CN107330936B - Monocular vision-based double-circular marker positioning method and system - Google Patents

Monocular vision-based double-circular marker positioning method and system Download PDF

Info

Publication number
CN107330936B
CN107330936B CN201710474508.9A CN201710474508A CN107330936B CN 107330936 B CN107330936 B CN 107330936B CN 201710474508 A CN201710474508 A CN 201710474508A CN 107330936 B CN107330936 B CN 107330936B
Authority
CN
China
Prior art keywords
marker
position sensor
plane
circular
point
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201710474508.9A
Other languages
Chinese (zh)
Other versions
CN107330936A (en
Inventor
吴剑
苏宇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shenzhen Graduate School Tsinghua University
Original Assignee
Shenzhen Graduate School Tsinghua University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Shenzhen Graduate School Tsinghua University filed Critical Shenzhen Graduate School Tsinghua University
Priority to CN201710474508.9A priority Critical patent/CN107330936B/en
Publication of CN107330936A publication Critical patent/CN107330936A/en
Application granted granted Critical
Publication of CN107330936B publication Critical patent/CN107330936B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/70—Determining position or orientation of objects or cameras
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/13—Edge detection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/80—Analysis of captured images to determine intrinsic or extrinsic camera parameters, i.e. camera calibration
    • G06T7/85—Stereo camera calibration
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30244—Camera pose

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Length Measuring Devices By Optical Means (AREA)

Abstract

The invention discloses a monocular vision-based quick positioning method for double-circle markers, which can detect markers in real time and calculate the plane normal vector and the center position of a circle marker of the markers in a camera coordinate system. Aiming at two groups of solutions in the marker positioning adopting the single circular characteristic, the invention provides a double-circle discrimination method to remove false solutions, and simultaneously, in order to ensure the marker positioning real-time property, the invention combines a PSD position sensor to reduce the detection range of the marker in a camera plane, thereby reducing the total time of the positioning algorithm of the circular marker.

Description

Monocular vision-based double-circular marker positioning method and system
Technical Field
The invention relates to the field of object tracking and marking, in particular to a monocular vision-based double-circular marker positioning method and system.
Background
In the traditional object tracking, a moving object is tracked on the basis of pixel change in a camera plane, if the space position needs to be acquired, a binocular camera is required to be matched, the object tracking based on a circular marker cannot acquire object surface information, and the object position cannot be accurately calculated. Meanwhile, due to the complexity of calculation, the traditional method is difficult to realize real-time positioning.
In the prior art, referring to fig. 3, when spatial positioning is performed by a single circular ring, since two sets of solutions satisfying an equation are provided when a projection profile curve in a camera plane is back-projected, the position of a circular mark in space cannot be determined.
Disclosure of Invention
Aiming at the technical problems of object tracking of the circular mark, the invention provides a method and a system for accurately acquiring the space position of a tracked object and the normal vector information of the surface of the tracked object by the mark.
In order to achieve the purpose, the invention adopts the following technical scheme:
a method for positioning a double-circle mark based on monocular vision comprises the following steps:
(1) image pre-processing, comprising: histogram normalization, image denoising and image distortion removal;
(2) identifying the mark, namely, acquiring a positive sample and a negative sample in advance for training by adopting an image classification mode to obtain a characteristic parameter, and judging whether the image in the area is the mark or not according to the characteristic parameter;
(3) contour detection, which is used for extracting an elliptical contour to remove a background contour, setting a threshold value to extract the elliptical contour in a circularity mode in an area determined by a PSD position sensor, refining the contour in a sub-pixel interpolation mode for improving contour accuracy, and respectively processing the contour according to an extracted contour curve ellipse 1 and an extracted contour curve ellipse 2 to obtain a parameter 1 and a parameter 2, wherein the parameters 1 and 2 comprise two groups of directions obtained by resolving the ellipses 1 and 2, two groups of corresponding circular ring center coordinates and real point coordinates on a circular ring plane;
(4) removing false solution, projecting ellipse from the outer circle and the inner circle of the marked plane to obtain two real points respectively marked as e1,e2There is a certain linear direction in the marking plane
Figure BDA0001327910130000011
The two planes obtained by solving the outer ring are expressed as
Figure BDA0001327910130000012
Wherein
Figure BDA0001327910130000013
The center point of the circle and the normal vector of the plane are respectively satisfied with the equation solution. The criterion of the normal vector of the real plane is as follows:
Figure BDA0001327910130000014
a rapid marking positioning system with double circular features based on monocular vision comprises a development board, a marker, a two-dimensional PSD position sensor, an infrared L ED, an A/D converter, a single chip microcomputer and a camera, wherein the infrared L ED is fixed on the plane of the marker, L ED is fixed in the center of the marker, the two-dimensional PSD position sensor can form a point with higher energy on a receiving plane of the two-dimensional PSD position sensor through an optical lens arranged in the front of the two-dimensional PSD position sensor, the output voltage of the two-dimensional PSD position sensor changes along with the movement of the energy point, the AD converter converts the output voltage of the PSD position sensor from an analog signal to a digital signal, and the single chip microcomputer controls the data acquisition process of.
Preferably, the relative position of the camera and the two-dimensional PSD position sensor or other position detection means can be obtained by experimental or theoretical calculation.
Preferably, the tag has attached to it an item detectable by a PSD position sensor, in one implementation of the invention infrared L ED is used.
Preferably, the PSD two-dimensional position sensor model may be PSD 100-SPB.
Preferably, the AD converter model may be AD 7076.
Preferably, the single chip microcomputer is STM 32.
As a preference, the development board number may be Jetson TK 1.
Compared with the prior art, the technical scheme of the invention has the following advantages:
in the method, in the step (1), in order to improve the positioning accuracy, the optical lens is adjusted in a mode of combining the PSD position sensor and the optical lens within the detection range of the marker, so that the focal length of the camera is equal to the distance between the position sensor and the optical lens, after an analog signal output by the position sensor is obtained, the analog signal is converted into a digital signal through an A/D converter, and then the digital signal is processed by a single chip microcomputer to output infrared L ED (infrared ray) to project the position on the plane of the PSD position sensor.
In the method, the position relation between the PSD position sensor coordinate system and the camera coordinate system in the step (2) can be obtained by a direct measurement or calibration method, namely a rotation matrix and a translation matrix between the two coordinate systems are measured by an instrument, or a method similar to binocular camera calibration can be adopted to determine the relation between the two coordinate systems, the position of infrared L ED under the PSD position sensor coordinate system is known, the position under the camera coordinate system can be obtained by marking the position and solving, and a series of points are measured to obtain the rotation matrix and the displacement vector between the two coordinate systems by a least square method.
In the above method, the circularity method or other contour detection methods that can be used to distinguish different shapes can be used to extract the projected contour curve of the marker in step (3).
In the method, the step (4) provides a novel mark based on the double-circle feature, so that two groups of solution problems in single-circle calibration can be eliminated.
Drawings
Fig. 1 is a diagram of a positioning system.
Fig. 2 is a flow chart of a positioning algorithm.
FIG. 3 is a schematic diagram of parameter solving.
FIG. 4 is a diagram illustrating the use of PSD.
Detailed Description
The present invention is described in further detail below with reference to the attached drawings.
Referring to fig. 1, the rapid marking and positioning system based on the monocular vision circular feature comprises a development board, a marker, a two-dimensional PSD position sensor, infrared L ED, an A/D converter, a single chip microcomputer and a camera.
The infrared L ED is fixed on the plane of the marker, L ED is fixed in the center of the marker for convenient marking and rough positioning, the two-dimensional PSD position sensor can form a point with higher energy on the receiving plane through an optical lens arranged in the front of the two-dimensional PSD position sensor, the output voltage of the two-dimensional PSD position sensor changes along with the movement of the energy point, the AD converter converts the output voltage of the PSD position sensor from an analog signal to a digital signal, and the single chip microcomputer controls the data acquisition process of the AD converter and sends the sampling data to the development platform.
The camera is directly controlled by the development platform to acquire the image of the marker projected in the camera plane and finally locate the spatial position of the marker by combining the information of the PSD position sensor according to the algorithm provided by the invention, and the specific locating method and the implementation process are described with reference to fig. 2 and 3. Compared with a position sensor, the camera is a main device, namely the system can work normally even without the position sensor, but the marker positioning real-time performance of the system can be greatly improved by using the position sensor, and the specific principle and the operation method are illustrated in the figure 4.
Referring to fig. 2, the circular feature based indicia presented by the present invention is in the form of a double circle. The image preprocessing comprises the following steps: histogram normalization, image denoising and image distortion removal. The mark identification adopts an image classification mode, namely, positive and negative samples are collected in advance for training to obtain characteristic parameters, and whether the image in the area is a mark or not is judged according to the characteristic parameters. The contour detection is used for extracting an elliptical contour to remove a background contour, the background contour is greatly different from a circular projection contour in a mark in an area determined by a PSD position sensor, a circularity mode is adopted, a threshold value is set to extract the elliptical contour, and a sub-pixel interpolation mode is adopted to refine the contour in order to improve contour accuracy. And respectively processing the extracted contour curve ellipses 1 and 2 to obtain parameters 1 and 2, wherein the parameters 1 comprise two groups of directions obtained by resolving the ellipses 1, two groups of corresponding circular ring center coordinates and a real point coordinate on a circular ring plane, and the parameter 2 is similar to the parameter 1. And obtaining a vector in the marking plane by using the coordinates of the real points in the parameter 1 and the parameter 2, wherein the real normal vector in the parameter 1 is necessarily vertical to the vector of the plane, so that the vector is multiplied by the two groups of directions in the parameter 1 respectively, the direction with the right angle included angle is the real direction, and the corresponding coordinate of the real direction is the central coordinate of the real circular ring. The positioning result is the real normal vector in the parameter 1 and the circular center coordinate corresponding to the real normal vector.
Referring to fig. 3, when spatial localization is performed by a single circular ring, the position of the circular marker in space cannot be determined because the projected contour curve in the camera plane has two sets of solutions satisfying the equation when back-projected. However, the coordinates of a real point on the marker plane and two sets of normal vectors meeting the equation solution can be obtained by utilizing the back projection of a single circular marker, and the process implemented in the step is the process of solving the real point and the two sets of normal vectors of the marker plane.
Recording the back projection elliptic cone equation in the camera coordinate system as [ X [ ]cYcZc]Q[XcYcZc]T0, where Q is a coefficient matrix, which is orthogonally decomposed to obtain
Figure BDA0001327910130000031
(wherein Λ ═ diag { λ)1,λ2,λ3}，|λ1|＞|λ2Corresponding orthogonal matrix denoted as T1). Let [ X ]cYcZc]T＝T1*[X Y Z]TThe above elliptic cone equation can be transformed into a standard coordinate space with the corresponding equation of λ1X2+λ2Y2+λ3Z20. The lower plane Π of the camera coordinate system corresponds to the OXZ plane in the standard coordinate space. According to the method of r.safaee-Rad, et al, we can obtain two sets of labeled planes pi 1, pi 2 that correspond to the solutions of the equations in the camera coordinate system.
The solving process corresponding to the real points in fig. 3 is that L1 equals pi 1 ∩ pi, L2 equals pi 2 ∩ pi, and e equals L1 ∩L 2, if L equals pi 1 ∩ pi 2, there is e ∈L, and all points on the straight line L are real points, but considering the actual solving process, the e point is still selected as the real point.
The text adopts a double-circle ring form, so that two real points can be obtained by projecting ellipses from the outer circle and the inner circle on the marking plane and respectively marked as e1,e2There is a certain linear direction in the marking plane
Figure BDA0001327910130000032
The two planes obtained by solving the outer ring are expressed as
Figure BDA0001327910130000033
Wherein
Figure BDA0001327910130000034
The center point of the circle and the normal vector of the plane are respectively satisfied with the equation solution. The criterion of the normal vector of the real plane is as follows:
Figure BDA0001327910130000035
the rationality of the criterion, and its range of applicability, are demonstrated below.
And (3) proving that: without loss of generality, it is assumed that the camera planes are elliptical back-projected circular marks on the planes Π 1, Π 2, respectively, as shown in fig. 3. According to the theorem that the normal vector of a plane is perpendicular to any straight line on the plane, the normal vector must exist
Figure BDA0001327910130000041
So that
Figure BDA0001327910130000042
This is true.
I. Special cases are as follows: plane II 1 coincides with II 2, i.e.
Figure BDA0001327910130000043
Or
Figure BDA0001327910130000044
At this time
Figure BDA0001327910130000045
And is
Figure BDA0001327910130000046
If both are true, the criterion is invalid, which corresponds to the situation that the projection profile of the outer ring in the camera plane is a ring. However, the back projection circular marks are overlapped, and the real point corresponding to the center of the outer ring of the mark is overlapped with the false point, which shows that the mark center point and the plane normal vector obtained under the condition that the criterion is invalid are real.
II. General case: plane II 1 not coinciding with II 2, i.e.
Figure BDA0001327910130000047
And is
Figure BDA0001327910130000048
This time
Figure BDA0001327910130000049
And
Figure BDA00013279101300000410
only one is true, the true plane normal vector can be uniquely determined, and thus the circular marker center. Obtaining the real center point P (i is 1 or 2) of the marked outer ring, and the normal vector of the real plane
Figure BDA00013279101300000411
In conclusion, the uncertainty of the back projection direction of the circular mark can be effectively removed under the conditions that the criterion is effective and invalid by adopting a constraint mode that real points on the circular mark form a straight line which is perpendicular to a normal vector of a mark plane, so that the space parameters of the mark are obtained.
Referring to fig. 4, in order to realize rapid detection of the marker in the camera plane, a two-dimensional PSD position sensor is used in combination with the camera. The positioning process and algorithm thereof are briefly described as follows.
First, positioning of marker
The search of the projection area of the marker in the camera plane is divided into two steps, 1, solving the projection area corresponding to infrared L ED in the camera plane through distance constraint by using infrared L ED projection point coordinates obtained on a PSD position sensor, and 2, determining the projection area of the marker in the camera plane according to the infrared L ED projection area and the position relation between the infrared L ED and the marker in the space.
According to the projective geometrical knowledge, one point on a projection plane is back-projected to form a spatial straight line, so that the infrared L ED spatial point coordinates cannot be determined only by the point coordinates obtained on the PSD position sensor, but the infrared L ED spatial coordinates can be determined within a certain range by adding a distance constraint condition, and then a projection region of the infrared L ED in the camera plane is obtained, and an infrared L ED projection image must exist in the region.
For convenience of problem description, the method comprises three steps of I and projection relations, firstly, regarding the distance between infrared L ED and a PSD position sensor as a constant to obtain a general relation between a projection point of infrared L ED on a PSD position sensor plane and a projection point of the infrared L ED on a camera plane, II and L ED projection areas, determining the distance between infrared L ED and the PSD position sensor in a certain range, determining a projection area of infrared L ED on the camera plane according to the projection relation in I, III and a marker projection area, determining a marker projection area according to the infrared L ED projection area in II, and sequentially explaining the following steps.
I. Projective relationships
In combination with the three-dimensional representation of FIG. 4(a), point E in the figure represents the mid-infrared L ED, o in spacep-xpypzpIs a PSD position sensor coordinate system, oc-xcyczcThe camera coordinate system, the relationship between the two coordinate systems is: [ x ] ofcyczc]＝[xpypzp]+[t00](where t is a constant). Point P (X)p,Yp) Point E is in the PSD position sensor plane, i.e. op-xpypThe projection point on the plane can be obtained by the scheme, and the point C is the point E on the camera plane, namely oc-xcycThe inner projection point on the plane is the parameter to be solved in the step. The focal length of the front end optical lens of the PSD position sensor is fp＝|opFpI, the focal length of the front optical lens of the camera is fc＝|ocFc|，fpAnd fcThis is obtained experimentally when the distance | EP | ═ d between infrared L ED and the PSD position sensor coordinates is constant.
If the coordinates of the point C in the camera plane can be obtained from the point P and the distance d, the corresponding relationship between the point P and the point C can be obtained. For ease of problem description and solution, two projection views are obtained from the spatial view as shown in fig. 4(b) (c). The point C corresponds to the point C on the y-z axis projection diagram of FIG. 4(b)yI.e. point C at ycThe coordinate on the axis is that the point C corresponds to the point C on the projection diagram of the x-z axis of FIG. 4(C)xI.e. point C is at xcOn-axis coordinates, the point C projection coordinate problem in the camera plane is converted into a solution line segment | ocCyI and I ocCxThe length of l. (Note: F in the figure)cAnd FpIs a general representation only and does not indicate that fpAnd fcThe size relationship between the points E and E, and the similar solving process exists when the point E is at other positions)
Solving for | o in conjunction with FIG. 4(b), respectivelycCySolving for | o in conjunction with FIG. 4(c)cCx|。
|ocCyI solution
P in FIG. 4(b)y、EyAnd CyPoint P, point E and point C are at opzpypProjected point on plane, point eyIs EyAt ypProjection point on axis, | opPy|＝|Yp|，
Figure BDA0001327910130000051
Figure BDA0001327910130000052
Figure BDA0001327910130000053
According to the triangle similarity principle:
Figure BDA0001327910130000054
can be solved to obtain:
Figure BDA0001327910130000055
for convenience of description, let us note | ocCyG (d) indicates the projected point at o where the coordinates of the point in the camera plane are obtained, corresponding to point P on the PSD position sensor planecycThe upper coordinate is a function of the distance d.
|ocCxI solution
P in FIG. 4(c)x、ExAnd CxPoint P, point E and point C are at opzpxpProjected point on plane, point exIs ExAt xpProjection point on axis, | ocop|＝t，|opPx|＝|Xp|，
Figure BDA0001327910130000056
Figure BDA0001327910130000057
According to the triangle similarity principle:
Figure BDA0001327910130000061
can be solved to obtain:
Figure BDA0001327910130000062
|ocex|＝|ocop|-|opex|＝t-|opex|
according to the triangle similarity principle:
Figure BDA0001327910130000063
can be solved to obtain:
Figure BDA0001327910130000064
remember | ocCxH (d) indicates the projected point o at which the coordinates of the point in the camera plane are obtained, corresponding to point P on the PSD position sensor planecxcThe upper coordinate is a function of the distance d.
The above process shows that the coordinate P of a certain point obtained on the PSD position sensing plane can be expressed as (h), (d), g (d)) under the distance constraint condition, and the coordinate C of the point obtained on the camera plane can be expressed as (h), (d), g (d)).
To simplify the measurement and description process without loss of generality, infrared L ED is placed here in the center of a circular marker, assuming that the maximum projected dimension of the marker in the camera plane is W × H (how this dimension is obtained will be described in III) at distance d, and the projected image area of the marker in the camera plane is present inside an area of dimension W × H centered on point C.
Two, L ED projection area
Consider the general case where a marker moves within a range in the PSD position sensor coordinate system, namely d ∈ (d)min,dmax) Corresponding to point C (X)c,Yc) At ocxcOn-axis coordinate Xc∈(min(h(dmin),h(dmax)),max(h(dmin),h(dmax)))＝(Xmin,Xmax) At ocycOn-axis coordinate Yc∈(min(g(dmin),g(dmax)),max(g(dmin),g(dmax)))＝(Ymin,Ymax) Let Δ W ═Xmax-Xmin，ΔH＝Ymax-Ymin，
Figure BDA0001327910130000065
The infrared L ED projected point in the camera plane must exist as a point
Figure BDA0001327910130000066
Centered within a region of size Δ W × Δ H.
Third, marker projection area
Within a certain distance range, the maximum projection size of the marker in the camera plane is the size when the marker is closest to the origin of the PSD position sensor coordinate system and the projection outer contour in the camera plane is a circular ring, the size is marked as W × H, and the projection point of the infrared L ED in II in the camera plane exists in the form of a point
Figure BDA0001327910130000067
Centered, in a region of size Δ W × Δ H, and infrared L ED is centered in the marker, the projected image of the marker in the camera plane must exist at a point
Figure BDA0001327910130000068
Centered within a region of size (W + Δ W) × (H + Δ H).
Through the three steps, the projection area of the marker in the camera plane can be obtained by the point coordinates on the PSD position sensor plane and combining the distance constraint condition.
The present invention has been described in detail with reference to the specific embodiments, and the detailed description should not be construed as limiting the invention to only those descriptions. Any modifications and equivalents of the invention in light of the above teachings and in light of the common general knowledge in the art are intended to be included within the scope of the appended claims.

Claims (8)

1. A method for positioning a double-circle mark based on monocular vision is characterized by comprising the following steps:
(1) image pre-processing, comprising: histogram normalization, image denoising and image distortion removal;
(2) identifying the mark, namely, acquiring a positive sample and a negative sample in advance for training by adopting an image classification mode to obtain a characteristic parameter, and judging whether the image in the area is the mark or not according to the characteristic parameter;
(3) contour detection, which is used for extracting an elliptical contour to remove a background contour, setting a threshold value to extract the elliptical contour by adopting a circularity mode in an area determined by a PSD position sensor, refining the contour by adopting a sub-pixel interpolation mode to improve contour accuracy, and respectively processing according to extracted contour curve ellipses 1 and 2 to obtain parameters 1 and 2, wherein the parameters 1 comprise two groups of directions obtained by resolving the ellipses 1, two groups of corresponding circular center coordinates and real point coordinates located on a circular plane, and the parameters 2 comprise two groups of directions obtained by resolving the ellipses 2, two groups of corresponding circular center coordinates and real coordinates located on the circular plane;
(4) removing false solution, projecting ellipse from the outer circle and the inner circle of the marked plane to obtain two real points respectively marked as e1,e2There is a certain linear direction in the marking plane
Figure FDA0002482318380000011
The two planes obtained by solving the outer ring are expressed as
Figure FDA0002482318380000012
In which the sum of pi,
Figure FDA0002482318380000013
the criterion of setting a real plane normal vector according to the fact that a straight line formed by real points on the circular mark is perpendicular to a marked plane normal vector is as follows:
Figure FDA0002482318380000014
(i ═ 1 or 2).
2. A rapid positioning system for a mark with a double-circular feature based on monocular vision is characterized by comprising a development board, a marker, a two-dimensional PSD position sensor, an infrared L ED, an A/D converter, a single chip microcomputer and a camera, wherein the infrared L ED is fixed on the plane of the marker, L ED is fixed at the center of the marker, the two-dimensional PSD position sensor forms a point with higher energy on a receiving plane of the two-dimensional PSD position sensor through an optical lens placed in the front of the two-dimensional PSD position sensor, the output voltage of the two-dimensional PSD position sensor changes along with the movement of an energy point, the AD converter converts the output voltage of the PSD position sensor from an analog signal to a digital signal, the single chip microcomputer controls the data acquisition process of the AD converter and sends sampling data to a development platform, and the development platform is used for achieving.
3. The system for rapidly positioning the markers of the double-circular features based on monocular vision as claimed in claim 2, wherein the relative positions of the camera and the two-dimensional PSD position sensor or other position detection devices are obtained by experimental or theoretical calculation.
4. The rapid marker locating system based on monocular vision double-circle feature of claim 2, wherein the infrared L ED detected by PSD position sensor is attached to the marker.
5. The monocular vision based rapid marker locating system of a double-circle feature of claim 2, wherein the PSD two-dimensional position sensor model is PSD 100-SPB.
6. The rapid marker locating system based on monocular vision double-circular feature of claim 2, wherein the AD converter model is AD 7076.
7. The monocular vision based rapid marker locating system for double circular features of claim 2, wherein the single chip microcomputer model is STM 32.
8. The monocular vision based marker rapid positioning system of a double circle feature of claim 2, wherein the development board model is Jetson TK 1.
CN201710474508.9A 2017-06-21 2017-06-21 Monocular vision-based double-circular marker positioning method and system Active CN107330936B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710474508.9A CN107330936B (en) 2017-06-21 2017-06-21 Monocular vision-based double-circular marker positioning method and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710474508.9A CN107330936B (en) 2017-06-21 2017-06-21 Monocular vision-based double-circular marker positioning method and system

Publications (2)

Publication Number Publication Date
CN107330936A CN107330936A (en) 2017-11-07
CN107330936B true CN107330936B (en) 2020-07-17

Family

ID=60194199

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710474508.9A Active CN107330936B (en) 2017-06-21 2017-06-21 Monocular vision-based double-circular marker positioning method and system

Country Status (1)

Country Link
CN (1) CN107330936B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108548482B (en) * 2018-03-01 2019-11-15 华南农业大学 A kind of pattern recognition device and method based on monocular vision and laser sensing
CN110021035B (en) * 2019-04-12 2020-12-11 哈尔滨工业大学 Marker of Kinect depth camera and virtual marker tracking method based on marker

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101804521A (en) * 2010-04-15 2010-08-18 中国电子科技集团公司第四十五研究所 Galvanometer system correction device and correction method thereof
CN102915535A (en) * 2012-08-23 2013-02-06 深圳大学 Method and system for correcting circle center deviation of round mark points during camera projection transformation
CN105957090A (en) * 2016-05-17 2016-09-21 中国地质大学（武汉） Monocular vision pose measurement method and system based on Davinci technology

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101804521A (en) * 2010-04-15 2010-08-18 中国电子科技集团公司第四十五研究所 Galvanometer system correction device and correction method thereof
CN102915535A (en) * 2012-08-23 2013-02-06 深圳大学 Method and system for correcting circle center deviation of round mark points during camera projection transformation
CN105957090A (en) * 2016-05-17 2016-09-21 中国地质大学（武汉） Monocular vision pose measurement method and system based on Davinci technology

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
"融合距离图像和彩色图像的圆位姿测量方法";陈欣 等;《计算机应用与软件》;20160630;第33卷(第6期);第190-194页 *

Also Published As

Publication number Publication date
CN107330936A (en) 2017-11-07

Similar Documents

Publication Publication Date Title
CN110332887B (en) Monocular vision pose measurement system and method based on characteristic cursor points
CN108932475A (en) A kind of Three-dimensional target recognition system and method based on laser radar and monocular vision
CN101609504B (en) Method for detecting, distinguishing and locating infrared imagery sea-surface target
CN108022264B (en) Method and equipment for determining camera pose
CN106826815A (en) Target object method of the identification with positioning based on coloured image and depth image
CN110334678A (en) A kind of pedestrian detection method of view-based access control model fusion
CN107392929B (en) Intelligent target detection and size measurement method based on human eye vision model
CN110763204B (en) Planar coding target and pose measurement method thereof
CN113624225B (en) Pose resolving method for mounting engine positioning pins
CN107330936B (en) Monocular vision-based double-circular marker positioning method and system
CN111289226A (en) Line laser flatness detection method based on visual measurement technology
CN113740864B (en) Laser three-dimensional point cloud-based detector soft landing end-segment autonomous pose estimation method
CN112508885B (en) Method and system for detecting three-dimensional central axis of bent pipe
JP2015111128A (en) Position attitude measurement device, position attitude measurement method, and program
KR20170142379A (en) Apparatus for detect dimensional welding line of welding robot using image processing
CN108090931B (en) Anti-blocking and anti-interference marker identification and pose measurement method based on combination of circle and cross features
JP4935769B2 (en) Plane region estimation apparatus and program
CN115790539B (en) Cooperative target underwater photogrammetry method
US20240085448A1 (en) Speed measurement method and apparatus based on multiple cameras
CN113109762B (en) Optical vision guiding method for AUV (autonomous Underwater vehicle) docking recovery
CN109815966A (en) A kind of mobile robot visual odometer implementation method based on improvement SIFT algorithm
CN112862862B (en) Aircraft autonomous oil receiving device based on artificial intelligence visual tracking and application method
CN115457130A (en) Electric vehicle charging port detection and positioning method based on depth key point regression
JP5032415B2 (en) Motion estimation apparatus and program
JP4034003B2 (en) Flying object image processing apparatus and flying object image processing method

Legal Events

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