Registration fusion method for three-dimensional model and radiation information of nuclear power plant room

Abstract

The invention provides a nuclear power plant room three-dimensional model and radiation information registration fusion method, and relates to the technical field of radiation information and scene model fusion. The method comprises: constructing a nuclear power plant room space structure model; acquiring a panoramic image of the nuclear power plant room; constructing a three-dimensional model of the nuclear power plant room; deploying a near field communication tag and acquiring near field communication tag information; acquiring a real-time dose rate, a three-dimensional model coordinate point and timestamp information of a near field communication tag position; and associating the coordinate points of the three-dimensional model with near field communication label information, and mapping the real-time dose rate into the three-dimensional model to realize fusion of radiation monitoring data and the three-dimensional model of the nuclear power plant. According to the invention, the dose rate instrument is arranged in the plant for continuous data acquisition, and the monitoring result is transmitted to the processing system in real time to be fused with the three-dimensional model, so that the real-time dynamic monitoring of the radiation field is realized, the radiation distribution situation is visually identified according to the fused data, the radiation risk is rapidly judged, and the personnel safety is effectively guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of fusion of radiation information and scene models, and in particular to a method for merging a three-dimensional model of a nuclear power plant with radiation information. Background Art

[0002] Accurate measurement of radiation dose fields in nuclear power plants is a core requirement for ensuring the safe operation of nuclear facilities and is directly related to optimizing radiation protection and controlling occupational exposure risks. Current radiation monitoring methods commonly used in the nuclear power sector have significant limitations: they rely on manually carried single-point dosimeters for point-by-point inspections, which only capture radiation data at discrete times and isolated locations. This creates blind spots for inaccessible locations such as pipeline interlayers and high-dose areas, resulting in low radiation risk identification accuracy and delayed response. This discrete, non-continuous monitoring model makes it difficult to meet the requirements for rapid location of radioactive sources during nuclear accident emergency response. Furthermore, manual inspections themselves expose workers to additional radiation doses, severely hindering the implementation of the principle of optimizing radiation protection.

[0003] To address the technical bottlenecks of traditional radiation monitoring models, there is an urgent need to develop new technologies for real-time reconstruction of three-dimensional dose fields. To overcome this technical challenge, the project team successfully developed a gamma dose rate monitoring system with wireless communication capabilities, enabling autonomous and continuous monitoring of nuclear power plants. However, this system only provides discrete dose data, making it difficult to intuitively visualize the spatial distribution characteristics of the radiation field. Summary of the Invention

[0004] The present invention provides a method for registering and fusing a three-dimensional model of a nuclear power plant with radiation information, so as to solve the defects existing in the prior art.

[0005] The present invention provides a method for registering and fusing a three-dimensional model of a nuclear power plant with radiation information, comprising:

[0006] Use lidar to scan the three-dimensional point cloud data of the nuclear power plant and construct the spatial structure model of the nuclear power plant.

[0007] Multi-angle images of the nuclear power plant are collected and stitched and fused to obtain a panoramic image of the nuclear power plant.

[0008] The spatial structure model is divided into multiple regions, and the spatial information, surface features and regional images of each region are extracted and fused to obtain a three-dimensional model of the nuclear power plant.

[0009] Near-field communication technology is used to establish spatial reference points, and near-field communication tags are deployed in the nuclear power plant. The location information of the near-field communication tags is marked in the three-dimensional model to obtain the near-field communication tag information.

[0010] A dose rate meter is placed at the near-field communication tag location to monitor the dose rate and upload it to the data processing system according to a preset cycle to obtain the real-time dose rate, three-dimensional model coordinate point and timestamp information of the near-field communication tag location.

[0011] The coordinate points of the three-dimensional model are associated with the near-field communication tag information, and the real-time dose rate is mapped to the three-dimensional model to achieve the integration of radiation monitoring data and the three-dimensional model of the nuclear power plant.

[0012] According to a method for registering and fusing a three-dimensional model of a nuclear power plant with radiation information provided by the present invention, the process of constructing a spatial structure model of a nuclear power plant includes:

[0013] The acquired 3D point cloud data is preprocessed to remove noise points and outliers.

[0014] Based on the preprocessed three-dimensional point cloud data, a spatial partitioning algorithm is used to divide the point cloud data into multiple spatial regions.

[0015] Using geometric feature extraction algorithm, the basic geometric features of nuclear power plant buildings are extracted from multiple spatial regions. The basic geometric features include the boundaries and shape information of walls, floors, and roofs.

[0016] Based on basic geometric features, a three-dimensional modeling algorithm is used to construct a spatial structure model of a nuclear power plant.

[0017] According to a method for registering and fusing a three-dimensional model of a nuclear power plant with radiation information provided by the present invention, the process of stitching and fusing multi-angle images includes:

[0018] The multi-angle image is preprocessed to obtain a preprocessed multi-angle image.

[0019] Visual feature points in preprocessed multi-angle images are extracted based on the scale-invariant feature transformation algorithm, and the size, direction and scale information of the visual feature points are calculated.

[0020] The change matrix between the preprocessed multi-angle images is calculated, and the preprocessed multi-angle images are transformed into the same coordinate system using the change matrix based on the visual feature points. The preprocessed multi-angle images are panoramically stitched to obtain a panoramic image of the nuclear power plant.

[0021] According to a method for registering and fusing a three-dimensional model of a nuclear power plant with radiation information provided by the present invention, the process of extracting visual feature points from pre-processed multi-angle images includes:

[0022] In different scale spaces, the Gaussian difference function is used to perform convolution operations on the preprocessed multi-angle images to generate difference images of different scales.

[0023] Find local extreme points in the difference image as visual feature points.

[0024] The positions and scales of key points in local extreme points are determined by fitting a three-dimensional quadratic function.

[0025] Assign one or more directions to each key point to make the feature point rotation invariant. In the neighborhood of the key point, calculate the direction and amplitude of the image gradient, and statistically calculate the histogram of the gradient direction. The peak of the histogram represents the main direction of the key point.

[0026] According to a registration and fusion method for a three-dimensional model of a nuclear power plant and radiation information provided by the present invention, the process of calculating the change matrix between pre-processed multi-angle images includes:

[0027] Perform feature point matching on different preprocessed multi-angle images.

[0028] Compare the descriptors of feature points and find pairs of feature points with similar descriptors in different images.

[0029] A preset number of feature point pairs are randomly selected and an initial transformation matrix is ​​calculated using the least squares method.

[0030] The initial transformation matrix is ​​optimized using a random sampling consensus algorithm, and the transformation matrix with the smallest error is selected as the final transformation matrix.

[0031] According to a method for registering and fusing a three-dimensional model of a nuclear power plant with radiation information provided by the present invention, the process of fusing spatial information, surface features, and regional images includes:

[0032] According to the characteristics of the spatial structure model, the model surface is divided into multiple regions.

[0033] For each region, the corresponding image is selected as the map based on the spatial information, model surface features and panoramic image content, and the corresponding image content is matched with the model surface features.

[0034] The corresponding image is scaled, rotated, and translated to match the shape and size of the model surface.

[0035] The processed corresponding image is fused with the spatial information and surface features of the area, the display effect of the area in the three-dimensional model is updated, and a three-dimensional model of the nuclear power plant is obtained.

[0036] According to a method for registering and fusing a three-dimensional model of a nuclear power plant with radiation information provided by the present invention, a process of deploying near-field communication tags in a nuclear power plant includes:

[0037] Select representative and stable locations in the nuclear power plant as spatial reference points.

[0038] A near field communication reading and writing device is installed at the spatial reference point and a communication connection is established with the data processing system.

[0039] Initialize the near field communication reading and writing device, including setting the communication frequency, data transmission format and communication protocol to meet the system requirements.

[0040] Use a measuring device to measure the three-dimensional coordinates of the reference point and enter the three-dimensional coordinate information into the data processing system.

[0041] According to a method for registering and fusing a three-dimensional model of a nuclear power plant with radiation information provided by the present invention, the process of marking the position information of a near-field communication tag in the three-dimensional model includes:

[0042] When deploying the NFC tags, use a measurement tool to record the actual three-dimensional coordinate position of each NFC tag.

[0043] The actual three-dimensional coordinate position is uniformly converted into the coordinate system of the three-dimensional model of the nuclear power plant.

[0044] According to the converted coordinate information, a near field communication tag is added to the corresponding position in the three-dimensional model.

[0045] Each tag is associated with the unique device identifier of its corresponding near-field communication tag and related information, including tag type and installation time.

[0046] According to a method for registering and fusing a three-dimensional model of a nuclear power plant with radiation information provided by the present invention, the process of associating the coordinate points of the three-dimensional model with near-field communication tag information includes:

[0047] Extract 3D model coordinate point information and timestamp information.

[0048] In the near field communication tag information database, the position information of the near field communication tag that matches the extracted three-dimensional coordinate point within a preset error range is searched.

[0049] If matching NFC tag location information is found, the real-time dose rate uploaded by the dose rate meter is associated with the unique device identifier of the corresponding NFC tag.

[0050] According to a method for registering and fusing a three-dimensional model of a nuclear power plant with radiation information provided by the present invention, the process of mapping the real-time dose rate to the three-dimensional model includes:

[0051] Establish an association table to record the real-time dose rate, the corresponding near-field communication tag UID code, the three-dimensional model coordinate point and the timestamp information.

[0052] According to the information in the association table, the real-time dose rate is mapped to the corresponding position in the three-dimensional model, realizing the fusion of radiation monitoring data and the three-dimensional model of the nuclear power plant.

[0053] This invention provides a registration and fusion method for a three-dimensional model of a nuclear power plant with radiation information. By establishing a spatial mapping relationship between radiation data and the plant's physical structure, this method achieves the coordinate positioning of discrete monitoring points within the three-dimensional model and the continuous visualization of the radiation field distribution. By dynamically matching real-time data from wireless dose rate meters with the plant's three-dimensional model coordinates, the radiation data acquires spatial attributes, forming a visual representation of the radiation field distribution. This enables continuous visual monitoring of the plant and significantly reduces the additional radiation risks associated with manual inspections. The intuitive display of radiation hotspot distribution through a three-dimensional visualization interface allows workers to quickly grasp radiation field dynamics without entering high-risk areas, effectively supporting radiation protection decision-making, improving the efficiency of radioactive source location during nuclear accident emergency response, and reducing the frequency of manual inspections, fundamentally reducing occupational exposure doses and complying with the principle of radiation protection optimization. By deploying dose rate meters throughout the plant for continuous data collection and transmitting the monitoring results in real time to a processing system for integration with the three-dimensional model, real-time dynamic monitoring of the radiation field is achieved. Based on the integrated data, workers can intuitively identify radiation distribution trends, quickly assess radiation risks, and effectively ensure personnel safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the technical solutions in the present invention or the prior art, a brief introduction is given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0055] Figure 1 This is a flow chart of a method for registering and fusing a three-dimensional model of a nuclear power plant with radiation information, provided by an embodiment of the present invention;

[0056] Figure 2 is a schematic diagram of a process for obtaining an optical panoramic image in an embodiment of the present invention;

[0057] Figure 3 It is a schematic diagram of the process of integrating radiation information with a plant model in an embodiment of the present invention. DETAILED DESCRIPTION

[0058] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0059] The following combination Figure 1-Figure 3The present invention describes a method for registering and fusing a three-dimensional model of a nuclear power plant with radiation information.

[0060] Figure 1 The present invention provides a flowchart of a method for merging a three-dimensional model of a nuclear power plant with radiation information.

[0061] like Figure 1 As shown, an embodiment of the present invention provides a method for registering and fusing a three-dimensional model of a nuclear power plant with radiation information, the method comprising:

[0062] Use lidar to scan the three-dimensional point cloud data of the nuclear power plant and construct the spatial structure model of the nuclear power plant.

[0063] The process of building a nuclear power plant spatial structure model includes:

[0064] The acquired 3D point cloud data is preprocessed to remove noise points and outliers.

[0065] Based on the preprocessed three-dimensional point cloud data, a spatial partitioning algorithm is used to divide the point cloud data into multiple spatial regions.

[0066] Using geometric feature extraction algorithm, the basic geometric features of nuclear power plant buildings are extracted from multiple spatial regions. The basic geometric features include the boundaries and shape information of walls, floors, and roofs.

[0067] Based on basic geometric features, a three-dimensional modeling algorithm is used to construct a spatial structure model of a nuclear power plant.

[0068] The spatial partitioning algorithms include octree partitioning and KD tree partitioning. Octree partitioning divides the entire point cloud space into a cube, and then recursively divides the cube into eight sub-cubes until a preset termination condition is met, such as the number of points in each sub-cube is less than a certain threshold.

[0069] The process involves determining the bounding box of the point cloud and using it as the root node. The center point of the root node is calculated and then divided into eight sub-cubes. For each sub-cube, the number of points is checked. If the number of points exceeds a threshold, the sub-cube continues to be divided; otherwise, the sub-cube is marked as a leaf node.

[0070] KD tree partitioning constructs a binary tree by alternately partitioning the point cloud space along different dimensions. Each node represents a hyperrectangular region, and the point is divided into the left or right subtree by comparing its coordinates along a certain dimension.

[0071] The process involves selecting a dimension, finding the median point along that dimension, and dividing the point cloud into two parts. Recursively partitioning the left and right parts until a termination condition is met, such as when the number of points in each region falls below a certain threshold.

[0072] Geometric feature extraction can be performed using either plane fitting or boundary extraction. Plane fitting involves fitting a plane to each spatial region using the least squares method. Points are then determined to belong to the plane by calculating their distances from the plane. The process involves: selecting an initial set of points for each spatial region. Fitting a plane using the least squares method. Calculating the distance from each point to the plane. A distance threshold is set, and points with distances less than the threshold are considered to be on the plane.

[0073] Boundary extraction: For a fitted plane, the boundary is determined by finding edge points on the plane. Either a normal-based or curvature-based method can be used. The process involves calculating the normal for each point on the plane. Points with significant changes in normal direction are identified, indicating that these points are likely boundary points. By connecting these boundary points, the boundaries of the wall, floor, and roof are determined.

[0074] Based on the basic geometric characteristics, the process of constructing the spatial structure model of a nuclear power plant includes:

[0075] The extracted geometric features (such as the boundaries of walls, floors, and roofs) are converted into triangulated meshes. For each planar region, its boundary points are triangulated. The triangulated meshes of all planar regions are combined to obtain a spatial structural model of the entire nuclear power plant.

[0076] Based on the extracted geometric features, a parametric model is constructed. For example, a wall can be represented by a rectangle or polygon, and parameters (such as length, width, and angle) can be adjusted to fit the actual geometric shape.

[0077] Analyze the extracted geometric features and determine the model parameters. Based on these parameters, construct basic geometric shapes (such as walls, floors, and roofs). Combine these basic geometric shapes to obtain the spatial structural model of the nuclear power plant.

[0078] Figure 2 FIG. 4 is a flow chart of obtaining an optical panoramic image in an embodiment of the present invention.

[0079] like Figure 2 As shown, the process of obtaining an optical panoramic image includes: collecting multi-angle images of the nuclear power plant, and stitching and fusing the multi-angle images to obtain a panoramic image of the nuclear power plant.

[0080] The process of stitching and fusing multi-angle images includes:

[0081] The multi-angle image is preprocessed to obtain a preprocessed multi-angle image.

[0082] Visual feature points in preprocessed multi-angle images are extracted based on the scale-invariant feature transformation algorithm, and the size, direction and scale information of the visual feature points are calculated.

[0083] The change matrix between the preprocessed multi-angle images is calculated, and the preprocessed multi-angle images are transformed into the same coordinate system using the change matrix based on the visual feature points. The preprocessed multi-angle images are panoramically stitched to obtain a panoramic image of the nuclear power plant.

[0084] The process of extracting visual feature points from pre-processed multi-angle images includes:

[0085] In different scale spaces, the Gaussian difference function is used to perform convolution operations on the preprocessed multi-angle images to generate difference images of different scales.

[0086] Find local extreme points in the difference image as visual feature points.

[0087] The positions and scales of key points in local extreme points are determined by fitting a three-dimensional quadratic function.

[0088] Assign one or more directions to each key point to make the feature point rotation invariant. In the neighborhood of the key point, calculate the direction and amplitude of the image gradient, and statistically calculate the histogram of the gradient direction. The peak of the histogram represents the main direction of the key point.

[0089] The process of calculating the change matrix between preprocessed multi-angle images includes:

[0090] Perform feature point matching on different preprocessed multi-angle images.

[0091] Compare the descriptors of feature points and find pairs of feature points with similar descriptors in different images.

[0092] A preset number of feature point pairs are randomly selected and an initial transformation matrix is ​​calculated using the least squares method.

[0093] The initial transformation matrix is ​​optimized using a random sampling consensus algorithm, and the transformation matrix with the smallest error is selected as the final transformation matrix.

[0094] Multi-angle image preprocessing includes removing noise from the image and adjusting brightness and contrast to improve the accuracy of subsequent feature extraction and matching. It includes:

[0095] Converting color images to grayscale images reduces the amount of data while preventing color information from interfering with feature extraction, as many feature extraction algorithms perform better on grayscale images.

[0096] Use filtering algorithms to remove noise from images. Common filtering methods include Gaussian filtering and median filtering. Gaussian filtering smoothes the image by taking a weighted average, effectively removing Gaussian noise. Median filtering replaces the current pixel value with the median value of the neighborhood, effectively removing salt and pepper noise.

[0097] Adjust the brightness and contrast of the image to make the grayscale distribution of the image more uniform. By adjusting the image histogram, the overall visual effect of the image is enhanced and the details are clearer.

[0098] Generate difference images in different scale spaces, including:

[0099] Convolution operations are performed on the preprocessed multi-angle images using Gaussian kernel functions of different scales, resulting in a series of blurred images of varying scales. The larger the Gaussian kernel size, the more blurred the image. This step simulates the effect of the human eye observing an object at different focal lengths, enabling feature points to be detected at all scales.

[0100] Subtracting Gaussian blurred images at adjacent scales produces differential images at different scales. The differential images highlight local changes at different scales in the image and help detect feature points in the image.

[0101] Finding local extreme points in the difference image involves comparing each pixel in each difference image with its 26 neighboring pixels (including 8 neighboring pixels at the same scale and 18 neighboring pixels at adjacent scales). If the value of the pixel is greater than or less than the values ​​of all its neighboring pixels, it is considered a local extreme point. These local extreme points are potential visual feature points.

[0102] Determine the location and scale of key points. For each local extreme point, use a three-dimensional quadratic function to fit the differential image values ​​within its neighborhood. By solving the extreme points of the quadratic function, the precise location and scale of the key points can be obtained. This improves the positioning accuracy of feature points and reduces errors caused by discrete sampling and noise.

[0103] Assigning a direction to a keypoint involves calculating the image's gradient direction and magnitude within the keypoint's neighborhood. The gradient direction indicates the direction of change in the image at that point, while the magnitude indicates the intensity of that change. The gradient direction within the neighborhood is divided into several intervals, and the sum of the gradient magnitudes within each interval is calculated to form a histogram of the gradient directions. The peak of the histogram represents the primary direction of the keypoint. Furthermore, if a secondary peak reaches a certain ratio of the primary peak, it can be used as an auxiliary direction for the keypoint, making the feature point rotationally invariant.

[0104] Calculating the change matrix between preprocessed multi-angle images includes feature point matching, calculating the initial transformation matrix and optimizing the initial transformation matrix.

[0105] Feature point matching includes: Descriptor calculation: A feature descriptor is generated for each key point. This descriptor contains information such as the key point's size, orientation, and scale. The scale-invariant feature transformation algorithm uses a 128-dimensional vector as a feature descriptor to describe the local characteristics of the key point.

[0106] Feature Point Pair Search: Compare the descriptors of feature points in different images and find pairs of feature points with similar descriptors in different images. Use Euclidean distance to measure the similarity between descriptors, with smaller distances indicating greater similarity.

[0107] Calculate the initial transformation matrix: Randomly select feature point pairs: Randomly select a preset number of feature point pairs from the matched feature point pairs.

[0108] Least Squares Calculation: Use the least squares method to calculate an initial transformation matrix based on the selected feature point pairs. The least squares method solves the transformation matrix by minimizing the projection error between the feature point pairs, so that the feature points in different images can be aligned as much as possible.

[0109] The initial transformation matrix is ​​optimized, and a random sampling consensus algorithm is used to find correct matching point pairs from among feature point pairs containing noise and mismatches. In each iteration, a random set of feature point pairs is selected to calculate the transformation matrix, and then the projection error of all feature point pairs under this transformation matrix is ​​calculated. Feature point pairs with an error below a certain threshold are considered inliers, and the rest are considered outliers. The transformation matrix with the largest number of inliers is selected as the final transformation matrix, effectively removing the impact of mismatches and improving the accuracy of the transformation matrix.

[0110] Panoramic stitching includes the following: Image transformation: Using the final transformation matrix, the preprocessed multi-angle images are transformed into the same coordinate system. By transforming each pixel in the image, images from different perspectives can be aligned on the same plane. Image fusion: The transformed images are stitched and fused together to create a panoramic image of the nuclear power plant. During the stitching process, overlapping areas between images need to be processed to avoid stitching artifacts. Common fusion methods include linear fusion and fade-in and fade-out fusion. By taking a weighted average of the pixels in the overlapping areas, the stitched image transitions smoothly.

[0111] The spatial structure model is divided into multiple regions, and the spatial information, surface features and regional images of each region are extracted and fused to obtain a three-dimensional model of the nuclear power plant.

[0112] The process of fusing spatial information, surface features, and regional images includes:

[0113] According to the characteristics of the spatial structure model, the model surface is divided into multiple regions.

[0114] For each region, the corresponding image is selected as the map based on the spatial information, model surface features and panoramic image content, and the corresponding image content is matched with the model surface features.

[0115] The corresponding image is scaled, rotated, and translated to match the shape and size of the model surface.

[0116] The processed corresponding image is fused with the spatial information and surface features of the area, the display effect of the area in the three-dimensional model is updated, and a three-dimensional model of the nuclear power plant is obtained.

[0117] Near-field communication technology is used to establish spatial reference points, and near-field communication tags are deployed in the nuclear power plant. The location information of the near-field communication tags is marked in the three-dimensional model to obtain the near-field communication tag information.

[0118] The process of deploying NFC tags within a nuclear power plant building includes:

[0119] Select representative and stable locations in the nuclear power plant as spatial reference points.

[0120] A near field communication reading and writing device is installed at the spatial reference point and a communication connection is established with the data processing system.

[0121] Initialize the near field communication reading and writing device, including setting the communication frequency, data transmission format and communication protocol to meet the system requirements.

[0122] Use a measuring device to measure the three-dimensional coordinates of the reference point and enter the three-dimensional coordinate information into the data processing system.

[0123] The process of marking the location information of the near-field communication tag in the 3D model includes:

[0124] When deploying the NFC tags, use a measurement tool to record the actual three-dimensional coordinate position of each NFC tag.

[0125] The actual three-dimensional coordinate position is uniformly converted into the coordinate system of the three-dimensional model of the nuclear power plant.

[0126] According to the converted coordinate information, a near field communication tag is added to the corresponding position in the three-dimensional model.

[0127] Each tag is associated with the unique device identifier of its corresponding near-field communication tag and related information, including tag type and installation time.

[0128] A dose rate meter is placed at the near-field communication tag location to monitor the dose rate and upload it to the data processing system according to a preset cycle to obtain the real-time dose rate, three-dimensional model coordinate point and timestamp information of the near-field communication tag location.

[0129] Figure 3 It is a schematic diagram of the process of integrating radiation information with a plant model in an embodiment of the present invention.

[0130] like Figure 3 As shown in FIG, the process of fusing radiation information with the plant model includes: associating the coordinate points of the three-dimensional model with the near-field communication tag information, mapping the real-time dose rate into the three-dimensional model, and realizing the fusion of radiation monitoring data with the three-dimensional model of the nuclear power plant.

[0131] The process of associating 3D model coordinate points with NFC tag information includes:

[0132] Extract 3D model coordinate point information and timestamp information.

[0133] In the near field communication tag information database, the position information of the near field communication tag that matches the extracted three-dimensional coordinate point within a preset error range is searched.

[0134] If matching NFC tag location information is found, the real-time dose rate uploaded by the dose rate meter is associated with the unique device identifier of the corresponding NFC tag.

[0135] The process of mapping the real-time dose rate into the 3D model involves:

[0136] Establish an association table to record the real-time dose rate, the corresponding near-field communication tag UID code, the three-dimensional model coordinate point and the timestamp information.

[0137] According to the information in the association table, the real-time dose rate is mapped to the corresponding position in the three-dimensional model, realizing the fusion of radiation monitoring data and the three-dimensional model of the nuclear power plant.

[0138] In summary, this embodiment provides a method for registering and fusing a three-dimensional model of a nuclear power plant with radiation information. By establishing a spatial mapping relationship between radiation data and the plant's physical structure, this method achieves the coordinate positioning of discrete monitoring points within the three-dimensional model and the continuous visualization of the radiation field distribution. By dynamically matching real-time data from wireless dose rate meters with the coordinates of the plant's three-dimensional model, the radiation data acquires spatial attributes, forming a visual representation of the radiation field distribution. This enables continuous visual monitoring of the plant and significantly reduces the additional radiation risks associated with manual inspections. The intuitive display of radiation hotspot distribution through a three-dimensional visualization interface allows workers to quickly grasp radiation field dynamics without entering high-risk areas, effectively supporting radiation protection decision-making. This improves the efficiency of locating radioactive sources during nuclear accident emergency response, reduces the frequency of manual inspections, and fundamentally reduces occupational exposure doses, in line with the principle of optimal radiation protection. By deploying dose rate meters throughout the plant for continuous data collection and transmitting the monitoring results in real time to a processing system for integration with the three-dimensional model, real-time dynamic monitoring of the radiation field is achieved. Based on the fused data, workers can intuitively identify radiation distribution trends, quickly assess radiation risks, and effectively ensure personnel safety.

[0139] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, or of course, by hardware. Based on this understanding, the essence of the above technical solution or the part that contributes to the existing technology can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or certain parts of the embodiments.

[0140] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A registration and fusion method for a nuclear power plant three-dimensional model and radiation information, characterized in that: include: Use LiDAR to scan the 3D point cloud data of nuclear power plants and build a spatial structure model of the nuclear power plants; Collecting multi-angle images of the nuclear power plant and stitching and fusing the multi-angle images to obtain a panoramic image of the nuclear power plant; Segmenting the spatial structure model into multiple regions, extracting spatial information, surface features, and regional images of each region, and fusing them to obtain a three-dimensional model of the nuclear power plant; Using near-field communication technology to establish spatial reference points, deploying near-field communication tags in the nuclear power plant, marking the location information of the near-field communication tags in the three-dimensional model, and obtaining near-field communication tag information; A dose rate meter is deployed at the position of the near-field communication tag to monitor the dose rate and upload the information to the data processing system according to a preset period to obtain the real-time dose rate, three-dimensional model coordinate point and time stamp information of the near-field communication tag position; The three-dimensional model coordinate points are associated with the near-field communication tag information, and the real-time dose rate is mapped into the three-dimensional model to achieve the fusion of radiation monitoring data and the three-dimensional model of the nuclear power plant.

2. The registration and fusion method of a nuclear power plant three-dimensional model and radiation information according to claim 1 is characterized in that: The process of building a nuclear power plant spatial structure model includes: Preprocess the acquired 3D point cloud data to remove noise points and outliers; Based on the pre-processed 3D point cloud data, a spatial partitioning algorithm is used to divide the point cloud data into multiple spatial regions; Extracting basic geometric features of the nuclear power plant from the plurality of spatial regions using a geometric feature extraction algorithm, wherein the basic geometric features include boundary and shape information of walls, floors, and roofs; According to the basic geometric features, a three-dimensional modeling algorithm is used to construct a spatial structural model of the nuclear power plant.

3. The registration and fusion method of a nuclear power plant three-dimensional model and radiation information according to claim 1 is characterized in that: The process of stitching and fusing multi-angle images includes: Preprocessing the multi-angle image to obtain a preprocessed multi-angle image; extracting visual feature points from the preprocessed multi-angle image based on a scale-invariant feature transformation algorithm, and calculating the size, direction, and scale information of the visual feature points; The change matrix between the preprocessed multi-angle images is calculated, the preprocessed multi-angle images are transformed into the same coordinate system using the change matrix based on the visual feature points, and the preprocessed multi-angle images are panoramically stitched to obtain a panoramic image of the nuclear power plant.

4. The registration and fusion method of a nuclear power plant three-dimensional model and radiation information according to claim 3 is characterized in that: The process of extracting visual feature points from pre-processed multi-angle images includes: In different scale spaces, the Gaussian difference function is used to perform convolution operations on the preprocessed multi-angle images to generate difference images of different scales; Finding local extreme points in the differential image as visual feature points; Determine the position and scale of the key point in the local extreme point by fitting a three-dimensional quadratic function; One or more directions are assigned to each key point to make the feature point rotationally invariant. In the neighborhood of the key point, the direction and magnitude of the image gradient are calculated, and a histogram of the gradient direction is statistically generated. The peak of the histogram represents the main direction of the key point.

5. The registration and fusion method of a nuclear power plant three-dimensional model and radiation information according to claim 3 is characterized in that: The process of calculating the change matrix between preprocessed multi-angle images includes: Perform feature point matching on different pre-processed multi-angle images; Compare the descriptors of feature points and find pairs of feature points with similar descriptors in different images; Randomly select a preset number of feature point pairs and use the least squares method to calculate an initial transformation matrix; The initial transformation matrix is ​​optimized using a random sampling consensus algorithm, and the transformation matrix with the smallest error is selected as the final change matrix.

6. The registration and fusion method of a nuclear power plant three-dimensional model and radiation information according to claim 1 is characterized in that: The process of fusing spatial information, surface features, and regional images includes: According to the characteristics of the spatial structure model, the model surface is divided into multiple regions; For each region, a corresponding image is selected as a map based on spatial information, model surface features, and panoramic image content, wherein the corresponding image content matches the model surface features; Scaling, rotating, and translating the corresponding image so that the corresponding image matches the shape and size of the model surface; The processed corresponding image is fused with the spatial information and surface features of the area, the display effect of the area in the three-dimensional model is updated, and a three-dimensional model of the nuclear power plant is obtained.

7. The registration and fusion method of a nuclear power plant three-dimensional model and radiation information according to claim 1 is characterized in that: The process of deploying NFC tags within a nuclear power plant building includes: Select representative and stable locations in the nuclear power plant as spatial reference points; Install a near field communication reading and writing device at the location of the spatial reference point and establish a communication connection with the data processing system; Initialize the near-field communication reading and writing equipment, including setting the communication frequency, data transmission format and communication protocol to meet the system requirements; The three-dimensional coordinates of the reference point are measured using a measuring device, and the three-dimensional coordinate information is entered into a data processing system.

8. The registration and fusion method of a nuclear power plant three-dimensional model and radiation information according to claim 1 is characterized in that: The process of marking the location information of the near-field communication tag in the 3D model includes: When deploying NFC tags, use measurement tools to record the actual three-dimensional coordinate position of each NFC tag; Perform a unified conversion between the actual 3D coordinate position and the coordinate system of the 3D model of the nuclear power plant; Adding a near field communication tag to a corresponding position in the three-dimensional model according to the converted coordinate information; Each tag is associated with a unique device identifier of its corresponding near field communication tag and related information, including tag type and installation time.

9. The method for registering and fusing a three-dimensional model of a nuclear power plant with radiation information according to claim 1, characterized in that: The process of associating 3D model coordinate points with NFC tag information includes: Extract 3D model coordinate point information and timestamp information; Searching, in a near field communication tag information database, for position information of a near field communication tag that matches the extracted three-dimensional coordinate point within a preset error range; If matching NFC tag location information is found, the real-time dose rate uploaded by the dose rate meter is associated with the unique device identifier of the corresponding NFC tag.

10. The registration and fusion method of a nuclear power plant three-dimensional model and radiation information according to claim 1, characterized in that: The process of mapping the real-time dose rate into the 3D model involves: Establish an association table to record the real-time dose rate, the corresponding near-field communication tag UID code, the three-dimensional model coordinate point and the timestamp information; According to the information in the association table, the real-time dose rate is mapped to the corresponding position in the three-dimensional model, realizing the fusion of radiation monitoring data and the three-dimensional model of the nuclear power plant.

Images