Product surface appearance obtaining method based on three-dimensional scanning analysis

Abstract

The invention relates to a product surface appearance obtaining method based on three-dimensional scanning analysis. The method comprises the following steps: manually holding a portable optical three-dimensional laser scanning and light pen measuring system to scan a product, unfolding a phase image by using a multi-frequency heterodyne technology, and respectively obtaining continuous phase shift images of two cameras; the corresponding pixel points of the same three-dimensional point in the left camera and the right camera can be conveniently obtained by utilizing epipolar constraint and phase constraint, unordered point clouds are subjected to registration association, and a key frame extraction method is adopted, so that on the premise that the image quality is not influenced, the calculation power is effectively reduced, the calculation process is reduced, and the calculation efficiency is improved. And finally, the three-dimensional coordinate of the point in a camera coordinate system is calculated by using a binocular stereoscopic vision principle. The actual final assembly effect of the product can be effectively displayed, basic guarantee is provided for subsequent assembly path planning, and implementation of the automatic assembly process is facilitated; the manual operation frequency is reduced; the operation process is simple and convenient.

Description

Technical field:

[0001] The present invention belongs to the technical field of computer-aided optimization, and in particular relates to a method for obtaining product surface morphology based on three-dimensional scanning analysis. Background technology:

[0002] At present, the entire assembly process of products is operated manually. Complex assembly links such as large equipment and engine docking are manually assembled by personnel using tooling. Often, a docking process takes 2-3 days, which seriously increases the production cycle of the product. The manual operation mode cannot meet the needs of subsequent emergency batch production. The use of 3D scanning technology can obtain the surface morphology characteristics of the product, and the motion trajectory of the product assembly can be obtained by computer-assisted analysis, replacing the complexity of manual visual operation. At the same time, it can effectively reduce the assembly time and protect the internal equipment of the product from damage due to human error.

[0003] The actual assembly effect of the product is quite different from the original three-dimensional design state in terms of the distribution of spatial parts (especially flexible components such as cables). Compared with the traditional three-dimensional reconstruction method of industrial parts based on external positioning assistance, directly accurately aligning the surface structured light three-dimensional measurement data to achieve high-precision three-dimensional reconstruction of the overall shape of industrial parts is a practical and innovative method with great advantages. Summary of the invention:

[0004] The present invention proposes a method for obtaining the three-dimensional surface morphology of a product by expanding the phase image using multi-frequency heterodyne technology. After obtaining the continuous phase shift images of two cameras respectively, the corresponding pixel points of the same three-dimensional point in the left and right cameras can be conveniently obtained by using the epipolar constraint and the phase constraint, and finally the three-dimensional coordinates of the point in the camera coordinate system can be calculated by using the binocular stereo vision principle.

[0005] The present invention is achieved by the following technical solutions:

[0006] A method for obtaining product surface morphology based on three-dimensional scanning analysis comprises the following steps:

[0007] 1) A series of sinusoidal grating stripes are projected onto the surface of the object to be measured. Two cameras simultaneously capture the grating stripe pattern modulated by the surface shape of the object to be measured. The captured grating pattern can be expressed as:

[0008] I i (x, y) = A i (x,y)+B i (x,y)cos(φ(x,y)+δ i ),i=1,2,...,n,

[0009] Where (x, y) is the pixel coordinate on the captured image, Ii (x, y) represents the gray value of the pixel, A i (x, y) represents the average grayscale of the raster image, B i (x, y) represents the modulation grayscale, δ i is the fixed phase shift, n is the total phase shift, φ(x, y) is the phase shift to be determined;

[0010] 2) To measure the curvature consistency of the registration results of the two point cloud data, it is necessary to first calculate the curvature information of each 3D point. For any given 3D point on the 3D space surface, its principal curvature represents the curvature of the 3D surface along two orthogonal main directions at that point. Based on the principal curvature information, the Gaussian curvature and mean curvature of the point can be calculated accordingly.

[0011] 3) When registering unordered point clouds, it is necessary to perform nearest neighbor search in three-dimensional space to determine corresponding points. Different from registering unordered point clouds, for the registration between depth images, the corresponding relationship can be established more quickly and efficiently through projection matching. The process of establishing data association between depth images using projection matching requires the use of camera reverse projection model and forward projection model in sequence;

[0012] 4) When performing 3D scanning and reconstruction of large and complex parts, it is necessary to scan different areas of the object to be measured from multiple different viewpoints to obtain complete 3D point cloud data of the surface of the object to be measured. However, the key frame technology is used to select only a limited number of measurement data as key frames, and loop detection is performed between key frames to control the calculation scale;

[0013] 5) Establish output 3D model to obtain the 3D surface morphology of the product.

[0014] The present invention can effectively demonstrate the actual assembly effect of the product, provide a basic guarantee for subsequent assembly path planning, and is conducive to the implementation of the automated assembly process; it reduces the number of manual operations; and the operation process is simple and convenient. Description of the drawings:

[0015] Figure 1 It is a flowchart of the method of the present invention;

[0016] Figure 2 It is a schematic flow chart of the method of the present invention;

[0017] Figure 3 It is a schematic diagram of the process of manually scanning a product according to the method of the present invention. Specific implementation method:

[0018] Reference Figure 1 , 2 The technical invention is further described in detail with implementation cases.

[0019] Figure 1 , 2 The figure shows the overall process of the method for accurate registration by directly using the surface structured light three-dimensional measurement data of the present invention. Figure 3 ) Use portable optical 3D laser scanning and light pen measurement three-coordinate system to perform 3D scanning on the required product. The scanning form is blue light irradiation, and there is no need to process the surface of the scanned product. The data results are imported into the computer, and the phase image is expanded and analyzed using multi-frequency heterodyne technology to obtain the continuous phase shift images of the two cameras. After key frame extraction and processing, the 3D surface morphology required by the product is obtained.

[0020] The present invention proposes a method for obtaining product surface morphology based on 3D scanning analysis, which scans the product through a manually handheld portable optical 3D laser scanning and light pen measurement system, expands the phase image with multi-frequency heterodyne technology, aligns and associates the disordered point cloud, and adopts a key frame extraction method. Under the premise of not affecting the image quality, the computing power and the computing process can be effectively reduced, and the actual assembly effect of the product can be effectively displayed, providing a basic guarantee for subsequent assembly path planning, which is conducive to the implementation of the automated assembly process; reducing the number of manual operations; and the operation process is simple and convenient.

Claims

1. A method for obtaining product surface morphology based on three-dimensional scanning analysis, characterized in that: The following steps are involved: 1) A series of sinusoidal grating stripes are projected onto the surface of the object to be measured. Two cameras simultaneously capture the grating stripe pattern modulated by the surface shape of the object to be measured. The captured grating pattern can be expressed as: Yo i (x, y)=A i (x, y)+B i (x, y)cos(φ(x, y)+δ i ),i,1,2,...,n, Where (x, y) is the pixel coordinate on the captured image, I i (x, y) represents the gray value of the pixel, A i (x, y) represents the average grayscale of the raster image, B i (x, y) represents the modulation grayscale, δ i is the fixed phase shift, n is the total phase shift, φ(x, y) is the phase shift to be determined; 2) To measure the curvature consistency of the registration results of the two point cloud data, it is necessary to first calculate the curvature information of each 3D point. For any given 3D point on the 3D space surface, its principal curvature represents the curvature of the 3D surface along two orthogonal main directions at that point. Based on the principal curvature information, the Gaussian curvature and mean curvature of the point can be calculated accordingly. 3) When registering unordered point clouds, it is necessary to perform nearest neighbor search in three-dimensional space to determine corresponding points. Different from registering unordered point clouds, for the registration between depth images, the corresponding relationship can be established more quickly and efficiently through projection matching. The process of establishing data association between depth images using projection matching requires the use of camera reverse projection model and forward projection model in sequence; 4) When performing 3D scanning and reconstruction of large and complex parts, it is necessary to scan different areas of the object to be measured from multiple different viewpoints to obtain complete 3D point cloud data of the surface of the object to be measured. However, the key frame technology is used to select only a limited number of measurement data as key frames, and loop detection is performed between key frames to control the calculation scale; 5) Establish output 3D model to obtain the 3D surface morphology of the product.

Images