Dimension measurement method and system based on three-dimensional digital space

Abstract

The invention provides a dimension measurement method and system based on a three-dimensional digital space, and the method comprises the following steps: S1, obtaining the depth information of all shooting points in a real physical space, and converting the depth information into point cloud information; s2, splicing and fusing the point cloud information of all the shooting points to obtain a global point cloud digital three-dimensional space; s3, performing registration gridding on the global point cloud digital three-dimensional space to obtain a grid digital three-dimensional space; s4, fusing the global point cloud digital three-dimensional space and the grid digital three-dimensional space to obtain a fused digital three-dimensional space; s5, calling point cloud data or gridding data fused with the digital three-dimensional space according to actual measurement requirements; s6, measuring points are selected to complete size measurement. According to the invention, distance measurement in various complex three-dimensional digital spaces under different precision requirements can be met, and the measurement convenience and the measurement precision are effectively improved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and particularly to a method and system for dimension measurement based on a three-dimensional digital space. Background Art

[0002] With the development of social technology, three-dimensional digital twin technology has been applied to various industries, such as house sales, building decoration, cloud exhibition halls, online shopping, intelligent security, etc. However, in the actual application process, sometimes it is necessary to accurately measure certain dimensions in a three-dimensional digital space. For example, in house decoration, it is necessary to accurately measure the space to meet services such as cabinet customization; for another example, in the model of a large factory building, it is necessary to conduct large-scale and long-distance measurements; for another example, in the case where the measurement distance is too long and there are obstacles (such as walls) between two measurement points, the existing measurement methods need to perform multiple segmented measurements and then add them up, resulting in low measurement efficiency and large errors. In addition, in the actual application of real life, there are often multiple measurement requirements described above in the same three-dimensional model, and the current existing technologies cannot quickly and accurately meet the dimension measurement requirements of the three-dimensional digital space under different conditions. Summary of the Invention

[0003] In order to overcome the deficiencies of the prior art, the present invention provides a method and system for dimension measurement based on a three-dimensional digital space, which can meet the distance measurements in various complex three-dimensional digital spaces under different accuracy requirements, and effectively improves the convenience and accuracy of measurement.

[0004] The technical solution adopted by the present invention to solve its technical problems is as follows:

[0005] The present invention first provides a method for dimension measurement based on a three-dimensional digital space, including the following steps:

[0006] S1. Obtain the depth information of all shooting points in the real physical space, and convert the depth information of each shooting point into initial point cloud information;

[0007] S2. Stitch and fuse the initial point cloud information of all shooting points to obtain a global point cloud digital three-dimensional space of the real physical space;

[0008] S3. Register and grid the obtained global point cloud digital three-dimensional space to obtain a grid digital three-dimensional space;

[0009] S4. Fuse the global point cloud digital three-dimensional space and the grid digital three-dimensional space to obtain a fused digital three-dimensional space that contains both point clouds and grids;

[0010] S5. Call the point cloud data or grid data that integrates the digital three-dimensional space according to the actual measurement requirements;

[0011] S6. Select measurement points to complete dimension measurement.

[0012] As a further preferred implementation, the form of obtaining the depth information in step S1 is: obtaining by scanning and photographing through an RGBD camera or a laser scanning device, or obtaining by performing depth prediction on the panoramic image taken by a color camera.

[0013] As a further preferred implementation, step S5 specifically includes the following operations:

[0014] When performing high-precision measurement, call the single-point cloud data that integrates the digital three-dimensional space for measurement;

[0015] When performing medium-precision measurement, call the global point cloud data that integrates the digital three-dimensional space for measurement;

[0016] When performing low-precision measurement, call the grid data that integrates the digital three-dimensional space for measurement;

[0017] As a further preferred implementation, the specific operation of step S6 is:

[0018] Select the first measurement point by clicking with the mouse / touch screen to obtain the three-dimensional coordinate information of the first measurement point;

[0019] Select the second measurement point by clicking with the mouse / touch screen to obtain the three-dimensional coordinate information of the second measurement point;

[0020] Calculate the distance between the first measurement point and the second measurement point through the three-dimensional coordinate information of the first measurement point and the three-dimensional coordinate information of the second measurement point.

[0021] The present invention also provides a dimension measurement system based on a three-dimensional digital space, including:

[0022] An information acquisition and conversion module, configured to acquire the depth information of all shooting points in the real physical space, and convert the depth information of each shooting point into initial point cloud information;

[0023] A point cloud stitching module, configured to stitch and fuse the initial point cloud information of all shooting points to obtain a global point cloud digital three-dimensional space of the real physical space;

[0024] A grid module, configured to register and grid the obtained global point cloud digital three-dimensional space to obtain a grid digital three-dimensional space;

[0025] A model fusion module for fusing the global point cloud digital three-dimensional space and the grid digital three-dimensional space to obtain a fused digital three-dimensional space that contains both point clouds and grids;

[0026] A data calling module for selectively calling the point cloud data or the meshed data of the fused digital three-dimensional space according to actual measurement requirements;

[0027] A measurement module for selecting measurement points to complete dimension measurement.

[0028] The positive effects of the present invention: According to the user's usage requirements, the present invention allows the user to freely choose the measurement method. If the distance to be measured is short and high measurement accuracy is required, to improve the measurement accuracy, the high-precision measurement mode provided by the present invention can be selected at this time (measuring in a single-point reflectivity point cloud, with an error of ±1 cm); if the distance to be measured is large and medium measurement accuracy is required, the medium-precision measurement mode provided by the present invention can be selected (measuring in the global point cloud, with an error of ±2 cm); if the accuracy requirement for the distance to be measured is not high, but rapid measurement and convenient use are desired, the fast low-precision measurement mode provided by the present invention can be selected (measuring in a mesh grid model, with an error of ±4 cm). In short, compared with the prior art, the present invention can meet the distance measurement in various complex three-dimensional digital spaces under different accuracy requirements, greatly improving the measurement convenience and measurement accuracy compared with the prior art. Description of the Drawings

[0029] Figure 1 is a flowchart of the dimension measurement method of the present invention. Detailed Embodiments

[0030] The preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0031] Refer to Figure 1 , the preferred embodiment of the present invention provides a dimension measurement method based on a three-dimensional digital space, including the following steps:

[0032] S1. Obtain the depth information of all shooting points in the real physical space, and convert the depth information of each shooting point into initial point cloud information;

[0033] The form of obtaining the depth information is: obtaining by scanning and shooting at multiple scanning points with an RGBD camera. Of course, it can also be obtained by scanning and shooting at multiple scanning points with a laser scanning device, or by obtaining depth prediction for the panoramic image taken by a color camera.

[0034] S2. Stitch and fuse the initial point cloud information of all shooting points to obtain the global point cloud digital three-dimensional space of the real physical space;

[0035] S3. Register and grid the obtained global point cloud digital three-dimensional space to obtain a grid digital three-dimensional space;

[0036] S4. Integrate the global point cloud digital three-dimensional space with the grid digital three-dimensional space to obtain an integrated digital three-dimensional space that contains both point clouds (including single-point point clouds and global point clouds) and grids;

[0037] S5. According to the actual measurement requirements, call the point cloud data or grid data of the integrated digital three-dimensional space; specifically, the following operations are included:

[0038] When performing high-precision measurement, call the single-point point cloud data of the integrated digital three-dimensional space for measurement; the single-point point cloud obtained by the imaging device in step S1 can be a point cloud with reflectivity or a colored point cloud with color information. Preferably, it is a point cloud with reflectivity because the point cloud with reflectivity is easier to assist the user in selecting the target point during measurement compared to the point cloud with color information.

[0039] When performing medium-precision measurement, call the global point cloud data of the integrated digital three-dimensional space for measurement;

[0040] When performing low-precision measurement, call the grid data of the integrated digital three-dimensional space for measurement;

[0041] More specifically, assume N single-point point clouds Pi (i = 1, 2,..., N), and the transformation matrix from each single-point point cloud Pi to the world coordinate system W is Mi (i = 1, 2,..., N). Integrating the captured single-point point clouds Pi is to determine a transformation matrix Mi for Pi and transform the corresponding Pi to the same world coordinate system W through Mi to achieve the integration of the global point cloud. Mi is generally a 4x4 matrix, including a 3x3 rotation matrix Ri and a 3*1 translation matrix Ti:

[0042]

[0043] Assume that the transformation matrix from the camera Cam (used when walking in the virtual space) for preview to the world coordinate system W is C. C is generally a 4x4 matrix, including a 3x3 rotation matrix R and a 3*1 translation matrix T:

[0044]

[0045] Then for the given Pi, the transformation matrix Mi_C from it to the camera coordinate system is

[0046] Mi_C = C_inv × Mi

[0047] Here, C_inv is the inverse matrix of C, and the matrix C will be obtained in real time during model preview.

[0048] The global point cloud Pg is formed by fusing Pi transformed to W according to Mi. Therefore, the transformation matrix from Pg to the world coordinate system W is the identity matrix I, and I is generally a 4x4 matrix:

[0049]

[0050] The Mesh model is formed by triangulating the global point cloud Pg. Therefore, the transformation matrix from Mesh to the world coordinate system W is also the identity matrix I.

[0051] Therefore, the transformation matrix Pg_C from Pg to the camera coordinate system is

[0052] Pg_C = C_inv × I = C_inv

[0053] The transformation matrix Mesh_C from Mesh (meshed model) to the camera coordinate system is Mesh_C = C_inv × I = C_inv

[0054] High-precision measurement, medium-precision measurement, and fast measurement (low-precision measurement) are all carried out in the camera coordinate system W. The transformation matrix of the preview camera is C. When switching the measurement method, the camera transformation matrix C remains unchanged, thus playing the role of fixing the perspective and enabling smooth switching. Specifically, when high-precision measurement is selected, the data of the single point Pi where the current measurement point is located is transformed to the camera coordinate system according to Mi_C for measurement; when switching to medium-precision measurement, the Pg data is transformed to the camera coordinate system according to Pg_C for measurement; when switching to fast measurement, the Mesh data is transformed to the camera coordinate system according to Mesh_C for measurement.

[0055] That is to say, the measurement modes based on point cloud data include two types: medium-precision measurement based on the global point cloud and high-precision measurement based on single-point point cloud. Because there is a registration error in the global point cloud, which affects the measurement accuracy to a certain extent, the single-point point cloud well avoids this problem. However, the single-point point cloud measurement can only provide two-point measurement within the corresponding scanned point positions, and the two-point measurement across scanned point positions can only be carried out in the global point cloud.

[0056] When performing high-precision measurement based on the single-point point cloud, only one single-point point cloud (i.e., the point cloud data of one shooting position) is highlighted, and the point cloud data of other scanned point positions is grayed out or semi-transparent (indicating non-operable), and the single-point point cloud to be operated can be activated through click interaction.

[0057] S6. Select measurement points to complete dimension measurement. The specific operation is as follows:

[0058] Select the first measurement point by clicking with the mouse / touch screen to obtain the three-dimensional coordinate information of the first measurement point;

[0059] Select a second measuring point by clicking the mouse or touch screen to obtain the three-dimensional coordinate information of the second measuring point;

[0060] Connecting the first measuring point and the second measuring point into a straight line;

[0061] The distance between the first measuring point and the second measuring point is calculated using the three-dimensional coordinate information of the first measuring point and the three-dimensional coordinate information of the second measuring point.

[0062] The specific calculation method is:

[0063] Assume that the coordinates of the first measuring point are (x1, y1, z1) and the coordinates of the second measuring point are (x2, y2, z2); then the distance d between the first measuring point and the second measuring point is:

[0064]

[0065] When performing fast measurement, the Mesh local zoom method is used, and the measurement point is determined by clicking the mouse / touch screen to intersect the triangular mesh, and then the coordinate information of the measurement point is obtained. That is, assuming that the coordinates of the three vertices of the triangular mesh are V1 (x1, y1, z1), V2 (x2, y2, z2), and V3 (x3, y3, z3); ​​assuming that the coordinates of the screen mouse click are (x, y), the position Pos_c and the direction Dir_c of the camera in the world coordinate system can be obtained by measuring the camera transformation matrix C used. The three-dimensional coordinates of the intersection point are obtained by intersecting Dir_c with the triangles V1, V2, and V3.

[0066] When measuring in the point cloud data model, the area near each point to be measured is displayed in the form of a patch. By zooming in the camera angle, the area of ​​the patch can be appropriately enlarged to facilitate the selection of the measurement point. After selecting the measurement point, the 3D coordinates of the point are directly read as the coordinate value of the measurement point.

[0067] In summary, the present invention provides three measurement modes, specifically:

[0068] 1. High-precision measurement

[0069] Measuring in a single point cloud (suitable for measuring short distances and requiring high accuracy, with an error of ±1cm);

[0070] 2. Medium precision measurement

[0071] Measure in the global point cloud (suitable for measuring long distances with medium accuracy requirements, with an error of ±2cm);

[0072] 3. Fast measurement (low-precision measurement)

[0073] Measure in the mesh model of the global space (suitable for low-precision requirements, but convenient and fast to use, with an error of ±4 cm).

[0074] This embodiment also provides a dimension measurement system based on a three-dimensional digital space, including:

[0075] An information acquisition and conversion module, configured to acquire the depth information of all shooting points in the real physical space, and convert the depth information of each shooting point into initial point cloud information;

[0076] A point cloud stitching module, configured to stitch and fuse the initial point cloud information of all shooting points to obtain a global point cloud digital three-dimensional space of the real physical space;

[0077] A meshing module, configured to register and mesh the obtained global point cloud digital three-dimensional space to obtain a meshed digital three-dimensional space;

[0078] A model fusion module, configured to fuse the global point cloud digital three-dimensional space and the meshed digital three-dimensional space to obtain a fused digital three-dimensional space that contains both point clouds and meshes;

[0079] A data calling module, configured to select and call the point cloud data or meshed data of the fused digital three-dimensional space according to actual measurement requirements;

[0080] A measurement module, configured to select measurement points to complete dimension measurement.

[0081] The above are only the preferred embodiments of the present invention. It should be understood that the description of the above embodiments is only used to help understand the method and its core idea of the present invention, and is not used to limit the protection scope of the present invention. Any modifications, equivalent replacements, etc. made within the spirit and principle of the present invention should be included within the protection scope of the present invention.

Claims

1. A dimension measurement method based on three-dimensional digital space, characterized in that: The steps include: S1. Obtain the depth information of all shooting points in the real physical space, and convert the depth information of each shooting point into initial point cloud information; S2. stitching and fusing the initial point cloud information of all shooting points to obtain the global point cloud digital three-dimensional space of the real physical space; S3. aligning and gridding the obtained global point cloud digitized three-dimensional space to obtain a grid digitized three-dimensional space; S4. Fusing the global point cloud digitized three-dimensional space with the grid digitized three-dimensional space to obtain a fused digitized three-dimensional space containing both the point cloud and the grid; S5. According to the actual measurement requirements, call the point cloud data or grid data of the fused digital three-dimensional space; S6. Select the measuring point to complete the dimension measurement.

2. A dimension measurement method based on three-dimensional digital space according to claim 1, characterized in that: The depth information in step S1 is obtained by scanning and photographing with an RGBD camera or a laser scanning device, or by predicting the depth of a panoramic image taken by a color camera.

3. The dimension measurement method based on three-dimensional digital space according to claim 1, characterized in that: Step S5 specifically includes the following operations: When performing high-precision measurements, call for single-point point cloud data integrated into the digital three-dimensional space for measurement; When performing medium-precision measurement, the global point cloud data integrated with the digital three-dimensional space is called for measurement; When performing low-precision measurements, grid data that is integrated into the digitized three-dimensional space is called for measurement.

4. The dimension measurement method based on three-dimensional digital space according to claim 3, characterized in that: The single-point point cloud is a point cloud with reflectivity information.

5. The dimension measurement method based on three-dimensional digital space according to claim 1, characterized in that: The specific operations of step S6 are: Select the first measuring point by clicking the mouse or touch screen to obtain the three-dimensional coordinate information of the first measuring point; Select a second measuring point by clicking the mouse or touch screen to obtain the three-dimensional coordinate information of the second measuring point; The distance between the first measuring point and the second measuring point is calculated using the three-dimensional coordinate information of the first measuring point and the three-dimensional coordinate information of the second measuring point.

6. A dimension measurement system based on three-dimensional digital space, characterized in that: include: The information acquisition and conversion module is used to obtain the depth information of all shooting points in the real physical space and convert the depth information of each shooting point into initial point cloud information; Point cloud stitching module, used to stitch and fuse the initial point cloud information of all shooting points to obtain the global point cloud digital three-dimensional space of the real physical space; A gridding module is used to register and grid the obtained global point cloud digitized three-dimensional space to obtain a grid digitized three-dimensional space; A model fusion module, used for fusing the global point cloud digitized three-dimensional space and the grid digitized three-dimensional space to obtain a fused digitized three-dimensional space including both point cloud and grid; The data calling module is used to select and call the point cloud data or grid data integrated with the digital three-dimensional space according to the actual measurement requirements; The measurement module is used to select measurement points to complete dimension measurement.

Images