Point cloud data efficient acquisition method for complex structural member

Abstract

The invention discloses a method for efficiently collecting point cloud data of a complex structural component, and the method comprises the following steps: S1, target arrangement and scanning: arranging a target disc at the periphery of the component, carrying out the scanning around the component, and transmitting the data to a cloud end; s2, intelligent splicing: performing intelligent splicing by using a cloud technology, and automatically extracting a spliced component point cloud; and S3, data optimization processing: point cloud and image heterogeneous data fusion is adopted to reduce the processing difficulty, a deep neural network is utilized to carry out target identification auxiliary splicing, and a bounding box is generated based on a site position to automatically extract target point cloud data. According to the point cloud data efficient acquisition method of the complex structural component, the data processing method of the electronic information technology based on big data improves the acquisition efficiency, reduces the influence of human factors, and is suitable for scenes of factory streamlined single component scanning extraction and outdoor multi-component simultaneous scanning extraction.

Description

Technical Field

[0001] The present invention relates to the field of point cloud data acquisition methods, and in particular to an efficient point cloud data acquisition method for complex structural components. Background Art

[0002] Point cloud data acquisition is a key step in 3D modeling, which aims to obtain point cloud data on the surface of an object. Point cloud data acquisition methods are applicable to a variety of scenarios and needs, mainly including: large-scale terrain mapping, 3D modeling, unmanned driving, robot navigation and medical image analysis. Point cloud data acquisition methods can be selected according to specific scenarios and needs to meet the application requirements of different fields.

[0003] However, the existing efficient point cloud data collection methods for complex structural components have certain drawbacks when used. The traditional efficient point cloud data collection methods for complex structural components have low collection efficiency and are greatly affected by human factors. They are not suitable for factory assembly line single component scanning and extraction and outdoor multi-component simultaneous scanning and extraction scenarios. Summary of the invention

[0004] The main purpose of the present invention is to provide a method for efficiently collecting point cloud data of complex structural components, which can effectively solve the problems in the background technology.

[0005] To achieve the above object, the technical solution adopted by the present invention is:

[0006] A method for efficiently collecting point cloud data of complex structural components, the steps of the data processing method based on big data electronic information technology are as follows:

[0007] S1: Target placement and scanning: Target disks are placed around the component, and scan around the component to transmit data to the cloud;

[0008] S2: Intelligent splicing: Use cloud technology for intelligent splicing and automatically extract the point cloud of the spliced ​​components;

[0009] S3: Data optimization processing: Use point cloud and image heterogeneous data fusion to reduce processing difficulty, use deep neural network for target recognition and auxiliary stitching, and generate bounding boxes based on site locations to automatically extract target point cloud data;

[0010] S4: Application scenario: used for factory assembly line single component scanning and extraction, outdoor multi-component simultaneous scanning and extraction.

[0011] Preferably, in step S1, the target arrangement involves placing high-precision reference targets in the scanning area, which are spherical targets and planar targets, have specific geometric shapes and reflective properties, and can be accurately identified and located by the scanner. During the scanning process, it is necessary to ensure that the target is within the scanning range of the three-dimensional scanner, and check whether the battery power and memory space of the instrument are sufficient. During scanning, the instrument starts working after leveling to obtain array geometric image data of the object surface.

[0012] Preferably, in step S2, intelligent stitching is a technology that uses advanced technology and algorithms to automatically combine multiple data into a complete and coherent whole, and can automatically identify and match image features to achieve seamless stitching.

[0013] Preferably, in step S3, the fusion of heterogeneous point cloud and image data is a technology that combines point cloud data obtained by lidar scanning and image data captured by a camera. During the fusion process, it is usually necessary to perform feature extraction on the point cloud data, and at the same time, perform multi-scale feature extraction on the image data to obtain image features at different levels. Then, the point cloud features and image features are fused through a feature fusion algorithm to obtain a fused feature representation.

[0014] Preferably, in step S3, target recognition-assisted stitching is a key technology for improving measurement accuracy and efficiency.

[0015] Preferably, in the step S3, the process of automatically generating a bounding box is: determining the site location, generating a bounding box, extracting the target point cloud, and optimizing and verifying.

[0016] Preferably, in the step S3, determining the site location refers to inputting the site coordinate information and clarifying the scanning area, and generating the bounding box refers to automatically generating the bounding box according to the site location and the preset range. The bounding box needs to fit the target area closely to reduce irrelevant data.

[0017] Preferably, in the step S3, extracting the target point cloud refers to cropping the original point cloud data using a bounding box, retaining only the point cloud within the bounding box as the target data, and optimizing and verifying refers to performing a quality check on the extracted point cloud data, adjusting the bounding box if necessary, and optimizing the extraction results.

[0018] Compared with the prior art, the present invention has the following beneficial effects:

[0019] This data processing method based on big data electronic information technology improves the collection efficiency and reduces the influence of human factors. It is suitable for the scenes of factory assembly line single component scanning and extraction and outdoor multi-component simultaneous scanning and extraction. Among them, target arrangement and scanning can obtain comprehensive and accurate three-dimensional data through reasonable target arrangement and precise scanning process, providing a reliable foundation for subsequent applications. Intelligent splicing can improve creation efficiency and work quality. The fusion technology in data optimization processing can make full use of the high-precision three-dimensional geometric information of point cloud data and the rich texture and color information of image data to improve the accuracy and robustness of environmental perception. Target recognition and auxiliary splicing in the field of three-dimensional scanning and measurement, the target as a public reference point, through its high-precision positioning and recognition capabilities, can significantly reduce system errors and improve the absolute accuracy of measurement. Automatically generate a bounding box can automatically generate a bounding box based on the site location and effectively extract the target point cloud data, providing an accurate data basis for subsequent processing and analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 A flowchart of the steps of a method for efficiently collecting point cloud data of a complex structural component of the present invention

[0021] Figure 2 A data optimization processing flow chart of a method for efficiently collecting point cloud data of a complex structural component of the present invention

[0022] Figure 3 A flowchart of automatically generating a bounding box for a method for efficiently collecting point cloud data of a complex structural component of the present invention; DETAILED DESCRIPTION

[0023] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0024] like Figure 1-3 As shown, a method for efficiently collecting point cloud data of complex structural components, the steps of the data processing method based on big data electronic information technology are as follows:

[0025] S1: Target placement and scanning: Target disks are placed around the component, and scan around the component to transmit data to the cloud;

[0026] S2: Intelligent splicing: Use cloud technology for intelligent splicing and automatically extract the point cloud of the spliced ​​components;

[0027] S3: Data optimization processing: Use point cloud and image heterogeneous data fusion to reduce processing difficulty, use deep neural network for target recognition and auxiliary stitching, and generate bounding boxes based on site locations to automatically extract target point cloud data;

[0028] S4: Application scenario: used for factory assembly line single component scanning and extraction, outdoor multi-component simultaneous scanning and extraction.

[0029] In step S1, target layout involves placing high-precision reference targets in the scanning area. They are spherical targets and plane targets with specific geometric shapes and reflective properties, which can be accurately identified and located by the scanner. The precise position of the target is crucial for the registration of multi-site data and helps to improve the accuracy of point cloud stitching. During the scanning process, it is necessary to ensure that the target is within the scanning range of the 3D scanner and check whether the battery power and memory space of the instrument are sufficient. During scanning, the instrument is leveled and then starts to work to obtain array geometric image data of the object surface. Through reasonable target layout and precise scanning process, comprehensive and accurate The three-dimensional data can provide a reliable foundation for subsequent applications; in step S2, intelligent stitching is a technology that uses advanced technology and algorithms to automatically combine multiple data into a complete and coherent whole, which can automatically identify and match image features, achieve seamless stitching, and improve creation efficiency and work quality; in step S3, point cloud and image heterogeneous data fusion is a technology that combines point cloud data obtained by laser radar scanning and image data captured by the camera. In the fusion process, it is usually necessary to extract features from the point cloud data and perform multi-scale feature extraction on the image data to obtain image features at different levels. Then, through feature fusion, The fusion algorithm fuses the point cloud features and image features to obtain the fused feature representation, which can make full use of the high-precision 3D geometric information of the point cloud data and the rich texture and color information of the image data to improve the accuracy and robustness of environmental perception; in step S3, target recognition and auxiliary stitching is a key technology to improve measurement accuracy and efficiency. In the field of 3D scanning and measurement, the target, as a common reference point, can significantly reduce system errors and improve the absolute accuracy of measurement through its high-precision positioning and recognition capabilities; in step S3, the process of automatically generating a bounding box is: determining the site location, generating a bounding box, extracting the target point cloud, and optimizing and verifying. The bounding box can be automatically generated based on the site location, and the target point cloud data can be effectively extracted, providing an accurate data basis for subsequent processing and analysis; in the S3 step, determining the site location refers to inputting the site coordinate information and clarifying the scanning area, and generating the bounding box refers to automatically generating the bounding box based on the site location and the preset range. The bounding box needs to fit the target area closely to reduce irrelevant data; in the S3 step, extracting the target point cloud refers to using the bounding box to crop the original point cloud data and only retaining the point cloud within the bounding box as the target data. Optimization and verification refers to quality checking the extracted point cloud data, adjusting the bounding box when necessary, and optimizing the extraction results.

[0030] It should be noted that the present invention is an efficient point cloud data acquisition method for complex structural components. This method improves the acquisition efficiency and reduces the influence of human factors. It is suitable for factory assembly line single component scanning and extraction and outdoor multi-component simultaneous scanning and extraction scenarios. Among them, target arrangement and scanning can obtain comprehensive and accurate three-dimensional data through reasonable target arrangement and precise scanning process, providing a reliable foundation for subsequent applications. Intelligent splicing can improve creation efficiency and work quality. The fusion technology in data optimization processing can make full use of the high-precision three-dimensional geometric information of point cloud data and the rich texture and color information of image data to improve the accuracy and robustness of environmental perception. Target recognition and assisted splicing in the field of three-dimensional scanning and measurement, the target as a common reference point, through its high-precision positioning and recognition capabilities, can significantly reduce system errors and improve the absolute accuracy of measurement. Automatically generate a bounding box can automatically generate a bounding box based on the site location and effectively extract the target point cloud data, providing an accurate data basis for subsequent processing and analysis.

[0031] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A method for efficiently collecting point cloud data of complex structural components, characterized by: The steps of the data processing method based on big data electronic information technology are as follows: S1: Target placement and scanning: Target disks are placed around the component, and scan around the component to transmit data to the cloud; S2: Intelligent splicing: Use cloud technology for intelligent splicing and automatically extract the point cloud of the spliced ​​components; S3: Data optimization processing: Use point cloud and image heterogeneous data fusion to reduce processing difficulty, use deep neural network for target recognition and auxiliary stitching, and generate bounding boxes based on site locations to automatically extract target point cloud data; S4: Application scenario: used for factory assembly line single component scanning and extraction, outdoor multi-component simultaneous scanning and extraction.

2. The method for efficiently collecting point cloud data of a complex structural component according to claim 1, characterized in that: In step S1, the target arrangement involves placing high-precision reference targets in the scanning area, which are spherical targets and plane targets with specific geometric shapes and reflection characteristics, and can be accurately identified and located by the scanner. During the scanning process, it is necessary to ensure that the target is within the scanning range of the three-dimensional scanner and check whether the battery power and memory space of the instrument are sufficient. During scanning, the instrument starts working after leveling to obtain array geometric image data of the object surface.

3. The method for efficiently collecting point cloud data of a complex structural component according to claim 1, characterized in that: In step S2, intelligent stitching is a technology that uses advanced technology and algorithms to automatically combine multiple data into a complete and coherent whole, and can automatically identify and match image features to achieve seamless stitching.

4. The method for efficiently collecting point cloud data of a complex structural component according to claim 1, characterized in that: In step S3, the fusion of heterogeneous point cloud and image data is a technology that combines point cloud data obtained by laser radar scanning and image data captured by a camera. During the fusion process, it is usually necessary to perform feature extraction on the point cloud data. At the same time, multi-scale feature extraction is performed on the image data to obtain image features at different levels. Then, the point cloud features and image features are fused through a feature fusion algorithm to obtain a fused feature representation.

5. The method for efficiently collecting point cloud data of a complex structural component according to claim 4, characterized in that: In step S3, target recognition-assisted stitching is a key technology to improve measurement accuracy and efficiency.

6. The method for efficiently collecting point cloud data of a complex structural component according to claim 5, characterized in that: In the step S3, the process of automatically generating a bounding box is: determining the site location, generating a bounding box, extracting the target point cloud, and optimizing and verifying.

7. The method for efficiently collecting point cloud data of a complex structural component according to claim 6, characterized in that: In the step S3, determining the site location refers to inputting the site coordinate information and clarifying the scanning area, and generating the bounding box refers to automatically generating the bounding box according to the site location and the preset range. The bounding box needs to fit the target area closely to reduce irrelevant data.

8. The method for efficiently collecting point cloud data of a complex structural component according to claim 7, characterized in that: In the S3 step, extracting the target point cloud refers to cropping the original point cloud data using the bounding box, retaining only the point cloud within the bounding box as the target data, and optimizing and verifying refers to performing a quality check on the extracted point cloud data, adjusting the bounding box when necessary, and optimizing the extraction results.

Images