CN111380481A - Automobile morphology monocular active vision reconstruction system and method based on five-point identification - Google Patents
Automobile morphology monocular active vision reconstruction system and method based on five-point identification Download PDFInfo
- Publication number
- CN111380481A CN111380481A CN201911426733.0A CN201911426733A CN111380481A CN 111380481 A CN111380481 A CN 111380481A CN 201911426733 A CN201911426733 A CN 201911426733A CN 111380481 A CN111380481 A CN 111380481A
- Authority
- CN
- China
- Prior art keywords
- point
- camera
- laser
- dimensional
- coordinate system
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 42
- 239000011159 matrix material Substances 0.000 claims abstract description 26
- 238000013519 translation Methods 0.000 claims abstract description 11
- 229910000831 Steel Inorganic materials 0.000 claims description 31
- 239000010959 steel Substances 0.000 claims description 31
- 230000012447 hatching Effects 0.000 claims description 7
- 230000000007 visual effect Effects 0.000 claims description 7
- 238000000354 decomposition reaction Methods 0.000 claims description 6
- 239000000126 substance Substances 0.000 claims description 6
- 238000006243 chemical reaction Methods 0.000 claims description 4
- 230000009466 transformation Effects 0.000 claims description 4
- 238000003466 welding Methods 0.000 claims description 3
- 238000001514 detection method Methods 0.000 abstract description 5
- 238000012876 topography Methods 0.000 description 8
- 238000005259 measurement Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000000691 measurement method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006748 scratching Methods 0.000 description 1
- 230000002393 scratching effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
- G01B11/24—Measuring arrangements characterised by the use of optical techniques for measuring contours or curvatures
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention discloses an automobile morphology monocular active vision reconstruction system and method based on five-point identification, and aims to solve the problem of automobile morphology monocular active vision detection based on five-point identification. The automobile morphology monocular active vision reconstruction system based on the five-point identification mainly comprises a camera (1), a laser plane projector (2), a two-dimensional target (3), a two-dimensional target fixing seat (4), a connecting clamp (5) and a tripod (6). The automobile morphology monocular active vision reconstruction method based on the five-point identification is composed of the steps of image acquisition, solving of an internal parameter matrix of a camera (1), solving of an external parameter rotation matrix and a translation vector of the camera (1), reconstruction of automobile body characteristic points and the like, and the automobile morphology monocular active vision reconstruction system and method based on the five-point identification are simple in structure and reliable in performance.
Description
Technical Field
The invention relates to automobile detection equipment and a detection method, in particular to an automobile appearance monocular active vision reconstruction system and method based on five-point identification.
Background
The automobile morphology detection is an important link in the field of traffic detection, and has important application prospects in aspects of automobile shape and position deviation, wheelbase difference, morphology reconstruction, automobile type classification, overload overrun, illegal modification and the like. In order to avoid scratching the surface of the object to be measured, non-contact measurement methods are frequently used. However, non-contact measurement usually requires the target to be used and the camera to be calibrated by moving the target to a plurality of positions, which is time consuming and complicates the process. How to simplify the calibration process of the camera and how to realize the reconstruction of the three-dimensional information of the object is very important. In order to realize higher-precision characteristic point reconstruction of the automobile appearance, a monocular active vision reconstruction system and method for the automobile appearance based on five-point identification are designed.
Disclosure of Invention
Aiming at solving the problems of complex device, complicated process and the like in the prior art in the automobile appearance reconstruction process, the invention designs an automobile appearance monocular active vision reconstruction system and method based on five-point identification to realize the three-dimensional free reconstruction of the automobile appearance characteristic points. The method mainly comprises an industrial camera, a two-dimensional target consisting of regular fan-shaped patterns, a laser projection plane instrument with a fixed relative position with the two-dimensional target and a two-dimensional target fixing seat, and single-image camera calibration is carried out by applying projective transformation of five-point marks on intersecting straight lines, so that three-dimensional reconstruction of the appearance of the automobile is realized.
The invention is realized by adopting the following technical scheme by combining the attached drawings of the specification:
the automobile morphology monocular active vision reconstruction system based on the five-point identification comprises a camera, a laser projection instrument, a two-dimensional target fixing seat, a connecting clamp and a tripod;
the tripod is placed subaerial, the camera passes through the screw hole of bottom and the bolt thread fixed connection at tripod top, two-dimensional target fixing base is placed subaerial at the level, recess tight fit fixed connection on two-dimensional target bottom and the two-dimensional target fixing base, the two-dimensional target is put into the long and thin recess of the U-shaped steel sheet of connecting clamp, the bolt penetrates the through-hole and the nut thread fixed connection of connecting clamp and two-dimensional target, the laser throws the face appearance and inserts in the hole of laser instrument connecting clamp steel pipe, the screw hole of holding screw in connecting clamp steel pipe side, the holding screw tip throws face of a face appearance cylinder contact tight fit with the laser, the laser plane that the laser throws the face appearance and sends and two.
The camera in the technical scheme is a wide-angle industrial camera provided with a narrow-band filter.
The laser projection instrument in the technical scheme is a cylindrical part capable of emitting laser, and the laser wavelength emitted by the laser projection instrument is consistent with the band-pass wavelength of a narrow-band filter of a camera.
The two-dimensional target in the technical scheme is a part made of a rectangular steel plate with a through hole, and radial black-white alternating patterns capable of generating collinear point characteristics are adhered to the surface of the two-dimensional target.
In the technical scheme, the two-dimensional target fixing seat is a cuboid part processed by a steel plate, and a long and thin groove is processed on the upper surface of the cuboid part.
The connecting clamp in the technical scheme is a part formed by welding a U-shaped steel plate and a steel pipe, a through hole is processed on the side surface of the U-shaped steel plate, and a threaded hole is processed on the side surface of the steel pipe.
The tripod in the technical scheme is a tripod with adjustable height.
The automobile morphology monocular active vision reconstruction method based on the five-point identification specifically comprises the following steps:
the first step is as follows: the method comprises the following steps of image acquisition of an automobile morphology monocular active vision reconstruction method based on five-point identification:
the camera is fixed on the tripod, the tripod is placed on the ground, the laser projector is inserted into an inner hole of the steel pipe of the connecting clamp, the laser projector is opened, the laser projected by the laser projector and the two-dimensional target are coplanar and the laser projector is fixed by adjusting a screw on the steel pipe of the connecting clamp, the camera acquires an image, and the image comprises the two-dimensional target and a laser projection point where the laser projector and the surface of the vehicle are intersected;
the second step is that: solving the intrinsic parameter matrix of the camera:
respectively establishing a one-dimensional coordinate system on each straight line of the two-dimensional target, wherein the line segment endpoint Z on the straight lineq,1、Zq,2、Zq,3Expressed as a homogeneous two-dimensional vector (0, 1)T、(aq1,1)TAnd (a)q1+aq2,1)TCoordinates z of the projected points of each point on the image planeq,1、zq,2、zq,3Can also be expressed as a homogeneous two-dimensional vector (0, 1)T、(dq1,1)TAnd (d)q1+dq2,1)TThen the projection relationship between the points is
Hq,1Zq,i=zq,i
Wherein, the logarithm q of the intersecting line is 1, 2, n, i is 1, 2, 3, at least five points of two pairs of intersecting straight lines are needed, and Z can be obtained by SVD decompositionq,1、Zq,2、Zq,3Conversion to zq,1、zq,2、zq,31D projective transformation matrix Hq,1;
Infinite point projection point coordinate on each straight line under one-dimensional coordinate system
Wherein V∞=(1,0)TIs a one-dimensional point at infinity,is V under a one-dimensional coordinate system∞Homogeneous coordinates of the projected points of (a);
according to the homogeneous coordinate of the known point on the straight line in the one-dimensional coordinate system and the homogeneous coordinate in the image coordinate system, the slope of the q-th straight line is
Wherein (d)qi,1)TIn the form of a homogeneous two-dimensional vector of the coordinates of the projection points in a one-dimensional coordinate system on the image plane, zq,i=(xq,i,yq,i,1)TFor the homogeneous coordinate of the known point in the two-dimensional coordinate system, further obtaining the coordinate of the vanishing point projection point in the image coordinate system as
Wherein the content of the first and second substances,the vanishing point is a homogeneous coordinate in a one-dimensional coordinate system, and the relationship between the vanishing point and the vanishing line is
vq,∞ Tw∞=0
Wherein w∞The coordinates of the vanishing lines are obtained by all vanishing points on the vanishing lines
[v1,1v1,2v2,1v2,2… vq,1vq,2]Tw∞=0
The coordinates w of the hatching can be obtained by SVD∞;
From the origin O of the imageiTo the coordinates w of the hatching∞Making a vertical line, and setting the foot to be PvThe origin O of the camera can be obtained according to the common side of the right trianglecThe distance to the foot is
Wherein the content of the first and second substances,PV=(xv,yv,1)T,vq,1=(xq,1,yq,1,1)T,vq,2=(xq,2,yq,2,1)Tand β is vq,1OcAnd vq,2OcThe focal length of the camera can be further calculated according to the right-angled triangle Pythagorean theorem
OiPvCan be directly measured in the image, and the internal parameters of the camera can be obtained according to the focal length and the condition that the assumed principal point is positioned at the center of the image
Wherein u is0And v0Is half the resolution of the image in two perpendicular directions;
the third step: solving an external parameter rotation matrix and a translation vector of the camera:
with OcIs the origin, Ocvq,1Is Ocl axis, Ocvq,2Is Ocm-axis creating a new three-dimensional coordinate system Oc-lmn, according to fiThe rotation matrix between the new coordinate system and the coordinate system of the camera 1 is obtained
Wherein v isq,1=(xq,1,yq,1,1)T,vq,2=(xq,2,yq,2,1)T,Ocn=(nx,ny,nz)TAs the third axis direction unit vector of the new coordinate system, according to Ocn=Ocl×Ocm is obtained, SVD decomposition is carried out on the multiplication result of the two rotation matrixes, elements in the middle diagonal matrix are squared, and then the multiplication result is multiplied with the orthogonal matrixes at the two sides to obtain the optimal rotation matrix ROC;
Selecting a world coordinate system O with a known length under the world coordinate systemoXoThe line segment AB with parallel directions is taken as the parallel line AB' of the line segment through the projection point a of an end point A of the line segment according to the delta OCAB and Δ OCaB' is a similar triangle, and O can be obtainedcNorm of A is
Then according to OCA and OCFinding the vector with the same direction of a
Obtaining a translation vector according to the obtained rotation matrix
tOC=RCOOCA
The fourth step: rebuilding the characteristic points of the automobile morphology monocular active vision rebuilding method based on the five-point identification:
according to the projection point of the laser surface collected in the image on the vehicle bodyImage coordinates ofAccording toThe internal parameters, the rotation matrix and the translation vector of the camera obtained in the first step of calibration process can be obtained
By laser plane piTThe coordinate under the two-dimensional target coordinate system is piT=(0 0 1 1)TObtaining the coordinates of the laser plane in the camera coordinate system
Then according to the point to be reconstructed positioned on the laser plane emitted by the laser projection instrument
Wherein, piqFor a laser plane under a camera coordinate system, the two formulas are combined to calculate the three-dimensional coordinate of the intersection line point of the laser plane vehicle body under a two-dimensional target coordinate system;
in the process of reconstructing and detecting the automobile appearance, the fourth step is a step of circularly carrying out;
the intersection line point of the laser plane and the automobile body at other positions can be obtained by moving the two-dimensional target within the visual field range of the camera, so that the coordinates of the intersection line point of the automobile and the laser plane under the coordinate system of the two-dimensional target are determined, and the monocular active visual reconstruction of the automobile appearance based on the five-point identification is completed.
The invention has the beneficial effects that:
1. the invention adopts a monocular active vision reconstruction method for the appearance of the automobile based on five-point identification. Firstly, the two-dimensional target 3 is fixedly connected with the laser projection instrument 2, and the laser plane and the two-dimensional reference plane are ensured to be coplanar. And (3) according to the conversion of five points on the intersecting line on the target pattern between the linear coordinate system and the image coordinate system, obtaining the coordinates of the vanishing points and the vanishing lines, further solving the internal and external parameters of the camera, and realizing the calibration of the single-image camera. And then, reconstructing the coordinates of the laser intersection line points under the two-dimensional target 3 coordinate system by using the properties of the laser intersection line points on the laser plane. And finally, the global three-dimensional reconstruction of the target morphology can be realized by moving the position of the object to be detected.
2. The camera calibration method utilizes five-point identification on the intersecting straight line to calibrate the camera, theoretically, the two vanishing points can solve the vanishing line, the method adopts a plurality of vanishing points to solve, and under the condition that the coordinate solution of the individual vanishing points has errors, the influence of the vanishing points on the final result can be reduced, thereby being beneficial to realizing the high-precision reconstruction of the target object.
3. The measurement system provided by the invention solves the problems that the existing measurement mode is complicated, the structure is complex, a plurality of images need to be acquired for camera calibration and the like. The calibration of the camera can be completed through a single picture, and the reconstruction of the three-dimensional characteristic point is further realized by utilizing the condition of the laser intersection line point on the laser plane.
Drawings
FIG. 1 is an isometric view of an automotive topography monocular active vision reconstruction system based on five-point identification;
fig. 2 is an axonometric view of the camera 1 in the automotive topography monocular active vision reconstruction system based on five-point identification;
FIG. 3 is an isometric view of a laser projection plane 2 in an automotive topography monocular active vision reconstruction system based on five-point identification;
FIG. 4 is an isometric view of a two-dimensional target 3 in a monocular active vision reconstruction system for vehicle topography based on five-point identification;
FIG. 5 is an axonometric view of a two-dimensional target holder 4 in a monocular active vision reconstruction system for automobile morphology based on five-point identification;
FIG. 6 is an isometric view of a joining clip 5 in an automotive topography monocular active vision reconstruction system based on five-point identification;
FIG. 7 is an isometric view of a tripod 6 in an automotive topography monocular active vision reconstruction system based on five-point identification;
FIG. 8 is a schematic diagram of solving camera intrinsic parameters in a monocular active vision reconstruction method for automobile morphology based on five-point identification;
FIG. 9 is a schematic diagram of a camera translation vector solving method in an automobile morphology monocular active vision reconstruction method based on five-point identification;
FIG. 10 is a flowchart of a calibration process in the monocular active vision reconstruction method for vehicle morphology based on five-point identification;
FIG. 11 is a flow chart for solving the external parameters of the camera 1 in the automobile morphology monocular active vision reconstruction method based on five-point identification;
FIG. 12 is a flow chart of a reconstruction process in a monocular active vision reconstruction method for vehicle topography based on five-point identification;
in the figure: 1. the camera, 2. the laser appearance of throwing in the face, 3. the two-dimensional target, 4. the two-dimensional target fixing base, 5. the connecting clamp, 6. the tripod.
Detailed Description
The invention is described in further detail below with reference to the attached drawing figures:
referring to fig. 1 to 7, the automobile morphology monocular active vision reconstruction system based on the five-point identification comprises a camera 1, a laser projection instrument 2, a two-dimensional target 3, a two-dimensional target fixing seat 4, a connecting clamp 5 and a tripod 6;
the tripod 6 is a tripod with adjustable height, the tripod 6 is placed on the ground, the camera 1 is a wide-angle industrial camera with a narrow-band filter, the camera 1 is fixedly connected with a bolt thread at the top of the tripod 6 through a threaded hole at the bottom, the two-dimensional target fixing seat 4 is a cuboid part processed by a steel plate, a long and thin groove is processed on the upper surface, the two-dimensional target fixing seat 4 is placed on the horizontal ground, the two-dimensional target 3 is a part made of a rectangular steel plate with a through hole, radial black and white alternate patterns capable of generating collinear point characteristics are pasted on the surface of the two-dimensional target 3, the bottom of the two-dimensional target 3 is fixedly connected with the groove on the two-dimensional target fixing seat 4 in a close fit manner, the connecting clamp 5 is a part formed by welding a U-shaped steel plate and a steel pipe, the side of the U-shaped steel plate is processed with, the bolt penetrates through the through hole of the connecting clamp 5 and the through hole of the two-dimensional target 3 and is fixedly connected with the nut thread, the laser projection instrument 2 is inserted into the inner hole of the steel pipe of the laser connecting clamp 5, the fastening screw is screwed into the threaded hole in the side face of the steel pipe of the connecting clamp 5, the end part of the fastening screw is in close fit with the cylindrical surface of the laser projection instrument 2 in contact, the laser projection instrument 2 is a cylindrical part capable of emitting laser, the laser wavelength emitted by the laser projection instrument 2 is consistent with the band-pass wavelength of the narrow-band filter of the camera 1, and the laser plane emitted by the laser projection instrument 2 is coplanar with the plane where the two.
Referring to fig. 8 to 12, the automobile topography monocular active vision reconstruction method based on the five-point identification provided by the present invention includes the following four steps:
the first step is as follows: the method comprises the following steps of image acquisition of an automobile morphology monocular active vision reconstruction method based on five-point identification:
the camera 1 is fixed on the tripod 6, the tripod 6 is placed on the ground, the laser projection device 2 is inserted into an inner hole of a steel pipe of the connecting clamp 5, the laser projection device 2 is opened, laser projected by the laser projection device 2 and the two-dimensional target 3 are coplanar by adjusting a screw on the steel pipe of the connecting clamp 5, the laser projection device 2 is fixed, the camera 1 acquires an image, and the image comprises the two-dimensional target 3 and a laser projection point at which the laser projection device 2 intersects with the surface of a vehicle;
the second step is that: solving the parameter matrix in the camera 1:
respectively establishing a one-dimensional coordinate system on each straight line of the two-dimensional target 3, wherein the line segment endpoint Z on the straight lineq,1、Zq,2、Zq,3Expressed as a homogeneous two-dimensional vector (0, 1)T、(aq1,1)TAnd (a)q1+aq2,1)TCoordinates z of the projected points of each point on the image planeq,1、zq,2、zq,3Can also be expressed as a homogeneous two-dimensional vector (0, 1)T、(dq1,1)TAnd (d)q1+dq2,1)TThen the projection relationship between the points is
Hq,1Zq,i=zq,i
Wherein, the logarithm q of the intersecting line is 1, 2, n, i is 1, 2, 3, at least five points of two pairs of intersecting straight lines are needed, and Z can be obtained by SVD decompositionq,1、Zq,2、Zq,3Conversion to zq,1、zq,2、zq,31D projective transformation matrix Hq,1;
Infinite point projection point coordinate on each straight line under one-dimensional coordinate system
Wherein V∞=(1,0)TIs a one-dimensional point at infinity,is V under a one-dimensional coordinate system∞Homogeneous coordinates of the projected points of (a);
according to the homogeneous coordinate of the known point on the straight line in the one-dimensional coordinate system and the homogeneous coordinate in the image coordinate system, the slope of the q-th straight line is
Wherein (d)qi,1)TIn the form of a homogeneous two-dimensional vector of the coordinates of the projection points in a one-dimensional coordinate system on the image plane, zq,i=(xq,i,yq,i,1)TFor the homogeneous coordinate of the known point in the two-dimensional coordinate system, further obtaining the coordinate of the vanishing point projection point in the image coordinate system as
Wherein the content of the first and second substances,the vanishing point is a homogeneous coordinate in a one-dimensional coordinate system, and the relationship between the vanishing point and the vanishing line is
vq,∞ Tw∞=0
Wherein w∞To eliminateThe coordinates of the hatching are obtained by all vanishing points on the vanishing line
[v1,1v1,2v2,1v2,2… vq,1bq,2]Tw∞=0
The coordinates w of the hatching can be obtained by SVD∞;
From the origin O of the imageiTo the coordinates w of the hatching∞Making a vertical line, and setting the foot to be PvThe origin O of the camera 1 can be obtained according to the common side of the right trianglecThe distance to the foot is
Wherein the content of the first and second substances,PV=(xv,yv,1)T,vq,1=(xq,1,yq,1,1)T,vq,2=(xq,2,yq,2,1)Tand β is vq,1OcAnd vq,2OcThe focal length of the camera 1 can be further calculated according to the right-angled triangle Pythagorean theorem
OiPvCan be directly measured in the image, and the internal parameters of the camera 1 can be obtained according to the focal length and the condition that the assumed principal point is positioned at the center of the image
Wherein u is0And v0Is half the resolution of the image in two perpendicular directions;
the third step: solving an external parameter rotation matrix and a translation vector of the camera 1:
with OcIs the origin, Ocvq,1Is Ocl axis, Ocvq,2Is Ocm-axis creating a new three-dimensional coordinate system Oc-lmn, according to fiThe rotation matrix between the new coordinate system and the coordinate system of the camera 1 is obtained
Wherein v isq,1=(xq,1,yq,1,1)T,vq,2=(xq,2,yq,2,1)T,Ocn=(nx,ny,nz)TAs the third axis direction unit vector of the new coordinate system, according to Ocn=Ocl×Ocm is obtained, SVD decomposition is carried out on the multiplication result of the two rotation matrixes, elements in the middle diagonal matrix are squared, and then the multiplication result is multiplied with the orthogonal matrixes at the two sides to obtain the optimal rotation matrix ROC;
Selecting a world coordinate system O with a known length under the world coordinate systemoXoThe line segment AB with parallel directions is taken as the parallel line AB' of the line segment through the projection point a of an end point A of the line segment according to the delta OCAB and Δ OCaB' is a similar triangle, and O can be obtainedcNorm of A is
Then according to OCA and OCFinding the vector with the same direction of a
Obtaining a translation vector according to the obtained rotation matrix
tOC=RCOOCA
The fourth step: rebuilding the characteristic points of the automobile morphology monocular active vision rebuilding method based on the five-point identification:
according to the projection point of the laser surface collected in the image on the vehicle bodyImage coordinates ofObtaining the internal parameters, the rotation matrix and the translation vector of the camera 1 according to the calibration process in the first step
By laser plane piTThe coordinate under the 3 coordinate system of the two-dimensional target is piT=(0 0 1 1)TObtaining the coordinates of the laser plane in the coordinate system of the camera 1
Then according to the point to be reconstructed positioned on the laser plane emitted by the laser projection instrument 2
Wherein, piqFor a laser plane under a coordinate system of a camera 1, the two formulas are combined to calculate the three-dimensional coordinate of the intersection line point of the laser plane body under a coordinate system of a two-dimensional target 3;
in the process of reconstructing and detecting the automobile appearance, the fourth step is a step of circularly carrying out;
by moving the two-dimensional target 3 in the visual field range of the camera 1, the laser intersection line point of the laser plane and the automobile body at other positions can be obtained, so that the coordinates of the intersection line point of the automobile and the laser plane under the coordinate system of the two-dimensional target 3 are determined, and the monocular active vision reconstruction of the automobile appearance based on the five-point identification is completed.
Claims (8)
1. A monocular active vision reconstruction system for automobile morphology based on five-point identification is characterized by comprising a camera (1), a laser projection instrument (2), a two-dimensional target (3), a two-dimensional target fixing seat (4), a connecting clamp (5) and a tripod (6);
the tripod (6) is placed on the ground, the camera (1) is fixedly connected with the bolt thread at the top of the tripod (6) through the threaded hole at the bottom, the two-dimensional target fixing seat (4) is placed on the horizontal ground, the bottom of the two-dimensional target (3) is fixedly connected with the groove on the two-dimensional target fixing seat (4) in a tight fit manner, the two-dimensional target (3) is placed in the long and thin groove of the U-shaped steel plate of the connecting clamp (5), a bolt penetrates through the through holes of the connecting clamp (5) and the two-dimensional target (3) to be fixedly connected with the nut threads, the laser projection instrument (2) is inserted into the inner hole of the steel pipe of the laser connecting clamp (5), the fastening screw is screwed into a threaded hole in the side face of the steel pipe of the connecting clamp (5), the end part of the fastening screw is in contact tight fit with the cylindrical surface of the laser projection instrument (2), and a laser plane emitted by the laser projection instrument (2) is coplanar with the plane of the graph of the two-dimensional target (3).
2. The system for monocular active visual reconstruction of vehicle morphology based on five-point identification according to claim 1, characterized in that the video camera (1) is a wide-angle industrial camera equipped with a narrow-band filter.
3. The automobile morphology monocular active vision reconstruction system based on five-point identification as set forth in claim 1, characterized in that the laser projection instrument (2) is a cylindrical part capable of emitting laser, and the laser wavelength emitted by the laser projection instrument (2) is consistent with the band-pass wavelength of the narrow-band filter of the camera (1).
4. The automobile morphology monocular active vision reconstruction system based on five-point identification as claimed in claim 1, characterized in that the two-dimensional target (3) is a part made of a rectangular steel plate with through holes, and a radial black and white alternating pattern capable of generating collinear point features is adhered on the surface of the two-dimensional target (3).
5. The automobile morphology monocular active vision reconstruction system based on five-point identification as set forth in claim 1, characterized in that the two-dimensional target holder (4) is a rectangular parallelepiped component machined from a steel plate, and a long and thin groove is machined on the upper surface.
6. The automobile morphology monocular active vision reconstruction system based on five-point identification as claimed in claim 1, characterized in that the connecting clamp (5) is a part formed by welding a U-shaped steel plate and a steel pipe, a through hole is processed on the side surface of the U-shaped steel plate, and a threaded hole is processed on the side surface of the steel pipe.
7. The system for monocular active visual reconstruction of vehicle morphology based on five point identification according to claim 1, characterized in that the tripod (6) is a tripod with adjustable height.
8. The reconstruction method of the automobile morphology monocular active vision reconstruction system based on the five-point identification as claimed in claims 1 to 7, characterized by comprising the following specific steps:
the first step is as follows: the method comprises the following steps of image acquisition of an automobile morphology monocular active vision reconstruction method based on five-point identification:
the camera (1) is fixed on the tripod (6), the tripod (6) is placed on the ground, the laser plane projector (2) is inserted into an inner hole of a steel pipe of the connecting clamp (5), the laser plane projector (2) is opened, laser projected by the laser plane projector (2) is coplanar with the two-dimensional target (3) by adjusting a screw on the steel pipe of the connecting clamp (5), the laser plane projector (2) is fixed, the camera (1) acquires an image, and the image comprises the two-dimensional target (3) and a laser projection point where the laser plane projector (2) is intersected with the surface of a vehicle;
the second step is that: solving a parameter matrix in the camera (1):
a one-dimensional coordinate system is respectively established on each straight line of the two-dimensional target (3), and the line segment endpoint Z on the straight lineq,1、Zq,2、Zq,3Expressed as a homogeneous two-dimensional vector (0, 1)T、(aq1,1)TAnd (a)q1+aq2,1)TCoordinates z of the projected points of each point on the image planeq,1、zq,2、zq,3Can also be expressed as a homogeneous two-dimensional vector (0, 1)T、(dq1,1)TAnd (d)q1+dq2,1)TThen the projection relationship between the points is
Hq,1Zq,i=zq,i
Wherein, the logarithm q of the intersecting line is 1, 2, n, i is 1, 2, 3, at least five points of two pairs of intersecting straight lines are needed, and Z can be obtained by SVD decompositionq,1、Zq,2、Zq,3Conversion to zq,1、zq,2、zq,31D projective transformation matrix Hq,1;
Infinite point projection point coordinate on each straight line under one-dimensional coordinate system
Wherein V∞=(1,0)TIs a one-dimensional point at infinity,is V under a one-dimensional coordinate system∞Homogeneous coordinates of the projected points of (a);
according to the homogeneous coordinate of the known point on the straight line in the one-dimensional coordinate system and the homogeneous coordinate in the image coordinate system, the slope of the q-th straight line is
Wherein (d)qi,1)TIn the form of a homogeneous two-dimensional vector of the coordinates of the projection points in a one-dimensional coordinate system on the image plane, zq,i=(xq,i,yq,i,1)THomogeneous coordinates of known points in a two-dimensional coordinate systemFurther, the coordinates of the projected point of the vanishing point in the image coordinate system are obtained as
Wherein the content of the first and second substances,the vanishing point is a homogeneous coordinate in a one-dimensional coordinate system, and the relationship between the vanishing point and the vanishing line is
vq,∞ Tw∞=0
Wherein w∞The coordinates of the vanishing lines are obtained by all vanishing points on the vanishing lines
[v1,1v1,2v2,1v2,2… vq,1vq,2]Tw∞=0
The coordinates w of the hatching can be obtained by SVD∞;
From the origin O of the imageiTo the coordinates w of the hatching∞Making a vertical line, and setting the foot to be PvAccording to the common side of the right triangle, the origin O of the camera (1) can be obtainedcThe distance to the foot is
Wherein the content of the first and second substances,PV=(xv,yv,1)T,vq,1=(xq,1,yq,1,1)T,vq,2=(xq,2,yq,2,1)Tand β is vq,1OcAnd vq,2OcThe focal length of the camera (1) can be further calculated according to the right-angled triangle Pythagorean theorem
OiPvCan be directly measured in the image, and the internal parameter of the camera (1) can be obtained according to the focal length and the condition that the assumed principal point is positioned at the center of the image
Wherein u is0And v0Is half the resolution of the image in two perpendicular directions;
the third step: solving an external parameter rotation matrix and a translation vector of the camera (1):
with OcIs the origin, Ocvq,1Is Ocl axis, Ocvq,2Is Ocm-axis creating a new three-dimensional coordinate system Oc-lmn, according to fiThe rotation matrix between the new coordinate system and the coordinate system of the camera (1) is obtained as
Wherein v isq,1=(xq,1,yq,1,1)T,vq,2=(xq,2,yq,2,1)T,Ocn=(nx,ny,nz)TAs the third axis direction unit vector of the new coordinate system, according to Ocn=Ocl×Ocm is obtained, SVD decomposition is carried out on the multiplication result of the two rotation matrixes, elements in the middle diagonal matrix are squared, and then the multiplication result is multiplied with the orthogonal matrixes at the two sides to obtain the optimal rotation matrix ROC;
Selecting a world coordinate system O with a known length under the world coordinate systemoXoThe line segment AB with parallel directions is taken as the parallel line AB' of the line segment through the projection point a of an end point A of the line segment according to the delta OCAB and Δ OCaB' is a similar triangle, and O can be obtainedcNorm of A is
Then according to OCA and OCFinding the vector with the same direction of a
Obtaining a translation vector according to the obtained rotation matrix
tOC=RCOOCA
The fourth step: rebuilding the characteristic points of the automobile morphology monocular active vision rebuilding method based on the five-point identification:
according to the projection point of the laser surface collected in the image on the vehicle bodyImage coordinates ofObtaining the internal parameters, the rotation matrix and the translation vector of the camera (1) according to the calibration process in the first step
By laser plane piTThe coordinate under the coordinate system of the two-dimensional target (3) is piT=(0 0 1 1)TObtaining the coordinates of the laser plane in the coordinate system of the camera (1)
Then according to the point to be reconstructed, the point is positioned on the laser plane emitted by the laser projection instrument (2)
Wherein, piqFor a laser plane under a coordinate system of a camera (1), the two formulas are combined to calculate the three-dimensional coordinate of the intersection line point of the laser plane body under the coordinate system of a two-dimensional target (3);
in the process of reconstructing and detecting the automobile appearance, the fourth step is a step of circularly carrying out;
by moving the two-dimensional target (3) in the visual field range of the camera (1), the laser intersection line point of the laser plane and the automobile body at other positions can be obtained, so that the coordinates of the intersection line point of the automobile and the laser plane in the coordinate system of the two-dimensional target (3) are determined, and the monocular active visual reconstruction of the automobile appearance based on the five-point identification is completed.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911426733.0A CN111380481B (en) | 2019-12-31 | 2019-12-31 | System and method for monocular active visual reconstruction of automobile morphology based on five-point identification |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911426733.0A CN111380481B (en) | 2019-12-31 | 2019-12-31 | System and method for monocular active visual reconstruction of automobile morphology based on five-point identification |
Publications (2)
Publication Number | Publication Date |
---|---|
CN111380481A true CN111380481A (en) | 2020-07-07 |
CN111380481B CN111380481B (en) | 2024-06-07 |
Family
ID=71215236
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911426733.0A Active CN111380481B (en) | 2019-12-31 | 2019-12-31 | System and method for monocular active visual reconstruction of automobile morphology based on five-point identification |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111380481B (en) |
Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1481727A2 (en) * | 2003-05-30 | 2004-12-01 | Agilent Technologies Inc | Feature extraction methods and systems |
US7177740B1 (en) * | 2005-11-10 | 2007-02-13 | Beijing University Of Aeronautics And Astronautics | Method and apparatus for dynamic measuring three-dimensional parameters of tire with laser vision |
DE102005043753A1 (en) * | 2005-09-13 | 2007-03-15 | Deutsche Mechatronics Gmbh | Test sample surface point spatial coordinates determining method, involves computing spatial coordinates as projective reconstruction for all reference points, and converting coordinates of points into metric coordinates of surface points |
EP1770356A2 (en) * | 2005-09-30 | 2007-04-04 | Topcon Corporation | Three-dimensional measurement system with projection device |
CN101865675A (en) * | 2010-05-21 | 2010-10-20 | 北京化工大学 | Method for acquiring three-dimensional visual detection circle structure light calibration point |
CN102980528A (en) * | 2012-11-21 | 2013-03-20 | 上海交通大学 | Calibration method of pose position-free constraint line laser monocular vision three-dimensional measurement sensor parameters |
CN105302337A (en) * | 2013-06-04 | 2016-02-03 | 李文傑 | High resolution and high sensitivity three-dimensional (3D) cursor maneuvering system, device and motion detection method |
CN109540034A (en) * | 2019-01-14 | 2019-03-29 | 吉林大学 | The global free reconstructing system of the consistent monocular active vision automobile pattern of closed loop and method |
CN109540042A (en) * | 2019-01-14 | 2019-03-29 | 吉林大学 | Based on the automobile pattern Active visual inspection System and method for without constraint optical plane |
CN110120078A (en) * | 2019-05-08 | 2019-08-13 | 云南大学 | Camera marking method and system based on diameter Yu conjugate value property |
CN211060855U (en) * | 2019-12-31 | 2020-07-21 | 吉林大学 | Automobile morphology monocular active vision reconstruction system based on five-point identification |
-
2019
- 2019-12-31 CN CN201911426733.0A patent/CN111380481B/en active Active
Patent Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1481727A2 (en) * | 2003-05-30 | 2004-12-01 | Agilent Technologies Inc | Feature extraction methods and systems |
DE102005043753A1 (en) * | 2005-09-13 | 2007-03-15 | Deutsche Mechatronics Gmbh | Test sample surface point spatial coordinates determining method, involves computing spatial coordinates as projective reconstruction for all reference points, and converting coordinates of points into metric coordinates of surface points |
EP1770356A2 (en) * | 2005-09-30 | 2007-04-04 | Topcon Corporation | Three-dimensional measurement system with projection device |
US7177740B1 (en) * | 2005-11-10 | 2007-02-13 | Beijing University Of Aeronautics And Astronautics | Method and apparatus for dynamic measuring three-dimensional parameters of tire with laser vision |
CN101865675A (en) * | 2010-05-21 | 2010-10-20 | 北京化工大学 | Method for acquiring three-dimensional visual detection circle structure light calibration point |
CN102980528A (en) * | 2012-11-21 | 2013-03-20 | 上海交通大学 | Calibration method of pose position-free constraint line laser monocular vision three-dimensional measurement sensor parameters |
CN105302337A (en) * | 2013-06-04 | 2016-02-03 | 李文傑 | High resolution and high sensitivity three-dimensional (3D) cursor maneuvering system, device and motion detection method |
CN109540034A (en) * | 2019-01-14 | 2019-03-29 | 吉林大学 | The global free reconstructing system of the consistent monocular active vision automobile pattern of closed loop and method |
CN109540042A (en) * | 2019-01-14 | 2019-03-29 | 吉林大学 | Based on the automobile pattern Active visual inspection System and method for without constraint optical plane |
CN110120078A (en) * | 2019-05-08 | 2019-08-13 | 云南大学 | Camera marking method and system based on diameter Yu conjugate value property |
CN211060855U (en) * | 2019-12-31 | 2020-07-21 | 吉林大学 | Automobile morphology monocular active vision reconstruction system based on five-point identification |
Non-Patent Citations (1)
Title |
---|
王颖;张圆;: "一种基于平面靶标的圆结构光标定方法", 红外与激光工程, no. 1, 25 June 2013 (2013-06-25) * |
Also Published As
Publication number | Publication date |
---|---|
CN111380481B (en) | 2024-06-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
DE69636925T2 (en) | METHOD AND DEVICE FOR DETERMINING THE ORIENTATION OF MOTORCYCLE WHEELS | |
AT506110B1 (en) | DEVICE AND METHOD FOR DETECTING BODY MEASURE DATA AND CONTOUR DATA | |
CN110057295B (en) | Monocular vision plane distance measuring method without image control | |
US8086026B2 (en) | Method and system for the determination of object positions in a volume | |
CN101526336B (en) | Calibration method of linear structured light three-dimensional visual sensor based on measuring blocks | |
CN109855603B (en) | Focus measurement method and terminal | |
CN110672020A (en) | Stand tree height measuring method based on monocular vision | |
CN109540033B (en) | Closed-loop consistent binocular active vision automobile morphology global free reconstruction system and method | |
CN103994732B (en) | A kind of method for three-dimensional measurement based on fringe projection | |
CN106643567B (en) | A kind of method of calibration and system of lane shift system producing line scaling board | |
CN111412856A (en) | Automobile morphology monocular active vision reconstruction system and method based on double-rectangle identification | |
US11692815B2 (en) | Imaging system with calibration target object | |
CN108180888A (en) | A kind of distance detection method based on rotating pick-up head | |
Zheng et al. | Calibration of linear structured light system by planar checkerboard | |
CN209399945U (en) | Based on without constraint homocentric beam race automobile pattern Active visual inspection system | |
CN111380481A (en) | Automobile morphology monocular active vision reconstruction system and method based on five-point identification | |
CN211060855U (en) | Automobile morphology monocular active vision reconstruction system based on five-point identification | |
CN211060859U (en) | Automobile morphology reconstruction system based on triangular array affine invariant | |
CN111412854A (en) | Point and circle invariant based active vision reconstruction system and method for automobile morphology | |
CN109540042B (en) | Automobile morphology active vision detection system and method based on unconstrained light plane | |
CN109540041B (en) | Binocular active vision vector orthogonal normalization automobile morphology measurement system and method | |
CN111242901A (en) | Space point-based global calibration system and method for automobile detection camera without common view field | |
CN111380480B (en) | Automobile morphology reconstruction system and method based on affine invariant of triangular array | |
CN211060857U (en) | Automobile feature point vision reconstruction system based on one-dimensional reference object | |
CN111174726A (en) | Shaft end grounding binocular reconstruction system and method for concentric quadratic curve polar line geometry |
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 |