Large part point cloud splicing method based on handheld three-dimensional scanner measurement

Abstract

The invention discloses a large part point cloud splicing method based on handheld three-dimensional scanner measurement, and the method comprises the steps: setting a plurality of CCT codes on a housing of a handheld three-dimensional laser scanner, and employing two cameras to form a positioning tracking device; at different positions, a handheld three-dimensional laser scanner is used for collecting point cloud data at the positions, and a positioning tracking device is used for shooting CCT codes on a shell of the scanner at the positions; obtaining a spatial three-dimensional coordinate of the center point of each CCT coding pattern at each position through CCT coding and decoding and binocular three-dimensional reconstruction; acquiring relative spatial positions of any two positions by using a least square method, and splicing point cloud data at the two positions according to the relative spatial positions of the two positions; and outputting the spliced point cloud data, and constructing a three-dimensional model of a complete part. According to the invention, real-time and automatic splicing of three-dimensional measurement point clouds of large parts is realized without auxiliary point pasting, the splicing speed is improved, and the splicing precision is ensured.

Description

Technical Field

[0001] The present invention belongs to the field of machinery and automation, and particularly relates to a method for point cloud stitching of large parts based on a binocular laser scanning measurement system. Background Art

[0002] Large equipment structural parts are usually structural components with complex geometric shapes and high-precision requirements. The accuracy and efficiency of their dimensional measurement are directly related to the performance and reliability of the overall system. Any slight deviation may lead to assembly instability. Traditional contact measurement methods are not only inefficient but also difficult to ensure measurement accuracy when dealing with complex curved surfaces. Three-dimensional laser scanning technology, with its large measurement range, high precision, non-contact, not affected by light, fast scanning speed, high degree of automation, etc., has gradually become one of the main means for measuring large components. Through high-density point cloud data, every detail on the surface of large components can be comprehensively and detailedly captured, which not only improves the measurement accuracy and efficiency but also reduces measurement errors caused by human factors.

[0003] Due to the large size of the components, common three-dimensional laser scanning devices are difficult to complete the overall scanning and complete measurement of large components at one station. Therefore, it is necessary to scan at multiple stations and then construct a complete three-dimensional model of the object surface through three-dimensional point cloud stitching. Therefore, three-dimensional point cloud stitching technology is a key technology in three-dimensional laser scanning measurement, which determines the measurement efficiency, and the stitching accuracy directly affects the overall measurement accuracy of the system.

[0004] Currently, the common three-dimensional point cloud stitching methods are mainly the following two: The first is the non-assisted point cloud stitching using registration algorithms such as ICP. The principle of this method is not applicable to multi-line lasers, and when the point cloud of a large-size object is dense, it will cause complex algorithm processing, slow execution speed, and easily lead to problems of increased measurement errors in the coordinate conversion process, affecting the measurement accuracy; the second method is the assisted stitching method using marker points pasted on the object to be measured for point cloud stitching. This method not only consumes a lot of time for pasting marker points but also cannot paste on components with complex textures, will block their own textures, and it is difficult to achieve automatic stitching, affecting the measurement accuracy and efficiency.

[0005] Currently, the common scanners on the market are divided into two types. One is a fixed mobile laser scanner, whose disadvantage is that it cannot perform three-dimensional measurements of large-scale point cloud stitching, and its fixed turntable cannot be held by hand and can only be used on specific turntables or within a specific range; the other is a handheld point-pasting three-dimensional scanner. The principle of point cloud stitching of this scanner is based on a large number of marker points pasted on the object to be measured, which leads to not only wasting a lot of time during the pasting process but also being unable to paste on components with complex textures and will block their own textures. Summary of the Invention

[0006] The object of the present invention is to provide a method for point cloud stitching of large parts based on handheld 3D scanner measurement, which can realize real-time and automatic stitching of 3D measurement point clouds of large parts without auxiliary point sticking, improve the stitching speed and ensure the stitching accuracy.

[0007] In order to achieve the above object, the present invention provides a method for point cloud stitching of large parts based on handheld 3D scanner measurement. A number of CCT codes are set on the outer shell of the handheld 3D laser scanner, and a positioning and tracking device is composed of two cameras; at different positions, the handheld 3D laser scanner is used to collect the point cloud data at that position, and the positioning and tracking device is used to photograph the CCT codes on the outer shell of the handheld laser 3D scanner at that position; the spatial 3D coordinates of the center points of each CCT code pattern at each position are obtained through CCT code decoding and binocular 3D reconstruction; the relative spatial positions of any two positions are obtained by using the least squares method, and the point cloud data at these two positions are stitched according to the relative spatial positions of the two positions; the stitched point cloud data is output to construct a 3D model of the complete part.

[0008] The above-mentioned large-scale component point cloud stitching method based on handheld 3D scanner measurement includes: S1: using a handheld 3D laser scanner to scan at a first position to obtain point cloud data A at the first position; at the same time, the positioning and tracking device shoots the CCT code on the housing of the handheld laser 3D scanner at the first position; S2: decoding the CCT code shot at the first position, and calculating the spatial three-dimensional coordinates of the center point of each CCT code pattern according to the internal and external parameters of the positioning and tracking device; S3: using a handheld 3D laser scanner to scan at a second position to obtain point cloud data B at the second position; at the same time, the positioning and tracking device shoots the CCT code on the housing of the handheld laser 3D scanner at the second position; S4: decoding the CCT code shot at the second position, and calculating the spatial three-dimensional coordinates of each CCT code pattern center point according to the internal and external parameters of the positioning and tracking device. The spatial three-dimensional coordinates of the center point of each CCT coding pattern; S5: using the least squares method to calculate the transformation matrix of the three-dimensional points with the same CCT code at the first position and the second position, that is, the relative spatial position of the first position and the second position can be obtained; according to the relative spatial position of the first position and the second position, the point cloud data A and the point cloud data B are spliced; S6: at other positions, repeat steps S1 to S2 to obtain the point cloud data at each position and the spatial three-dimensional coordinates of the center point of each CCT coding pattern; keep changing the position until the entire component is scanned; S7: using the least squares method to obtain the relative spatial position of any two positions, splicing the point cloud data at the two positions according to the relative spatial position of the two positions; outputting the spliced ​​point cloud data, constructing a three-dimensional model of the complete component, and providing support for the subsequent calculation of the component measurement size.

[0009] In the above-mentioned large-scale component point cloud stitching method based on measurement by a handheld three-dimensional scanner, in the step S5, the transformation matrix is ​​a rotation matrix or a translation matrix.

[0010] In the above-mentioned large-scale component point cloud stitching method based on the measurement of the handheld three-dimensional scanner, in the step S5, the stitched point cloud data is defined as point cloud data AB, point cloud data AB=point cloud data A+point cloud data B*transformation matrix.

[0011] Compared with the prior art, the beneficial technical effects of the present invention are:

[0012] Since the present invention is based on positioning the CCT code on the housing of the handheld three-dimensional scanner, by calculating the three-dimensional coordinates of the coding pattern at different positions, the least squares method is used to obtain the transformation matrix (rotation matrix or translation matrix) between different positions, and then the rotation matrix or translation matrix between the point clouds at different positions is obtained. It can be seen that the method is not subject to the limitation of spatial range, and will not be disturbed by the complexity of the surface texture of the object being measured, and has strong practicality and advancement. Brief Description of the Drawings

[0013] The method for point cloud stitching of large parts based on handheld 3D scanner measurement of the present invention is given by the following embodiments and drawings.

[0014] Figure 1 It is a flowchart of the method for point cloud stitching of large parts based on handheld 3D scanner measurement according to an embodiment of the present invention. Detailed Description of the Invention

[0015] The following will further describe in detail Figure 1 the method for point cloud stitching of large parts based on handheld 3D scanner measurement of the present invention.

[0016] Figure 1 Shown is a flowchart of the method for point cloud stitching of large parts based on handheld 3D scanner measurement according to an embodiment of the present invention.

[0017] Referring to Figure 1 , for the method for point cloud stitching of large parts based on handheld 3D scanner measurement in this embodiment, a number of CCT codes are set (pasted) on the outer shell of the handheld 3D laser scanner, and a positioning and tracking device is composed of two cameras. The method includes:

[0018] S1: Use the handheld 3D laser scanner to scan at the first position (denoted as A) to obtain the point cloud data A at position A; at the same time, the positioning and tracking device takes pictures of the CCT codes on the outer shell of the handheld laser 3D scanner at position A;

[0019] S2: Decode the CCT codes photographed at position A, and calculate the three-dimensional spatial coordinates of the center point of each CCT code pattern according to the internal and external parameters of the positioning and tracking device;

[0020] S3: Use the handheld 3D laser scanner to scan at the second position (denoted as B) to obtain the point cloud data B at position B; at the same time, the positioning and tracking device takes pictures of the CCT codes on the outer shell of the handheld laser 3D scanner at position B;

[0021] S4: Decode the CCT codes photographed at position B, and calculate the three-dimensional spatial coordinates of the center point of each CCT code pattern according to the internal and external parameters of the positioning and tracking device;

[0022] S5: Use the least squares method to calculate the rotation matrix or translation matrix of the three-dimensional points (i.e., the center points of the CCT code patterns) with the same CCT code at positions A and B, so as to obtain the relative spatial position between positions A and B; stitch the point cloud data A and the point cloud data B according to the relative spatial position between positions A and B;

[0023] The spliced point cloud data is defined as point cloud data AB, and point cloud data AB = point cloud data A + point cloud data B * transformation matrix (i.e., rotation matrix or translation matrix);

[0024] S6: At other positions, repeat steps S1 - S2 to obtain the point cloud data at each position and the three-dimensional spatial coordinates of the center point of each CCT coding pattern; continuously change positions until the entire component is scanned;

[0025] S7: The least squares method can be used to obtain the relative spatial positions of any two positions, and the point cloud data at these two positions is spliced according to the relative spatial positions of the two positions; output the spliced point cloud data, construct a three-dimensional model of the complete component, and provide support for the subsequent calculation of the component measurement size.

Claims

1. A point cloud stitching method for large parts based on handheld 3D scanner measurement, characterized in that: A plurality of CCT codes are set on the housing of the handheld 3D laser scanner, and two cameras are used to form a positioning and tracking device; at different positions, the handheld 3D laser scanner is used to collect point cloud data at the position, and the positioning and tracking device is used to photograph the CCT code on the housing of the handheld laser 3D scanner at the position; The spatial three-dimensional coordinates of the center point of each CCT coding pattern at each position are obtained through CCT coding and decoding and binocular three-dimensional reconstruction; the relative spatial position of any two positions is obtained using the least squares method, and the point cloud data at the two positions are spliced ​​according to the relative spatial positions of the two positions; the spliced ​​point cloud data is output to construct a three-dimensional model of the complete component.

2. The method for assembling point clouds of large parts based on measurement by a handheld three-dimensional scanner as claimed in claim 1, characterized in that: include: S1: Use a handheld 3D laser scanner to scan at the first position to obtain point cloud data A at the first position; At the same time, the positioning and tracking device photographs the CCT code on the housing of the handheld laser 3D scanner at the first position; S2: Decode the CCT code captured at the first position, and calculate the spatial three-dimensional coordinates of the center point of each CCT code pattern according to the internal and external parameters of the positioning and tracking device; S3: Use the handheld 3D laser scanner to scan at the second position to obtain point cloud data B at the second position; at the same time, the positioning and tracking device photographs the CCT code on the housing of the handheld 3D laser scanner at the second position; S4: decoding the CCT code captured at the second position, and calculating the spatial three-dimensional coordinates of the center point of each CCT code pattern according to the internal and external parameters of the positioning and tracking device; S5: using the least square method to calculate the transformation matrix of the three-dimensional points with the same CCT code at the first position and the second position, that is, the relative spatial position of the first position and the second position can be obtained; and the point cloud data A and the point cloud data B are spliced ​​according to the relative spatial position of the first position and the second position; S6: Repeat steps S1 to S2 at other positions to obtain point cloud data at each position and the spatial three-dimensional coordinates of the center point of each CCT coding pattern; keep changing the position until the entire component is scanned; S7: Use the least squares method to obtain the relative spatial positions of any two positions, and splice the point cloud data at the two positions according to the relative spatial positions of the two positions; output the spliced ​​point cloud data, build a three-dimensional model of the complete component, and provide support for the subsequent calculation of the component measurement size.

3. The method for assembling point clouds of large parts based on measurement by a handheld three-dimensional scanner as claimed in claim 2, characterized in that: In step S5, the transformation matrix is ​​a rotation matrix or a translation matrix.

4. The method for assembling point clouds of large parts based on measurement by a handheld three-dimensional scanner as claimed in claim 3, characterized in that: In the step S5, the spliced ​​point cloud data is defined as point cloud data AB, where point cloud data AB=point cloud data A+point cloud data B*transformation matrix.

Images