A method for three-dimensional pose and visual measurement of building components based on planar markings

Abstract

This invention relates to a method for three-dimensional pose and visual measurement of building components based on planar markers, including a floor, building components, several planar markers, and a binocular camera system; the patterns of the planar markers include calibration plate patterns and numbering patterns, enabling automatic data classification after the planar markers are captured; the process for capturing planar markers is defined; a theoretical world coordinate system is established based on coordinate system transformation, avoiding the problems of overall coordinate system translation deviation, three-axis rotation deviation, and error accumulation caused by traditional methods; the coordinate system transformation unifies all camera coordinate systems to the theoretical world coordinate system, enabling coordinate stitching of large-sized components by small-format cameras.

Description

Technical Field

[0001] This invention relates to a three-dimensional pose and visual measurement method for building components based on flat markings, belonging to binocular visual measurement methods and pose measurement of building components. Background Technology

[0002] The detection of dimensional deviations and deformations in building concrete components can be divided into detection items such as cross-sectional dimensions and deviations, tilt, deflection, and settlement. Traditional on-site position detection of building components mainly employs methods such as ruler measurement, leveling, theodolite measurement, prism-free total station measurement, plumb line measurement, and 3D laser scanning. Thus, current measurement methods can be broadly categorized into two types: one is purely manual ruler and plumb line measurement, which heavily relies on human observation, resulting in insufficient accuracy and reliability; the other is equipment-based methods, using instruments such as levels, theodolites, and total stations. These instruments perform point-to-point measurements, measuring only one point at a time, offering high accuracy but being time-consuming, labor-intensive, and dependent on manual operation. While 3D laser scanning is commonly used for irregularly shaped components, it is fast and accurate, but the equipment is expensive, and the amount of data collected is enormous, posing significant challenges to subsequent data processing and analysis.

[0003] Since the permissible dimensional deviation of building components is normally within ±5mm, machine vision measurement methods have emerged, especially those based on binocular camera systems for measuring the pose of building components. This measurement method has the following advantages: ① Low equipment cost: The complete cost of a binocular camera system is less than 4,000 yuan, lower than the cost of dedicated measuring equipment; ② Satisfactory measurement accuracy: For conventional building components on construction sites, the span is normally within 5m, and the measurement accuracy of the binocular camera measurement system within this distance range can reach within 1.5mm, achieving the same accuracy as a total station; ③ Automatic multi-point acquisition: By attaching specific identifiable patterns to the building components and setting up machine recognition programs, the three-dimensional coordinates of several points can be automatically acquired, resulting in high efficiency; ④ Convenient for secondary development: Automatic data acquisition and analysis processing can be achieved, and the required dimensional deviation of the building components can be directly calculated.

[0004] Patent application number 201910867867.X proposes a vision-based measurement system for large building components. However, it calculates the size of the building component by counting the number of pixels and the actual size corresponding to each pixel. This application has the following problems: ① It only calculates the size of the building component by counting the number of pixels, using a similarity-based method. This patent does not require measuring the row and column coordinates of the pixels or calculating the three-dimensional coordinates in the world coordinate system, thus failing to solve the problem of measuring three-dimensional spatial coordinates; ② It calculates the number of pixels and their corresponding world coordinate system dimensions to obtain the actual component size by analogy, achieving pixel-level accuracy (i.e., the unit of measurement is pixels). However, for visual measurement calculations of three-dimensional coordinates in space, pixel-level errors are very large (the error of a monocular camera is about 2 millimeters; if the three-dimensional coordinate measurement of a binocular camera is considered, the accuracy is at the centimeter level); ③ The patent does not consider the image stitching problem during camera movement, especially for the three-dimensional coordinate measurement of building components. The stitching of three-dimensional coordinates cannot be achieved through image stitching; it can only be achieved through the transformation from a local coordinate system to a global coordinate system.

[0005] Therefore, there is an urgent need to provide a new visual measurement method to solve the above problems. Summary of the Invention

[0006] This invention provides a method for three-dimensional pose and visual measurement of building components based on flat markings. It avoids the problems of overall coordinate system translation deviation, three-axis rotation deviation and error accumulation caused by traditional methods. By transforming the coordinate system, it realizes the unification of all camera coordinate systems to the theoretical world coordinate system, and can realize coordinate stitching of large-sized components by small-format cameras.

[0007] The technical solution adopted by this invention to solve its technical problem is:

[0008] A three-dimensional pose of a building component based on planar markings includes a floor, a building component, several planar markings, and a binocular camera system; the building component is vertically installed on the floor, and several planar markings are affixed to both the floor and the surface of the building component.

[0009] Several leveling marks are grouped together. Each group contains leveling marks that are pasted sequentially from the top of the building component to the bottom of the building component and then to the floor, starting from the building component and moving along the direction perpendicular to the floor. Only one leveling mark in each group is pasted on the floor. The leveling mark pasted on the building component is defined as the component leveling mark, and the leveling mark pasted on the floor is defined as the floor leveling mark.

[0010] The binocular camera system is installed on the floor and uses the line-of-sight intersection method to photograph building components or the floor, thereby acquiring the three-dimensional coordinate information of flat markings pasted on the building components or the floor. When the two cameras in the binocular camera system photograph the building components or the floor, the shooting range of the binocular camera system only includes two groups of flat markings that are vertically or horizontally adjacent, and the included angle between each camera and the corresponding object is equal.

[0011] As a further preferred embodiment of the present invention, the flat marking includes a single-sided frosted acrylic base plate and a UV-printed specific pattern, the pattern being arranged on the acrylic base plate.

[0012] The pattern includes a calibration plate pattern and a numbering pattern. The calibration plate pattern is a black right-angled triangle or a black marking circle, and a black border is provided on the outer edge of the black right-angled triangle or the black marking circle. The binocular camera system captures the center of the black right-angled triangle or the center of the black marking circle on the flat mark.

[0013] As a further preferred embodiment of the present invention, the calibration plate pattern is arranged in a grid pattern on the acrylic base plate, and the numbering pattern is arranged in the same vertical direction on the acrylic base plate.

[0014] The visual measurement method for the three-dimensional pose of building components based on planar markings specifically includes the following steps:

[0015] Step S1: Design of leveling marks; Determine the size and number of leveling marks based on the dimensions of the building components and the shooting range of the binocular camera system;

[0016] Step S2: Placement of leveling marks; Marker No. 1 in each group of leveling marks is set as the floor leveling mark and is pasted on the floor, while Marker No. 2 and subsequent marks in each group are set as component leveling marks and are pasted on the building components.

[0017] Step S3: Start the binocular camera system and take pictures in a fixed order; the binocular camera system only takes pictures of two flat marks that are adjacent horizontally or vertically upward each time; define the left, right, up, and down in the main view; first take pictures of the two flat marks that are adjacent vertically on the far left from bottom to top, and then take pictures of the flat marks that are adjacent vertically on the second left from left to right from bottom to top; after the flat marks in the first and second vertical directions have been taken, take pictures of the two flat marks that are adjacent horizontally upward, take pictures of the two flat marks that are adjacent vertically on the same horizontal direction from bottom to top, and then take pictures of the two flat marks that are adjacent vertically on the second and third vertical directions on the same horizontal direction from bottom to top, until all pictures are taken;

[0018] Step S4: By capturing images from the binocular camera system in step S3, collect the three-dimensional coordinate information of all flat marks, which are the coordinates of the center of the black right-angled triangle or the center of the black mark circle in the calibration plate pattern on the flat mark under the camera coordinate system.

[0019] Step S5: Establish the transformation between the theoretical world coordinate system and the coordinate system of the first shot; establish the theoretical world coordinate system during the design based on the building component establishment requirements; solve for the theoretical three-dimensional coordinates of all points on the floor surface markings at the far left and bottom of the building component under the theoretical world coordinate system, i.e., the coordinates under the camera coordinate system.

[0020] By capturing the first image in the leftmost vertical direction, the coordinates of the floor level mark in the camera coordinate system can be obtained; through coordinate system transformation, the translation vector and rotation matrix from the first camera coordinate system to the theoretical world coordinate system can be solved.

[0021] Step S6: Based on the coordinate system transformation, transform the coordinates of the second flat mark in the first image; based on the establishment of the theoretical world coordinate system and the coordinate system transformation of the first acquired image, solve for the theoretical world coordinate system coordinates of the second flat mark in the first image after the translation vector and rotation matrix of the first image have been solved; the second flat mark is the first component flat mark from bottom to top;

[0022] Step S7: Transform the coordinates of the second flat mark in the second shot; then, based on the camera coordinates of the first component flat mark from bottom to top in the second shot and the theoretical world coordinates obtained above, solve for the translation vector and rotation matrix of the camera coordinates to the theoretical world coordinates in the second captured image; repeat steps S6-S7 to convert all points within the flat marks to the theoretical world coordinates;

[0023] Step S8: Calculate the surface flatness of the building components by using the established theoretical world coordinate system and the coordinates of the black right triangles or black marked circles on all the flatness marks on the building components obtained in step S7.

[0024] Step S9: Using the camera coordinates of the black right triangle or black marked circle in the top row of the flat markings pasted on the top of the building component obtained in Step S7, draw a perpendicular line to the horizontal plane of the theoretical world coordinate system, and then calculate the distance difference between the foot of the perpendicular and the camera coordinates of the black right triangle or black marked circle in the bottom row of the flat markings pasted on the bottom of the building component. The vertical deviation of the building component can then be calculated.

[0025] As a further preferred embodiment of the present invention, in step S1, the calibration plate pattern is arranged in a grid pattern with at least 7 rows and 7 columns, and the ratio of the distance between the centers or circles of adjacent calibration plate patterns is less than or equal to 0.5.

[0026] The numbering pattern is constructed using polygons, and its quantity matches the vertical classification of the calibration plate pattern;

[0027] As a further preferred embodiment of the present invention, in step S2, each flat mark is attached to the floor or building component with a traceless double-sided adhesive tape with a thickness of less than or equal to 1 mm.

[0028] As a further preferred embodiment of the present invention, in step S3, when the binocular camera system takes pictures in a fixed order, there is a common flat mark in adjacent vertical images and adjacent horizontal images.

[0029] For adjacent groups of flat markings on adjacent surfaces of the same building component, both cameras of the binocular camera system take pictures at a 45° angle to the adjacent flat marking surfaces;

[0030] For the same set of bottom component leveling marks and floor leveling marks on building components, the two cameras of the binocular camera system take pictures of the two leveling marks at a 45° angle.

[0031] As a further preferred embodiment of the present invention, the specific method for establishing the transformation between the theoretical world coordinate system and the coordinate system of the first photograph in step S5 is to establish the translation vector and the rotation matrix using three translation parameters (x, y, z) and three rotation angles (α, β, γ) around the coordinate axes.

[0032] The coordinate system transformation uses a global optimization algorithm to solve for the six parameters in the translation vector and rotation matrix;

[0033] The error function of the global optimization algorithm is set to be equal to the difference in Euclidean space distance between the three-dimensional coordinates of each corresponding black right triangle or black marked circle on the first flat mark, after translation vector and rotation matrix, and the three-dimensional coordinates of the corresponding black right triangle or black marked circle on the second flat mark, and then the sum of the differences in Euclidean space distances of all points.

[0034] By employing the above technical solutions, the present invention has the following beneficial effects compared to the prior art:

[0035] 1. The three-dimensional pose of building components based on flat markings provided by the present invention includes a calibration plate pattern and a numbering pattern for the flat markings. Not only can the three-dimensional coordinates of points be measured through the calibration plate pattern, but the numbering pattern can also be used to number and classify the flat markings, so that the flat markings can be automatically classified after being photographed.

[0036] 2. The three-dimensional pose of building components based on flat markings provided by this invention determines the spacing of the calibration plate patterns, which can avoid recognition failures caused by excessively small spacing of the calibration plate patterns and continuous grayscale of the gaps; it determines the spacing of adjacent flat markings, which can minimize the use of flat markings; and it determines the shooting angle of different surfaces, which can improve the shooting calculation accuracy of the binocular camera system.

[0037] 3. The visual measurement method for the three-dimensional pose of building components based on flat markings provided by this invention proposes a global optimization method for coordinate system transformation of all points of one flat marking to all points of another flat marking, which can avoid the problems of overall coordinate system translation deviation, three-axis rotation deviation and error accumulation caused by traditional methods.

[0038] 4. The visual measurement method for the three-dimensional pose of building components based on flat markings provided by the present invention establishes a theoretical world coordinate system according to coordinate system transformation. It can realize the establishment of any world coordinate system determined during the design and can change the theoretical world coordinate system according to the usage requirements. It has the advantage of customizability in the determination of the world coordinate system.

[0039] 5. The visual measurement method for the three-dimensional pose of building components based on flat markings provided by this invention determines that, for different camera coordinate systems in different shooting sessions, the coordinate system transformation can unify all camera coordinate systems to the theoretical world coordinate system. This can solve the problem of small shooting area and large building component size in traditional binocular cameras. It can stitch the three-dimensional coordinates of any large building component into a theoretical world coordinate system through the coordinate system transformation method. Attached Figure Description

[0040] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0041] Figure 1 This is a schematic diagram of the three-dimensional pose of a building component based on flat markings provided by the present invention;

[0042] Figure 2 This is a schematic diagram of the structure of the flat marking provided by the present invention;

[0043] Figure 3 This is a schematic diagram of the pattern provided by the present invention;

[0044] Figure 4 This is a schematic diagram illustrating the transformation between the camera coordinate system and the theoretical world coordinate system provided by this invention.

[0045] In the diagram: 1 is the floor surface, 2 is the building component, 3 is the leveling mark, 301 is the floor surface leveling mark, 302 is the component leveling mark, 4 is the binocular camera system, 5 is the acrylic base plate, 6 is the pattern, 601 is the calibration plate pattern, and 602 is the numbering pattern.

[0046] Figure 4 In this context, the CCS coordinate system refers to the camera coordinate system, while the WCS coordinate system refers to the theoretical world coordinate system. Detailed Implementation

[0047] The present invention will now be described in further detail with reference to the accompanying drawings. In the description of this application, it should be understood that the terms "left side," "right side," "upper part," "lower part," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. "First," "second," etc., do not indicate the importance of the components, and therefore should not be construed as a limitation of the present invention. The specific dimensions used in this embodiment are only for illustrating the technical solution and do not limit the scope of protection of the present invention.

[0048] In existing technologies, the application of binocular camera systems to the pose measurement of building components on construction sites still presents several challenges: ① Binocular camera systems can only capture the field of view through the lens, failing to achieve 360° coverage like a total station. Therefore, the surface of the building component needs to be divided into several local blocks, with the binocular camera system capturing only one block at a time, followed by coordinate system transformation for stitching. ② The pattern of the flat markings on the building component needs design, including not only measurable markers but also numbering of each flat marking to automatically identify its assigned number. ③ The spacing of the flat markings on the building component needs to be determined. ④ A separate theoretical world coordinate system needs to be established. ⑤ The coordinate system transformation method needs to be defined. ⑥ The overall operational steps need to be determined.

[0049] To better apply binocular camera systems in the field of building component pose measurement, this application proposes a three-dimensional pose measurement method for building components based on planar markers, including a floor 1, a building component 2, several planar markers 3, and a binocular camera system; the building component is vertically installed on the floor, and several planar markers are pasted on both the floor and the surface of the building component; the planar markers are grouped, and each group contains planar markers pasted sequentially from the top of the building component, the bottom of the building component, to the floor along the direction perpendicular to the floor, starting from the building component, and only one planar marker in each group is pasted on the floor. The planar marker pasted on the building component is defined as the component planar marker, and the planar marker pasted on the floor is defined as the floor planar marker 301;

[0050] The binocular camera system is installed on the floor and uses the line-of-sight intersection method to photograph building components or the floor, thereby acquiring the three-dimensional coordinate information of flat markings pasted on the building components or the floor. When the two cameras in the binocular camera system photograph the building components or the floor, the shooting range of the binocular camera system only includes two groups of flat markings that are adjacent in the vertical or horizontal direction, and the included angle between each camera and the corresponding object is equal.

[0051] Regarding flatness markings, such as Figure 2 As shown, the flat marking includes a single-sided frosted acrylic base plate 5 and a UV-printed specific pattern 6, with the pattern laid on the acrylic base plate. Figure 3 As shown, the pattern includes a calibration plate pattern 601 and a numbering pattern 602, enabling automatic data classification of the flat markings after shooting. On the acrylic base plate, the calibration plate pattern is arranged in a grid pattern, and the numbering patterns are arranged in the same vertical direction on the acrylic base plate. The calibration plate pattern is a black right-angled triangle or a black marking circle, and a black border is provided on the outer edge of the black right-angled triangle or black marking circle. The numbering pattern consists of a certain number of shapes with obvious graphic edge features. The number indicates the numbering of the flat markings pasted on the building component from bottom to top. The binocular camera system 4 shoots the center of the black right-angled triangle or the center of the black marking circle on the flat marking.

[0052] Here, we need to explain the reasons behind the design of the leveling marks. The black right-angled triangles or black marking circles in the leveling marks are defined as a grid composed of several rows and columns, used to define the angle deviation function during coordinate system transformation. Since the error function during coordinate system transformation can be defined as the angle deviation of all rows and columns, it can avoid the overall rotation deviation that occurs during coordinate system transformation, effectively assisting in the realization of high-precision coordinate system transformation. The numbering pattern of the leveling marks must at least express the following information: which column of the building component the leveling mark is located in and at what height within that column. Therefore, it is necessary to specify at least the height number and the column number information, and thus it is defined as a certain number of shapes with obvious graphic edge features.

[0053] The spacing between adjacent flat markings should ensure that both adjacent flat markings are within the field of view captured by the binocular camera system, and that the same field of view should only include two adjacent flat markings; that is, when the binocular camera system captures images in a fixed order, the same flat marking is included in adjacent vertical images and adjacent horizontal images.

[0054] Simultaneously, the binocular camera system should capture images at approximately the same angle as the flat markings on adjacent surfaces. For the same set of bottom component flat markings (302) and floor flat markings on building components, the binocular camera system should capture images at approximately the same angle as both markings. This shooting angle specification is because, for adjacent surfaces, and for bottom component flat markings and floor flat markings, the binocular camera system should capture images at approximately the same angle as the adjacent flat markings to ensure the accuracy of visual recognition of the center of the black right-angled triangle or the center of the black marking circle. For example, if the two column surfaces are perpendicular to each other (left side and front side), the binocular camera should capture images at an angle of 135° to each of the two flat markings. If the angle with a flat marking on a surface is too large or too small, the center of the black right-angled triangle or the center of the black marking circle will become an ellipse, making it difficult to identify the center.

[0055] This application also provides a visual measurement method for the three-dimensional pose of building components based on planar markings, specifically including the following steps:

[0056] Step S1: Design of flat markings; Based on the dimensions of the building components and the shooting range of the binocular camera system, determine the size and number of flat markings; Among them, when the calibration plate pattern is arranged in a grid, it should be at least 7 rows and 7 columns, and the distance ratio between the centers or circles of adjacent calibration plate patterns should be less than or equal to 0.5; The numbering pattern adopts a polygonal structure, and its number matches the vertical classification of the calibration plate pattern.

[0057] Step S2: Placement of leveling marks; Marker No. 1 in each group of leveling marks is set as a floor leveling mark and is pasted on the floor, while Marker No. 2 and subsequent marks in each group are set as component leveling marks and are pasted on building components; Each leveling mark is pasted to the floor or building component with traceless double-sided tape with a thickness of less than or equal to 1mm.

[0058] Step S3: Start the binocular camera system and take pictures in a fixed order; the binocular camera system only takes pictures of two flat marks that are adjacent horizontally or vertically upward each time; define the left, right, up, and down in the main view; first take pictures of the two flat marks that are adjacent vertically on the far left from bottom to top, and then take pictures of the flat marks that are adjacent vertically on the second left from left to right from bottom to top; after the flat marks in the first and second vertical directions have been taken, take pictures of the two flat marks that are adjacent horizontally upward, take pictures of the two flat marks that are adjacent vertically on the same horizontal direction from bottom to top, and then take pictures of the two flat marks that are adjacent vertically on the second and third vertical directions on the same horizontal direction from bottom to top, until all pictures are taken;

[0059] For adjacent groups of flat markings on adjacent surfaces of the same building component, both cameras of the binocular camera system take pictures at a 45° angle to the adjacent flat marking surfaces;

[0060] For the same set of component leveling marks and floor leveling marks on the bottom of building components, the two cameras of the binocular camera system take pictures at a 45° angle to both leveling marks.

[0061] Step S4: By capturing images from the binocular camera system in step S3, collect the three-dimensional coordinate information of all flat marks, which are the coordinates of the center of the black right-angled triangle or the center of the black mark circle in the calibration plate pattern on the flat mark under the camera coordinate system.

[0062] Step S5: Establish the transformation between the theoretical world coordinate system and the coordinate system of the first shot; establish the theoretical world coordinate system during the design based on the building component establishment requirements; solve for the theoretical three-dimensional coordinates of all points on the floor surface markings at the far left and bottom of the building component under the theoretical world coordinate system, i.e., the coordinates under the camera coordinate system.

[0063] By capturing the first image in the leftmost vertical direction, the coordinates of the floor level mark in the camera coordinate system can be obtained; through coordinate system transformation, the translation vector and rotation matrix from the first camera coordinate system to the theoretical world coordinate system can be solved.

[0064] The specific method for establishing the transformation between the theoretical world coordinate system and the coordinate system of the first photograph is as follows: using three translation parameters (x, y, z) and three rotation angles (α, β, γ) around the coordinate axes, a translation vector and a rotation matrix are established. The coordinate system transformation uses a global optimization algorithm to solve for the six parameters in the translation vector and rotation matrix. The error function of the global optimization algorithm is set to be equal to the difference in Euclidean space distance between the three-dimensional coordinates of each corresponding black right-angled triangle or black marked circle on the first flat mark, after translation and rotation, and the corresponding three-dimensional coordinates of the black right-angled triangle or black marked circle on the second flat mark. Then, the differences in Euclidean space distances of all points are summed.

[0065] Step S6: Based on the coordinate system transformation, transform the coordinates of the second flat mark in the first shot; based on the establishment of the theoretical world coordinate system and the coordinate system transformation of the first acquired image, solve the theoretical world coordinate system coordinates of the second flat mark in the first image after the translation vector and rotation matrix of the first image have been solved; the second flat mark is the first component flat mark from bottom to top.

[0066] Step S7: Transform the coordinates of the second flat mark in the second shot; then, based on the camera coordinates of the first flat mark from bottom to top in the second shot and the theoretical world coordinates obtained above, solve for the translation vector and rotation matrix of the camera coordinates to the theoretical world coordinates in the second captured image; repeat steps S6-S7 to convert all points within the flat marks to the theoretical world coordinates.

[0067] Step S8: Calculate the surface flatness of the building components by using the established theoretical world coordinate system and the coordinates of the black right-angled triangles or black marked circles on all the flatness marks on the building components obtained in step S7.

[0068] Step S9: Using the camera coordinates of the black right-angled triangle or black marked circle in the top row of the flat markings pasted on the top of the building component obtained in Step S7, draw a perpendicular line to the horizontal plane of the theoretical world coordinate system. Then calculate the distance difference between the foot of the perpendicular and the camera coordinates of the black right-angled triangle or black marked circle in the bottom row of the flat markings pasted on the bottom of the building component. The vertical deviation of the building component can then be calculated.

[0069] In summary, this application establishes a theoretical world coordinate system based on coordinate system transformation, which avoids the problems of overall coordinate system translation deviation, three-axis rotation deviation, and error accumulation caused by traditional methods. It can also establish any world coordinate system determined during the design phase. This application also demonstrates that coordinate system transformation can unify all camera coordinate systems to the theoretical world coordinate system, enabling coordinate stitching of large-sized components by small-format cameras.

[0070] Those skilled in the art will understand that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. It should also be understood that terms such as those defined in general dictionaries should be understood to have the meaning consistent with their meaning in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless defined as herein.

[0071] The meaning of "and / or" as used in this application includes situations where each exists alone or both exist simultaneously.

[0072] The term "connection" as used in this application can mean a direct connection between components or an indirect connection between components through other components.

[0073] Based on the above-described preferred embodiments of the present invention, and through the foregoing description, those skilled in the art can make various changes and modifications without departing from the inventive concept. The technical scope of this invention is not limited to the contents of the specification, but must be determined according to the scope of the claims.

Claims

1. A visual measurement method for the three-dimensional pose of a building component based on planar markers, wherein the three-dimensional pose of the building component based on planar markers includes a floor surface, a building component, several planar markers, and a binocular camera system; The building components are installed vertically on the floor, and several flat markings are affixed to both the floor and the surface of the building components. Several leveling marks are grouped together. Each group contains leveling marks that are pasted sequentially from the top of the building component to the bottom of the building component and then to the floor, starting from the building component and moving along the direction perpendicular to the floor. Only one leveling mark in each group is pasted on the floor. The leveling mark pasted on the building component is defined as the component leveling mark, and the leveling mark pasted on the floor is defined as the floor leveling mark. The binocular camera system is installed on the floor and uses the line-of-sight intersection method to photograph building components or the floor, thereby acquiring the three-dimensional coordinate information of flat markings pasted on the building components or the floor. When the two cameras in the binocular camera system photograph the building components or the floor, the shooting range of the binocular camera system only includes two groups of flat markings that are vertically or horizontally adjacent, and the included angle between each camera and the corresponding object is equal. The flat markings consist of a single-sided frosted acrylic base and a UV-printed pattern, which is placed on the acrylic base. The pattern includes a calibration plate pattern and a numbering pattern. The calibration plate pattern is a black right-angled triangle or a black marking circle, and a black border is provided on the outer edge of the black right-angled triangle or the black marking circle. The binocular camera system captures the center of the black right-angled triangle or the center of the black marking circle on the flat mark. On the acrylic base plate, the calibration plate pattern is arranged in a grid pattern, and the numbering pattern is arranged in the same vertical direction on the acrylic base plate. The visual measurement method is characterized by the following steps: Step S1: Design of leveling marks; Determine the size and number of leveling marks based on the dimensions of the building components and the shooting range of the binocular camera system; Step S2: Placement of leveling marks; Marker No. 1 in each group of leveling marks is set as the floor leveling mark and is pasted on the floor, while Marker No. 2 and subsequent marks in each group are set as component leveling marks and are pasted on the building components. Step S3: Start the binocular camera system and take pictures in a fixed order; the binocular camera system only takes pictures of two flat marks that are adjacent horizontally or vertically upward each time; define the left, right, up, and down in the main view; first take pictures of the two flat marks that are adjacent vertically on the far left from bottom to top, and then take pictures of the flat marks that are adjacent vertically on the second left from left to right from bottom to top; after the flat marks in the first and second vertical directions have been taken, take pictures of the two flat marks that are adjacent horizontally upward, take pictures of the two flat marks that are adjacent vertically on the same horizontal direction from bottom to top, and then take pictures of the two flat marks that are adjacent vertically on the second and third vertical directions on the same horizontal direction from bottom to top, until all pictures are taken; Step S4: By capturing images from the binocular camera system in step S3, collect the three-dimensional coordinate information of all flat marks, which are the coordinates of the center of the black right-angled triangle or the center of the black mark circle in the calibration plate pattern on the flat mark under the camera coordinate system. Step S5: Establish the transformation between the theoretical world coordinate system and the coordinate system of the first shot; establish the theoretical world coordinate system during the design based on the building component establishment requirements; solve for the theoretical three-dimensional coordinates of all points on the floor surface markings at the far left and bottom of the building component under the theoretical world coordinate system, i.e., the coordinates under the camera coordinate system. The coordinates of the floor level mark in the camera coordinate system are obtained from the first captured image in the leftmost vertical direction; by transforming the coordinate system, the translation vector and rotation matrix from the first camera coordinate system to the theoretical world coordinate system can be solved. Step S6: Based on the coordinate system transformation, transform the coordinates of the second flat mark in the first image; based on the establishment of the theoretical world coordinate system and the coordinate system transformation of the first acquired image, solve for the theoretical world coordinate system coordinates of the second flat mark in the first image after the translation vector and rotation matrix of the first image have been solved; the second flat mark is the first component flat mark from bottom to top; Step S7: Transform the coordinates of the second flat mark in the second shot; then, based on the camera coordinates of the first component flat mark from bottom to top in the second shot and the theoretical world coordinates obtained above, solve for the translation vector and rotation matrix of the camera coordinates to the theoretical world coordinates in the second captured image; repeat steps S6-S7 to convert all points within the flat marks to the theoretical world coordinates; Step S8: Calculate the surface flatness of the building components by using the established theoretical world coordinate system and the coordinates of the black right triangles or black marked circles on all the flatness marks on the building components obtained in step S7. Step S9: Using the camera coordinates of the black right-angled triangle or black marked circle in the top row of the flat markings pasted on the top of the building component obtained in Step S7, draw a perpendicular line to the horizontal plane of the theoretical world coordinate system. Then calculate the distance difference between the foot of the perpendicular and the camera coordinates of the black right-angled triangle or black marked circle in the bottom row of the flat markings pasted on the bottom of the building component. The vertical deviation of the building component can then be calculated.

2. The method according to claim 1, characterized in that: In step S2, each flat mark is attached to the floor or building component with a traceless double-sided adhesive tape with a thickness of less than or equal to 1 mm.

3. The method according to claim 1, characterized in that: In step S3, when the binocular camera system takes pictures in a fixed order, there is a common flat mark in adjacent vertical images and adjacent horizontal images. For adjacent groups of flat markings on adjacent surfaces of the same building component, both cameras of the binocular camera system take pictures at a 45° angle to the adjacent flat marking surfaces; For the same set of component leveling marks and floor leveling marks on the bottom of building components, the two cameras of the binocular camera system take pictures at a 45° angle to both leveling marks.

4. The method according to claim 1, characterized in that: The specific method for establishing the transformation between the theoretical world coordinate system and the coordinate system of the first shot in step S5 is to establish the translation vector and rotation matrix using three translation parameters (x, y, z) and three rotation angles (α, β, γ) around the coordinate axes. The coordinate system transformation uses a global optimization algorithm to solve for the six parameters in the translation vector and rotation matrix; The error function of the global optimization algorithm is set to be equal to the difference in Euclidean space distance between the three-dimensional coordinates of each corresponding black right triangle or black marked circle on the first flat mark, after translation vector and rotation matrix, and the three-dimensional coordinates of the corresponding black right triangle or black marked circle on the second flat mark, and then the sum of the differences in Euclidean space distances of all points.

Images