Super-density stripe structured light three-dimensional surface scanning method and system

Abstract

The invention belongs to the technical field of structured light three-dimensional measurement, and particularly relates to a super-density stripe structured light three-dimensional surface scanning method and system, and the method comprises the steps: obtaining camera parameters; projecting a super-density stripe structure light pattern formed by alternating wide bright and dark stripes to an object; acquiring an image by using a camera, and extracting a central point pixel coordinate of the stripe-structured light pattern by using a sub-pixel precision positioning method; calibrating a light plane equation of each stripe structure light pattern under a camera coordinate system by means of a plane target; reversely projecting the pixel coordinate of the central point of the stripe-structured light pattern into a ray and intersecting the ray with a corresponding light plane to obtain the three-dimensional coordinate of the ray and form an initial three-dimensional point cloud; and finally, carrying out gray scale estimation and filling on missing data caused by signal saturation or difficult positioning based on a radiation brightness physical model to obtain complete and accurate three-dimensional surface shape data. According to the invention, high-precision, high-density and high-integrity three-dimensional surface scanning is realized, and the reliability and applicability of a measurement result are remarkably improved.

Description

Technical Field

[0001] This invention belongs to the field of structured light three-dimensional measurement technology, specifically relating to a method and system for ultra-density striped structured light three-dimensional surface scanning. Background Technology

[0002] For a long time, traditional contact measurement methods have been widely used due to their high measurement accuracy. However, with technological advancements and practical engineering needs, non-contact 3D reconstruction measurement methods have developed rapidly, better meeting various requirements. Non-contact 3D reconstruction measurement methods refer to obtaining and measuring the 3D spatial structure of a target without physical contact. Choosing the appropriate method for reconstruction and measurement, considering factors such as the target's structural characteristics, measurement environment, and accuracy requirements, is a key challenge in solving practical problems. Among these methods, structured light 3D measurement technology, with its advantages of high accuracy and rapid measurement, has always been a research hotspot in the field of computer vision.

[0003] Currently, most structured light 3D measurement technologies are active projection structured light, based on the principle of triangulation. The system typically consists of a projector and an industrial camera. The projector projects a series of coded structured light patterns onto the object's surface, while the camera captures the pattern deformed by the object's surface shape from another angle. The 3D coordinates of each point on the object's surface are calculated by combining the calibrated geometric relationship between the camera and the projector. The phase-shifting method is commonly used for calculation and unfolding to obtain the absolute phase value. However, for applications with high real-time requirements, such as dynamic scene capture, the complexity and computational power of the phase-decomposition algorithm limit the number of stripes, resulting in lower stripe density, poorer performance, and insufficient coverage. This leads to the problem of missing 3D reconstruction data due to insufficient stripe quantity, a problem common in traditional stripe technology. Summary of the Invention

[0004] To address the problems existing in the prior art, the present invention aims to provide a method and system for ultra-density striped structured light three-dimensional surface scanning, which is applicable to the measurement of light structures with ultra-high stripe density, enabling more detailed capture of minute changes on the surface of objects, thereby improving the accuracy and completeness of three-dimensional reconstruction, realizing the extension from line scanning to surface scanning, providing better scanning coverage, and reducing measurement errors.

[0005] The technical solution of this invention is: A method and system for three-dimensional surface scanning with ultra-density striped structured light includes the following steps: The camera was calibrated using Zhang's calibration method to obtain camera parameters; A striped structured light pattern is projected onto the surface of the object being tested by a light source projector. The striped structured light pattern is a striped light pattern formed by alternating bright and dark stripes to cover the surface of the object. The striped structured light pattern is captured by a camera. The pixel coordinates of the center point of the striped structured light pattern are extracted from the image captured by the camera. A sub-pixel precision positioning method is adopted. The zero intersection point of the striped structured light pattern is calculated by convolution response value, thereby improving the positioning accuracy from the pixel level to the sub-pixel level. The light plane of each stripe structured light pattern is calibrated using a calibration plate, and the light plane equation of each light plane in the camera coordinate system is obtained. The pixel coordinates of the center point of the striped structured light pattern are back-projected into a ray originating from the optical center of the camera, and the intersection of this ray with the corresponding light plane equation obtained by calibration is obtained to obtain the three-dimensional coordinates of the center point of the striped structured light pattern in the camera coordinate system. The three-dimensional coordinates of all the center points of the striped structured light pattern are collected to form the initial three-dimensional point cloud of the measured object. Based on radiance, grayscale values ​​are estimated for data points missing in the initial 3D point cloud due to signal saturation in wide bright stripe regions or difficulty in center positioning, thus filling in the missing points and obtaining complete and accurate 3D surface data of the measured object.

[0006] Preferably, the camera parameters include intrinsic parameters, extrinsic parameters, and distortion parameters; wherein: Intrinsic parameters are used to describe the camera's internal geometry and optical characteristics, including the camera's focal length and principal point coordinates; The extrinsic parameters are used to describe the camera's attitude and position in the world coordinate system, including rotation matrices and translation vectors. The rotation matrix represents the rotation transformation from the world coordinate system to the camera coordinate system, and the translation vector represents the offset of the origin of the world coordinate system relative to the origin of the camera coordinate system.

[0007] Preferably, the bright and dark stripes in the striped structured light pattern are super-density bright and dark stripes with the same pixel width, or wide bright and dark stripes with a pixel width ratio of 15:1; wherein the stripe density of the super-density bright and dark stripes is greater than 2 stripes / mm.

[0008] Preferably, the zero-intersection point of the striped structured light pattern is determined according to the following formula: , In the formula, where, j Integer pixel position indicating where the sign change occurs, i.e., the index of the convolution response value that is negative; R ( j) Indicates position j The convolution response value at that point is negative. R ( j+ 1) Indicates location j The convolution response value at +1 is positive. subPixelPos This indicates the calculated zero-crossing point position with sub-pixel precision.

[0009] Preferably, the method for obtaining the equation of the light plane in the camera coordinate system includes the following steps: Using the camera's focal length, principal point coordinates, and distortion parameters, the pixel coordinates of the center point of the extracted striped structured light pattern are normalized to obtain the corrected pixel coordinates. The corrected pixel coordinates are then back-projected as rays originating from the camera's optical center; Based on the rotation matrix and translation vector, the plane equation of the calibration plate is transformed from the world coordinate system to the camera coordinate system, and the plane equation of the calibration plate in the camera coordinate system is obtained. Obtain the intersection point of the ray equation and the calibration plate plane equation to obtain the three-dimensional coordinates of the center point of the striped structured light pattern in the camera coordinate system; The three-dimensional coordinates of the center points of all striped structured light patterns are collected, and the least squares method is used to fit and form the light plane equation.

[0010] Preferably, the radiance physical model decomposes the light propagation process into two stages: the first stage is the reflection stage after the light emitted from the light source illuminates the object surface, and the second stage is the imaging stage after the reflected light from the object surface enters the camera; the grayscale value estimation is determined according to the following formula: , in, , In the formula, Grayscale value; The radiant power of the light source; The electro-optical conversion efficiency of the light source is constant when the light source is determined. The photoelectric conversion coefficient of a pixel; Image distance; when the camera position is determined, , as well as All are constant values; when the surface material of an object is determined, This is a physical property of the object's surface, and it is a constant value. The pixel on the object's surface corresponds to the center of the striped structured light pattern. This refers to a tiny area on the surface of the corresponding object. The pixel area of ​​the pixel on the imaging surface; The area of ​​a tiny region; The distance between a tiny area and the light source; The distance between the tiny region and the center point of the camera lens; The angle between a small region and the surface directly opposite the solid angle; The angle between the pixel plane of the pixel on the imaging surface and the face angle of the opposite solid angle; It is the angle between the tiny region and the pixel on the corresponding pixel plane of the imaging surface.

[0011] Preferably, it also includes: processing missing data in the initial 3D point cloud caused by occlusion or measurement error by optimizing measurement settings or model matching repair; wherein, the optimized measurement settings include aligning the camera optical center and the ring structured light emission point emitted by the light source projector along the same straight line to reduce occlusion.

[0012] Preferably, before projecting the super-density fringe structured light pattern onto the surface of the object being measured, an envelope space estimation is performed to ensure that the fringe structured light pattern is not occluded on the object surface: By using camera intrinsic parameters and light plane calibration, an unobstructed spatial region is determined to adjust the number or spacing of the striped structured light pattern and reduce occlusion; the envelope space is composed of the intersection line between the striped structured light pattern acquisition plane and the base plane, as well as the intersection line between the light planes.

[0013] Preferably, it also includes determining the occlusion of the striped structured light pattern on the object surface, including the following steps: Get the pixel coordinate information of the object and store it in the array container V. Create a three-channel window and set the background image to gray. Draw the laser line in red based on the input laser position, number of stripes, and coverage area; Iterate through the pixels stored in container V, determine whether the pixel is on the laser line, and if so, connect the pixel to the camera location with a blue line. Repeatedly traverse the pixels in container V and check if the number of blue lines connecting the pixels to the camera is equal in both cases. If they are not equal, it is determined that a stripe occlusion phenomenon has occurred.

[0014] A high-density striped structured light three-dimensional surface scanning system for implementing the above method includes: Camera, used to capture images of striped structured light patterns; A light source projector is used to project ultra-density striped structured light patterns onto the surface of the object being measured. Processing unit, used to perform the following operations: The camera was calibrated using Zhang's calibration method. The light source projector projects a super-density striped structured light pattern, and the pixel coordinates of the center point of the striped structured light pattern are extracted from the image for sub-pixel precision positioning. The light plane containing each stripe structured light pattern is calibrated using a calibration plate, and the plane equation of each light plane in the camera coordinate system is obtained. The pixel coordinates are back-projected into rays, and the intersection points with the light plane equation are found to generate the initial 3D point cloud of the object being measured. The missing point cloud data was filled in using a radiance physical model to obtain complete 3D surface data.

[0015] Compared with the prior art, the ultra-density striped structured light three-dimensional surface scanning method and system of the present invention have the following beneficial effects: 1. By utilizing ultra-dense wide bright and dark fringes, and keeping the dark fringes constant, the point cloud gaps in the wide bright fringe areas are filled by grayscale estimation based on illumination imaging. This yields complete and accurate 3D surface data of the measured object, enabling more detailed capture of minute changes on the object's surface. This improves the accuracy and completeness of 3D reconstruction, reduces the problem of missing 3D reconstruction data caused by insufficient fringe count in traditional fringe techniques, and expands from line scanning to surface scanning. This results in better scanning coverage and reduced measurement errors.

[0016] 2. Enables dynamic object scanning: Utilizes ultra-dense stripes and wide bright and dark stripes to effectively increase data coverage; effectively captures surface information of dynamic objects, reduces measurement errors caused by object movement, and achieves data coverage exceeding the required measurement accuracy.

[0017] 3. Estimate stripe occlusion and restore the overall surface shape of the object: The concept of envelope space is proposed to more intuitively determine the occlusion of stripe structured light patterns. Attached Figure Description

[0018] Figure 1 This is an overall technical circuit diagram in an embodiment of the present invention; Figure 2 This is a schematic diagram of the system structure in an embodiment of the present invention, where 1 is a camera; 2 is a light source projector; 3 is a processing unit; 4 is an object; 5 is the imaging plane of the camera; and 6 is the light plane. Figure 3 This is a schematic diagram of the calibration plate structure in an embodiment of the present invention; Figure 4 This is a stripe center detection image from an embodiment of the present invention; Figure 5 This is a schematic diagram of the Gaussian first derivative method for extracting three-dimensional point clouds in an embodiment of the present invention; Figure 6 This is a schematic diagram of wide bright and dark stripes and high-density bright and dark stripes in an embodiment of the present invention; Figure 7 This is a schematic diagram of optical plane calibration in an embodiment of the present invention; Figure 8 This is a schematic diagram of the light propagation process in an embodiment of the present invention; Figure 9 This is a schematic diagram showing the general placement of the camera and structured light in an embodiment of the present invention; Figure 10 This is a schematic diagram illustrating excessively large three-dimensional coordinates in an embodiment of the present invention; Figure 11This is a schematic diagram showing the camera optical center and the ring structure light emission point aligned in a straight line in an embodiment of the present invention; Figure 12 This is a schematic diagram of the striped structured light pattern occlusion phenomenon in an embodiment of the present invention; Figure 13 This is a schematic diagram of the envelope space in an embodiment of the present invention; Figure 14 This is a schematic diagram of the three-dimensional point cloud data of the PVC pipe in an embodiment of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0020] Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of this invention.

[0021] Furthermore, the technical solutions of the various embodiments of the present invention can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such combination of technical solutions does not exist and is not within the scope of protection claimed by the present invention.

[0022] See Figures 1 to 2 As shown, in order to improve the accuracy and completeness of 3D reconstruction, an extension from line scanning to surface scanning is achieved, resulting in better scanning coverage and reduced measurement errors. This embodiment provides a super-density striped structured light 3D surface scanning method and system. The system includes a camera 1, a light source projector 2, and a processing unit 3. The camera is used to acquire images of striped structured light patterns. The light source projector projects ultra-dense striped structured light patterns onto the surface of the object being measured. The processing unit performs the following operations: calibrates the camera using Zhang's calibration method; controls the light source projector to project ultra-dense striped structured light patterns and extracts the pixel coordinates of the center point of the striped structured light pattern from the image for sub-pixel accuracy positioning; calibrates the light plane containing each striped structured light pattern using a calibration plate to obtain the plane equations of each light plane in the camera coordinate system; projects the pixel coordinates back into rays and finds their intersection points with the light plane equations to generate the initial 3D point cloud of the object being measured; and applies a radiance physical model to fill in the missing data in the point cloud to obtain complete 3D surface data.

[0023] The method for performing three-dimensional surface scanning of ultra-density striped structured light based on the above system includes the following steps: S1. Calibrate the camera using Zhang's calibration method to obtain camera parameters.

[0024] S2. Project a striped structured light pattern onto the surface of the object being measured using a light source projector to cover the surface of the object. The striped structured light pattern is formed by alternating wide bright stripes and dark stripes, and the bright stripes and dark stripes are ultra-dense bright and dark stripes with the same pixel width, or wide bright and dark stripes with a pixel width ratio of 15:1; wherein, the stripe density of the ultra-dense bright and dark stripes is greater than 2 stripes / mm.

[0025] S3. Use a camera to capture the striped structured light pattern, extract the pixel coordinates of the center point of the striped structured light pattern from the image captured by the camera, and use a sub-pixel precision positioning method to calculate the zero intersection point of the striped structured light pattern through convolution response value, thereby improving the positioning accuracy from the pixel level to the sub-pixel level.

[0026] S4. Use the calibration plate to calibrate the light plane where each stripe structured light pattern is located, and obtain the light plane equation of each light plane in the camera coordinate system.

[0027] S5. Project the pixel coordinates of the center point of the striped structured light pattern in reverse as a ray originating from the optical center of the camera, and find the intersection point of this ray with the corresponding light plane equation obtained by calibration to obtain the three-dimensional coordinates of the center point of the striped structured light pattern in the camera coordinate system. Gather the three-dimensional coordinates of all the center points of the striped structured light patterns to form the initial three-dimensional point cloud of the measured object.

[0028] S6. Based on radiance, for data points missing in the initial three-dimensional point cloud due to signal saturation in the wide bright stripe region or difficulty in center positioning, grayscale values ​​are estimated to fill in the missing points in the point cloud and obtain complete and accurate three-dimensional surface data of the measured object.

[0029] The detailed explanation of the above steps is as follows: Furthermore, camera parameters include intrinsic parameters, extrinsic parameters, and distortion parameters. Intrinsic parameters describe the camera's internal geometry and optical characteristics, including the camera's focal length and principal point coordinates. Extrinsic parameters describe the camera's attitude and position in the world coordinate system, including rotation matrices and translation vectors. The rotation matrix represents the rotation transformation from the world coordinate system to the camera coordinate system, and the translation vector represents the offset of the origin of the world coordinate system relative to the origin of the camera coordinate system.

[0030] Specifically, a multi-line structured light monocular vision measurement system is employed, mainly composed of a projector, an industrial camera, and a computer. The camera calibration method uses the classic Zhang's calibration method, a practical method proposed by Zhang Zhengyou that utilizes a planar checkerboard for camera calibration. Its basic principle involves using a planar calibration plate (a solid circular calibration plate), such as... Figure 3Multiple images are captured from different camera perspectives. Feature points such as corner points are extracted, and the homography matrix corresponding to each image is calculated. Then, the camera's intrinsic and extrinsic parameters are solved. Intrinsic parameters include the camera's focal length and principal point coordinates; extrinsic parameters are the camera's position and orientation relative to the calibration board (world coordinates). Finally, the reprojection error is minimized through nonlinear optimization to obtain high-precision camera parameters. Because Zhang's calibration method has advantages such as ease of operation, strong applicability, and high accuracy, this invention selects this method as the core step of system calibration.

[0031] The current step uses OpenCV The library implements camera calibration code, using an industrial camera to capture multiple images and store them in a folder. glob () Read the directory .bmp File, then use findCirclesGrid The function () detects a 7x7 circular array, saves the corresponding point set after successful detection, and sets the plane where the calibration plate is located as the world coordinate system. XY flat( Z =0), and set the world coordinates of the first center point to the origin (0, 0), initialize the camera parameter matrix and distortion coefficient matrix, and call calibrateCamera () Calculate the camera intrinsic matrix and distortion coefficients, as well as the rotation matrix and translation vector for each image.

[0032] 2. Extract the center of the striped structured light pattern to obtain the zero intersection point of the striped structured light pattern, thereby improving the positioning accuracy from the pixel level to the sub-pixel level.

[0033] The center position of the fringe structured light pattern is extracted using the Gaussian first derivative method. For the wide-fringe structured light used in this invention, the pattern does not conform to a Gaussian distribution due to the change in fringe width; however, the dark fringe areas conform to a reverse Gaussian distribution. The basic principle of the Gaussian first derivative method is... createGaussianKernel The function generates a Gaussian first-order derivative kernel, the shape of which resembles an antisymmetric peak-trough curve, with its center corresponding to the inflection point of the Gaussian function. When this kernel is convolved with the striped structured light pattern image, zero-crossing points are generated at the locations where the stripe brightness changes most strongly, eliminating invalid points. Sub-pixel coordinates are calculated and the positions are marked through interpolation. This invention mainly employs an image processing method based on convolution and linear interpolation to achieve sub-pixel-level detection of the center of the striped structured light, mainly including image preprocessing, convolution operation, sign transformation detection, and linear interpolation calculation.

[0034] 1. The definition of the Gaussian function (two-dimensional) is: , in, Indicates at coordinate point The Gaussian function value calculated at the location; The x and y coordinates of a pixel in a striped structured light pattern image; Pi; It is a natural constant; The standard deviation controls the width and smoothness of the Gaussian distribution. The edges of bright and dark stripes represent abrupt changes in grayscale values, and the first derivative effectively balances the rate of grayscale change. When the light intensity follows a Gaussian distribution, the derivative exhibits a sign jump at the edges. In the process of extracting the center of the striped structured light pattern, this invention generates a suitable range of independent variable values ​​to define the first-order Gaussian derivative kernel. The expression for the kernel is:

[0035] , in, This indicates the position of the original Gaussian function. The first derivative value at the point. After the first derivative kernel is convolved with the original signal, an extremum will be generated at the edge of the stripe (where the gray level changes the fastest). This can accurately detect the edges of bright and dark stripes while suppressing noise, thus achieving precise positioning of the stripe center.

[0036] 2. After obtaining the edge positions of the bright and dark fringes, the center position of the fringes can be determined by calculating the midpoints of adjacent edge positions. This involves comparing the product of adjacent elements in the convolution result to see if it is less than 0 to identify the location where the sign changes. If the product of two adjacent elements is less than 0, it indicates that their signs are different, meaning a sign change has occurred. To further improve detection accuracy, sub-pixel optimization is performed on the sign change points using linear interpolation. The interpolation formula is:

[0037] , In the formula, where, j Integer pixel position indicating where the sign change occurs, i.e., the index of the convolution response value that is negative; R ( j) Indicates position j The convolution response value at that point is negative. R ( j+ 1) Indicates location j The convolution response value at +1 is positive. subPixelPos This indicates the calculated zero-crossing point position with sub-pixel precision.

[0038] Through the above steps, high-precision detection of the center of structured light patterns with wide dark stripes and uniform stripes can be achieved. This method can improve the positioning accuracy from the pixel level to the sub-pixel level (e.g., 0.1 pixels), significantly improving the overall accuracy of the measurement system, as well as stripe center detection and 3D point cloud detection. Figure 4 and Figure 5 .in, Figure 4The scan data corresponding to each pixel in the medium-width bright stripe area can be obtained, with a coverage ratio of 15 / 16. Figure 4 The number of ultra-high density stripes can reach 2.4 stripes / mm, reaching the ultra-high density benchmark.

[0039] III. Calibrate the plane equations of each striped structured light pattern. Furthermore, the method for obtaining the equation of the light plane in the camera coordinate system includes the following steps: Using the camera's focal length, principal point coordinates, and distortion parameters, the pixel coordinates of the center point of the extracted striped structured light pattern are normalized to obtain the corrected pixel coordinates. The corrected pixel coordinates are then back-projected as rays originating from the camera's optical center; Based on the rotation matrix and translation vector, the plane equation of the calibration plate is transformed from the world coordinate system to the camera coordinate system, and the plane equation of the calibration plate in the camera coordinate system is obtained. Obtain the intersection of the ray equation and the calibration plate plane equation to get the three-dimensional coordinates of the center point of the striped structured light pattern in the camera coordinate system; The three-dimensional coordinates of the center points of all striped structured light patterns are collected, and the least squares method is used to fit and form the light plane equation.

[0040] Specifically, in line structured light measurement, a single light stripe is projected onto an object. The area of ​​the projected light stripe in space can be extracted into a plane, intersecting with the surface of the object to form a characteristic striped structured light pattern. By establishing a line structured light measurement model, the pixel coordinates of the center of the striped structured light pattern are obtained from images captured by a camera through light stripe extraction. Then, by adding constraints from the light plane equation, the corresponding three-dimensional coordinate points can be calculated from the image points. This invention uses a planar target, i.e., Zhang's calibration method, to complete the calibration of multi-line structured light. The stripes are selected as high-density bright and dark stripes and 15 sets of wide bright and dark stripes (bright stripes become wider, while the number of pixels in the dark stripes remains unchanged), such as... Figure 6 As shown, the pixel ratio of the wide bright stripes to the dark stripes is 15:1.

[0041] The calibration process is as follows Figure 7 As shown, the light source projector projects multiple light planes, forming multiple straight stripes on the camera's imaging plane.

[0042] In the above steps, the camera's intrinsic parameter matrix has been obtained. Let the coordinates of the origin of the image coordinate system in the pixel coordinate system be... Rotation matrix R Translation vector T In step two, the pixel coordinates of the center of the striped structured light pattern have already been obtained. u strip , v stripUsing the camera intrinsic parameter matrix and distortion parameters obtained in step one, the pixel coordinates are normalized to obtain the corrected pixel coordinates. u corrected , v corrected Using the camera model, the pixel is back-projected as a ray originating from the camera's optical center. The equation of the ray is:

[0043] , in, Indicates the point in the camera coordinate system Z Coordinates, i.e., the distance from the point to the optical center of the camera; f x and f y These represent the focal lengths along the X and Y axes, respectively.

[0044] Based on the external parameters obtained in step one R , T The nodal transformation of the calibration plate can be applied to the camera coordinate system, and the plane equation of the calibration plate is: , , , These are the components of the normal vector forming the calibration plate plane along the X, Y, and Z axes of the camera coordinate system. The ray equations are then... Substituting into the plane equation, we can finally obtain the 3D camera coordinates of the intersection point:

[0045] in, Let be the three-dimensional coordinate vector of a point on the ray in the camera coordinate system. This is the proportionality coefficient. , , Direction vector d The three components are located on the X, Y, and Z axes of the camera coordinate system. The center points of all the striped structured light patterns are gathered together, and the least squares method is used to fit the plane. The plane equation is: ,make C =1, for all the center points of the striped structured light pattern ( x i ,y i ,z i From this, we can conclude that: , written as:

[0046]

[0047] Solve using the least squares method Parameters A, B, and D can be obtained, and thus the equation of the light plane can be derived.

[0048] Fourth, further, the radiance physical model decomposes the propagation process of light into two stages: the first stage is the reflection stage after the light emitted from the light source shines on the object surface, and the second stage is the imaging stage after the light reflected from the object surface enters the camera. like Figure 8 As shown, C Indicates observing the camera, L Indicates the light source, P Represents the imaging plane. S This indicates the illuminated area on the surface of the observed object. A Represents a specific pixel. This represents a tiny area on the surface of the object corresponding to A. Light rays travel from... L Sent, via S After reflection, it enters the camera. C .

[0049] 1. In the first stage, the main physical quantities and relationships involved are as follows: Radiant energy of the light source: Q .

[0050] Light source radiant power: .

[0051] Electro-optical conversion efficiency of the light source: .

[0052] Luminous flux: .

[0053] When from L When the light emitted from a point is modulated by a lens and radiates outward in a hemispherical wave shape: sphericity:

[0054] Light intensity emitted by the light source: , When a tiny area Area is The distance from the light source is The angle between the face and the face facing the solid angle is hour: exist Luminous flux: , The corresponding illuminance , Assumption Surface reflectivity : Reflected light intensity: , Reflected illuminance: , Sunlit area S It can be decomposed into many disjoint similar The sub-regions, each with its own luminous flux, illuminance, and emissivity, can be modeled. Calculation of the form of the region.

[0055] 2. The main physical quantities and relationships involved in the second stage are as follows: like Figure 7 As shown, from The light rays will reach the image point A The corresponding pixel is then converted into an electrical signal, which is subsequently recorded as a grayscale value. The distance from the center point of the camera lens is Image distance is The pixel area is The angle between the pixel plane and the face angle of the opposite solid is :

[0056] Like a dot A Luminous flux at the corresponding pixel: , Observable illuminance: , If we This represents the photoelectric conversion coefficient of a pixel. Represents grayscale value, when it is small and A The angle formed by the corresponding pixel plane is ,and When, the calculation yields: , In the formula, Grayscale value; The radiant power of the light source; The radiant energy of the light source; The electro-optical conversion efficiency of the light source is constant when the light source is determined. The photoelectric conversion coefficient of a pixel; Image distance; when the camera position is determined, , as well as All are constant values; when the surface material of an object is determined, This is a physical property of the object's surface, and it is a constant value. The pixel on the object's surface corresponds to the center of the striped structured light pattern. This refers to a tiny area on the surface of the corresponding object. The pixel area of ​​the pixel on the imaging surface; The area of ​​a tiny region; The distance between a tiny area and the light source; The distance between the tiny region and the center point of the camera lens; The angle between a small region and the surface directly opposite the solid angle; The angle between the pixel plane of the pixel on the imaging surface and the face angle of the opposite solid angle; It is the angle between the tiny region and the pixel on the corresponding pixel plane of the imaging surface.

[0057] When the system consisting of the light source and the camera is determined... k It is only related to the surface reflectivity of the object being measured. When the pixels are observed... A After obtaining the grayscale value and the above parameter values ​​through calculation, substituting them into the formula, the result can be calculated. k For relative values, the same method is used to calculate the values ​​for all pixels corresponding to all striped structured light pattern regions on the image.

[0058] 5. When solving for the 3D point cloud data of the scanned object, the placement of the measurement settings may cause issues. Figure 9 And the situation where the three-dimensional coordinate values ​​are too large due to the influence of angle ( ) Figure 10 This results in missing data in the initial 3D point cloud due to occlusion or measurement errors. Figure 9 In the image, 11 represents the striped structured light pattern formed by the ring-shaped structured light projected onto the object; 12 represents the image formed by the striped structured light pattern on the imaging plane; 13 represents the striped structured light pattern formed by the intersection of the structured light and the plane; 5 represents the imaging plane of the camera; (dot) C For the camera's optical center, point L For structured light emission points, point P A point in the world coordinate system is a stripe pattern formed by structured light projected onto the scanned object. Q For point P The projection point on the camera plane, point G For structured light emission points and points P The intersection point of the line and the imaging plane. And the solution point. P The three-dimensional coordinates in the camera coordinate system are the solution for the straight line. CQP Intersection with structured light cone surface P The three-dimensional coordinates of a line can be simply described as a straight line. CQP With a straight line LGP The coordinates of the intersection point. It can be seen that the angle between the two lines at this point is... When the included angle The smaller the solution point P The larger the three-dimensional coordinates, the larger the obtained three-dimensional coordinate values ​​will be in the solution.

[0059] The above-described problems can be addressed by optimizing measurement settings or fixing model matching issues: 1. Optimize measurement settings: This includes aligning the camera optical center and the ring structured light emission point emitted by the light source projector along the same straight line to reduce occlusion.

[0060] Figure 11 In the image, 5 represents the camera's imaging plane; 111 represents the image of the striped structured light pattern on the imaging plane; and 112 represents the striped structured light pattern formed by the ring structured light projected onto the object. C For the camera's optical center, point L The point is the ring-shaped structured light emission point. H The center of the striped structured light pattern formed by the ring structured light projected onto the object is the point. H The projection point on the imaging plane, point P A point in the world coordinate system is a point on the striped structured light pattern formed by the ring structured light projected onto the scanned object. P The projection point on the imaging plane.

[0061] As can be seen from the above, a suitable included angle is sufficient to find the correct point. P Coordinate values.

[0062] By adjusting the ring structured light emission point L Wako center point C The angle can be adjusted by adjusting the distance between the ring structured light and the optical center. Therefore, the measurement settings can be optimized during equipment setup, and the scanning results can be kept within the correct range by adjusting the distance between the ring structured light and the optical center.

[0063] 2. Model matching repair: Based on Figure 9 By constructing a geometric model using the principle of structured light measurement, a straight line can be... CQP With a straight line LGP The angle is quantified, and the derivation shows that it contains a tangent function. Therefore, the resulting three-dimensional coordinate values ​​may be too large. This can be solved by the following method: (1) Geometric model constraints Constrain the calculated 3D coordinates using known geometric models or prior knowledge.

[0064] (2) Data interpolation and repair For lost 3D coordinate data, interpolation repair can be performed using neighborhood information. A convolutional interpolation algorithm can be used, leveraging the pixel values ​​within the floating-point coordinate neighborhood to enhance the resolution of high-frequency components through convolution operations, thereby recovering the lost data.

[0065] (3) Optimization algorithm In the process of solving three-dimensional coordinates, optimization algorithms, such as the iterative linear least squares triangle algorithm, are introduced to improve the calculation accuracy by minimizing the reprojection error.

[0066] (4) Error detection and removal During the calculation process, outliers are detected in real time, and statistical methods are used to identify and remove these outliers, thereby improving the overall accuracy of the data.

[0067] VI. Furthermore, before projecting the super-density fringe structured light pattern onto the surface of the object being measured, the envelope space of the fringe structured light pattern on the object surface without occlusion is estimated: 1. Estimate the stripe occlusion situation For uniformly arranged striped structured light patterns, when the surface of the object has many bumps and depressions and the height difference exceeds a certain range, stripe occlusion is likely to occur, such as... Figure 12 The diagram shows ①②③. In the diagram, 1 is the camera; 2 is the light source projector; and 5 is the camera's imaging plane.

[0068] For multi-line structured light measurement systems, matching the center of the fringe structured light pattern is a critical step. However, occlusion of the fringe structured light pattern is one of the main reasons for matching difficulties, which can seriously increase the error rate and data loss rate, affecting subsequent processing and analysis. Therefore, this invention aims to estimate the unoccluded spatial region of the fringe structured light pattern. By using camera intrinsic parameters and light plane calibration, an unoccluded spatial region is determined to adjust the number or spacing of the fringe structured light pattern, reduce occlusion, and obtain a fringe structured light pattern that meets accuracy requirements and is adapted to the measured object.

[0069] (1) Establishing the envelope space like Figure 13 As shown, to reduce occlusion of the fringe structured light pattern, this study proposes to determine an unoccluded spatial region, namely the envelope space, based on the field of view, the number of fringes, and the camera-projector pose relationship by estimating the occlusion area and reducing the fringe density. Within this space, the probability of fringe occlusion is greatly reduced. Through camera intrinsic parameters and light plane calibration, the planar equations of each fringe structured light pattern in the camera coordinate system are obtained. When the fringes are projected onto the base plane, the fringe structured light pattern is not occluded and does not distort, allowing for rapid matching along the horizontal axis according to the sequence number. Figure 13 In the diagram, 1 is the image acquisition device (camera); 2 is the light source projector; 131 is the light plane; 132 is the striped structured light pattern acquisition plane; 133 is the striped structured light pattern; and 134 is the base plane.

[0070] Suppose that the light emitted from the light source k The equations for the light planes are as follows: ... , in, X , Y , Z Represents the coordinates of a point in three-dimensional space. Represents the first light plane. Indicates the first k A light plane, The decision was made k The position of a light plane in space. and The coefficient that controls the tilt of the light plane relative to the XZ and YZ planes is the coefficient that determines the light plane; both values ​​cannot be 0 at the same time.

[0071] The camera observed the corresponding k The equation for the acquisition plane of the striped structured light pattern is: ... , in, C Represents camera, and The orientation of the acquisition plane is determined. This determined its location. and It is automatically calculated through a camera-laser plane joint calibration process. Therefore, the envelope space can be measured. k The area enclosed by +1 straight line (including the baseline) constitutes ( Figure 11 (Only the cross-section of the straight lines is shown). The first and last lines are the intersection lines of the corresponding striped structured light pattern acquisition plane equation and the base plane, respectively. Figure 11 The grid area is shown. The middle part is the first... m Striped structured light pattern acquisition plane and the first m +1 intersection line of light planes, with constraints as shown in the equation:

[0072] , in, k This represents the total number of striped structured light patterns projected by the structured light projector. , , Together they defined the first emission emitted by the light source. m +1 The position and orientation of the light plane in three-dimensional space, all three cannot be 0 at the same time; , , Together they defined the first [value] observed by the camera. m The position and orientation of the plane corresponding to the striped structured light pattern in three-dimensional space.

[0073] After solving: , in, t The parameters represent the parameters of the parametric equation of a straight line.

[0074] When finding the plane formed by the m-th line and the (m+1)-th line, to simplify the expression as much as possible, we take a point on each of the two lines with the parameter t=0 and substitute it into the plane equation ax + by + c = z. The solution is: , For the sake of simplicity, These are not new, independent parameters, but rather the original parameters whose complex, common computational parts are replaced with these letters. , , , , , The combination The values ​​are shown below. , (2) Determining the occlusion of the striped structured light pattern on the object surface Based primarily on the input data of the object being measured, the number of stripes, laser coordinates, and camera coordinates, the algorithm simulates the actual cross-section of the object during scanning. Occlusion is determined by the number of intersections between the camera line and the laser line with the object. The image processing algorithm is as follows, including the following steps:

[0075] Step 1. Obtain the pixel coordinate information of the object and store it in the array container V. Create a three-channel window and set the background image to gray. Step 2. Draw the laser line in red based on the basic information input by the user, such as the laser position, number of stripes, and coverage area; Step 3. Iterate through the pixels stored in container V, set the green channel value to 255, and at the same time check whether the red channel of the point is 255. If so, store the point in container V1 and connect the point to the camera with a blue line. Step 4. Iterate through the pixels in container V again. If the blue channel value is 255, store the pixel in container V2. Step 5. Determine if the number of containers V1 and V2 are equal. If they are not equal, stripe occlusion has occurred.

[0076] 2. Match the center of the striped structured light pattern Step 1 estimates the fringe occlusion. In practical applications, limitations such as accuracy requirements and the object being measured may cause distortion and breakage in the acquired fringe structured light pattern, increasing the difficulty of matching the center of the fringe structured light pattern. The following matching method is proposed:

[0077] (1) Sequential matching method The main process is as follows: Each row is scanned sequentially, and the extracted points are temporarily placed in a temporary array. After scanning the entire row, the points are compared. If the number of points equals the number of stripes, they are sequentially stored in the arrays corresponding to each stripe. If they are not equal, the points scanned in that row are discarded, and the process continues to scan the next row until the entire image or region of interest has been scanned. In special cases, such as when the entire stripe pattern is tilted due to surface features or projector angle, resulting in fewer stripes in each row at the ends of the pattern than the total number of stripes, if there are few noise points, the scanning order at the ends of the pattern is not used for comparison; instead, the extraction order is directly used as the sequence number of the stripe structured light pattern.

[0078] (2) Elevation difference matching method Main steps: a. After fixing the position of the laser and camera and completing the preliminary calibration, place the whiteboard on the base surface and the surface where the highest point of the object is located, and take two images.

[0079] b. The striped structured light patterns of the two images are extracted and matched using the gray-scale centroid method and the number of stripes matching method to determine the range of each stripe on the image.

[0080] c. During actual measurement, the extracted points are sequentially matched to their ranges and assigned to the corresponding stripes. If a match fails, the points are discarded as noise.

[0081] (3) Breadth-first search matching method Main steps: The input image is the image to be matched using structured light patterns, and the output is the matched pixels. The region of interest is traversed by row scanning. If the number of points obtained equals the actual number of stripes, the points in that row are retained and sequentially added to the corresponding stripe array; otherwise, it is impossible to determine which stripe they belong to, and they are added to the unmatched array. Then, the matched points are pushed into a queue, and the points in the queue are matched sequentially. X 0 , Y 0 Search for points to be matched. X , Y )satisfy and ( m and n (Set the search range according to the stripe spacing), then the point will be assigned to the search range. X 0 , Y 0 The stripe containing the structured light pattern is pushed into the queue, and the search continues downward until the queue is empty. Then, the above operation is repeated for the next stripe where the structured light pattern has been matched.

[0082] 3. Restore the shape of the entire surface of the object. The utilization rate of a single image is improved by extracting the center point and edge point of dark stripes and by using B-spline surface fitting, so as to restore the global surface shape of the object.

[0083] (1) Extract the center point and edge points of the dark stripes: Set the bright stripe threshold td1 (usually 200~255) and the dark stripe thresholds td2 and td3 (usually 70~110). This allows you to locate the area between adjacent bright and dark stripes during line scanning. The point where the absolute value of the difference between adjacent pixels in this area is the maximum value is considered the stripe edge point.

[0084] (2) B-spline surface fitting: The three-dimensional coordinates between adjacent stripes are calculated using a fourth-order quasi-uniform B-spline fitting method, thereby recovering the surface shape of the measured object and improving the utilization rate and data density of a single image. The main formula steps are as follows: B-Spline curve expression: , in, , , This indicates the curve corresponding to the parameter u The coordinates of the point, For each control point, For Bernstein polynomials; The value represents the parameter u Control point For the final curve point The contribution weights; the recursive definition of B-Spline given by De Boor and Cox is computationally stable, simple to represent, and universal, and can represent various forms of B-Spline. It is currently the most widely used, as shown in the following formula: , in, Indicates the index of the basis function, used to distinguish different basis functions; The order of the B-spline determines the smoothness of the curve and the range of its influence. Represents the first node in the node vector i Each node value; and These are two linear weights, whose function is to smoothly merge two low-order basis functions together, thereby producing high-order continuity. Represents the first i indivual k Rank B spline basis functions in parametersu The value at point A is the same as the values ​​elsewhere.

[0085] Based on the above formulas, let the parameter to be solved be... u The number of control points is num order is k Based on the properties of basis functions, this invention first obtains the current parameters through a binary search. u Given the node interval, two parameters, parm1 and parm2, are introduced according to the recursive rule to solve for the non-zero values ​​of the k basis functions; The subsequent process involves fitting a B-spline surface, where the striped structured light pattern includes... n The number of 3D points extracted and calculated from each striped structured light pattern during actual object measurement is: m The fitting process is as follows: Algorithm 2: Fitting Process for B-spline Surfaces Input: Partial 3D points obtained after scanning the object being measured; Output: A 4th-order B-spline surface representation of the object under test; Step 1. Fit the three-dimensional points on each stripe structured light pattern to a B-spline curve to obtain control points; Step 1.1. For each stripe structured light pattern, set its parameter values ​​according to the chord length method; Step 1.2. Based on the number of control points and the B-spline degree p, set up the node table U according to the following method: multiplication at both ends (multiplicity of 4) and uniform distribution in the middle; Step 1.3. Calculate according to the algorithm for solving basis functions; Step 1.4. Use the least squares method to obtain the control point matrix P, which gives the coordinates of a control point. Solve for other striped structured light patterns using the same method.

[0086] Step 2. Fit the j-th control point of the n B-spline curves obtained in Step 1 to a B-spline curve. The resulting control points form the control grid of the surface. The specific solution steps are the same as in Step 1. Step 3. Project the data points onto the surface generated in step 3, calculate the error. If the average error is greater than the user-specified tolerance, increase the number of control points and regenerate the surface until the average error is less than the tolerance.

[0087] Step 7: Optimize the image grayscale to obtain the 3D scan result of the measured object: Based on the results obtained from the previous steps, camera calibration, light plane calibration, and calculation of image stripe region information are performed to establish a 3D scanning model. This allows for the acquisition of 3D point cloud data of the PVC pipe surface. The same steps can be used to obtain 3D point cloud data of other objects, generating results such as... Figure 14 As shown.

[0088] In summary: 1. The ultra-dense bright stripes and wide bright-dark stripes proposed in this invention capture minute changes on the object surface in greater detail, thereby improving the accuracy and completeness of 3D reconstruction. This reduces the problem of missing 3D reconstruction data caused by insufficient stripe quantity in traditional stripe technology, and realizes the expansion from line scanning to surface scanning. For the widened bright stripe area, the information of the entire bright stripe area is obtained by estimating the grayscale of the illumination imaging, which effectively increases the data coverage. This solves the drawback of only being able to use lower stripe density and the complexity of calculation in dynamic objects, and improves the accuracy and robustness of structured light 3D scanning.

[0089] 2. The Gaussian first derivative method provided by this invention enables high-precision detection of the center of wide dark stripes and uniform stripe structured light patterns. This method can improve the positioning accuracy from the pixel level to the sub-pixel level, significantly improving the overall accuracy of the measurement system. 3. Based on radiance, a grayscale estimation method is proposed to calculate some information in the wide bright stripe region, which solves the problem of grayscale imbalance in the stripe structured light pattern region. At the same time, the information in the wide bright stripe region can effectively improve the data coverage, such as 1. 4. In step five, by optimizing measurement settings or model matching and repairing, the problem of excessively large values ​​in some three-dimensional coordinates due to the influence of angle is solved, thereby improving the accuracy of the overall data; 5. In step six, by estimating the occlusion of stripes, the center point and edge point of dark stripes are extracted to improve the utilization rate of a single image. The overall picture of the object is calculated by using a fourth-order quasi-uniform B-spline surface fitting method, thereby increasing the accuracy of the data.

[0090] Therefore, by selecting two stripe models, using grayscale estimation methods, and estimating stripe occlusion and fitting to obtain the full picture of the object, this invention overcomes the limitations of existing technologies and provides a more accurate, lower-cost, and more applicable 3D scanning solution.

[0091] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for three-dimensional surface scanning using ultra-density striped structured light, characterized in that, Includes the following steps: The camera was calibrated using Zhang's calibration method to obtain camera parameters; A striped structured light pattern is projected onto the surface of the object being tested by a light source projector. The striped structured light pattern is formed by alternating bright and dark stripes to cover the surface of the object. The striped structured light pattern is captured by a camera. The pixel coordinates of the center point of the striped structured light pattern are extracted from the image captured by the camera. A sub-pixel precision positioning method is adopted to calculate the zero intersection point of the striped structured light pattern by convolution response value, thereby improving the positioning accuracy from the pixel level to the sub-pixel level. The light plane of each stripe structured light pattern is calibrated using a calibration plate, and the light plane equation of each light plane in the camera coordinate system is obtained. The pixel coordinates of the center point of the striped structured light pattern are back-projected into a ray originating from the optical center of the camera, and the intersection of this ray with the corresponding light plane equation obtained by calibration is obtained to obtain the three-dimensional coordinates of the center point of the striped structured light pattern in the camera coordinate system. The three-dimensional coordinates of all the center points of the striped structured light pattern are collected to form the initial three-dimensional point cloud of the measured object. Based on radiance, grayscale values ​​are estimated for data points missing in the initial 3D point cloud due to signal saturation in bright stripe areas or difficulty in center localization, filling in the missing points and obtaining complete and accurate 3D surface shape data of the measured object.

2. The ultra-density striped structured light three-dimensional surface scanning method according to claim 1, characterized in that, The camera parameters include intrinsic parameters, extrinsic parameters, and distortion parameters; wherein: Intrinsic parameters are used to describe the camera's internal geometry and optical characteristics, including the camera's focal length and principal point coordinates; The extrinsic parameters are used to describe the camera's attitude and position in the world coordinate system, including rotation matrices and translation vectors. The rotation matrix represents the rotation transformation from the world coordinate system to the camera coordinate system, and the translation vector represents the offset of the origin of the world coordinate system relative to the origin of the camera coordinate system.

3. The ultra-density striped structured light three-dimensional surface scanning method according to claim 1, characterized in that, The bright and dark stripes in the striped structured light pattern are either super-density bright and dark stripes with the same pixel width, or wide bright and dark stripes with a pixel width ratio of 15:1 for bright and dark stripes; wherein the stripe density of the super-density bright and dark stripes is greater than 2 stripes / mm.

4. The ultra-density striped structured light three-dimensional surface scanning method according to claim 1, characterized in that, The zero-crossing point of the striped structured light pattern is determined according to the following formula: ， In the formula, where, j Integer pixel position indicating where the sign change occurs, i.e., the index of the convolution response value that is negative; R ( j) Indicates position j The convolution response value at that point is negative. R ( j+ 1) Indicates location j The convolution response value at +1 is positive. subPixelPos This indicates the calculated zero-crossing point position with sub-pixel precision.

5. The ultra-density striped structured light three-dimensional surface scanning method according to claim 2, characterized in that, The method for obtaining the equation of the light plane in the camera coordinate system includes the following steps: Using the camera's focal length, principal point coordinates, and distortion parameters, the pixel coordinates of the center point of the extracted striped structured light pattern are normalized to obtain the corrected pixel coordinates. The corrected pixel coordinates are then back-projected as rays originating from the camera's optical center; Based on the rotation matrix and translation vector, the plane equation of the calibration plate is transformed from the world coordinate system to the camera coordinate system, and the plane equation of the calibration plate in the camera coordinate system is obtained. Obtain the intersection point of the ray equation and the calibration plate plane equation to obtain the three-dimensional coordinates of the center point of the striped structured light pattern in the camera coordinate system; The three-dimensional coordinates of the center points of all striped structured light patterns are collected, and the least squares method is used to fit and form the light plane equation.

6. The ultra-density striped structured light three-dimensional surface scanning method according to claim 1, characterized in that, The radiance physical model decomposes the propagation process of light into two stages: the first stage is the reflection stage after light emitted from the light source hits the object's surface, and the second stage is the imaging stage after the reflected light from the object's surface enters the camera; the grayscale value estimation is determined according to the following formula: ， in, , In the formula, Grayscale value; The radiant power of the light source; The electro-optical conversion efficiency of the light source is constant when the light source is determined. The photoelectric conversion coefficient of a pixel; Image distance; when the camera position is determined, , as well as All are constant values; when the surface material of an object is determined, This is a physical property of the object's surface, and it is a constant value. The pixel on the object's surface corresponds to the center of the striped structured light pattern. This refers to a tiny area on the surface of the corresponding object. The pixel area of ​​the pixel on the imaging surface; The area of ​​a tiny region; The distance between a tiny area and the light source; The distance between the tiny region and the center point of the camera lens; The angle between a small region and the surface directly opposite the solid angle; The angle between the pixel plane of the pixel on the imaging surface and the face angle of the opposite solid angle; It is the angle between the tiny region and the pixel on the corresponding pixel plane of the imaging surface.

7. The ultra-density striped structured light three-dimensional surface scanning method according to claim 1, characterized in that, Also includes: For missing data in the initial 3D point cloud caused by occlusion or measurement error, the data is processed by optimizing the measurement settings or model matching repair; wherein, the optimized measurement settings include aligning the camera optical center and the ring structured light emission point emitted by the light source projector along the same straight line to reduce occlusion.

8. The ultra-density striped structured light three-dimensional surface scanning method according to claim 1, characterized in that, Before projecting the super-density fringe structured light pattern onto the surface of the object under test, the envelope space of the fringe structured light pattern on the object surface without occlusion is estimated: By using camera intrinsic parameters and light plane calibration, an unobstructed spatial region is determined to adjust the number or spacing of the striped structured light pattern and reduce occlusion; the envelope space is composed of the intersection line between the striped structured light pattern acquisition plane and the base plane, as well as the intersection line between the light planes.

9. The ultra-density striped structured light three-dimensional surface scanning method according to claim 1, characterized in that, It also includes determining the occlusion of the striped structured light pattern on the object surface, including the following steps: Get the pixel coordinate information of the object and store it in the array container V. Create a three-channel window and set the background image to gray. Draw the laser line in red based on the input laser position, number of stripes, and coverage area; Iterate through the pixels stored in container V, determine whether the pixel is on the laser line, and if so, connect the pixel to the camera location with a blue line. Repeatedly traverse the pixels in container V and check if the number of blue lines connecting the pixels to the camera is equal in both cases. If they are not equal, it is determined that a stripe occlusion phenomenon has occurred.

10. A high-density striped structured light three-dimensional surface scanning system, used to implement the method described in claims 1-9, characterized in that, include: Camera, used to capture images of striped structured light patterns; A light source projector is used to project ultra-density striped structured light patterns onto the surface of the object being measured. Processing unit, used to perform the following operations: The camera was calibrated using Zhang's calibration method. The light source projector projects a super-density striped structured light pattern, and the pixel coordinates of the center point of the striped structured light pattern are extracted from the image for sub-pixel precision positioning. The light plane containing each stripe structured light pattern is calibrated using a calibration plate, and the plane equation of each light plane in the camera coordinate system is obtained. The pixel coordinates are back-projected into rays, and the intersection points with the light plane equation are found to generate the initial 3D point cloud of the object being measured. The missing point cloud data was filled in using a radiance physical model to obtain complete 3D surface data.

Images