CN109373912B - Binocular vision-based non-contact six-degree-of-freedom displacement measurement method - Google Patents

Binocular vision-based non-contact six-degree-of-freedom displacement measurement method Download PDF

Info

Publication number
CN109373912B
CN109373912B CN201811567836.4A CN201811567836A CN109373912B CN 109373912 B CN109373912 B CN 109373912B CN 201811567836 A CN201811567836 A CN 201811567836A CN 109373912 B CN109373912 B CN 109373912B
Authority
CN
China
Prior art keywords
dimensional
camera
cylinder
coordinate
coordinates
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201811567836.4A
Other languages
Chinese (zh)
Other versions
CN109373912A (en
Inventor
陈志聪
吴丽君
苏忆艳
陈疏影
程树英
徐森
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Fuzhou University
Original Assignee
Fuzhou University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Fuzhou University filed Critical Fuzhou University
Priority to CN201811567836.4A priority Critical patent/CN109373912B/en
Publication of CN109373912A publication Critical patent/CN109373912A/en
Application granted granted Critical
Publication of CN109373912B publication Critical patent/CN109373912B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01BMEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
    • G01B11/00Measuring arrangements characterised by the use of optical means
    • G01B11/02Measuring arrangements characterised by the use of optical means for measuring length, width or thickness
    • G01B11/03Measuring arrangements characterised by the use of optical means for measuring length, width or thickness by measuring coordinates of points
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01BMEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
    • G01B11/00Measuring arrangements characterised by the use of optical means
    • G01B11/26Measuring arrangements characterised by the use of optical means for measuring angles or tapers; for testing the alignment of axes

Abstract

The invention relates to a binocular vision-based non-contact six-degree-of-freedom displacement measurement method. Firstly, designing a two-dimensional coding sequence and generating a chessboard code, and fixing the chessboard code on the surface of an object to be detected to obtain a cylinder to be detected; then, calibrating the internal and external parameters of the binocular camera; then, dividing the targets of two groups of images obtained by shooting by the left camera and the right camera respectively, and extracting characteristic angular points of the targets; then, converting the two-dimensional image coordinates of the characteristic corner points into three-dimensional world coordinates based on the internal and external parameter matrixes and the imaging model of the camera; and finally, performing cylinder fitting on the three-dimensional coordinate points of the characteristic angular point group, calculating the translation and rotation information of the cylinder, and finally calculating to obtain the displacement information with six degrees of freedom. According to the invention, the information of six degrees of freedom such as translation, rotation and the like can be calculated by using the information of the characteristic points of the target object in two groups of input images shot by the left camera and the right camera without contacting the target.

Description

Binocular vision-based non-contact six-degree-of-freedom displacement measurement method
Technical Field
The invention relates to a binocular vision-based non-contact six-degree-of-freedom displacement measurement method.
Background
In the conventional displacement measurement technology architecture, a multi-degree-of-freedom detection platform or a coordinate measuring machine is generally used for measurement. Coordinate measuring machines are the main common instruments for measuring the contour, form and position dimensions, etc. of a workpiece. The application scenarios of the degree-of-freedom detection platform are limited due to the fact that the degree-of-freedom detection platform is complex in structure, large in system, high in price and required to be in direct contact with a detected object. Therefore, it is highly desirable to develop a high-performance and cost-effective non-contact displacement measurement method to replace the multi-degree-of-freedom detection platform, so as to be conveniently applicable in various scenes. The displacement measuring method based on vision is a non-contact measuring method established on the digital image processing technology, can replace the traditional driving and positioning system consisting of a driving mechanism and a grating ruler on a working reference surface, and has the characteristics of flexibility and high cost performance.
Disclosure of Invention
In view of this, the present invention aims to provide a binocular vision-based non-contact six-degree-of-freedom displacement measurement method, in which six-degree-of-freedom displacement information such as displacement and rotation angle can be calculated by attaching a two-dimensional coding cylinder to the surface of an object to be measured and using feature point information of a target object in two sets of input images captured by left and right cameras.
In order to achieve the purpose, the invention adopts the following technical scheme:
a binocular vision-based non-contact six-degree-of-freedom displacement measurement method comprises the following steps:
step S1: designing a two-dimensional coding sequence and generating a chessboard code, and fixing the chessboard code on the surface of an object to be detected to obtain a cylinder to be detected;
step S2: installing a binocular camera and calibrating internal and external parameters of the binocular camera;
step S3: acquiring a cylinder image to be detected through a binocular camera, then respectively carrying out cylinder target segmentation on the image, and extracting two-dimensional image coordinates of a characteristic angular point in a target on the basis of segmentation;
step S4: converting two-dimensional image coordinates of the characteristic corner points into three-dimensional world coordinates based on an internal and external parameter matrix of the binocular camera and a camera imaging model;
step S5: and performing cylinder fitting on the three-dimensional world coordinates of the characteristic angular points, and calculating the center point and the direction vector of the cylinder to obtain translation and rotation information of the object to be detected, namely six-degree-of-freedom displacement information.
Further, the step S3 is specifically:
step S31: obtaining im by carrying out binocular image acquisition on cylinder to be measuredl1,imr1,iml2,imr2
Step S32: automatically segmenting a target x according to images acquired by a binocular cameral1、xr1And xl2、xr2
Step S33: for the divided object xl1、xr1And xl2、xr2Extracting characteristic angular points to obtain image two-dimensional coordinates x of each characteristic angular pointl11,…,xl1n、xr11,…,xr1nAnd xl21,…,xl2n、xr21,…,xr2nAnd a two-dimensional coordinate set X ofl1、Xr1And Xl2、Xr2
Step S34: for image two-dimensional coordinate set Xl1,Xr1And Xl2,Xr2Optimizing the checkerboard constraint to generate a final two-dimensional coordinate set Yl1、Yr1And Yl2、Yr2
Step S35: restoring binary coded information B by utilizing gray threshold1、B2
Further, the step S4 is specifically:
step S41: obtaining binocular camera internal parameter M by binocular calibrationleft、MrightAn external reference matrix H;
step S42: according to the obtained internal and external parameter matrix of the camera, a two-dimensional coordinate set Y is utilizedl1、Yr1And Yl2、Yr2Reconstructing a three-dimensional world coordinate set G1And G2
Step S43: for discrete three-dimensional world coordinate set G1And G2Optimizing the linear constraint to generate a coordinate set group1And group2
Further, the step S5 is specifically:
step S51: according to the obtained binary code B1And B2Judging the coordinate set group1And group2Set of coordinates g of common feature corners present in1And g2
Step S52: using coordinate set group1And group2Respectively performing cylinder fitting to obtain central coordinates C of the two1、C2And a radius R1、R2And a direction vector V1,V2
Step S53: using centre coordinates C1、C2The difference calculation is carried out to obtain the displacement information, and meanwhile, the direction vector V is utilized1,V2And the Euler angle of rotation can be calculated according to the Euler angle formula of the rotation matrix.
Compared with the prior art, the invention has the following beneficial effects:
when the six-degree-of-freedom displacement information is processed, the information of the degree of freedom such as displacement, a rotation angle and the like can be calculated by only utilizing the information of the characteristic points of the target object in two groups of input images shot by the binocular camera, and the measuring device equipment does not need to be in direct contact with the target object and can be applied under the condition of rotary motion even if shielding occurs. The method is easy to operate and realize, and can meet different requirements through less parameter adjustment.
Drawings
FIG. 1 is a general flow diagram of the present invention.
Fig. 2 is a schematic view of a six-degree-of-freedom displacement measurement according to an embodiment of the present invention.
FIG. 3 is a schematic diagram of a binary coded cylinder according to an embodiment of the present invention.
FIG. 4 is a schematic diagram of a target segmentation process according to an embodiment of the present invention.
Fig. 5 is a schematic view of a corner point detection process according to an embodiment of the present invention.
Fig. 6 is a relationship diagram of four coordinate systems of a camera coordinate system, an image coordinate system, a pixel coordinate system and a world coordinate system.
FIG. 7 is a reconstructed target cylinder with feature points according to an embodiment of the invention.
Detailed Description
The invention is further explained below with reference to the drawings and the embodiments.
Referring to fig. 1 and 2, the present invention provides a binocular vision-based non-contact six-degree-of-freedom displacement measurement method, and fig. 1 and 2 show an implementation example of the present invention, which includes the following steps:
step S1: designing a two-dimensional coding sequence and generating a chessboard code, and fixing the chessboard code on the surface of an object to be detected as a chessboard coding cylinder. Attaching a non-repeated binary code to the surface of the cylinder to code the position information representing the surface of the cylinder and obtain a cylinder C;
in order to represent the surface position of the rolling object by using specific available information, a coding cylinder with obvious characteristic corner points is designed, and the coding cylinder uses non-repeated binary coding on the surface of a cylinder to represent the position information of the cylinder surface. In addition to the binary coding of the present invention, other binary coding may be substituted. As shown in fig. 3, a white lattice represents 0, a black lattice represents 1, one code may correspond to a row of black and white lattices, and there may be obvious corner points between black and white lattices formed by two adjacent codes, and we take these points as characteristic corner points of a target. Due to the non-repeatability of the binary code, a certain code can be used for representing a certain specific position of the surface of the cylinder. During rotation and movement, the rotation and displacement information of the cylinder can be obtained according to the change of the common characteristic point on the target object in the front group of images and the rear group of images captured by the binocular camera.
Step S2: calibrating the internal and external parameters of the binocular camera;
step S3: acquiring images of the cylinder according to the calibrated binocular camera, then respectively segmenting the targets of two groups of images of the cylinder obtained by shooting through the binocular camera, and extracting two-dimensional image coordinates of characteristic corner points of the targets;
step S31: obtaining im by carrying out binocular image acquisition on the cylinder Cl1,imr1,iml2,imr2
Step S32: automatically segmenting a target x according to images acquired by a binocular cameral1、xr1And xl2、xr2
Since the target to be detected is a cylinder with distinct black and white grid characteristics and its gray scale difference is significant, the target segmentation can be performed using (but not limited to) a conventional threshold-based segmentation method. After the gray level image is converted into the binary image through threshold judgment, the binary image needs to be corroded and expanded, the corrosion can delete some pixels of the target boundary to remove small connected regions, and the expansion can add pixels to the target boundary in the image to remove holes. Then, the position of the cylinder in the picture can be determined and divided by searching the connected region and the shape characteristics of the cylinder, as shown in fig. 4.
Step S33: for the divided object xl1、xr1And xl2、xr2Extracting characteristic angular points to obtain image two-dimensional coordinates x of each characteristic angular pointl11,…,xl1n、xr11,…,xr1nAnd xl21,…,xl2n、xr21,…,xr2nAnd a two-dimensional coordinate set X ofl1、Xr1And Xl2、Xr2
Step S34: for image two-dimensional coordinate set Xl1,Xr1And Xl2,Xr2Optimizing the checkerboard constraint to generate a final two-dimensional coordinate set Yl1、Yr1And Yl2、Yr2
The corner positions extracted in step S35 still have slight errors due to the influence of scrolling and the like, and for such errors, we first adopt corner fusion and screening to fuse the corners whose adjacent distances are smaller than the grid edge distance 1/2, select points with obvious features, and delete the redundant corners which are the black edges nearest to the grid. Then, according to the characteristic that the feature points of the elliptic section and each row and each column are on the same straight line in the cylinder rolling process, an arc-shaped checkerboard is fitted by using the fused and screened discrete points, as shown in fig. 5. On the basis, the angular points of the arc-shaped checkerboard are set as final accurate angular points.
Step S36: restoring binary coded information B by utilizing gray threshold1、B2
For two-dimensional coordinate set Yl1And Yl2Judging whether black and white grids surrounding the point at angles of 45 and 135 degrees are known by using a gray threshold method, and restoring binary coded information B on the basis of the black and white grids1、B2The reverse process is shown in fig. 2.
Step S4: converting two-dimensional image coordinates of the characteristic corner points into three-dimensional world coordinates based on an internal and external parameter matrix and an imaging model of the binocular camera;
step S41: obtaining binocular camera internal parameter M by binocular calibrationleft、MrightAn external reference matrix H;
the MATLAB carries out binocular calibration on calibration plate images acquired by the left camera and the right camera by utilizing a self-contained binocular calibration tool box, and respective internal parameters M of the left camera and the right camera can be obtainedleft、MrightAnd the external parameter H between the two cameras is composed of a rotation matrix and a displacement matrix.
Step S42: according to the obtained internal and external parameter matrix of the camera, a two-dimensional coordinate set Y is utilizedl1、Yr1And Yl2、Yr2Reconstructing a three-dimensional world coordinate set G1And G2
The conversion process from two-dimensional image coordinates to three-dimensional image coordinates can be known by using the relationship among the four coordinate systems of the camera coordinate system, the image coordinate system, the pixel coordinate system, and the world coordinate system in fig. 6. And calculating the two-dimensional coordinates of the images passing through the characteristic corner points of the left camera and the right camera and an equation of two space straight lines of the corresponding camera optical center, wherein the coordinates of the intersection point of the two straight lines are the three-dimensional world coordinates of the characteristic points. Because various errors may exist in actual calculation and two straight lines are not intersected possibly, the discrete three-dimensional world coordinate set G can be solved and reconstructed by using the least square method1And G2
Step S43: for discrete three-dimensional world coordinate set G1And G2Optimizing the linear constraint to generate a coordinate set group1And group2
Step S5: and performing cylinder fitting on the three-dimensional world coordinates of the characteristic angular points, and calculating the translation and rotation information of the cylinder to obtain the displacement information with six degrees of freedom.
Step S51: according to the obtained binary code B1And B2Judging the coordinate set group1And group2Set of coordinates g of common feature corners present in1And g2
The binary coded information B is restored in step S261And B2Then, the current output can be knownThe cylindrical object in the image appears as a surface location relative to the camera. After two groups of input images shot by a binocular camera are respectively subjected to characteristic corner detection and binary coding restoration, if the binary coding B of the two groups of images can be judged1And B2After at least one identical code is included in the two sets of coordinates, the feature corner points corresponding to the identical code can be regarded as a common feature corner point, or two sets of coordinates can be grouped1And group2Finding the coordinate set g of the common feature corner points corresponding to the two groups of images1And g2
Step S52: using coordinate set group1And group2Respectively performing cylinder fitting to obtain central coordinates C of the two1、C2Radius R1、R2And a direction vector V1,V2
Step S53: using centre coordinates C1、C2The difference calculation can obtain the displacement information, and meanwhile, the displacement information is based on the direction vector V1,V2And calculating the Euler angle of rotation according to the Euler angle formula of the rotation matrix.
Example 1:
this embodiment contemplates a non-repeating binary-coded cylinder as shown in fig. 3, having a radius of about 3.4140cm, a height of 15cm, and a surface covered by 8 binary codes. The acquisition hardware used a CCD camera model ALLIED Gige GT1910C with a resolution of 1920 × 1080 and a maximum frame rate of 57fps, and a lens model computer FA M2518 with 500 million pixels with a focal length F of 25mm and an aperture F of 1.8. The four-degree-of-freedom measurement result of the invention is compared with the actual rotation angle and displacement. Table 1 shows the displacement measurement method of the present invention in comparison to displacement at a known actual rotation angle.
Table 1:
as can be seen from Table 1, the method of the present invention can measure 4-degree-of-freedom information consisting of displacements in three directions and rotation angles in a specific direction. Therefore, the method provided by the invention can calculate the information of the degrees of freedom such as displacement, rotation angle and the like only by utilizing the information of the characteristic points of the target object in two groups of input images shot by the left camera and the right camera, is also applicable to the condition that the rotation angle is blocked, and can be popularized to the displacement measurement of six degrees of freedom including three-dimensional translation and three-dimensional rotation.
The above description is only a preferred embodiment of the present invention, and all equivalent changes and modifications made in accordance with the claims of the present invention should be covered by the present invention.

Claims (2)

1. A non-contact six-degree-of-freedom displacement measurement method based on binocular vision is characterized by comprising the following steps:
step S1: designing a two-dimensional coding sequence and generating a chessboard code, and fixing the chessboard code on the surface of an object to be detected to obtain a cylinder to be detected;
step S2, installing the binocular camera and calibrating the internal and external parameters of the binocular camera;
step S3, acquiring a to-be-detected cylinder image through a binocular camera, then respectively carrying out cylinder target segmentation on the image, and extracting two-dimensional image coordinates of a characteristic corner point in a target on the basis of segmentation;
step S4, converting the two-dimensional image coordinates of the characteristic corner points into three-dimensional world coordinates based on the internal and external parameter matrixes of the binocular camera and the camera imaging model;
step S5, performing cylinder fitting on the three-dimensional world coordinates of the characteristic angular points, and calculating the center point and the direction vector of the cylinder to obtain translation and rotation information of the object to be detected, namely displacement information with six degrees of freedom; the step S3 specifically includes:
step S31, binocular image acquisition is carried out on the cylinder to be measured to obtain iml1,imr1,iml2,imr2
Step S32, automatically segmenting the target x according to the images collected by the binocular cameral1、xr1And xl2、xr2
Step S33: for the divided object xl1、xr1And xl2、xr2Extracting characteristic angular points to obtain image two-dimensional coordinates x of each characteristic angular pointl11,…,xl1n、xr11,…,xr1nAnd xl21,…,xl2n、xr21,…,xr2nAnd a two-dimensional coordinate set X ofl1、Xr1And Xl2、Xr2
Step S34: for image two-dimensional coordinate set Xl1,Xr1And Xl2,Xr2Optimizing the checkerboard constraint to generate a final two-dimensional coordinate set Yl1、Yr1And Yl2、Yr2
Step S35: restoring binary coded information B by utilizing gray threshold1、B2
The step S4 specifically includes:
step S41, obtaining binocular camera internal parameter M by binocular calibrationleft、MrightAn external reference matrix H;
step S42: according to the obtained internal and external parameter matrix of the camera, a two-dimensional coordinate set Y is utilizedl1、Yr1And Yl2、Yr2Reconstructing a three-dimensional world coordinate set G1And G2
Step S43: for discrete three-dimensional world coordinate set G1And G2Optimizing the linear constraint to generate a coordinate set group1And group2
2. The binocular vision-based non-contact six-degree-of-freedom displacement measurement method according to claim 1, wherein: the step S5 specifically includes:
step S51, coding B according to the obtained binary value1And B2Judging the coordinate set group1And group2Set of coordinates g of common feature corners present in1And g2
Step S52: using coordinate set group1And group2Respectively performing cylinder fitting to obtain central coordinates C of the two1、C2And a radius R1、R2And a direction vector V1,V2
Step S53: using centre coordinates C1、C2The difference calculation is carried out to obtain the displacement information, and meanwhile, the direction vector V is utilized1,V2And the Euler angle of rotation can be calculated according to the Euler angle formula of the rotation matrix.
CN201811567836.4A 2018-12-21 2018-12-21 Binocular vision-based non-contact six-degree-of-freedom displacement measurement method Active CN109373912B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811567836.4A CN109373912B (en) 2018-12-21 2018-12-21 Binocular vision-based non-contact six-degree-of-freedom displacement measurement method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811567836.4A CN109373912B (en) 2018-12-21 2018-12-21 Binocular vision-based non-contact six-degree-of-freedom displacement measurement method

Publications (2)

Publication Number Publication Date
CN109373912A CN109373912A (en) 2019-02-22
CN109373912B true CN109373912B (en) 2021-03-02

Family

ID=65371227

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811567836.4A Active CN109373912B (en) 2018-12-21 2018-12-21 Binocular vision-based non-contact six-degree-of-freedom displacement measurement method

Country Status (1)

Country Link
CN (1) CN109373912B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113284195A (en) * 2021-07-08 2021-08-20 上海海栎创科技股份有限公司 On-line compensation method and system for binocular stereoscopic vision imaging

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108765494A (en) * 2018-05-21 2018-11-06 南昌航空大学 A kind of polyphaser scaling method for demarcating object based on cylinder

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1141670C (en) * 2001-07-27 2004-03-10 南开大学 3D tracking and measurement method of moving objects by 2D code
CN101576379B (en) * 2009-05-12 2011-04-27 四川大学 Fast calibration method of active projection three dimensional measuring system based on two-dimension multi-color target
CA2843892C (en) * 2011-08-02 2019-08-27 The Government Of The United States Of America, As Represented By The Secretary Of The Navy System and method for remote full field three-dimensional displacement and strain measurements
CN103591906A (en) * 2012-08-13 2014-02-19 上海威塔数字科技有限公司 A method for carrying out three dimensional tracking measurement on a moving object through utilizing two dimensional coding
CN103256896B (en) * 2013-04-19 2015-06-24 大连理工大学 Position and posture measurement method of high-speed rolling body

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108765494A (en) * 2018-05-21 2018-11-06 南昌航空大学 A kind of polyphaser scaling method for demarcating object based on cylinder

Also Published As

Publication number Publication date
CN109373912A (en) 2019-02-22

Similar Documents

Publication Publication Date Title
Chen et al. High-accuracy multi-camera reconstruction enhanced by adaptive point cloud correction algorithm
Johannsen et al. On the calibration of focused plenoptic cameras
JP5029618B2 (en) Three-dimensional shape measuring apparatus, method and program by pattern projection method
CN109035320B (en) Monocular vision-based depth extraction method
JP2015203652A (en) Information processing unit and information processing method
WO2007015059A1 (en) Method and system for three-dimensional data capture
TW201724026A (en) Generating a merged, fused three-dimensional point cloud based on captured images of a scene
CN109146980B (en) Monocular vision based optimized depth extraction and passive distance measurement method
Dufour et al. Integrated digital image correlation for the evaluation and correction of optical distortions
JP5412757B2 (en) Optical system distortion correction method and optical system distortion correction apparatus
CN109559355B (en) Multi-camera global calibration device and method without public view field based on camera set
CN107809610A (en) Camera parameter set calculating apparatus, camera parameter set calculation method and program
Wenzel et al. High-resolution surface reconstruction from imagery for close range cultural Heritage applications
CN109373912B (en) Binocular vision-based non-contact six-degree-of-freedom displacement measurement method
CN109974618B (en) Global calibration method of multi-sensor vision measurement system
CN111981982B (en) Multi-directional cooperative target optical measurement method based on weighted SFM algorithm
JP2018179577A (en) Position measuring device
CN102881040A (en) Three-dimensional reconstruction method for mobile photographing of digital camera
CN107424194A (en) The detection method of keyboard profile tolerance
KR101673144B1 (en) Stereoscopic image registration method based on a partial linear method
CN108362205B (en) Space distance measuring method based on fringe projection
Oniga et al. Metric and Non-Metric Cameras Calibration for the Improvement of Real-Time Monitoring Process Results.
CN104157008A (en) Depth image matching method based on ASIFT (Affine Scale-invariant Feature Transform)
TWI234640B (en) Contactless three dimensional measure method
JP2006135621A (en) Camera calibrating device

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant