equipment image of the fire-fighting equipment in a three-dimensional space according to the equipment image set, and generating a three-dimensional space background where the fire-fighting equipment is located according to the background image set; based on the smart city CIM, constructing a city three-dimensional model of the target city in a three-dimensional space, mapping a three-dimensional space background into the city three-dimensional model, and mapping the complete equipment image into the three-dimensional space background of the city three-dimensional model; the three-dimensional shape and structure of the fire-fighting equipment are accurately reflected, and powerful support is provided for fine management of the fire-fighting equipment.

Description

Technical Field

[0001] The present invention belongs to the technical field of image processing, and particularly relates to a visualization real-time imaging method and system based on Smart City CIM. Background Art

[0002] With the acceleration of the urbanization process, the complexity and management difficulty of cities are constantly increasing. Traditional urban management models face many challenges in dealing with urban public safety, resource management, etc. For example, in the field of fire protection, the deployment and management of fire-fighting equipment are crucial, but traditional monitoring methods often only provide a two-dimensional and partial perspective, making it difficult to comprehensively and real-time grasp the status of fire-fighting equipment, its specific location in the urban space, and the surrounding environment information. In addition, there are also deficiencies in the visual management of the overall urban space, lacking an efficient visualization means that can closely integrate urban infrastructure with the three-dimensional space.

[0003] In this context, the visualization technology based on Smart City CIM (City Information Modeling) has emerged. It can combine various information of the city with a three-dimensional space model, providing more comprehensive and intuitive decision-making support for urban management and emergency response. Summary of the Invention

[0004] In view of the above technical problems, the present invention provides a visualization real-time imaging method and system based on Smart City CIM. The present application provides a visualization real-time imaging method based on Smart City CIM. Fire-fighting equipment and N cameras are deployed in the public space of the target city, where N is a positive integer greater than or equal to 1, and the N cameras are deployed at different positions in the public space. The method includes: Obtaining N initial images collected in real time by the N cameras; Dividing the N initial images into a reference image set and a background image set according to whether the fire-fighting equipment is recorded in the initial images, and extracting equipment images of the fire-fighting equipment at different angles from the reference image set to obtain an equipment image set. The reference image set includes M reference images that record the fire-fighting equipment, where M is a positive integer greater than or equal to 1, and the background image set includes (N - M) background images that do not record the fire-fighting equipment; Generating a complete equipment image of the fire-fighting equipment in the three-dimensional space according to the equipment image set, and generating a three-dimensional space background where the fire-fighting equipment is located according to the background image set; Constructing a city three-dimensional model of the target city in the three-dimensional space based on Smart City CIM, mapping the three-dimensional space background onto the city three-dimensional model, and mapping the complete equipment image onto the three-dimensional space background of the city three-dimensional model.

[0005] Further, divide the N initial images into a reference image set and a background image set according to whether fire-fighting equipment is recorded in the initial images, including: Scale the N initial images to fit the size of the device recognition model, and through the formula Normalize the pixel values of the N initial images, is the pixel value of the initial image, is the mean value of the N initial images, is the standard deviation of the N initial images; Input the N initial images into the device recognition model in sequence, and obtain the initial images, reference coordinates, and coordinate confidence levels with corresponding relationships output by the device recognition model. The reference coordinates are used to indicate the position of the fire-fighting equipment in the initial images, and the coordinate confidence levels are used to indicate the accuracy of the device recognition model in detecting the reference coordinates; Compare the coordinate confidence level of each initial image with the confidence level threshold; Determine the initial images with coordinate confidence levels greater than or equal to the confidence level threshold as reference images and add them to the reference image set; Determine the initial images with coordinate confidence levels less than the confidence level threshold as background images and add them to the background image set.

[0006] Further, train the device recognition model through the following steps: Construct an initial recognition model, a first training set, and a second training set. The first training set includes multiple first training images with the image coordinates of fire-fighting equipment marked, and the second training set includes multiple second training images without fire-fighting equipment recorded; Input the first training set and the second training set into the initial recognition model in sequence, and obtain the training images, training coordinates, and training confidence levels with corresponding relationships output by the initial recognition model; Adopt as the loss function of the initial recognition model, for balancing and weight, , C is the number of categories, that is, 2, is used to indicate that the i-th component including fire-fighting equipment in the training image is 1 and the other components are 0, is the probability of the i-th category predicted by the initial recognition model, , , is the difference between the training coordinates corresponding to the i-th training image output by the initial recognition model and the true image coordinates corresponding to the i-th training image; Through the formula Backpropagate to calculate the gradient of the loss function with respect to the model parameters, is The gradient of the parameters of the initial recognition model, is the gradient of the parameters of the initial recognition model; Use an optimizer to update the parameters of the initial recognition model according to the gradient, where ɑ is the learning rate used to control the parameter update step size; Repeat the above steps until the initial recognition model completes the training rounds for the first training set and the second training set to obtain the device recognition model.

[0007] Furthermore, extract device images of fire-fighting equipment at different angles from the reference image set to obtain a device image set, including: Obtain the reference coordinates corresponding to each background image in the background image set; Extract device images from each background image according to the reference coordinates corresponding to each background image to obtain a device image set.

[0008] Furthermore, generate a complete device image of the fire-fighting equipment in three-dimensional space based on the device image set, including: Extract feature points from each reference image in the device image set; Through Match the feature points of different reference images in the device image set. By calculating the similarity measure between the feature points, determine the corresponding feature point pairs between different reference images, thereby determining the spatial relationship and relative position between different reference images. and are feature points in two reference images respectively, m is the dimension of the feature points, and are feature points and at the coordinate values in the k-th dimension respectively; According to the feature point pairs, through calculate the point cloud data of the fire-fighting equipment in three-dimensional space. The point cloud data includes the three-dimensional coordinate information of the fire-fighting equipment, and the three-dimensional coordinate information is used to reflect the three-dimensional shape and structure of the fire-fighting equipment. and are the rotation matrices of two reference images respectively; Through perform model fitting on the point cloud data, and fit the point cloud data into a complete three-dimensional model to obtain a complete device image of the fire-fighting equipment in three-dimensional space. is the point cloud data, is the weight coefficient used to represent the contribution degree of each point.

[0009] Furthermore, build a three-dimensional urban model of the target city in three-dimensional space based on the smart city CIM, including: Collect geographical information data of the target city, where the geographical information data includes terrain elevation data, building outline data, and road network data; Convert the processed data into a 3D model. Generate a terrain grid through the terrain elevation data, construct a building model using the building outline data and height information, generate a road model in combination with the road network data, and finally integrate to obtain the 3D city model of the target city; Map the 3D space background onto the 3D city model, including: Merge the 3D space background and the geometric shape of the 3D city model through Boolean operations to ensure seamless docking in space; Map the complete device image onto the 3D space background of the 3D city model, including: Obtain the coordinates of the complete device image in the 3D space , is the 2D coordinate of the fire-fighting equipment, is the depth information of the fire-fighting equipment, and K is the internal parameter of the camera; Map the calculated 3D position onto the 3D city model; Use the texture fusion algorithm to smoothly fuse the texture of the fire-fighting equipment with the 3D space background.

[0010] This application provides a visualization real-time imaging system based on the Smart City CIM. Fire-fighting equipment and N cameras are deployed in the public space of the target city, where N is a positive integer greater than or equal to 1, and the N cameras are deployed at different positions in the public space. The system includes: An acquisition module for acquiring N initial images collected in real time by the N cameras; A division module for dividing the N initial images into a reference image set and a background image set according to whether the fire-fighting equipment is recorded in the initial images, and extracting the equipment images of the fire-fighting equipment at different angles from the reference image set to obtain an equipment image set. The reference image set includes M reference images that record the fire-fighting equipment, where M is a positive integer greater than or equal to 1, and the background image set includes (N - M) background images that do not record the fire-fighting equipment; A generation module for generating a complete device image of the fire-fighting equipment in the 3D space according to the equipment image set, and generating the 3D space background where the fire-fighting equipment is located according to the background image set; A construction module for constructing the 3D city model of the target city in the 3D space based on the Smart City CIM, mapping the 3D space background onto the 3D city model, and mapping the complete device image onto the 3D space background of the 3D city model.

[0011] The beneficial effects of the present invention compared with the prior art are as follows: (1) By collecting images in real time through multiple cameras deployed in urban public spaces and dividing them into a reference image set and a background image set, it is possible to comprehensively monitor the status of fire-fighting equipment from different angles, realize real-time visualization management of urban fire-fighting equipment, and promptly discover equipment abnormalities or potential risks; (2) Using an equipment recognition model to accurately identify the position of fire-fighting equipment in the image, and through technologies such as feature point extraction, matching, and point cloud data calculation, a complete equipment image of fire-fighting equipment in three-dimensional space is generated, which accurately reflects the three-dimensional shape and structure of fire-fighting equipment, providing strong support for the refined management of fire-fighting equipment; (3) Collecting geographical information data of the target city and converting it into a three-dimensional model, seamlessly docking the three-dimensional space background with the urban three-dimensional model, and mapping the complete equipment image into the three-dimensional space background, realizing the efficient integration of the overall urban space and fire-fighting equipment information, providing an intuitive and comprehensive three-dimensional visualization platform for urban managers, and helping to optimize the urban space layout and resource allocation; (4) In case of emergencies such as fires, through this system, it is possible to quickly and accurately obtain the position and status information of fire-fighting equipment, as well as the surrounding three-dimensional space environment, providing more accurate navigation and decision-making basis for emergency rescue personnel, thereby improving the emergency response speed and rescue efficiency and reducing disaster losses. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 is a flowchart of a visualization real-time imaging method based on smart city CIM of the present invention; Figure 2 is a schematic structural diagram of a visualization real-time imaging system based on smart city CIM of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0014] Example 1: As Figure 1 shown, a visualization real-time imaging method based on smart city CIM, in which fire-fighting equipment and N cameras are deployed in the public space of the target city, N is a positive integer greater than or equal to 1, and the N cameras are deployed at different positions in the public space, including: Step S101: Obtain N initial images collected in real time by N cameras; Step S102: Divide the N initial images into a reference image set and a background image set according to whether the fire-fighting equipment is recorded in the initial images, and extract the equipment images of the fire-fighting equipment at different angles from the reference image set to obtain an equipment image set. The reference image set includes M reference images recording the fire-fighting equipment, M is a positive integer greater than or equal to 1, and the background image set includes (N - M) background images not recording the fire-fighting equipment; Step S103: Generate a complete device image of the fire-fighting equipment in three-dimensional space based on the device image set, and generate a three-dimensional space background where the fire-fighting equipment is located based on the background image set; Step S104: Build a city three-dimensional model of the target city in three-dimensional space based on the Smart City CIM, map the three-dimensional space background onto the city three-dimensional model, and map the complete device image onto the three-dimensional space background of the city three-dimensional model.

[0015] In this embodiment, the N initial images are divided into a reference image set and a background image set according to whether the fire-fighting equipment is recorded in the initial images, including: scaling the N initial images to fit the size of the device recognition model, and through the formula Normalize the pixel values of the N initial images, is the pixel value of the initial image, is the mean of the N initial images, is the standard deviation of the N initial images; input the N initial images into the device recognition model in sequence, and obtain the initial images, reference coordinates, and coordinate confidence levels with corresponding relationships output by the device recognition model. The reference coordinates are used to indicate the position of the fire-fighting equipment in the initial image, and the coordinate confidence level is used to indicate the accuracy of the device recognition model in detecting the reference coordinates; compare the coordinate confidence level of each initial image with the confidence threshold; determine the initial images with coordinate confidence levels greater than or equal to the confidence threshold as reference images and add them to the reference image set; determine the initial images with coordinate confidence levels less than the confidence threshold as background images and add them to the background image set.

[0016] In this embodiment, the device recognition model is trained through the following steps: construct an initial recognition model, a first training set, and a second training set. The first training set includes multiple first training images with the image coordinates of the fire-fighting equipment marked, and the second training set includes multiple second training images without the fire-fighting equipment recorded; input the first training set and the second training set into the initial recognition model in sequence, and obtain the training images, training coordinates, and training confidence levels with corresponding relationships output by the initial recognition model; use as the loss function of the initial recognition model, for balancing and weights, , C is the number of categories, that is, 2, is used to indicate that the i-th component including the fire-fighting equipment in the training image is 1 and the other components are 0, is the probability of the i-th category predicted by the initial recognition model, , , is the difference between the training coordinates corresponding to the i-th training image output by the initial recognition model and the true image coordinates corresponding to the i-th training image; through the formula Backpropagation is used to calculate the gradient of the loss function with respect to the model parameters, is the gradient of the parameters of the initial recognition model, is the gradient of the parameters of the initial recognition model; use the optimizer to update the parameters of the initial recognition model according to the gradient, where ɑ is the learning rate used to control the parameter update step size; repeat the above steps until the initial recognition model completes the training rounds for the first training set and the second training set to obtain the device recognition model.

[0017] In this embodiment, device images of the fire-fighting equipment at different angles are extracted from the reference image set to obtain a device image set, including: obtaining the reference coordinates corresponding to each background image in the background image set; extracting the device images from each background image according to the reference coordinates corresponding to each background image to obtain the device image set..

[0018] In this embodiment, a complete device image of the fire-fighting equipment in three-dimensional space is generated according to the device image set, including: extracting feature points from each reference image in the device image set; through matching the feature points of different reference images in the device image set, and determining the corresponding feature point pairs between different reference images by calculating the similarity measure between the feature points, so as to determine the spatial relationship and relative position between different reference images, and are the feature points in two reference images respectively, m is the dimension of the feature points, and are the feature points and respectively at the coordinate values of the k-th dimension; according to the feature point pairs, through calculating the point cloud data of the fire-fighting equipment in three-dimensional space, where the point cloud data includes the three-dimensional coordinate information of the fire-fighting equipment, and the three-dimensional coordinate information is used to reflect the three-dimensional shape and structure of the fire-fighting equipment, and are the rotation matrices of two reference images respectively; through performing model fitting on the point cloud data to fit the point cloud data into a complete three-dimensional model to obtain a complete device image of the fire-fighting equipment in three-dimensional space, is the point cloud data, is the weight coefficient used to represent the contribution degree of each point.

[0019] Optionally, the three-dimensional space background where the fire-fighting equipment is located can be generated according to the background image set in the following ways, but not limited to: 1. Extract features from each background image in the background image set: Extract features from each background image, including feature information such as texture, edges, and colors, for subsequent 3D reconstruction and background synthesis.

[0020] 2. Estimate the depth of the background image through the following formula: Assume the background image is I(x, y), and use a monocular depth estimation network or a stereo vision method to estimate the depth value D(x, y) of each pixel point: where f is the depth estimation function and θ is the model parameter. Depth estimation can be achieved through a pre-trained deep learning model, such as a monocular depth estimation network or a stereo matching network.

[0021] 3. Generate a 3D point cloud based on the depth estimation result: Use the depth value D(x, y) and the camera internal parameters K (including the focal length and the optical center coordinates) to convert each pixel point (x, y) into a point P(x, y, z) in 3D space: where, and are the focal lengths of the camera in the x and y directions respectively, and are the optical center coordinates. Through the above formula, map the 2D pixel points into 3D space to generate the 3D point cloud data of the background.

[0022] 4. Filter and denoise the generated 3D point cloud: Use voxel grid filtering or statistical filtering methods to filter and denoise the point cloud, removing noise points and outliers to improve the quality of the point cloud: where P is the original point cloud, is the filtered point cloud, leaf_size is the size of the voxel grid, mean_k is the number of neighborhood points used to calculate the average value in statistical filtering, and std_dev_mul_thresh is the threshold multiple of the standard deviation.

[0023] 5. Model the filtered point cloud through surface fitting or voxelization methods: Perform surface fitting or voxelization on the filtered point cloud data to generate a 3D model of the background. For example, use polynomial surface fitting or voxelization methods: where, is the fitted surface model, is the voxelized model, degree is the polynomial order of surface fitting, and voxel_size is the size of the voxel.

[0024] 6. Integrate the generated three-dimensional background model with the urban three-dimensional model: The generated three-dimensional background model Or Map it into the urban three-dimensional model to seamlessly dock it with other parts of the urban three-dimensional model, forming a complete three-dimensional background environment: Among them, Is the finally integrated urban three-dimensional model, Is the original urban three-dimensional model.

[0025] Through the above steps, generate the three-dimensional space background where the fire-fighting equipment is located, and integrate it with the urban three-dimensional model to form a complete three-dimensional visualization environment.

[0026] In this embodiment, an urban three-dimensional model of the target city in three-dimensional space is constructed based on the smart city CIM, including: collecting geographic information data of the target city, where the geographic information data includes terrain elevation data, building outline data, and road network data; converting the processed data into a three-dimensional model, generating a terrain grid through the terrain elevation data, constructing a building model using the building outline data and height information, generating a road model in combination with the road network data, and finally integrating to obtain the urban three-dimensional model of the target city.

[0027] In this embodiment, mapping the three-dimensional space background in the urban three-dimensional model includes: merging the three-dimensional space background with the geometric shape of the urban three-dimensional model through Boolean operations to ensure seamless docking in space.

[0028] In this embodiment, mapping the complete equipment image in the three-dimensional space background of the urban three-dimensional model includes: obtaining the coordinates of the complete equipment image in three-dimensional space , Is the two-dimensional coordinate of the fire-fighting equipment, Is the depth information of the fire-fighting equipment, and K is the internal parameter of the camera; mapping the calculated three-dimensional position Into the urban three-dimensional model; using a texture fusion algorithm to smoothly fuse the texture of the fire-fighting equipment with the three-dimensional space background.

[0029] Embodiment 2: As Figure 2 Shown, a visualization real-time imaging system based on the smart city CIM, where fire-fighting equipment and N cameras are deployed in the public space of the target city, N is a positive integer greater than or equal to 1, and the N cameras are deployed at different positions in the public space, including: An acquisition module for acquiring N initial images collected in real time by the N cameras; A partitioning module, configured to partition N initial images into a reference image set and a background image set according to whether fire-fighting equipment is recorded in the initial images, and extract equipment images of the fire-fighting equipment at different angles from the reference image set to obtain an equipment image set. The reference image set includes M reference images that record the fire-fighting equipment, where M is a positive integer greater than or equal to 1, and the background image set includes (N - M) background images that do not record the fire-fighting equipment; A generating module, configured to generate a complete equipment image of the fire-fighting equipment in a three-dimensional space according to the equipment image set, and generate a three-dimensional space background where the fire-fighting equipment is located according to the background image set; A constructing module, configured to construct a city three-dimensional model of the target city in a three-dimensional space based on the smart city CIM, map the three-dimensional space background in the city three-dimensional model, and map the complete equipment image in the three-dimensional space background of the city three-dimensional model.

[0030] In the description of this specification, the description of reference terms such as "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example.

Claims

1. A visualization real-time imaging method based on the CIM of a smart city, characterized in that, Firefighting equipment and N cameras are deployed in the public space of the target city, where N is a positive integer greater than or equal to 1, and the N cameras are deployed at different positions in the public space. The method includes: Obtain N initial images collected in real time by the N cameras; Divide the N initial images into a reference image set and a background image set according to whether the firefighting equipment is recorded in the initial images, and extract equipment images of the firefighting equipment at different angles from the reference image set to obtain an equipment image set. The reference image set includes M reference images that record the firefighting equipment, where M is a positive integer greater than or equal to 1, and the background image set includes (N - M) background images that do not record the firefighting equipment; Generate a complete equipment image of the firefighting equipment in three-dimensional space according to the equipment image set, and generate a three-dimensional space background where the firefighting equipment is located according to the background image set; Based on the smart city CIM, construct a city three-dimensional model of the target city in three-dimensional space, map the three-dimensional space background to the city three-dimensional model, and map the complete equipment image to the three-dimensional space background of the city three-dimensional model.

2. The visual real-time imaging method based on the smart city CIM according to claim 1, wherein: The step of dividing the N initial images into a reference image set and a background image set according to whether the firefighting equipment is recorded in the initial images includes: Scale the N initial images to fit the size of the device recognition model and normalize the pixel values of the N initial images through the formula where is the pixel value of the initial image, is the mean of the N initial images, is the standard deviation of the N initial images; Input the N initial images into the equipment recognition model in sequence, and obtain the initial images, reference coordinates, and coordinate confidence levels with corresponding relationships output by the equipment recognition model. The reference coordinates are used to indicate the position of the firefighting equipment in the initial image, and the coordinate confidence level is used to indicate the accuracy of the equipment recognition model in detecting the reference coordinates; Compare the coordinate confidence level of each initial image with a confidence threshold; Determine the initial images with the coordinate confidence level greater than or equal to the confidence threshold as the reference images and add them to the reference image set; Determine the initial images with the coordinate confidence level less than the confidence threshold as the background images and add them to the background image set.

3. The visualization real-time imaging method based on the CIM of a smart city according to claim 2, wherein Train the equipment recognition model through the following steps: Construct an initial recognition model, a first training set, and a second training set. The first training set includes multiple first training images with the image coordinates of the firefighting equipment marked, and the second training set includes multiple second training images that do not record the firefighting equipment; Input the first training set and the second training set into the initial recognition model in sequence, and obtain the training images, training coordinates, and training confidence levels with corresponding relationships output by the initial recognition model; Adopt as the loss function of the initial recognition model, for balancing and weights, , where C is the number of categories, i.e., 2, used to indicate that the i-th component including the fire-fighting equipment in the training image is 1 and the remaining components are 0, is the probability of the i-th category predicted by the initial recognition model, , , is the difference between the training coordinates corresponding to the i-th training image output by the initial recognition model and the true image coordinates corresponding to the i-th training image; Calculate the gradient of the loss function with respect to the model parameters through the formula by backpropagation, where is the gradient of the parameters of the initial recognition model, and is the gradient of the parameters of the initial recognition model; Use an optimizer Update the parameters of the initial recognition model according to the gradient, where ɑ is the learning rate used to control the parameter update step size; Repeat the above steps until the initial recognition model completes the training rounds for the first training set and the second training set to obtain the equipment recognition model.

4. The visual real-time imaging method based on the CIM of the smart city according to claim 2, wherein The step of extracting equipment images of the firefighting equipment at different angles from the reference image set to obtain an equipment image set includes: Obtain the reference coordinates corresponding to each background image in the background image set; Extract the device images from each of the background images according to the reference coordinates corresponding to each of the background images to obtain the device image set.

5. A visualization real-time imaging method based on the CIM of a smart city according to claim 1, characterized in that, Generating the complete device image of the fire-fighting device in three-dimensional space according to the device image set, including: Extract feature points from each of the reference images in the device image set; By matching the feature points of different reference images in the device image set, and determining the corresponding feature point pairs between different reference images by calculating the similarity metric between the feature points, so as to determine the spatial relationship and relative position between different reference images and are feature points in two reference images respectively, m is the dimension of the feature point and are the coordinate values of the feature points and in the k-th dimension respectively Based on the pair of feature points, by calculating the point cloud data of the fire fighting equipment in the three-dimensional space, the point cloud data includes the three-dimensional coordinate information of the fire fighting equipment, and the three-dimensional coordinate information is used to reflect the three-dimensional shape and structure of the fire fighting equipment, and are respectively the rotation matrices of two reference images; By performing model fitting on the point cloud data, fitting the point cloud data into a complete three-dimensional model to obtain the complete device image of the fire-fighting equipment in three-dimensional space, is the point cloud data, is the weight coefficient used to represent the contribution degree of each point.

6. The method for visual real-time imaging based on smart city CIM according to claim 1, wherein: Constructing the urban three-dimensional model of the target city in three-dimensional space based on smart city CIM, including: Collect geographical information data of the target city, where the geographical information data includes terrain elevation data, building contour data, and road network data; Convert the processed data into a three-dimensional model, generate a terrain grid through the terrain elevation data, construct building models using the building contour data and height information, generate road models in combination with the road network data, and finally integrate to obtain the urban three-dimensional model of the target city; Mapping the three-dimensional space background in the urban three-dimensional model, including: Merge the three-dimensional space background with the geometric shape of the urban three-dimensional model through Boolean operations to ensure seamless docking in space; Mapping the complete device image in the three-dimensional space background of the urban three-dimensional model, including: Obtain the coordinates of the complete device image in the three-dimensional space , is the two-dimensional coordinate of the fire-fighting device, is the depth information of the fire-fighting device, and K is the internal parameter of the camera; Map the calculated three-dimensional position to the three-dimensional model of the city; Use a texture fusion algorithm to smoothly fuse the texture of the fire-fighting device with the three-dimensional space background.

7. A visualization real-time imaging system based on the CIM of a smart city, characterized in that, Fire-fighting devices and N cameras are deployed in the public space of the target city, where N is a positive integer greater than or equal to 1, and the N cameras are deployed at different positions in the public space. The system includes: An acquisition module for acquiring N initial images collected in real time by the N cameras; A division module for dividing the N initial images into a reference image set and a background image set according to whether the fire-fighting device is recorded in the initial images, and extracting device images of the fire-fighting device at different angles from the reference image set to obtain a device image set. The reference image set includes M reference images recording the fire-fighting device, where M is a positive integer greater than or equal to 1, and the background image set includes (N - M) background images not recording the fire-fighting device; A generation module for generating the complete device image of the fire-fighting device in three-dimensional space according to the device image set, and generating the three-dimensional space background where the fire-fighting device is located according to the background image set; A construction module for constructing the urban three-dimensional model of the target city in three-dimensional space based on smart city CIM, mapping the three-dimensional space background in the urban three-dimensional model, and mapping the complete device image in the three-dimensional space background of the urban three-dimensional model.

Images