WO2018171031A1 - 三相机组特征点匹配方法、测量方法及三维检测装置 - Google Patents

三相机组特征点匹配方法、测量方法及三维检测装置 Download PDF

Info

Publication number
WO2018171031A1
WO2018171031A1 PCT/CN2017/084814 CN2017084814W WO2018171031A1 WO 2018171031 A1 WO2018171031 A1 WO 2018171031A1 CN 2017084814 W CN2017084814 W CN 2017084814W WO 2018171031 A1 WO2018171031 A1 WO 2018171031A1
Authority
WO
WIPO (PCT)
Prior art keywords
matching
point
image plane
image
points
Prior art date
Application number
PCT/CN2017/084814
Other languages
English (en)
French (fr)
Inventor
曹亮
周之琪
尹兴
龚婷
Original Assignee
北京清影机器视觉技术有限公司
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 北京清影机器视觉技术有限公司 filed Critical 北京清影机器视觉技术有限公司
Publication of WO2018171031A1 publication Critical patent/WO2018171031A1/zh

Links

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

Definitions

  • the invention relates to the field of optical electronic measurement technology, in particular to a three-phase unit feature point matching method, a measuring method and a three-dimensional detecting device.
  • three-dimensional stereo vision measurement generally adopts the method of line laser light screenshot measurement technology or binocular measurement plus structured light illumination.
  • the main reason that line laser or structured light is widely used in three-dimensional measurement is: by line laser or structured light indication
  • the corresponding matching points are defined in the imaging, which reduces the ambiguity of the matching and achieves a certain and unique match.
  • binocular matching cannot avoid the ambiguity of multi-point matching, and thus cannot meet the measurement requirements.
  • the line laser or structured light can only be used for the line laser or structured light imaging part. Measurements define the scope of application of the technology, and the use of line lasers or structured light can adversely affect the object being tested, such as a person.
  • the current binocular matching often uses a way of marking points on the surface of the object, and this method is also used to improve the accuracy of the matching.
  • this method is also used to improve the accuracy of the matching.
  • there is a disadvantage in that the manner in which the object is marked on the surface of the object needs to be manually processed and interfered with in advance.
  • the matching method includes using one of the four image planes as the base image plane, the base image a feature point on the plane finds all matching points on the image plane adjacent to the base image plane in the lateral direction that match the feature point; and finds the feature point on the base image plane in the longitudinal direction and the base All matching points on the image plane adjacent to the plane that match the feature point; re-match all the matching points in the horizontal and vertical directions to find all the sub-match point groups; find the diagonal position image plane Matching points corresponding to the feature points on the base image plane and all the found sub-match point groups; determining the unique matching point groups corresponding to the same viewpoint in the four image planes.
  • the three-dimensional space coordinates of the viewpoint can be calculated according to the image coordinates of the matching point group and the parameters of the camera system itself. As long as the acquired image is clear enough under any lighting conditions, for any object that is imaged on the image of the four-camera planar array and has certain image features, the same matching method and measurement method can be used to achieve the object. Three-dimensional measurement.
  • the four-camera group planar array feature point matching method uses a lot of cameras, which increases the application cost. At the same time, due to the large number of cameras, the processing difficulty and the computational complexity are also increased, and the measurement accuracy is also affected.
  • an object of the embodiments of the present invention is to provide a three-phase unit feature point matching method, a measurement method, and a three-dimensional detecting device; to reduce the complexity of the existing multi-camera matching method and the measuring method, and simplify the space size calculation process. , reduce system measurement error.
  • a three-phase unit feature point matching method includes the following steps:
  • A1 One of the lateral image planes of a group of three-phase units is used as the base image plane, and a feature point on the base image plane is found on the image plane adjacent to the base image plane in the lateral direction. All matching points that match the feature point;
  • Matching point groups finding matching points matching the pair of matching point groups on the third image plane, and removing matching point groups that do not satisfy the matching condition according to the matching condition, forming the matching point groups existing on the three image planes a matching point group that satisfies the matching condition;
  • A4 Perform full-image verification on the matching points corresponding to the other two non-base image planes in all the matching point groups, and determine that the three image planes correspond to the unique matching point group of the same viewpoint.
  • step a1 for a feature point on the base image plane, all matching points matching the feature point on the image plane adjacent to the base image plane in the lateral direction are found according to the matching condition 1);
  • the matching condition 1) when an image is imaged on three image planes of a group of three-phase units, the image points on two image planes adjacent in the lateral direction satisfy: the viewpoint is in the left image plane The imaging point on the image point on the corresponding right image plane of the viewpoint is located on the same line parallel to the horizontal coordinate axis, and the level of the image point on the left image plane relative to the coordinate origin of the left image plane The offset is greater than the horizontal offset of the imaged origin on the right image plane relative to the coordinate origin of the right image plane.
  • step a2 finding a matching point matching the pair of matching point groups on the third image plane according to the matching condition 2) and the matching condition 3);
  • the matching condition 2) is that when one viewpoint is imaged on three image planes of a group of three-phase units, an image point on the image plane in the other height direction is located adjacent to the image plane The intersection of the imaged point on the two image planes on the left and right sides with the line connecting the imaged point on the image plane, and the two lines are respectively parallel to the center point on the image plane where the viewpoint is located a line connecting the center points on the two left and right image planes on the corresponding horizontal axis;
  • the matching condition 3) is that a necessary condition that a viewpoint is satisfied in a corresponding matching point group on three image planes of a group of three-phase units is that the three imaging points form a triangle, and the three sides of the triangle are The three sides of the triangle composed of the three image plane focal points are parallel, and the triangle is similar to the triangle composed of the focus of the three image planes.
  • step a4 if two or more matching point groups on the base image plane corresponding to the same matching point appear on the other two non-base image planes, the matching points are used according to the matching conditions 1), 2), and 3 ), and steps a1, a2, and a3, sequentially search for corresponding matching points on the base image plane, determine whether there is another matching point group at the matching point, and if so, delete the matching relationship until a match occurs.
  • the relationship is the only matching relationship corresponding to the two pairs; if there is always a point-to-multipoint, and the two planes on the non-base image plane have completed the full-image verification, the matching relationship is deleted.
  • the unique matching point group corresponding to all the feature points in the base image plane is found by using the three-phase unit feature point matching method as described above;
  • step b2 According to the spatial position coordinates of the viewpoint obtained in the step b2, three-dimensional point cloud data is formed, and a three-dimensional point cloud graphic is created to reproduce the three-dimensional stereoscopic image.
  • the step b1 includes:
  • One of the lateral image planes of the three-phase unit is used as the base image plane, and a feature point on the base image plane is found on the image plane adjacent to the base image plane in the lateral direction. All matching points that match the feature point;
  • the matching point group For the feature points on the base image plane in the step b11 and all matching points on the image plane adjacent to the base image plane in the lateral direction that match the feature points, the matching point group, according to the matching condition 2) and the matching condition 3), find a matching point matching the pair of matching point groups on the third image plane, and according to the matching condition, remove the matching point group that does not satisfy the matching condition, Forming a matching point group that satisfies the matching condition 3) existing on the three image planes;
  • the three-phase unit is arranged in an arbitrary triangle, including a camera and b camera respectively located at two sides of the horizontal axis of the space coordinate, and a c camera located at an upper position of the vertical coordinate axis, and the focal points Oa, Ob, Oc of the three cameras are on the same plane, And the focal lengths of the three cameras are the same as f, the optical axes of the three cameras are perpendicular to the plane, the viewpoint P (Px, Py, Pz) and the point on the a image plane, the b image plane and the c image plane
  • the imaging points are Pa (Pax, Pay), Pb (Pbx, Pby), Pc (Pcx, Pcy), m is the length of OaOb, u is the target length of the image sensor on each image plane, and v is the camera's
  • the target surface width of the image sensor, P' point, Pa' point and Pb' point are the projection points of P point, Pa point and Pb point on the O
  • ⁇ a is the image plane pixel projection angle
  • ⁇ b is the b-image plane pixel projection angle
  • the pixel projection angle ⁇ ( ⁇ a or ⁇ b) is for a fixed focal length
  • Do pixel projection angle is an image acquisition system with known parameters, the angle between the line of each pixel on the image and its focus and the center of the optical axis.
  • a three-dimensional detecting device adopts the above measuring method, including a three-camera lens array.
  • the method further includes a first substrate and a second substrate parallel to each other, a control circuit board disposed on a side of the second substrate facing away from the first substrate, and accommodating the first substrate, the second substrate, and the The fixed rear case of the control circuit board;
  • the three-lens lens array is disposed on the first substrate, the axes of the respective lenses of the three-camera lens array are perpendicular to a plane of the first substrate, and the second substrate is disposed toward a surface of the first substrate There is an array of photosensitive elements, and each of the photosensitive elements in the array of photosensitive elements is disposed in one-to-one correspondence with each of the lenses.
  • the photosensitive element array is soldered on the second substrate or integrally formed with the second substrate, and the scanning lines of the two photosensitive elements at the bottom of the triangle formed by the photosensitive element array are horizontally parallel and completely aligned up and down, and at the same time, the photosensitive
  • the scanning line of the photosensitive element on the top of the element array is parallel to the scanning lines of the two photosensitive elements at the bottom, and the three photosensitive elements form an isosceles triangle with a bottom edge as a base or a right angle with a bottom edge as a right angle side triangle.
  • An image generation circuit board a fixed substrate for fixing the image generation circuit board, and an image acquisition circuit board located behind the fixed substrate;
  • the image generation circuit board is provided with three micro lenses, a micro photosensitive element corresponding to three micro lenses, and an image generation circuit disposed under the micro photosensitive elements, and the three micro lenses constitute a three camera lens. Array.
  • the micro photosensitive element array is soldered on the image forming circuit board or integrally formed with the image generating circuit board, and the scanning lines of the two micro photosensitive elements at the bottom of the triangle formed by the micro photosensitive element array are horizontally parallel and vertically aligned. Meanwhile, the scanning line of the micro photosensitive element on the top of the micro photosensitive element array is parallel to the scanning lines of the two micro photosensitive elements at the bottom, and the three micro photosensitive elements form an isosceles triangle with a bottom edge as the bottom or A right triangle with a bottom edge as a right angle.
  • the three-dimensional detecting device may be used as a basic measuring unit to form a three-dimensional detecting device plane detecting camera array of more than three cameras.
  • a planar three-phase unit feature point matching method capable of quickly matching the viewed point in three image planes according to the position of an imaged point on a three image plane of a set of planar three-phase units of a viewed object Corresponding unique imaging point group, which achieves versatility and unique matching of the viewpoints that can be imaged on all three cameras;
  • the same measurement method is adopted for any object to be observed.
  • the three-dimensional measurement of the object can be realized, and the measurement method does not need to perform any calibration on the field of view, and the measurement accuracy and resolution are only related to the measurement system, and independent of the object, the autonomic measurement can be completely realized.
  • the method uses fewer cameras, which reduces the application cost. At the same time, due to the small number of cameras, the processing difficulty and the computational complexity are reduced, and the measurement accuracy is improved.
  • Figure 1 is a schematic diagram of a spatial coordinate system established based on a three-phase unit
  • FIG. 2 is a schematic diagram of imaging of a certain viewpoint P(Px, Py, Pz) of the object and the point on the a image plane, the b image plane, and the c image plane;
  • Figure 3 is a schematic plan view of an arbitrary view point P on a planar three-phase unit
  • FIG. 4 is a schematic perspective view of an arbitrary view P on a and b camera sets
  • Figure 5 is a projection view of the imaging point of the P point on the a, b camera group on the OXZ coordinate plane;
  • FIG. 6 is a schematic diagram of pixel projection angle and P point Y value calculation
  • Figure 7 is an exploded view of one of the three-dimensional detecting devices provided in Embodiment 3.
  • Figure 8 is a schematic view of the assembly of Figure 7;
  • Figure 9 is a schematic structural view of the fixed rear case of Figure 7;
  • Figure 10 is a cross-sectional view of Figure 8.
  • Figure 11 is an exploded view of another three-dimensional detecting device provided in Embodiment 3.
  • Figure 12 is a schematic view of the assembly of Figure 11;
  • Figure 13 is a flow chart of a measurement method based on a three-phase unit feature point matching method.
  • Icon 100-first substrate; 101-lens; 102-mounting through hole; 200-second substrate; 201-photosensitive element; 300-control circuit board; 400-fixed rear case; 401-first power input port; - first data communication interface; 301 - fixed column; 500 - front case; 501 - lens hole; 600 - image generation circuit board; 601 - miniature photosensitive element; 602 - mounting hole; 603 - micro lens; 604 - image generation circuit 700-fixed substrate; 800-image acquisition circuit board; 900-back shell; 901-second power input port; 902-second data communication interface.
  • connection In the description of the present invention, it should be noted that the terms “installation”, “connected”, and “connected” are to be understood broadly, and may be fixed or detachable, for example, unless otherwise explicitly defined and defined. Connected, or integrally connected; can be mechanical or electrical; can be directly connected, or indirectly connected through an intermediate medium, can be the internal communication of the two components.
  • Connected, or integrally connected can be mechanical or electrical; can be directly connected, or indirectly connected through an intermediate medium, can be the internal communication of the two components.
  • the specific meaning of the above terms in the present invention can be understood in a specific case by those skilled in the art.
  • FIG. 1 is a schematic diagram of a spatial coordinate system established based on a three-phase unit
  • FIG. 2 is a viewpoint P (Px, Py, Pz) of the object being viewed and the point is on the a image plane, the b image plane, and the c image plane.
  • FIG. 3 is a schematic diagram of an imaging plane of any viewpoint P on a planar three-phase unit
  • FIG. 4 is a schematic perspective view of any viewpoint P on a and b camera groups
  • FIG. 5 is a point P at a, b a projection view of the imaging point on the camera group on the OXZ coordinate plane
  • FIG. 6 is a schematic diagram of pixel projection angle and P point Y value calculation;
  • FIG. 1 is a schematic diagram of a spatial coordinate system established based on a three-phase unit
  • FIG. 2 is a viewpoint P (Px, Py, Pz) of the object being viewed and the point is on the a image plane, the b image plane, and the c image
  • FIG. 7 is an exploded view of one of the three-dimensional detection devices provided in Embodiment 3;
  • FIG. 9 is a schematic structural view of the fixed rear case 900 of FIG. 7;
  • FIG. 10 is a cross-sectional view of FIG. 8;
  • FIG. 11 is an exploded view of another three-dimensional detecting device provided by Embodiment 3.
  • FIG. It is a schematic view of the assembly of Figure 11.
  • the imaged point refers to the image of the corresponding pixel position of the viewed object on the image plane (or image), and each of the viewed points of the object is respectively on the three image planes of a group of three-phase units.
  • a matching point refers to an image point on an image plane, and an image feature found on the image plane or other image plane that satisfies a certain matching condition with the imaging point and with the imaging point (eg Texture, color or grayscale, etc.) Similar imaging points. There may be one or more matching points corresponding to one imaging point.
  • a feature point refers to an image point corresponding to one or more matching points, the image point having image features different from other image points, such as different textures, colors or gray values, etc., in the embodiment of the present invention It is not specifically limited, and different image features can be selected according to actual conditions as a basis for judgment.
  • the image points corresponding to the viewpoints at the edge of the object to be measured or the transition zone of the texture have sharp image features
  • FIG. 13 is a flowchart of the measurement method based on the feature point matching method of the three-phase unit.
  • the embodiment relates to a method for matching feature points of a three-phase unit.
  • Figure 1 is a schematic diagram of a spatial coordinate system established based on a three-phase unit.
  • the focus of the three-phase unit be arranged in an arbitrary triangle on the same plane, including the a camera and the b camera respectively located at the two sides of the horizontal axis of the space coordinate, and the c camera located at the upper position of the vertical coordinate axis.
  • the configurations of the three cameras are identical.
  • the parameters of the lens 101, image sensor, and focal length of the three cameras are the same.
  • the focus of the three cameras Oa, Ob, Oc On the same plane, and the focal lengths of the three cameras are the same as f
  • the optical axes of the three cameras are perpendicular to the plane, and Oa, Ob, Oc form the bottom edge as m, the height is h, and the two bottom angles are ⁇ .
  • any triangle of ⁇ the midpoint of the bottom edge of the triangle is set to O, and a three-dimensional Cartesian coordinate system is established with O as the origin.
  • the X axis is on the bottom side of the triangle, the direction of the X axis is called the length direction or the horizontal direction or the lateral direction, and the X axis can be extended in the left and right direction (the imaging direction of the camera image plane in the Z coordinate direction, the X axis The direction indicated by the arrow is to the right, defined here as the positive direction);
  • the Y axis is perpendicular to the bottom edge of the triangle and is in the same plane as Oa, Ob, Oc, and the direction of the Y axis is called the width direction or the vertical direction or
  • the Y-axis can extend in the up-and-down direction (the direction indicated by the arrow of the Y-axis is upward, here defined as the positive direction); the Z-axis is perpendicular to the plane of the focus and parallel
  • a plane parallel to the plane in which the focal points Oa, Ob, Oc is located and the distance from the plane is the focal length f
  • three image planes a, b, and c are disposed on the plane, according to the imaging principle
  • the center of each image plane is the passing point of the optical axis of the corresponding camera
  • the plane coordinate origin of each image plane is the upper left corner of the image plane, which are Oa', Ob', and Oc', respectively.
  • the two-dimensional Cartesian coordinate axis is set to Oa'XaYa
  • the two-dimensional Cartesian coordinate axis of the b image plane is set to Ob'XbYb
  • the two-dimensional Cartesian coordinate axis of the c image plane is set to Oc'XcYc
  • the length of each image plane is set (
  • the target surface length of the image sensor corresponding to each camera is u
  • the width (corresponding to the target surface width of the image sensor of each camera) is v (not shown).
  • the horizontal scanning lines defining the three image planes a, b, and c are all parallel to the X-axis direction, and the horizontal scanning lines of the two image planes a and b are vertically aligned in the vertical direction. And the centers of the three image planes a, b, and c coincide with the passing points of the optical axes of the corresponding cameras.
  • the set of planar three-phase units adopting the above arrangement proposed in the present invention is a minimum basic three-dimensional measuring unit.
  • a multi-camera planar array three-dimensional measurement system composed of more cameras (positive integers greater than or equal to 3) can be constructed according to similar rules and methods, and a three-dimensional coordinate system corresponding to each group of three-phase units can be constructed according to similar rules and methods. Or build a unified three-dimensional coordinate system.
  • the matching method and the measuring method of the embodiment of the present invention are applicable to the matching and measurement of the three-dimensional coordinate position of the viewpoint of the object to be imaged which can be imaged on the three cameras of a group of three-phase units arranged as above.
  • FIG. 2 is a schematic diagram of a certain viewpoint P (Px, Py, Pz) of the object and the point on the a image plane, the b image plane, and the c image plane.
  • P the point
  • the P point is The imaging points on the a, b, and c image planes are Pa (Pax, Pay), Pb (Pbx, Pby), and Pc (Pcx, Pcy), respectively.
  • Figure 3 is a schematic plan view of any of the viewpoints P on a planar three-phase unit.
  • the straight line PaPb composed of the three imaging points Pa, Pb, and Pc is parallel to the straight line OaOb
  • the straight line PaPc is parallel to the straight line OaOc
  • the straight line PbPc is parallel to the straight line ObOc
  • the triangular PaPbPc is similar to the triangular OaObOc.
  • FIG. 4 is a schematic perspective view of any of the viewpoints P on the camera groups of a and b.
  • the projections of the P point, the Pa point, and the Pb point on the OXZ coordinate plane where the two focal points Oa and Ob are located are: P' point, Pa' point, and Pb' point.
  • the OaOb line is parallel to the imaging plane composed of two image planes, the triangle consisting of P point, Oa point and Ob point intersects the plane where the two image planes a and b intersect, and the intersection line is the line PaPb, so The PaPb line is parallel to the OaOb line.
  • the PbPc line is parallel to the ObOc line
  • the PaPc line is parallel to the OaOc line
  • the triangle PaPbPc is similar to the triangle OaObOc.
  • FIG. 5 is a projection view of the imaging point of the P point on the a, b camera group on the OXZ coordinate plane.
  • m is the length of OaOb
  • u is the target length of each image sensor
  • P' point, Pa' point and Pb' point are the projection points of P point, Pa point and Pb point on the OXZ coordinate plane, Pax and Pbx respectively The coordinate values of the Pa' point and the Pb' point in the X-axis direction on the a image plane and the b image plane.
  • Pax is larger than Pbx, that is, the horizontal offset of the image point of the P point in the a image plane relative to the coordinate origin of the a image plane is greater than the image point of the point in the b image plane relative to the b image plane.
  • the horizontal offset of the coordinate origin is larger than Pbx, that is, the horizontal offset of the image point of the P point in the a image plane relative to the coordinate origin of the a image plane is greater than the image point of the point in the b image plane relative to the b image plane.
  • Matching condition 1) When a viewpoint is imaged on three image planes of a group of three-phase units, the image points on the two image planes adjacent in the lateral direction satisfy: imaging of the viewpoint on the left image plane The image point on the right image plane corresponding to the viewpoint is located on the same line parallel to the horizontal coordinate axis, and the horizontal offset of the image point on the left image plane relative to the coordinate origin of the left image plane A horizontal offset greater than the coordinate origin of the imaged point on the right image plane relative to the right image plane.
  • Matching condition 2 When one viewpoint is imaged on three image planes of a group of three-phase units, the image point on the image plane in the other height direction is located at two sides on the left and right sides adjacent to the image plane The intersection of the imaged point on the image plane and the line connecting the imaged point on the image plane, and the two lines are respectively parallel to the center point on the image plane where the viewpoint is located and the corresponding horizontal axis The line connecting the center points on the left and right image planes.
  • Matching condition 3 A necessary condition for a corresponding matching point group on the three image planes of a group of three-phase units is satisfied that the three imaging points form a triangle, and the three sides of the triangle and the three image planes The three sides of the triangle consisting of the focus are parallel, and the triangle is similar to the triangle consisting of the focus of the three image planes.
  • the corresponding imaging points on the other two images can be matched according to the matching 3), and the corresponding matching points on the three images found constitute a pair. Match the point group.
  • FIG. 5 is a projection view of the imaging point of the P point on the a, b camera group on the OXZ coordinate plane.
  • m is the length of OaOb
  • u is the target length of each image sensor
  • P' point, Pa' point, and Pb' point are projection points of P point, Pa point, and Pb point on the OXZ coordinate plane, respectively.
  • Pax and Pbx are coordinate values of the Pa' point and the Pb' point in the X-axis direction on the a image plane and the b image plane, respectively.
  • Pax is larger than Pbx, that is, the horizontal offset of the image point of the P point in the a image plane relative to the coordinate origin of the a image plane is greater than the image point of the point in the b image plane relative to the b image plane.
  • the horizontal offset of the coordinate origin is larger than Pbx, that is, the horizontal offset of the image point of the P point in the a image plane relative to the coordinate origin of the a image plane is greater than the image point of the point in the b image plane relative to the b image plane.
  • Fig. 6 is a schematic diagram of pixel projection angle and P point Y value calculation; horizontal matching of known values of Px and Pz requires further derivation of the value of Py, and we introduce the concept of pixel projection angle ⁇ .
  • the pixel projection angle ⁇ refers to an image acquisition system whose fixed focal length and parameters are known, and the angle between the line of each pixel on the image and its focus and the center of the optical axis is called the pixel projection angle.
  • a pixel projection line is a ray consisting of an extension of a line connecting a focus to the pixel.
  • the angle ⁇ as shown in FIG. 6 is the pixel projection angle
  • OaP is the pixel projection line.
  • A1 Taking one image plane of a lateral image plane of a group of three-phase units as a base image plane, for a feature point on the base image plane, finding a position adjacent to the base image plane in the lateral direction according to the matching condition 1) All matching points on the image plane that match the feature point;
  • a matching point group is formed, Matching points matching the pair of matching points on the third image plane, according to matching conditions 2) and 3), removing matching point groups that do not satisfy the matching condition, forming a matching matching condition existing on all three image planes Match point group;
  • A4 Perform full-image verification on the corresponding matching points on the other two image planes of all matching point groups, and determine that the three image planes correspond to the unique matching point group of the same viewpoint.
  • step a1 for a feature point on the base image plane, all matching points matching the feature point on the image plane adjacent to the base image plane in the lateral direction are found according to the matching condition 1);
  • Matching condition 1) when a viewpoint is imaged on three image planes of a group of three-phase units, the image points on the two image planes adjacent in the lateral direction satisfy: the viewpoint is on the left image plane The imaged point and the imaged point on the corresponding right image plane of the viewpoint are located on the same line parallel to the horizontal coordinate axis, and the imaged point on the left image plane is relative to the left image The horizontal offset of the coordinate origin of the plane is greater than the horizontal offset of the origin of the image on the right image plane relative to the origin of the right image plane.
  • step a2 finding matching points matching the pair of matching points on the third image plane according to the matching condition 2) and the matching condition 3);
  • the matching condition 2) is that when one viewpoint is imaged on three image planes of a group of three-phase units, the image point on the image plane in the other height direction is located, and the left and right sides adjacent to the image plane The intersection of the imaged point on the two image planes of the side with the line connecting the imaged point on the image plane, and the two lines are respectively parallel to the center point on the image plane where the viewpoint is located and corresponding a line connecting the center points on the left and right image planes on the horizontal axis;
  • the matching condition 3) is that a necessary condition that a viewpoint is satisfied in a corresponding matching point group on three image planes of a group of three-phase units is that the three imaging points form a triangle, and three sides and three sides of the triangle The three sides of the triangle formed by the focal point of the plane are parallel, and the triangle is similar to the triangle composed of the focal points of the three image planes.
  • step a4 if the same matching point appearing on the other two non-base image planes corresponds to two or more matching point groups on the base image plane, the point is used as the base point, according to the matching conditions 1), 2) and 3), and steps a1, a2, and a3, sequentially search for corresponding matching points on the base image plane, and determine whether there is another matching point group on the matching point on the base image plane, and if so, match the match
  • the relationship is deleted until the matching relationship is the only matching relationship corresponding to the two pairs; if there is always a point-to-multipoint, and the two planes on the non-base image plane have completed the full-image verification, then The matching relationship is all deleted.
  • the embodiment relates to a three-camera planar array feature point measuring method.
  • Embodiment 1 After the image acquisition is completed, the method used in Embodiment 1 is used to find a unique matching point group corresponding to all feature points in the base image plane;
  • step b2 According to the spatial position coordinates of the viewpoint obtained in step b2, three-dimensional point cloud data is formed, and a three-dimensional point cloud graphic is created to reproduce the three-dimensional stereoscopic image.
  • Step b1 includes:
  • the matching point group is formed according to the matching condition. 2) and matching condition 3) finding matching points matching the pair of matching points on the third image plane, and removing matching point groups that do not satisfy the matching condition according to the matching condition, forming three image planes a matching point group that satisfies the matching condition 3);
  • B14 Perform full-image verification on the matching points on the other two image planes of all matching point groups. If two or more matching point groups on the base image plane corresponding to the same matching point appear on the graph, the point is For the base point, according to the matching principle and the above steps, the corresponding matching points are sequentially searched on the base image plane, and it is determined whether there is another matching point group at the matching point, and if so, the matching relationship is deleted. Until the matching relationship is the only matching relationship corresponding to the two pairs, if there is always a point-to-multipoint, and the two planes on the non-base image plane have completed the full-image verification, the matching relationship is obtained. delete all.
  • step b2
  • the three-phase unit is set in an arbitrary triangle, including a camera located at the left side of the spatial coordinate horizontal axis, b camera at the right position, and c camera at the upper position of the vertical coordinate axis.
  • the focal points Oa, Ob, Oc of the three cameras are On the same plane, and the focal lengths of the three cameras are the same as f, the optical axes of the three cameras are perpendicular to the plane, the viewpoint P (Px, Py, Pz) and the point in the a image plane, the b image plane and the c
  • the imaging points on the image plane are Pa (Pax, Pay), Pb (Pbx, Pby), Pc (Pcx, Pcy), m is the length of OaOb, and u is the target length of the image sensor on each image plane, v
  • the P' point, the Pa' point, and the Pb' point are the projection points of the P point, the Pa point, and the Pb point on the
  • the pixel projection angle is ⁇ a for the a image and the pixel projection angle ⁇ b for the b image:
  • the embodiment relates to a three-dimensional detecting device using the three-camera planar array feature point detecting method described in Embodiment 2.
  • the above three-dimensional measuring device comprises at least one set of three digital cameras with completely identical technical parameters.
  • the focal points of the three cameras form a plane, and the optical axis of the camera is perpendicular to the plane.
  • the invention includes a three-lens lens array, a first substrate 100 and a second substrate 200 that are parallel to each other, a control circuit board 300 disposed on a side of the second substrate 200 facing away from the first substrate 100, and a first substrate 100 and a second substrate. 200 and the fixed rear case 400 of the control circuit board 300.
  • the three-lens lens array is mounted on the first substrate 100.
  • the axes of the respective lenses 101 of the three-camera lens array are perpendicular to the plane of the first substrate 100, and the second substrate 200 is disposed with the array of photosensitive elements 201 toward the surface of the first substrate 100.
  • Each of the photosensitive elements 201 in the array of photosensitive elements 201 is disposed in one-to-one correspondence with each of the lenses 101.
  • the control circuit board 300 is connected to the second substrate 200 through the fixing post 301.
  • a first power input port 401 and a first data communication structure 402 are disposed on the fixed rear case 400.
  • the array of the photosensitive elements 201 composed of the plurality of photosensitive elements 201 is integrated on the same substrate, thereby ensuring the geometric accuracy of the mutual position of the photosensitive elements 201, the parallelism of the optical axes of the lenses 101, and the geometric accuracy of the array structure, and can also be reduced as needed.
  • the distance between adjacent lenses 101. The above measures ensure that the accuracy and accuracy of the three-dimensional image are higher, the measurement range is larger, the integration of the camera is higher, and the processing is facilitated, so that the device becomes a true three-dimensional image acquisition device.
  • the lens 101 described above is an industrial lens 101, and each lens 101 and a corresponding photosensitive element 201 constitute an industrial digital camera, thereby forming an industrial digital camera array formed by a plurality of industrial digital cameras, wherein
  • the digital camera is hereinafter referred to as a camera
  • the industrial digital camera array is hereinafter referred to as a camera array.
  • the lens 101 of each of the lens 101 arrays has the same model and type, and may be a fixed focus lens 101 or a zoom lens 101.
  • the functions of the respective photosensitive elements 201 in the array of the photosensitive elements 201 are to convert the optical signals captured by the lens 101 into electrical signals for subsequent image processing, which may be CCD (Charge-coupled Device) or CMOS. (Complementary Metal Oxide Semiconductor, complementary metal oxide semiconductor), of course, can also be used for other known electrical devices that can be applied to convert an optical signal into an electrical signal during image acquisition.
  • CCD Charge-coupled Device
  • CMOS complementary Metal Oxide Semiconductor,
  • the second substrate 200 may be a PCB (Printed Circuit Board) circuit board, and then the respective photosensitive elements 201 are soldered to the second substrate 200 in accordance with the layout requirements of the lens 101 described above.
  • the dedicated photosensitive member 201 can be used to position the fixture to ensure the positional accuracy of each photosensitive member 201 after welding.
  • the photosensitive element 201 may be integrally formed with the second substrate 200 and directly formed as a photosensitive array plate.
  • the scanning lines of the two photosensitive elements 201 at the bottom of the triangle formed by the array of the photosensitive elements 201 are horizontally parallel and completely aligned up and down, and at the same time, the scanning lines of the photosensitive elements 201 at the top of the array of photosensitive elements 201 and the two photosensitive portions at the bottom
  • the scanning lines of the elements 201 are parallel, and the three photosensitive elements 201 constitute an isosceles triangle with a bottom side as a base or a right triangle with a bottom side as a right angle side.
  • FIG. 11 and FIG. 12 As another structural form of the three-dimensional detecting device, please refer to FIG. 11 and FIG. 12 together.
  • the image generation circuit board 600 is provided with three micro lenses 603, a micro photosensitive element 601 corresponding to the three micro lenses 603, and an image generation circuit 604 disposed under the micro photosensitive element 601.
  • the image generation circuit board 600 is provided around the image generation circuit board 600.
  • Three miniature lenses 603 form a three-camera lens array.
  • a front case 500 and a rear case 900 are provided, and the image generation circuit board 600, the fixed substrate 700, and the image pickup circuit board 800 are integrated in a space surrounded by the front case 500 and the rear case 900.
  • the front housing 500 is provided with a lens hole 501 disposed in one-to-one correspondence with the micro lens 603, and the rear housing 900 is provided with a second power input port 901 and a second data communication interface 902.
  • the micro photosensitive elements 601 each include a CCD and/or a CMOS. According to the requirements of the field of view, measurement accuracy, and measurement speed, the micro photosensitive element 601 suitable for use can be selected, and then the corresponding image generation circuit board 600, the image generation circuit board 600, is designed. A PCB (Printed Circuit Board) circuit board can be used, and then the respective micro photosensitive elements 601 are soldered to the image generation circuit board 600 in accordance with the layout requirements of the aforementioned micro lens 603. When the welding method is adopted, in order to ensure the welding precision, the special photosensitive element 201 can be used to position the fixture to ensure the positional accuracy of each of the micro photosensitive elements 601 after welding. Further, the micro photosensitive element 601 may be integrally formed with the image forming circuit board 600 and directly formed as a miniature photosensitive array plate.
  • the scanning lines of the two micro photosensitive elements 601 at the bottom of the triangle formed by the array of the miniature photosensitive elements 601 are horizontally parallel and completely aligned up and down, and at the same time, the scanning lines of the micro photosensitive elements 601 on the top of the array of miniature photosensitive elements 601 and the bottom two
  • the scanning lines of the micro photosensitive elements 601 are parallel, and the three micro photosensitive elements 601 form an isosceles triangle with a bottom side as a base or a right triangle with a bottom side as a right angle side.

Abstract

一种三相机组特征点匹配方法、测量方法及三维检测装置，属于光学电子测量技术领域。其中测量方法包括以下步骤：b1.图像采集完成后，找出基像平面中的所有特征点对应的唯一性匹配点组；b2.根据步骤b1中获得的唯一性匹配点组的像坐标，计算被视点的空间位置坐标；b3.根据步骤b2中获得的被视点的空间位置坐标，形成三维点云数据，建立三维点云图形，重现三维立体图像。具备以下有益效果：第一、能够在三个相机上均成像的被视点的通用性、唯一性的匹配；第二、实现对被视物的三维测量；第三、快速实现三维感知和测量；第四、使用相机少，减少了应用成本。

Description

三相机组特征点匹配方法、测量方法及三维检测装置
本申请要求于2017年03月20日提交中国专利局的申请号为CN201710167851.9、名称为“三相机组特征点匹配方法、测量方法及三维检测装置”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本发明涉及光学电子测量技术领域，尤其是涉及一种三相机组特征点匹配方法、测量方法及三维检测装置。
背景技术
目前三维立体视觉测量一般采用线激光光截图测量技术或双目测量加结构光照明的方式，线激光或结构光在三维测量中得到广泛使用的主要原因是：通过线激光或结构光的指示使成像中明确了对应的匹配点，减少了匹配的歧义性，实现了确定且唯一的匹配。但是如果取消了线激光或结构光的指示，双目匹配就不能避免出现多点匹配的歧义性，从而不能满足测量要求，同时采用线激光或结构光只能对线激光或结构光成像部位进行测量，限定了该技术的适用范围，而且线激光或结构光的使用对被测物例如人会产生不良影响。
此外，目前双目匹配还经常采用一种在被视物表面贴标识点的方式，采用这种方式也是为了提高匹配的准确性。但是，在被视物表面贴标识点的方式存在需要提前对被测物进行人工处理和干预的缺点。
专利文件，“四相机组平面阵列特征点匹配方法及基于四相机组平面阵列特征点匹配方法的测量方法”，匹配方法包括以四个像平面中的一个像平面为基像平面，对基像平面上的一个特征点找出在横向方向上与该基像平面相邻的像平面上与该特征点匹配的所有匹配点；对于基像平面上的特征点找出在纵向方向上与该基像平面相邻的像平面上与该特征点匹配的所有匹配点；将找出的横纵两个方向上所有匹配点进行再匹配，找出所有子匹配点组；找出对角位置像平面上与基像平面上的特征点以及找出的所有子匹配点组对应的匹配点；确定四个像平面中对应于同一被视点的唯一性匹配点组。对于每组唯一性匹配点组，可根据该匹配点组的像坐标和相机系统本身的参数，计算被视点的三维空间坐标。在任何光照条件下只要采集的图像足够清晰对于任何在四相机组平面阵列的图像上成像且有一定的图像特征的被视物，采用完全相同的匹配方法和测量方法均可以实现被视物的三维测量。
四相机组平面阵列特征点匹配方法使用相机多，增加了应用成本，同时，由于相机多，也增加了加工难度和计算的复杂度，也影响了测量精度。
发明内容
有鉴于此，本发明实施例的目的在于提供一种三相机组特征点匹配方法、测量方法及三维检测装置；以降低现有的多相机匹配方法及测量方法的复杂度，简化空间尺寸计算过程，减小系统测量误差。
本发明提供了如下技术方案：
一种三相机组特征点匹配方法，包括以下步骤：
a1.以一组三相机组的横向像平面中的一个像平面为基像平面，对于所述基像平面上的一个特征点找出在横向方向上与该基像平面相邻的像平面上与该特征点匹配的所有匹配点；
a2.对于所述步骤a1中的所述基像平面上的所述特征点和对应的在横向方向上与所述基像平面相邻的像平面上与该特征点匹配的所有匹配点，组成的匹配点组，找出在第三个像平面上与该对匹配点组相匹配的匹配点，根据匹配条件，去除不满足匹配条件的匹配点组，形成由三个像平面上都存在的满足匹配条件的匹配点组；
a3.重复步骤a1和a2，找出基像平面上所有能够实现匹配的特征点及其对应的匹配点组，以及匹配点组中其它两个像平面上对应的匹配点；
a4.将所有匹配点组中位于其它两个非基像平面上对应的匹配点进行全图验证，确定出所述三个像平面对应于同一被视点的唯一性匹配点组。
更进一步地，
所述步骤a1中，对于所述基像平面上的一个特征点根据匹配条件1)找出在横向方向上与该基像平面相邻的像平面上与该特征点匹配的所有匹配点；
其中，
所述匹配条件1)，为一个被视点在一组三相机组的三个像平面上成像时，在横向方向上相邻的两个像平面上的成像点满足：该被视点在左像平面上的成像点与该被视点在对应的右像平面上的成像点位于平行于横向坐标轴的同一条直线上，并且该左像平面上的成像点相对于该左像平面的坐标原点的水平偏移量大于该右像平面上的成像点相对于该右像平面的坐标原点的水平偏移量。
更进一步地，
所述步骤a2中，根据匹配条件2)和匹配条件3)找出在第三个像平面上与该对匹配点组相匹配的匹配点；
其中，所述匹配条件2)为，一个被视点在一组三相机组的三个像平面上成像时，在另一个高度方向上的像平面上的成像点位于，与该像平面相邻的左右两侧的两个像平面上的成像点与该被视点在该像平面上的成像点的连线的交点上，且两条连线分别平行于该被视点所在该像平面上的中心点与对应的水平轴上两个左右像平面上的中心点的连线；
其中，所述匹配条件3)为，一个被视点在一组三相机组的三个像平面上对应的匹配点组满足的必要条件为该三个成像点组成一个三角形，该三角形的三条边与三个像平面的焦点组成的三角形的三条边平行，且该三角形与三个像平面的焦点组成的三角形相似。
更进一步地，
步骤a4中，如果在其他两个非基像平面上出现同一匹配点对应基像平面上的两个或多个匹配点组时，以该点为基点，按照匹配条件1)、2)和3)，以及步骤a1、a2和a3，在基像平面上依次寻找对应的匹配点，判断在该匹配点上是否还存在另外的匹配点组，如果有，就将本匹配关系删除，直到出现匹配关系为两两对应的唯一匹配关系为止；如果，一直存在一点对多点，且已经在非基像平面上的两个平面都完成了全图验证的情况下，就将该匹配关系全部删除。
一种基于三相机组特征点匹配方法的测量方法，
包括以下步骤：
b1.图像采集完成后，使用如上所述的三相机组特征点匹配方法找出所述基像平面中的所有特征点对应的唯一性匹配点组；
b2.根据所述步骤b1中获得的唯一性匹配点组的像坐标，计算被视点的空间位置坐标；
b3.根据所述步骤b2中获得的被视点的空间位置坐标，形成三维点云数据，建立三维点云图形，重现三维立体图像。
更进一步地，
所述步骤b1包括：
b11.以一组三相机组的横向像平面中的一个像平面为基像平面，对于所述基像平面上的一个特征点找出在横向方向上与该基像平面相邻的像平面上与该特征点匹配的所有匹配点；
b12.对于所述步骤b11中的所述基像平面上的所述特征点和对应的在横向方向上与所述基像平面相邻的像平面上与该特征点匹配的所有匹配点，组成的匹配点组，根据匹配条件2)和匹配条件3)找出在第三个像平面上与该对匹配点组相匹配的匹配点，根据匹配条件，去除不满足匹配条件的匹配点组，形成由三个像平面上都存在的满足匹配条件3)的匹配点组；
b13.重复步骤b11和b12，找出基像平面上所有能够实现匹配的特征点及其对应的匹配点组，以及匹配点组中位于其它两个像平面上对应的匹配点；
b14.将所有匹配点组中位于非基像平面的其它两个像平面上对应的匹配点进行全图验证，如果在图上出现同一匹配点对应基像平面上的两个或多个匹配点组时，以该点为基点，按照匹配原理以及以上步骤在基像平面上依次寻找对应的匹配点，判断在该匹配点上是否还存在另外的匹配点组，如果有，就将本匹配关系删除。直到出现匹配关系为两两对应的唯一匹配关系为止，如果，一直存在一点对多点，且已经在非基像平面上的两个平面都完成了全图验证的情况下，就将该匹配关系全部删除。
更进一步地，
所述步骤b2中，
三相机组以任意三角形设置，包括分别位于空间坐标水平轴两侧位置的a相机和b相机、位于垂直坐标轴上部位置的c相机，三个相机的焦点Oa、Ob、Oc在同一平面上，且三个相机的焦距相同设为f，三个相机的光轴均垂直于该平面，被视点P(Px，Py，Pz)和该点在a像平面、b像平面和c像平面上的成像点分别为Pa(Pax，Pay)、Pb(Pbx，Pby)、Pc(Pcx，Pcy)，m为OaOb的长度，u为各像平面上的图像传感器的靶面长度，v为各相机的图像传感器的靶面宽度，P＇点、Pa＇点和Pb＇点分别为P点、Pa点、Pb点在OXZ坐标平面上的投影点，Pax和Pbx分别为Pa＇点和Pb＇点在a像平面和b像平面上的X轴方向的坐标值；θa为a像平面像素投影角，θb为b像平面像素投影角，其中像素投影角θ(θa或θb)是指对于一个固定焦距和参数已知的图像采集系统，其图像上每个像素点与其焦点的连线与光轴中心的夹角叫做像素投影角；
P点的空间位置坐标的表达式为：
a，b像平面横向匹配时Px坐标计算公式：
Figure PCTCN2017084814-appb-000001
a，b像平面横向匹配时Pz坐标计算公式：
Figure PCTCN2017084814-appb-000002
由横向匹配已知的Px和Pz的值得出Py的值：
Figure PCTCN2017084814-appb-000003
Figure PCTCN2017084814-appb-000004
其中:
Figure PCTCN2017084814-appb-000005
Figure PCTCN2017084814-appb-000006
一种三维检测装置，采用了上述的测量方法，包括三相机镜头阵列。
更进一步地，
还包括相互平行的第一基板和第二基板、设置于所述第二基板背离所述第一基板一侧的控制电路板，以及用于容纳所述第一基板、所述第二基板和所述控制电路板的固定后壳；
所述三相机镜头阵列设置于第一基板上，所述三相机镜头阵列的各个镜头的轴线分别垂直于所述第一基板所在的平面，所述第二基板朝向所述第一基板的表面设置有感光元件阵列，所述感光元件阵列中的各个感光元件与各个所述镜头一一对应设置。
更进一步地，
所述感光元件阵列焊接在第二基板上或与第二基板一体成型，所述感光元件阵列形成的三角形底部的两个所述感光元件的扫描线水平平行且上下完全对齐，同时，所述感光元件阵列顶部的感光元件的扫描线与底部的两个所述感光元件的扫描线平行，且三个所述感光元件组成以底边为底的等腰三角形或者以底边为一个直角边的直角三角形。
更进一步地，
包括图像生成电路板、用于固定所述图像生成电路板的固定基板、以及位于所述固定基板后方的图像采集电路板；
所述图像生成电路板上设置有三个微型镜头、与三个所述微型镜头一一对应的微型感光元件、以及设置于所述微型感光元件下方的图像生成电路，三个微型镜头组成三相机镜头阵列。
更进一步地，
所述微型感光元件阵列焊接在图像生成电路板上或与图像生成电路板一体成型，所述微型感光元件阵列形成的三角形底部的两个所述微型感光元件的扫描线水平平行且上下完全对齐，同时，所述微型感光元件阵列顶部的微型感光元件的扫描线与底部的两个所述微型感光元件的扫描线平行，且三个所述微型感光元件组成以底边为底的等腰三角形或者以底边为一个直角边的直角三角形。
更进一步地，
根据被测物视场的范围，可以以所述三维检测装置为基本测量单元，形成多于三台相机的三维检测装置平面检测相机阵列。
本发明至少具备以下有益效果：
1.平面三相机组特征点匹配方法，能够根据被视物的一个被视点在一组平面三相机组的三个像平面上的成像点的位置，快速匹配出该被视点在三个像平面上对应的唯一性成像点组，实现对能够在三个相机上均成像的被视点的通用性、唯一性的匹配；
2.基于平面三相机组特征点匹配方法的测量方法，在任何光照条件下，只要采集的图像足够清晰，在未知被视物的情况下，对于任何被视物，采用完全相同的测量方法就可以实现对被视物的三维测量，并且该测量方法无需对视场做任何标定，其测量精度和分辨率只与测量系统有关，与被视物无关，能够完全实现自主性测量。
3.由于所述匹配方法和测量方法的通用性和可靠性，便于程序优化以及实现嵌入式级别和芯片级别的运算，从而快速实现三维感知和测量。
4.相对于四相机组平面阵列特征点匹配方法，该方法使用相机少，减少了应用成本，同时，由于相机少，也减少了加工难度和计算的复杂度，提高了测量精度。
附图说明
为了更清楚地说明本发明具体实施方式或现有技术中的技术方案，下面将对具体实施方式或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图是本发明的一些实施方式，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1是基于三相机组建立的空间坐标系的示意图；
图2是被视物的某个被视点P(Px，Py，Pz)和该点在a像平面、b像平面和c像平面上的成像示意图；
图3是任意被视点P在平面三相机组上的成像平面示意图；
图4是任意被视点P在a、b相机组上的成像立体示意图；
图5是P点在a、b相机组上的成像点在OXZ坐标平面上的投影图；
图6是像素投影角和P点Y值计算示意图；
图7是实施例3提供的其中一种三维检测装置的爆炸图；
图8是图7的总装示意图；
图9是图7中的固定后壳的结构示意图；
图10是图8的截面图；
图11是实施例3提供的另一种三维检测装置的爆炸图；
图12是图11的总装示意图；
图13是基于三相机组特征点匹配方法的测量方法的流程图。
图标：100-第一基板；101-镜头；102-安装通孔；200-第二基板；201-感光元件；300-控制电路板；400-固定后壳；401-第一电源输入口；402-第一数据通信接口；301-固定柱；500-前壳；501-镜头孔；600-图像生成电路板；601-微型感光元件；602-安装孔；603-微型镜头；604-图像生成电路；700-固定基板；800-图像采集电路板；900-后壳；901-第二电源输入口；902-第二数据通信接口。
具体实施方式
下面将结合附图对本发明的技术方案进行清楚、完整地描述，显然，所描述的实施例是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
在本发明的描述中，需要说明的是，术语“中心”、“上”、“下”、“左”、“右”、“竖直”、“水平”、“内”、“外”等指示的方位或位置关系为基于附图所示的方位或位置关系，仅是为了便于描述本发明和简化描述，而不是指示或暗示所指的装置或元件必须具有特定的方位、以特定的方位构造和操作，因此不能理解为对本发明的限制。此外，术语“第一”、“第二”、“第三”仅用于描述目的，而不能理解为指示或暗示相对重要性。
在本发明的描述中，需要说明的是，除非另有明确的规定和限定，术语“安装”、“相连”、“连接”应做广义理解，例如，可以是固定连接，也可以是可拆卸连接，或一体地连接；可以是机械连接，也可以是电连接；可以是直接相连，也可以通过中间媒介间接相连，可以是两个元件内部的连通。对于本领域的普通技术人员而言，可以具体情况理解上述术语在本发明中的具体含义。
下面结合附图对实施例1、实施例2和实施例3进行详细描述：
图1是基于三相机组建立的空间坐标系的示意图；图2是被视物的某个被视点P(Px，Py，Pz)和该点在a像平面、b像平面和c像平面上的成像示意图；图3是任意被视点P在平面三相机组上的成像平面示意图；图4是任意被视点P在a、b相机组上的成像立体示意图；图5是P点在a、b相机组上的成像点在OXZ坐标平面上的投影图；图6是像素投影角和P点Y值计算示意图；图7是实施例3提供的其中一种三维检测装置的爆炸图；图8是图7的总装示意图；图9是图7中的固定后壳900的结构示意图；图10是图8的截面图；图11是实施例3提供的另一种三维检测装置的爆炸图；图12是图11的总装示意图。
本发明中，成像点是指被视物的被视点在像平面(或图像)上对应像素位置的呈像，被视物的每个被视点在一组三相机组的三个像平面上分别对应有一个成像点；匹配点是指对于一个像平面上的成像点，在该像平面或其他像平面上寻找到的与该成像点满足某种匹配条件且与该成像点的图像特征(例如纹理、颜色或灰度等)相近的成像点。一个成像点对应的匹配点可能有一个或多个。
本发明中，特征点是指对应有一个或多个匹配点的如下成像点，该成像点具有区别于其他成像点的图像特征，例如不同的纹理、颜色或灰度值等，本发明实施例中不作具体限定，可以根据实际情况选择不同的图像特征作为判断依据。一般的，被测物边缘或纹理过渡区等位置的被视点对应的成像点都具有鲜明的图像特征，图13是基于三相机组特征点匹配方法的测量方法的流程图。
实施例1
本实施例涉及一种三相机组特征点匹配方法。
1、建立平面三相机组前像模型三维测量系统和三维坐标系。
为了便于阐述本发明实施例的平面三相机组特征点匹配方法和测量方法，需要建立一个满足要求的平面三相机组三维测量系统以及该平面三相机组三维测量系统所对应的三维坐标系，为了更方便应用本发明实施例的方法，使用前向投影模型，采用以如下方式建立的平面三相机组三维测量系统和对应的三维坐标系：
图1是基于三相机组建立的空间坐标系的示意图。
设三相机组的焦点在同一平面上以任意三角形设置，包括分别位于空间坐标水平轴两侧位置的a相机和b相机、位于垂直坐标轴上部位置的c相机，三个相机的配置完全相同，相同配置即三个相机的镜头101、图像传感器、焦距等参数完全相同。三个相机的焦点Oa、Ob、Oc 在同一平面上，且三个相机的焦距相同设为f，三个相机的光轴均垂直于该平面，Oa、Ob、Oc组成底边为m，高度为h，两个底角分别为α、β的任意三角形，三角形的底边的中点设为O，以O为原点建立一个三维直角坐标系。其中，X轴在三角形的底边上，X轴的方向称为长度方向或水平方向或横向方向，X轴可以在左右方向上延伸(以Z坐标方向为相机像平面成像观察方向，X轴的箭头所示的方向向右，在此定义为正方向)；Y轴垂直于三角形的底边，且与Oa、Ob、Oc在同一平面上，Y轴的方向称为宽度方向或竖直方向或纵向方向，Y轴可以在上下方向上延伸(Y轴的箭头所示的方向向上，在此定义为正方向)；Z轴垂直于焦点所在的平面且与三个相机的光轴方向平行，Z轴的方向称为深度方向(定义Z轴的箭头所示的方向为深度方向的正方向)。在Z轴的正方向上，设与焦点Oa、Ob、Oc所在的平面平行且与该平面的距离为焦距f的平面，在该平面上设a、b、c三个像平面，根据成像原理，每个像平面的中心为各自对应的相机的光轴的通过点，设每个像平面的平面坐标原点为该像平面的左上角，分别为Oa＇、Ob＇和Oc＇，a像平面的二维直角坐标轴设为Oa＇XaYa，b像平面的二维直角坐标轴设为Ob＇XbYb，c像平面的二维直角坐标轴设为Oc＇XcYc，并且设每个像平面的长度(对应于各相机的图像传感器的靶面长度)为u，宽度(对应于各相机的图像传感器的靶面宽度)为v(图中未示出)。同时，定义a、b、c三个像平面的横向扫描线都与X轴方向平行，且a、b二个像平面的横向扫描线在垂直方向上下对齐。且a、b、c三个像平面的中心与对应的相机的光轴的通过点重合。
本发明中所提出的采用以上布置结构的一组平面三相机组为一个最小的基本三维立体测量单元。可以根据类似的规则和方法构建由更多相机(大于等于3的正整数)组成的多相机组平面阵列三维测量系统，同时根据类似的规则和方法可以构建各组三相机组对应的三维坐标系，或者构建统一的三维坐标系。本发明的实施例的匹配方法和测量方法适用于能够在如上布置的一组三相机组的三个相机上均成像的被测物的被视点的三维坐标位置的匹配和测量。
2、平面三相机组特征点匹配条件的推导
图2是被视物的某个被视点P(Px，Py，Pz)和该点在a像平面、b像平面和c像平面上的成像示意图，参见图2，按照成像原理，P点在a、b、c像平面上的成像点分别为Pa(Pax，Pay)、Pb(Pbx，Pby)、Pc(Pcx，Pcy)。
图3是任意被视点P在平面三相机组上的成像平面示意图。
在三相机成像的像平面上，三个成像点Pa、Pb、Pc分别组成的直线PaPb和直线OaOb平行，直线PaPc和直线OaOc平行,直线PbPc和直线ObOc平行,三角形PaPbPc和三角形OaObOc相似。具体证明过程如下：
请参见图4，图4是任意被视点P在a、b相机组上的成像立体示意图。P点、Pa点、Pb点在两个焦点Oa、Ob所在的OXZ坐标平面上的投影分别为：P′点、Pa′点、Pb′点。
由于OaOb直线平行于a，b两个像平面组成的成像面，P点、Oa点和Ob点三点组成的三角形与a，b两个像平面所在的平面相交，交线为直线PaPb，所以PaPb直线与OaOb直线平行。
同理，也可以证明，PbPc直线与ObOc直线平行，PaPc直线与OaOc直线平行，同时，也证明了，三角形PaPbPc和三角形OaObOc相似。
请参见图5，图5是P点在a、b相机组上的成像点在OXZ坐标平面上的投影图。m为OaOb的长度，u为各图像传感器的靶面长度，P＇点、Pa＇点和Pb＇点分别为P点、Pa点、Pb点在OXZ坐标平面上的投影点，Pax和Pbx分别为Pa＇点和Pb＇点在a像平面和b像平面上的X轴方向的坐标值。
明显的，Pax大于Pbx，也就是说P点在a像平面中的成像点相对于a像平面的坐标原点的水平偏移量大于该点在b像平面中的成像点相对于b像平面的坐标原点的水平偏移量。
由以上证明，得出以下匹配条件：
匹配条件1)：一个被视点在一组三相机组的三个像平面上成像时，在横向方向上相邻的两个像平面上的成像点满足：该被视点在左像平面上的成像点与该被视点在对应的右像平面上的成像点位于平行于横向坐标轴的同一条直线上，并且该左像平面上的成像点相对于该左像平面的坐标原点的水平偏移量大于该右像平面上的成像点相对于该右像平面的坐标原点的水平偏移量。
由于PbPc直线与ObOc直线平行，PaPc直线与OaOc直线平行，三角形PaPbPc和三角形OaObOc相似，可以得出以下匹配条件：
匹配条件2)：一个被视点在一组三相机组的三个像平面上成像时，在另一个高度方向上的像平面上的成像点位于，与该像平面相邻的左右两侧的两个像平面上的成像点与该被视点在该像平面上的成像点的连线的交点上，且两条连线分别平行于该被视点所在该像平面上的中心点与对应的水平轴上两个左右像平面上的中心点的连线。
因而，根据前述的匹配条件1)和匹配条件2)可以得出以下匹配条件：
匹配条件3)：一个被视点在一组三相机组的三个像平面上对应的匹配点组满足的必要条件为该三个成像点组成一个三角形，该三角形的三条边与三个像平面的焦点组成的三角形的三条边平行，且该三角形与三个像平面的焦点组成的三角形相似。
按照以上匹配条件，当选定任意图像上的一个成像点后，其在另外两幅图像上对应的成像点，可按照匹配3)进行匹配，找到的三幅图像上对应的匹配点构成一对匹配点组。
3、平面三相机组特征点空间位置坐标计算。
请参见图5，图5是P点在a、b相机组上的成像点在OXZ坐标平面上的投影图。
参见图5，m为OaOb的长度，u为各图像传感器的靶面长度，P＇点、Pa＇点和Pb＇点分别为P点、Pa点、Pb点在OXZ坐标平面上的投影点，Pax和Pbx分别为Pa＇点和Pb＇点在a像平面和b像平面上的X轴方向的坐标值。
明显的，Pax大于Pbx，也就是说P点在a像平面中的成像点相对于a像平面的坐标原点的水平偏移量大于该点在b像平面中的成像点相对于b像平面的坐标原点的水平偏移量。
基于上述内容以及三角形相似原理，对于图5有：
Figure PCTCN2017084814-appb-000007
Figure PCTCN2017084814-appb-000008
根据①②推导出
Figure PCTCN2017084814-appb-000009
Figure PCTCN2017084814-appb-000010
Figure PCTCN2017084814-appb-000011
Figure PCTCN2017084814-appb-000012
根据①③推导出
Figure PCTCN2017084814-appb-000013
Figure PCTCN2017084814-appb-000014
将(Pax-Pbx)定义为横向匹配时b像平面上的成像点相对于a像平面上的成像点的横向偏移值，定义为Δx。那么，可以得出:
a，b像平面横向匹配时Px坐标计算公式(公式一)：
Figure PCTCN2017084814-appb-000015
a，b像平面横向匹配时Pz坐标计算公式(公式二)：
Figure PCTCN2017084814-appb-000016
按照双目视觉前述的两个公式，对于横向匹配我们可以直接得到X和Z的值，无法得出Y值。下面我们进一步推导出横向匹配的Y值的计算方法。
由图6所示，图6是像素投影角和P点Y值计算示意图；横向匹配已知Px和Pz的值，需要进一步推导得出Py的值，我们引入像素投影角θ的概念。
像素投影角θ是指对于一个固定焦距和参数已知的图像采集系统，其图像上每个像素与其焦点的连线与光轴中心的夹角叫做像素投影角。像素投影线是指焦点与该像素点的连线的延长线组成的一条射线。如附图6所示的θ角为像素投影角，OaP为像素投影线。当焦距和图像传感器确定后，其每个像素的像素投影角是个唯一确定的值，该像素上的呈像影像对应于像素投影线与被视物的交点图像。也就是说，每个像素的成像点对应的被视物都位于该像素投影线的延长线上。
由附图6，我们可以由横向匹配已知的Px和Pz的值，进一步推导得出Py的值，其中对于a图像设像素投影角为θa，对于b图像设像素投影角为θb，记为公式三：
Figure PCTCN2017084814-appb-000017
Figure PCTCN2017084814-appb-000018
其中，对于公式三：
Figure PCTCN2017084814-appb-000019
Figure PCTCN2017084814-appb-000020
4、平面三相机组特征点匹配方法。
根据前述匹配原理、匹配条件、匹配公式，得到一种平面三相机组特征点匹配方法，包括以下步骤：
a1.以一组三相机组的横向像平面中的一个像平面为基像平面，对于基像平面上的一个特征点，根据匹配条件1)找出在横向方向上与该基像平面相邻的像平面上与该特征点匹配的所有匹配点；
a2.对于步骤a1中的基像平面上的特征点和对应的在横向方向上与基像平面相邻的像平面上与该特征点匹配的所有匹配点，组成的匹配点组，找出在第三个像平面上与该对匹配点组相匹配的匹配点，根据匹配条件2)和3)，去除不满足匹配条件的匹配点组，形成由三个像平面上都存在的满足匹配条件的匹配点组；
a3.重复步骤a1和a2，找出基像平面上所有能够实现匹配的特征点及其对应的匹配点组，以及该匹配点组中在其它两个像平面上对应的匹配点；
a4.将所有匹配点组中其它两个像平面上对应的匹配点进行全图验证，确定出三个像平面对应于同一被视点的唯一性匹配点组。
步骤a1中，对于基像平面上的一个特征点根据匹配条件1)找出在横向方向上与该基像平面相邻的像平面上与该特征点匹配的所有匹配点；
其中，
匹配条件1)，为一个被视点在一组三相机组的三个像平面上成像时，在横向方向上相邻的两个像平面上的成像点满足：该被视点在左像平面上的成像点与该被视点在对应的右像平面上的成像点位于平行于横向坐标轴的同一条直线上，并且该左像平面上的成像点相对于该左像 平面的坐标原点的水平偏移量大于该右像平面上的成像点相对于该右像平面的坐标原点的水平偏移量。
步骤a2中，根据匹配条件2)和匹配条件3)找出在第三个像平面上与该对匹配点组相匹配的匹配点；
其中，匹配条件2)为，一个被视点在一组三相机组的三个像平面上成像时，在另一个高度方向上的像平面上的成像点位于，与该像平面相邻的左右两侧的两个像平面上的成像点与该被视点在该像平面上的成像点的连线的交点上，且两条连线分别平行于该被视点所在该像平面上的中心点与对应的水平轴上两个左右像平面上的中心点的连线；
其中，匹配条件3)为，一个被视点在一组三相机组的三个像平面上对应的匹配点组满足的必要条件为该三个成像点组成一个三角形，该三角形的三条边与三个像平面的焦点组成的三角形的三条边平行，且该三角形与三个像平面的焦点组成的三角形相似。
步骤a4中，如果出现在其它两个非基像平面上的同一匹配点对应基像平面上的两个或多个匹配点组时，以该点为基点，按照匹配条件1)、2)和3)，以及步骤a1、a2和a3，在基像平面上依次寻找对应的匹配点，判断在基像平面上的该匹配点上是否还存在另外的匹配点组，如果有，就将本匹配关系删除，直到出现匹配关系为两两对应的唯一匹配关系为止；如果，一直存在一点对多点，且已经在非基像平面上的两个平面都完成了全图验证的情况下，就将该匹配关系全部删除。
实施例2
本实施例涉及一种三相机平面阵列特征点测量方法。
包括以下步骤：
b1.图像采集完成后，使用实施例1中所采用的方法找出基像平面中的所有特征点对应的唯一性匹配点组；
b2.根据步骤b1中获得的唯一性匹配点组的像坐标，计算被视点的空间位置坐标；
b3.根据步骤b2中获得的被视点的空间位置坐标，形成三维点云数据，建立三维点云图形，重现三维立体图像。
步骤b1包括：
b11.以一组三相机组的横向像平面中的一个像平面为基像平面，对于基像平面上的一个特征点，根据匹配条件1)找出在横向方向上与该基像平面相邻的像平面上与该特征点匹配的所有匹配点；
b12.对于步骤b11中的基像平面上的特征点和对应的在横向方向上与基像平面相邻的像平面上与该特征点匹配的所有匹配点，组成的匹配点组，根据匹配条件2)和匹配条件3)找出在第三个像平面上与该对匹配点组相匹配的匹配点，根据匹配条件，去除不满足匹配条件的匹配点组，形成由三个像平面上都存在的满足匹配条件3)的匹配点组；
b13.重复步骤b11和b12，找出基像平面上所有能够实现匹配的特征点及其对应的匹配点组，以及匹配点组中位于其它两个像平面上对应的匹配点；
b14.将所有匹配点组中其它两个像平面上对应的匹配点进行全图验证，如果在图上出现同一匹配点对应基像平面上的两个或多个匹配点组时，以该点为基点，按照匹配原理以及以上步骤在基像平面上依次寻找对应的匹配点，判断在该匹配点上是否还存在另外的匹配点组，如果有，就将本匹配关系删除。直到出现匹配关系为两两对应的唯一匹配关系为止，如果，一直存在一点对多点，且已经在非基像平面上的两个平面都完成了全图验证的情况下，就将该匹配关系全部删除。
步骤b2中，
三相机组以任意三角形设置，包括分别位于空间坐标水平轴左侧位置的a相机、右侧位置的b相机、位于垂直坐标轴上部位置的c相机，三个相机的焦点Oa、Ob、Oc在同一平面上，且三个相机的焦距相同设为f，三个相机的光轴均垂直于该平面，被视点P(Px，Py，Pz)和该点在a像平面、b像平面和c像平面上的成像点分别为Pa(Pax，Pay)、Pb(Pbx，Pby)、Pc(Pcx，Pcy)，m为OaOb的长度，u为各像平面上的图像传感器的靶面长度，v为各相机的图像传感器的靶面宽度，P＇点、Pa＇点和Pb＇点分别为P点、Pa点、Pb点在OXZ坐标平面上的投影点，Pax和Pbx分别为Pa＇点和Pb＇点在a像平面和b像平面上的X轴方向的坐标值；θ为像素投影角，其中像素投影角θ是指对于一个固定焦距和参数已知的图像采集系统，其图像上每个像素与其焦点的连线与光轴中心的夹角叫做像素投影角；
P点的空间位置坐标的表达式为：
a，b像平面横向匹配时Px坐标计算公式：
Figure PCTCN2017084814-appb-000021
a，b像平面横向匹配时Pz坐标计算公式：
Figure PCTCN2017084814-appb-000022
由横向匹配已知的Px和Pz的值得出Py的值：
其中对于a图像设像素投影角为θa，对于b图像设像素投影角为θb：
Figure PCTCN2017084814-appb-000023
Figure PCTCN2017084814-appb-000024
其中：
Figure PCTCN2017084814-appb-000025
Figure PCTCN2017084814-appb-000026
实施例3
本实施例涉及一种采用了实施例2中述及的三相机平面阵列特征点检测方法的三维检测装置。
其中，上述的三维测量装置，至少包括一组由三台技术参数完全一致的数码相机组成，该测量系统中，三台相机的焦点组成一个平面，且相机的光轴垂直于该平面。
本实施例的可选方案中，作为三维检测装置的其中一种结构形式，请一并参照图7至图10。
包括三相机镜头阵列、相互平行的第一基板100和第二基板200、设置于第二基板200背离第一基板100一侧的控制电路板300，以及用于容纳第一基板100、第二基板200和控制电路板300的固定后壳400。
三相机镜头阵列安装在第一基板100上，三相机镜头阵列的各个镜头101的轴线分别垂直于第一基板100所在的平面，第二基板200朝向第一基板100的表面设置有感光元件201阵列，感光元件201阵列中的各个感光元件201与各个镜头101一一对应设置。另外，请参照图10，控制电路板300通过固定柱301与第二基板200连接。请参照图9，在固定后壳400上设置有第一电源输入口401和第一数据通信结构402。
将多个感光元件201组成的感光元件201阵列集成在同一块基板上，因此能够保证感光元件201相互位置的几何精度，以及镜头101光轴的平行以及阵列结构的几何精度，也可以根据需要缩小相邻的镜头101之间的距离。以上措施，保证了三维立体图像精确度和准确度更高，测量范围更大，相机的集成度更高，方便加工制作，使该装置成为真正意义上的三维图像采集装置。
较为优选地，以上述及的镜头101为工业镜头101，每个镜头101与对应的感光元件201组成一个工业数字相机，从而形成由多个工业数字相机形成的工业数字相机阵列，其中，工业 数字相机以下简称相机，工业数字相机阵列以下简称相机阵列，上述镜头101阵列中的各个镜头101型号、类型完全相同，其可以为定焦镜头101，也可以为变焦镜头101。上述感光元件201阵列中的各个感光元件201的作用是将镜头101捕获到的光信号转变为电信号，以便于后续图像处理，其可以为CCD(Charge-coupled Device，图像传感器)，也可以CMOS(Complementary Metal Oxide Semiconductor，互补金属氧化物半导体)，当然，也可以为其他公知的能够应用在图像获取过程中实现光信号转变为电信号的电器件。
第二基板200可以采用PCB(Printed Circuit Board，印制电路板)电路板的方式，然后将各个感光元件201按照前述的镜头101的布设要求焊接在第二基板200上。采用焊接方式时，为了保证焊接精度，可使用专用的感光元件201定位卡具，以保证焊接后各感光元件201的位置精度。另外，感光元件201也可以和第二基板200一体成型，直接为感光阵列板。
上述可选方案中，感光元件201阵列形成的三角形底部的两个感光元件201的扫描线水平平行且上下完全对齐，同时，感光元件201阵列顶部的感光元件201的扫描线与底部的两个感光元件201的扫描线平行，且三个感光元件201组成以底边为底的等腰三角形或者以底边为一个直角边的直角三角形。
作为三维检测装置的另一种结构形式，请一并参照图11和图12。
包括图像生成电路板600、用于固定图像生成电路板600的固定基板700、以及位于固定基板700后方的图像采集电路板800；
图像生成电路板600上设置有三个微型镜头603、与三个微型镜头603一一对应的微型感光元件601、以及设置于微型感光元件601下方的图像生成电路604，图像生成电路板600四周设置有安装孔602。三个微型镜头603组成三相机镜头阵列。
另外，还设置有前壳500和后壳900，图像生成电路板600、固定基板700和图像采集电路板800集成于由前壳500和后壳900围设而成的空间内。前壳500上设置有与微型镜头603一一对应设置的镜头孔501，后壳900上设置有第二电源输入口901和第二数据通信接口902。
微型感光元件601均包括CCD和/或CMOS，根据视场范围、测量精度和测量速度等要求，可以选择适合使用的微型感光元件601，然后设计对应的图像生成电路板600，图像生成电路板600可以采用PCB(Printed Circuit Board，印制电路板)电路板的方式，然后将各个微型感光元件601按照前述的微型镜头603的布设要求焊接在图像生成电路板600上。采用焊接方式时，为了保证焊接精度，可使用专用的感光元件201定位卡具，保证焊接后各微型感光元件601的位置精度。另外，微型感光元件601也可以和图像生成电路板600一体成型，直接为微型感光阵列板。
更进一步地，微型感光元件601阵列形成的三角形底部的两个微型感光元件601的扫描线水平平行且上下完全对齐，同时，微型感光元件601阵列顶部的微型感光元件601的扫描线与底部的两个微型感光元件601的扫描线平行，且三个微型感光元件601组成以底边为底的等腰三角形或者以底边为一个直角边的直角三角形。
最后应说明的是：以上各实施例仅用以说明本发明的技术方案，而非对其限制；尽管参照前述各实施例对本发明进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分或者全部技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本发明各实施例技术方案的范围。

Claims (13)

  1. 一种三相机组特征点匹配方法，其特征在于，包括以下步骤：
    a1.以一组三相机组的横向像平面中的一个像平面为基像平面，对于所述基像平面上的一个特征点找出在横向方向上与该基像平面相邻的像平面上与该特征点匹配的所有匹配点；
    a2.对于所述步骤a1中的所述基像平面上的所述特征点和对应的在横向方向上与所述基像平面相邻的像平面上与该特征点匹配的所有匹配点，组成的匹配点组，找出在第三个像平面上与该对匹配点组相匹配的匹配点，根据匹配条件，去除不满足匹配条件的匹配点组，形成由三个像平面上都存在的满足匹配条件的匹配点组；
    a3.重复步骤a1和a2，找出基像平面上所有能够实现匹配的特征点及其对应的匹配点组，以及匹配点组中其它两个像平面上对应的匹配点；
    a4.将所有匹配点组中位于其它两个非基像平面上对应的匹配点进行全图验证，确定出所述三个像平面对应于同一被视点的唯一性匹配点组。
  2. 根据权利要求1所述的三相机组特征点匹配方法，其特征在于，
    所述步骤a1中，对于所述基像平面上的一个特征点根据匹配条件1)找出在横向方向上与该基像平面相邻的像平面上与该特征点匹配的所有匹配点；
    其中，
    所述匹配条件1)，为一个被视点在一组三相机组的三个像平面上成像时，在横向方向上相邻的两个像平面上的成像点满足：该被视点在左像平面上的成像点与该被视点在对应的右像平面上的成像点位于平行于横向坐标轴的同一条直线上，并且该左像平面上的成像点相对于该左像平面的坐标原点的水平偏移量大于该右像平面上的成像点相对于该右像平面的坐标原点的水平偏移量。
  3. 根据权利要求2所述的三相机组特征点匹配方法，其特征在于，
    所述步骤a2中，根据匹配条件2)和匹配条件3)找出在第三个像平面上与该对匹配点组相匹配的匹配点；
    其中，所述匹配条件2)为，一个被视点在一组三相机组的三个像平面上成像时，在另一个高度方向上的像平面上的成像点位于，与该像平面相邻的左右两侧的两个像平面上的成像点与该被视点在该像平面上的成像点的连线的交点上，且两条连线分别平行于该被视点所在该像平面上的中心点与对应的水平轴上两个左右像平面上的中心点的连线；
    其中，所述匹配条件3)为，一个被视点在一组三相机组的三个像平面上对应的匹配点组满足的必要条件为该三个成像点组成一个三角形，该三角形的三条边与三个像平面的焦点组成的三角形的三条边平行，且该三角形与三个像平面的焦点组成的三角形相似。
  4. 根据权利要求3所述的三相机组特征点匹配方法，其特征在于，
    步骤a4中，如果在其它两个非基像平面上出现同一匹配点对应基像平面上的两个或多个匹配点组时，以该点为基点，按照匹配条件1)、2)和3)，以及步骤a1、a2和a3，在基像平面上依次寻找对应的匹配点，判断在该匹配点上是否还存在另外的匹配点组，如果有，就将本匹配关系删除，直到出现匹配关系为两两对应的唯一匹配关系为止；如果，一直存在一点对多点，且已经在非基像平面上的两个平面都完成了全图验证的情况下，就将该匹配关系全部删除。
  5. 一种基于三相机组特征点匹配方法的测量方法，其特征在于，
    包括以下步骤：
    b1.图像采集完成后，使用权利要求1-4任一项所述的匹配方法找出所述基像平面中的所有特征点对应的唯一性匹配点组；
    b2.根据所述步骤b1中获得的唯一性匹配点组的像坐标，计算被视点的空间位置坐标；
    b3.根据所述步骤b2中获得的被视点的空间位置坐标，形成三维点云数据，建立三维点云图形，重现三维立体图像。
  6. 根据权利要求5所述的测量方法，其特征在于：
    所述步骤b1包括：
    b11.以一组三相机组的横向像平面中的一个像平面为基像平面，对于所述基像平面上的一个特征点找出在横向方向上与该基像平面相邻的像平面上与该特征点匹配的所有匹配点；
    b12.对于所述步骤b11中的所述基像平面上的所述特征点和对应的在横向方向上与所述基像平面相邻的像平面上与该特征点匹配的所有匹配点，组成的匹配点组，根据匹配条件2)和匹配条件3)找出在第三个像平面上与该对匹配点组相匹配的匹配点，根据匹配条件，去除不满足匹配条件的匹配点组，形成由三个像平面上都存在的满足匹配条件3)的匹配点组；
    b13.重复步骤b11和b12，找出基像平面上所有能够实现匹配的特征点及其对应的匹配点组，以及匹配点组中位于其它两个像平面上对应的匹配点；
    b14.将所有匹配点组中位于非基像平面的其它两个像平面上对应的匹配点进行全图验证，如果在图上出现同一匹配点对应基像平面上的两个或多个匹配点组时，以该点为基点，按照匹配原理以及以上步骤在基像平面上依次寻找对应的匹配点，判断在该匹配点上是否还存在另外的匹配点组，如果有，就将本匹配关系删除，直到出现匹配关系为两两对应的唯一匹配关系为止，如果，一直存在一点对多点，且已经在非基像平面上的两个平面都完成了全图验证的情况下，就将该匹配关系全部删除。
  7. 根据权利要求5所述的测量方法，其特征在于：
    所述步骤b2中，
    三相机组以任意三角形设置，包括分别位于空间坐标水平轴两侧位置的a相机和b相机以 及位于垂直坐标轴上部位置的c相机，三个相机的焦点Oa、Ob和Oc在同一平面上，且三个相机的焦距相同设为f，三个相机的光轴均垂直于该平面，被视点P(Px，Py，Pz)和该点在a像平面、b像平面和c像平面上的成像点分别为Pa(Pax，Pay)、Pb(Pbx，Pby)和Pc(Pcx，Pcy)，m为OaOb的长度，u为各像平面上的图像传感器的靶面长度，v为各相机的图像传感器的靶面宽度，P＇点、Pa＇点和Pb＇点分别为P点、Pa点和Pb点在OXZ坐标平面上的投影点，Pax和Pbx分别为Pa＇点和Pb＇点在a像平面和b像平面上的X轴方向的坐标值；θ为a像平面的像素投影角(同理，可以据此推导出有关b像平面像素投影角的对应公式)，其中像素投影角θ是指对于一个固定焦距和参数已知的图像采集系统，其图像上每个像素点与其焦点的连线与光轴中心的夹角叫做像素投影角；
    P点的空间位置坐标的表达式为：
    a，b像平面横向匹配时Px坐标计算公式：
    Figure PCTCN2017084814-appb-100001
    a，b像平面横向匹配时Pz坐标计算公式：
    Figure PCTCN2017084814-appb-100002
    由横向匹配已知的Px和Pz的值得出Py的值：
    Figure PCTCN2017084814-appb-100003
    Figure PCTCN2017084814-appb-100004
    其中:
    Figure PCTCN2017084814-appb-100005
    Figure PCTCN2017084814-appb-100006
  8. 一种采用了如权利要求6-7任一项所述的测量方法的三维检测装置，其特征在于，包括三相机镜头阵列。
  9. 根据权利要求8所述的三维检测装置，其特征在于，
    还包括相互平行的第一基板和第二基板及设置于所述第二基板背离所述第一基板一侧的控制电路板，以及用于容纳所述第一基板、所述第二基板和所述控制电路板的固定后壳；
    所述三相机镜头阵列设置于第一基板上，所述三相机镜头阵列的各个镜头的轴线分别垂直于所述第一基板所在的平面，所述第二基板朝向所述第一基板的表面设置有感光元件阵列，所述感光元件阵列中的各个感光元件与各个所述镜头一一对应设置。
  10. 根据权利要求9所述的三维检测装置，其特征在于，
    所述感光元件阵列焊接在第二基板上或与第二基板一体成型，所述感光元件阵列形成的三角形底部的两个所述感光元件的扫描线水平平行且上下完全对齐，同时，所述感光元件阵列顶部的感光元件的扫描线与底部的两个所述感光元件的扫描线平行，且三个所述感光元件组成以底边为底的等腰三角形或者以底边为一个直角边的直角三角形。
  11. 根据权利要求8所述的三维检测装置，其特征在于，
    包括图像生成电路板、用于固定所述图像生成电路板的固定基板、以及位于所述固定基板后方的图像采集电路板；
    所述图像生成电路板上设置有三个微型镜头、与三个所述微型镜头一一对应的微型感光元件、以及设置于所述微型感光元件下方的图像生成电路，三个微型镜头组成三相机镜头阵列。
  12. 根据权利要求11所述的三维检测装置，其特征在于，
    所述微型感光元件阵列焊接在图像生成电路板上或与图像生成电路板一体成型，所述微型感光元件阵列形成的三角形底部的两个所述微型感光元件的扫描线水平平行且上下完全对齐，同时，所述微型感光元件阵列顶部的微型感光元件的扫描线与底部的两个所述微型感光元件的扫描线平行，且三个所述微型感光元件组成以底边为底的等腰三角形或者以底边为一个直角边的直角三角形。
  13. 根据权利要求8所述的三维检测装置，其特征在于，
    根据被测物视场的范围，能够以所述三维检测装置为基本测量单元，形成多于三台相机的三维检测装置平面检测相机阵列。
PCT/CN2017/084814 2017-03-20 2017-05-18 三相机组特征点匹配方法、测量方法及三维检测装置 WO2018171031A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201710167851.9A CN106813595B (zh) 2017-03-20 2017-03-20 三相机组特征点匹配方法、测量方法及三维检测装置
CN201710167851.9 2017-03-20

Publications (1)

Publication Number Publication Date
WO2018171031A1 true WO2018171031A1 (zh) 2018-09-27

Family

ID=59116292

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/084814 WO2018171031A1 (zh) 2017-03-20 2017-05-18 三相机组特征点匹配方法、测量方法及三维检测装置

Country Status (2)

Country Link
CN (1) CN106813595B (zh)
WO (1) WO2018171031A1 (zh)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109813251B (zh) * 2017-11-21 2021-10-01 蒋晶 用于三维测量的方法、装置以及系统
CN109443199B (zh) * 2018-10-18 2019-10-22 天目爱视（北京）科技有限公司 基于智能光源的3d信息测量系统
CN110148182B (zh) * 2019-05-08 2022-12-30 云南大学 一种标定摄像机的方法、存储介质、运算器和系统
CN112639390A (zh) * 2019-11-21 2021-04-09 北京机电研究所有限公司 用于三维尺寸的动态测量装置及其测量方法

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2005142765A (ja) * 2003-11-05 2005-06-02 Sony Corp 撮像装置及び方法
WO2012078126A1 (en) * 2010-12-08 2012-06-14 Thomson Licensing System and method for trinocular depth acquisition with triangular sensor
CN103743352A (zh) * 2013-12-18 2014-04-23 西安交通大学 一种基于多相机匹配的三维变形测量方法
CN105627926A (zh) * 2016-01-22 2016-06-01 尹兴 四像机组平面阵列特征点三维测量系统及测量方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2005142765A (ja) * 2003-11-05 2005-06-02 Sony Corp 撮像装置及び方法
WO2012078126A1 (en) * 2010-12-08 2012-06-14 Thomson Licensing System and method for trinocular depth acquisition with triangular sensor
CN103743352A (zh) * 2013-12-18 2014-04-23 西安交通大学 一种基于多相机匹配的三维变形测量方法
CN105627926A (zh) * 2016-01-22 2016-06-01 尹兴 四像机组平面阵列特征点三维测量系统及测量方法

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
GUAN, YEPENG ET AL.,: "Trinocular Stereo Matching Algorithm Based on Correlation between Disparities and Baseline", CHINESE JOURNAL OF SCIENTIFIC INSTRUMENT, vol. 27, no. 11, 30 November 2006 (2006-11-30) *
REN, ZHONG ET AL.,: "Research on Trinocular Matching Methods for Stereo Vision", ROBOT, vol. 23, no. 5, 30 September 2001 (2001-09-30), pages 456 - 458 *

Also Published As

Publication number Publication date
CN106813595B (zh) 2018-08-31
CN106813595A (zh) 2017-06-09

Similar Documents

Publication Publication Date Title
US10491883B2 (en) Image capturing device
US10001369B2 (en) Object-point three-dimensional measuring system using multi-camera array, and measuring method
WO2018171031A1 (zh) 三相机组特征点匹配方法、测量方法及三维检测装置
JP2007536652A5 (zh)
JP6209833B2 (ja) 検査用具、検査方法、ステレオカメラの生産方法及びシステム
JP2958458B1 (ja) 多眼画像センサ
JP2015049200A (ja) 計測装置、方法及びプログラム
US9594028B2 (en) Method and apparatus for determining coplanarity in integrated circuit packages
TW201348786A (zh) 鏡頭模組及其組裝方法
KR20230096057A (ko) 라이트 필드 카메라를 기반으로 한 결함 레이어링 검출 방법과 시스템 및 검출 생산 라인
JP2011099779A (ja) 距離画像取得装置及び距離画像取得処理方法
TW201326735A (zh) 寬度量測方法及系統
TWI571099B (zh) 深度估測裝置及方法
JP5493900B2 (ja) 撮像装置
JP2021135300A (ja) 基板計測システム及び基板計測方法
CN105928484A (zh) 基于双目视觉的电梯导轨自动测量系统
US20230408253A1 (en) Three-dimensional scanner having sensors with overlapping fields of view
JP2014041074A (ja) 画像処理装置及び検査装置
TWI597966B (zh) 全視角三維掃描器
JP2010187214A (ja) 撮像装置
JP4900951B2 (ja) 生産ラインの検査システム及び検査方法
CN112804515A (zh) 全向立体视觉的摄像机配置系统及摄像机配置方法
CN206905713U (zh) 三维检测装置
CN216905123U (zh) 一种图像获取装置和机器人
CN212163540U (zh) 全向立体视觉的摄像机配置系统

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17901804

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17901804

Country of ref document: EP

Kind code of ref document: A1