Reactor detection image data processing method and system based on 3D scanning
By using a 3D scanning-based method and leveraging distributed computation and integrated coefficient extraction, the problems of high computational load and loss of geometric features caused by redundancy in point cloud data were solved, achieving high-precision crack detection.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- YANGZHOU POLYTECHNIC INST
- Filing Date
- 2024-11-05
- Publication Date
- 2026-05-08
AI Technical Summary
Existing technologies that acquire point cloud data of buildings through 3D laser scanning suffer from excessive redundancy of data points, resulting in a large computational load and affecting response speed. Furthermore, the deletion of some data points leads to the loss of geometric features, affecting the accuracy of crack detection.
The 3D scanning-based method obtains the distribution extraction coefficient and feature extraction coefficient of data points in each grid through distribution calculation, integrates and calculates the comprehensive extraction coefficient, extracts the detection data points, covers the original point cloud distribution area collected by the scanner, avoids being too dense or too sparse, maintains geometric features, and reduces computational complexity.
It improves the accuracy of crack detection, preserves the geometric features and overall shape of the point cloud, and reduces the complexity of subsequent calculations.
Abstract
Description
Technical Field
[0001] This invention relates to the field of image data processing technology, and in particular to a method and system for processing reactor detection image data based on 3D scanning. Background Technology
[0002] To alleviate environmental pressures, it is proposed to upgrade the environmental protection and equipment at the iron ore raw material yard. The environmental upgrade will be designed to match the planned scale of iron ore production capacity. Simultaneously, given the current situation of dispersed control areas, low information sharing, and low levels of automation in the raw material system, the environmental upgrade should consider four aspects: fully automated process control, digital operation models, unmanned stacking and retrieving operations, and online inventory management. These measures aim to improve automation, increase production efficiency and management levels, and reduce production costs.
[0003] Titled "Prediction Method and System for Building Wall Cracks," the method includes: performing ground laser scanning on the target building to obtain high-precision 3D point cloud data; filtering the high-precision 3D point cloud data to obtain wall point cloud datasets and non-wall point cloud datasets; constructing a triangular irregular network dataset from the wall point cloud dataset; converting the triangular irregular network dataset into grating surface data using an inverse distance weighting algorithm; and extracting crack features from the grating surface data using a shape recognition algorithm to obtain a crack feature dataset. The current technology has a drawback: while using 3D laser scanning to obtain point cloud data of the building and then constructing a 3D model of the building from the point cloud data results in excessive redundancy of data points, leading to excessive computational load and affecting the subsequent processing response speed. While deleting some data points can reduce the number of data points and thus reduce subsequent computational load, it can easily lead to the loss of geometric features of the point cloud, thereby affecting the accuracy of subsequent crack detection and analysis.
[0004] To address the problems existing in the current technology, the company has designed a new reactor detection image data processing method and system based on 3D scanning, which can solve the technical problems existing in the current technology. Summary of the Invention
[0005] The purpose of this invention is to provide a method and system for processing reactor detection image data based on 3D scanning. This method and system for detecting cracks in buildings based on 3D scanning images obtains the distribution extraction coefficient and feature extraction coefficient of data points in each grid through distribution calculation, integrates them to calculate a comprehensive extraction coefficient, and calculates the extraction ratio of data points in each grid based on the comprehensive extraction coefficient. This achieves the integration and extraction of detection data points from the original data points in each grid from two dimensions: data point distribution and data point geometric feature richness. This ensures that the extracted detection data points can cover the distribution area of the original point cloud acquired by the scanner, avoiding overly dense or sparse areas that would affect the expression of the overall geometric features. At the same time, it can effectively maintain the geometric features of the original point cloud, thereby reducing the complexity of subsequent calculations while maintaining the geometric features and overall shape of the point cloud, and improving the accuracy of subsequent crack detection.
[0006] The technical solution of the present invention is as follows: A method for processing reactor detection image data based on 3D scanning is characterized by: using a 3D scanner to scan the surface of a building to acquire three-dimensional point cloud data of the building surface; filtering the scanned data to obtain a three-dimensional coordinate data file of the building surface, wherein the three-dimensional coordinate data file contains all the collected data points; integrating and analyzing the data points in the three-dimensional coordinate data file and extracting detection data points to obtain a set of detection data points; retrieving the detection data points from the set of detection data points to construct a three-dimensional model of the building, and smoothing the constructed three-dimensional model of the building to eliminate the jagged unevenness of the surface; identifying crack areas on the building surface using image processing technology, segmenting the identified crack areas from the surrounding normal surface to form independent crack objects and extracting the geometric feature parameters of the cracks; and processing the data in the three-dimensional coordinate data file... The process involves integrating and analyzing data points to extract detection data points, resulting in a set of detection data points. Specifically, this involves: distributing each data point from the 3D coordinate data file onto a 2D plane region based on 2D plane coordinates to obtain a 2D data point distribution map; dividing the 2D data point distribution map into multiple squares with predetermined length and width; retrieving one data point from a square and identifying its depth coordinates, importing it into a reference library, and recording the depth coordinate difference of the current square as 1; retrieving another data point from a square and identifying its depth coordinates, comparing the current depth coordinates with those in the reference library; if they are the same, not recording the current data point; if they are different, importing the depth coordinates of the current data point into the reference library and incrementing the depth coordinate difference of the current square by 1; repeating this process for all data points in the square to finally obtain the depth coordinate difference of the current square.
[0007] Furthermore, before constructing the 3D model of the building from the processed data points in the feature data point set, the validity analysis of the extracted detection data points is also included. Specifically, the validity analysis of the extracted detection data points involves: obtaining a detection point cloud composed of detection data points and acquiring the original point cloud by the scanner; using the ISS algorithm to extract feature data points from the original point cloud and the detection point cloud respectively; matching the feature data points in the detection point cloud with the feature data points in the original point cloud; collecting the number of matching feature data points and the average matching value in the detection point cloud; and performing correlation integration for validity evaluation analysis.
[0008] A reactor detection image data processing system based on 3D scanning, characterized by comprising: a scanning acquisition unit for scanning the surface of a building to acquire three-dimensional point cloud data of the building surface; an information processing module for filtering and denoising the three-dimensional point cloud data; an integration and extraction module for integrating and analyzing the three-dimensional point cloud data and extracting detection data points, and integrating them to obtain a set of detection data points; a model building module for retrieving detection data points from the set of detection data points to construct a three-dimensional model of the building; and an image processing module for identifying crack areas on the building surface, segmenting the identified crack areas from the surrounding normal surface to form independent crack objects, and extracting the geometric feature parameters of the cracks.
[0009] The beneficial effects of this invention are: This invention relates to a method and system for detecting building cracks based on 3D scanning images. It calculates a comprehensive extraction coefficient by integrating the distribution extraction coefficient and feature extraction coefficient of data points in each grid cell using distribution calculation. Based on this comprehensive extraction coefficient, it calculates the extraction ratio of data points in each grid cell. This allows for the integrated extraction of detection data points from both the data point distribution and the richness of geometric features. The extracted detection data points cover the distribution area of the original point cloud acquired by the scanner, avoiding overly dense or sparse areas that could affect the overall geometric feature representation. Simultaneously, it effectively preserves the geometric features of the original point cloud, reducing the complexity of subsequent calculations while maintaining the geometric features and overall shape of the point cloud, thus improving the accuracy of subsequent crack detection. Detailed Implementation
[0010] A method for processing reactor detection image data based on 3D scanning is characterized by: using a 3D scanner to scan the surface of a building to acquire three-dimensional point cloud data of the building surface; filtering the scanned data to obtain a three-dimensional coordinate data file of the building surface, wherein the three-dimensional coordinate data file contains all the collected data points; integrating and analyzing the data points in the three-dimensional coordinate data file and extracting detection data points to obtain a set of detection data points; retrieving the detection data points from the set of detection data points to construct a three-dimensional model of the building, and smoothing the constructed three-dimensional model of the building to eliminate the jagged unevenness of the surface; identifying crack areas on the building surface using image processing technology, segmenting the identified crack areas from the surrounding normal surface to form independent crack objects and extracting the geometric feature parameters of the cracks; and processing the data in the three-dimensional coordinate data file... The process involves integrating and analyzing data points to extract detection data points, resulting in a set of detection data points. Specifically, this involves: distributing each data point from the 3D coordinate data file onto a 2D plane region based on 2D plane coordinates to obtain a 2D data point distribution map; dividing the 2D data point distribution map into multiple squares with predetermined length and width; retrieving one data point from a square and identifying its depth coordinates, importing it into a reference library, and recording the depth coordinate difference of the current square as 1; retrieving another data point from a square and identifying its depth coordinates, comparing the current depth coordinates with those in the reference library; if they are the same, not recording the current data point; if they are different, importing the depth coordinates of the current data point into the reference library and incrementing the depth coordinate difference of the current square by 1; repeating this process for all data points in the square to finally obtain the depth coordinate difference of the current square.
[0011] A method and system for detecting building cracks based on 3D scanning images is proposed. This method calculates a comprehensive extraction coefficient by integrating the distribution extraction coefficient and feature extraction coefficient of data points in each grid cell. Based on this comprehensive extraction coefficient, the extraction ratio of data points in each grid cell is calculated. This approach integrates and correlates the data point distribution and geometric feature richness when extracting detection data points from the original data points in each grid cell. This ensures that the extracted detection data points cover the distribution area of the original point cloud acquired by the scanner, avoiding overly dense or sparse areas that could affect the overall geometric feature representation. Simultaneously, it effectively preserves the geometric features of the original point cloud, reducing the complexity of subsequent calculations while maintaining the geometric features and overall shape of the point cloud, thus improving the accuracy of subsequent crack detection. As a preferred option, before constructing the 3D model of the building from the processed data points in the feature data point set, the method also includes an effectiveness analysis of the extracted detection data points. Specifically, the effectiveness analysis of the extracted detection data points involves: obtaining a detection point cloud composed of detection data points and acquiring the original point cloud from the scanner; using the ISS algorithm to extract feature data points from the original point cloud and the detection point cloud respectively; matching the feature data points in the detection point cloud with the feature data points in the original point cloud; collecting the number of matching feature data points and the average number of matching points in the detection point cloud; and performing correlation and integration to conduct an effectiveness evaluation analysis.
[0012] A reactor detection image data processing system based on 3D scanning, characterized by comprising: a scanning acquisition unit for scanning the surface of a building to acquire three-dimensional point cloud data of the building surface; an information processing module for filtering and denoising the three-dimensional point cloud data; an integration and extraction module for integrating and analyzing the three-dimensional point cloud data and extracting detection data points, and integrating them to obtain a set of detection data points; a model building module for retrieving detection data points from the set of detection data points to construct a three-dimensional model of the building; and an image processing module for identifying crack areas on the building surface, segmenting the identified crack areas from the surrounding normal surface to form independent crack objects, and extracting the geometric feature parameters of the cracks.
[0013] A 3D scanner is used to scan the building surface, acquiring 3D point cloud data. The scanned data is filtered to obtain a 3D coordinate data file containing all acquired data points. The data points in the 3D coordinate data file are integrated and analyzed to extract detection data points, resulting in a detection data point set. The detection data points from this set are used to construct a 3D model of the building, which is then smoothed to eliminate jagged edges and unevenness. Image processing techniques are used to identify crack areas on the building surface. These crack areas are segmented from the surrounding normal surface to form independent crack objects, and their geometric feature parameters are extracted. Finally, the data points in the 3D coordinate data file are integrated and analyzed to extract detection data points. The specific steps to obtain the set of detection data points are as follows: Each data point in the 3D coordinate data file is distributed on a 2D plane region based on its 2D plane coordinates to obtain a 2D data point distribution map. The entire 2D data point distribution map is then divided into i squares with equal areas according to a preset length and width. One data point in each square is retrieved, its depth coordinates are identified, and it is imported into a reference library. The depth coordinate difference of the current square is recorded as 1. Another data point in the square is retrieved, its depth coordinates are identified, and the current depth coordinates are compared with those in the reference library. If they are the same, the current data point is not recorded; otherwise, its depth coordinates are imported into the reference library, and the depth coordinate difference of the current square is incremented by 1. This process is repeated for all data points in the squares to finally obtain the depth coordinate difference of the current square. Further explanation: The system also includes evenly distributed buffer mechanisms, each designed to provide cushioning for adjacent electric wheels. These buffer mechanisms are located within the movable worktable near the adjacent electric wheels. Each buffer mechanism includes a buffer shell, which is fixedly connected to the movable worktable near the adjacent electric wheel. The electric wheel is slidably connected to the adjacent buffer shell. A fixed plate is fixedly connected to the upper side of the electric wheel, and the fixed plate is slidably connected to the adjacent buffer shell. A rotating plug is rotatably connected to the upper side of the electric wheel, and the rotating plug is rotatably connected to the adjacent fixed plate. Both the rotating plug and the adjacent fixed plate have through holes of equal size. The rotating plug is slidably connected to the adjacent buffer shell. A fourth elastic element is installed between the rotating plug and the adjacent buffer shell. A fourth electric rotating shaft is fixedly connected to the upper side of the buffer shell via a bracket. The fourth electric rotating shaft is splinedly connected to the adjacent rotating plug. A blocking component is provided on the lower part of the movable worktable near the adjacent electric wheel to block debris from moving along the path of the adjacent electric wheel.
[0014] An intelligent stacker-reclaimer operation system receives various work tasks and instructions from a material yard intelligent management system, or input work tasks and control parameters. After confirmation or adjustment by the operator, the intelligent operation PLC sends equipment control parameter instructions to the PLC control systems of the stacker-reclaimer equipment, conveying equipment, etc., to achieve precise execution of work tasks / instructions. Simultaneously, it promptly senses the status of the aforementioned equipment and controls the interlocking actions between equipment. Production management personnel organize production and issue production tasks based on the panoramic information of the storage and transportation system displayed by the material yard intelligent management system. Task commands are sent to the material yard intelligent operation system, and after operator confirmation, the production commands are automatically executed. The video monitoring system will then display real-time images of relevant materials and key equipment. This achieves intelligent, unmanned, safe, and efficient production of stacker-reclaimer operations within the material yard.
[0015] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements or substitutions without departing from the principles of the present invention, and these improvements or substitutions should also be considered within the scope of protection of the present invention.
Claims
1. A method for processing reactor detection image data based on 3D scanning, characterized in that, A 3D scanner is used to scan the building surface to obtain 3D point cloud data. The scanned data is filtered to obtain a 3D coordinate data file containing all the collected data points. The data points in the 3D coordinate data file are integrated and analyzed to extract detection data points, resulting in a detection data point set. The detection data points in the detection data point set are used to construct a 3D model of the building, and the constructed 3D model is smoothed to eliminate jagged unevenness on the surface. Image processing technology is used to identify crack areas on the building surface, and the identified crack areas are segmented from the surrounding normal surface to form independent crack objects, and the geometric feature parameters of the cracks are extracted. The process of integrating, analyzing, and extracting detection data points from the 3D coordinate data file to obtain a set of detection data points involves: distributing each data point in the 3D coordinate data file onto a 2D plane region based on 2D plane coordinates to obtain a 2D data point distribution map; dividing the 2D data point distribution map into multiple squares with a preset length and width; retrieving one data point from a square and identifying its depth coordinates, importing it into a reference library, and recording the depth coordinate difference of the current square as 1; retrieving another data point from a square and identifying its depth coordinates, comparing the current depth coordinates with those in the reference library; if they are the same, not recording the current data point; if they are different, importing the depth coordinates of the current data point into the reference library and incrementing the depth coordinate difference of the current square by 1; repeating this process for all data points in the square to finally obtain the depth coordinate difference of the current square.
2. The reactor detection image data processing method based on 3D scanning according to claim 1, characterized in that: Before constructing a 3D model of a building from the processed data points in the feature data point set, the effectiveness analysis of the extracted detection data points is also included. Specifically, the effectiveness analysis of the extracted detection data points involves: obtaining a detection point cloud composed of detection data points and acquiring the original point cloud by a scanner; using the ISS algorithm to extract feature data points from the original point cloud and the detection point cloud respectively; matching the feature data points in the detection point cloud with the feature data points in the original point cloud; collecting the number of matching feature data points and the average matching value in the detection point cloud; and performing correlation and integration to conduct an effectiveness evaluation analysis.
3. The reactor detection image data processing system based on 3D scanning according to claim 1, characterized in that: include: The scanning and acquisition unit is used to scan the surface of a building and acquire three-dimensional point cloud data of the building surface; The information processing module is used to filter and denoise the 3D point cloud data; the integration and extraction module is used to integrate and analyze the 3D point cloud data, extract the detection data points, and integrate them to obtain a set of detection data points. The model building module is used to retrieve detection data points from the detection data point set to construct a 3D model of the building. The image processing module is used to identify cracked areas on the building surface, segment the identified cracked areas from the surrounding normal surface to form independent crack objects, and extract the geometric feature parameters of the cracks.