CN113240740B - Attitude measurement method based on phase-guided binocular vision dense marking point matching - Google Patents
Attitude measurement method based on phase-guided binocular vision dense marking point matching Download PDFInfo
- Publication number
- CN113240740B CN113240740B CN202110491674.6A CN202110491674A CN113240740B CN 113240740 B CN113240740 B CN 113240740B CN 202110491674 A CN202110491674 A CN 202110491674A CN 113240740 B CN113240740 B CN 113240740B
- Authority
- CN
- China
- Prior art keywords
- point
- image
- phase
- matching
- points
- 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
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
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
- G06T7/74—Determining position or orientation of objects or cameras using feature-based methods involving reference images or patches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
Abstract
The invention relates to the field of antenna attitude measurement based on binocular vision, in particular to an attitude measurement method based on phase-guided binocular vision dense marking point matching, which comprises the following steps: s1, collecting left and right images of the measured object by using a binocular camera, and respectively extracting mark points; s2, calculating to obtain a homonymy point set to be matched of the left image mark point in the right image according to the epipolar geometry principle by using the image coordinates of the left image mark point and the system parameters of the binocular camera; s3, acquiring the phase value of each mark point in the left and right images; then calculating the phase difference between each point in the homonymous point set to be matched in the right image and the mark point of the left image, wherein the mark point corresponding to the minimum phase difference is an unambiguous matching point; s4, carrying out coordinate optimization based on the unambiguous matching points to obtain the best matching point equivalent to the phase of the left image mark point; s5, calculating the three-dimensional coordinates and the rotation posture based on the best matching point. The problem of many matches appear in epipolar line matching is solved, and the extraction precision of the real homonymy point coordinates is improved.
Description
Technical Field
The invention relates to the field of antenna attitude measurement based on binocular vision, in particular to an attitude measurement method based on phase-guided binocular vision dense marking point matching.
Background
The attitude measurement can be used in the fields of satellite grabbing and positioning, attitude angle measurement of satellite antennas, lunar rover navigation and antenna attitude analysis, non-contact measurement of rocket engine tail nozzles, measurement of visual attitude of aircrafts in wind tunnels and the like. At present, a visual imaging system usually adopts characteristic points to realize attitude measurement, and has strong dependence on single characteristic points, so that the defect of poor robustness exists. The binocular vision stereo matching is adopted, partial three-dimensional point set data of the target are obtained according to the triangulation principle, the attitude of the three-dimensional coordinates of the feature points relative to a vision system is used for solving, high measurement robustness is guaranteed, and the method can be expanded and used for solving the relative attitude estimation problem of other non-cooperative targets. Binocular feature high-precision extraction and uniqueness homonymy point matching based on cooperative targets such as circles or angular points are the key points for success or failure of binocular vision attitude measurement.
The markers are generally divided into two categories: one type of the shape has more characteristics and is different from each other and is called a coding mark point, while the other type of the shape is single and is called a non-coding mark point. And the Galo M manually points a mark point center on the image, and then carries out binocular vision matching to realize the attitude measurement of the satellite antenna mark. However, the center of the marking point of the method can only reach the pixel-level precision, the reconstruction precision of the three-dimensional point is inevitably influenced, the labor and the time are wasted, and the automation cannot be realized. Richard Hartley explains the basic theory of epipolar geometry and binocular vision in detail in Multiple view geometry in computer vision, and pays more attention to matching of non-coding mark points through a matching method of multi-image epipolar constraint. Zhang et al propose a reference point-based matching method, which takes the coding mark point as a reference point, calculates the compatibility and other parameters of the non-coding mark point, and realizes the matching of the non-coding mark. Guo et al propose a matching technique based on the relaxation method that requires that the marker points appear on at least multiple pictures and is ineffective for marker points that appear only on two pictures.
A great deal of research has been carried out on binocular vision feature point matching at home and abroad, but when the number of marked feature points is large and the marked feature points have relatively regular pattern distribution, matching errors are easily caused in epipolar point matching based on epipolar lines, so that the attitude calculation fails, as shown in fig. 1, according to the multi-view geometric binocular vision homonymy point matching principle, a certain feature point of a left image is on a corresponding epipolar line of a corresponding homonymy point in a right image. As shown in the figure1, image points m and m ' corresponding to the space point P, it can be known that the left image point m is always in the corresponding polar line l ' at the same point of the right image according to the geometric principle of the polar line of the binocular vision system 'mHowever, when the feature marker point set is arranged in a large amount and regularly, as shown in the right graph (b) of fig. 1, the P point and the N point are located on the same polar plane, the corresponding epipolar lines of the two points are the same, and the left image point m has m ' and N ' on the corresponding epipolar line l ' of the right image, so that ambiguous matching occurs when searching for a matching point on the epipolar line, and in addition, more epipolar line matching error points are also caused due to image distortion, perspective projection and image noise.
Disclosure of Invention
The invention aims to solve the problems that ambiguous matching occurs to epipolar line search matching points when the arrangement quantity of a feature marker point set is large and regular in the prior art and more epipolar line matching error points are also caused by image distortion, perspective projection and image noise, and provides a posture measuring method based on phase-guided binocular vision dense marker point matching.
In order to achieve the above purpose, the invention provides the following technical scheme:
an attitude measurement method based on phase-guided binocular vision dense marking point matching comprises the following steps:
s1, acquiring a left image and a right image of the measured object by using a binocular camera, and extracting mark points from the left image and the right image respectively;
s2, calculating to obtain a homonymy point set to be matched of the left image mark point in the right image according to the epipolar geometry principle by using the image coordinates of the left image mark point and the system parameters of the binocular camera;
s3, acquiring phase values of the marking points in the left image and the right image; then calculating the phase difference between each point in the same-name point set to be matched in the right image and the mark point of the left image, wherein the point with the minimum phase difference is an unambiguous matching point;
s4, carrying out coordinate optimization based on the unambiguous matching points to obtain the best matching point equivalent to the phase of the left image mark point;
and S5, calculating and obtaining the three-dimensional coordinates and the rotation posture of the measured object based on the optimal matching point.
The method has the advantages that when the arrangement amount of the marker point sets is large and regular, the homonymy point set to be matched is found through polar line search, then, unambiguous matching points are found in the homonymy point set to be matched based on phase values and coordinate optimization is carried out, the optimal matching point of the accurate left image marker point is found, and the problems that in the prior art, when the arrangement amount of the marker point sets is large and regular, matching errors occur, and more polar line matching error points are caused by image distortion, perspective projection and image noise are solved.
Further, the acquiring phase values of the marker points in the left image and the right image specifically includes the following steps:
s31, obtaining a binocular stripe adjustment image of the measured object;
and S32, carrying out Fourier transform, frequency domain window fundamental frequency filtering, inverse Fourier transform and phase expansion on the binocular fringe adjustment image to obtain the phase value of each mark point in the left image and the right image.
Further, in step S4, the specific method for performing coordinate optimization based on the unambiguous matching point is to select a window formed by n × n points with the unambiguous matching point in the right image as a center point, and calculate and search the window for a best matching point having a phase equivalent to the mark point of the left image
As a preferred scheme of the present invention, the specific method for calculating and searching the best matching point equivalent to the phase of the left image mark point in the window is to traverse the phase values of n × n points in the window, wherein the point where the phase value is equal to the phase value of the left image mark point is the best matching point;
as a preferred embodiment of the present invention, the specific method for calculating and searching the best matching point equivalent to the phase of the left image mark point in the window is to calculate a correlation between the phase value of n × n points in the window and the phase value of the left image mark point, and when the correlation is the maximum, the corresponding point is the best matching point.
As a preferable aspect of the present invention, the value of n is an integer greater than or equal to 3.
Further, in step S1, the method for extracting the mark points includes extracting a circle center based on ellipse boundary fitting for the mark point type to obtain the image coordinates of the mark points.
Further, the specific step of S5 includes:
s51, calculating the three-dimensional coordinate of the measured object by a collinear equation according to the mark point of the left image, the best matching point in the right image and the system parameters of the binocular camera, wherein the collinear equation is as follows:
wherein r is1～r9The elements of a rotation matrix R in the system parameters of the binocular camera are shown, and the rotation matrix R is a 3 x 3 orthogonal matrix; t is tx、txAnd tzIs an element of a translation vector T, fμIs the focal length in the image width direction, f, in the system parameters of the binocular cameravIs the focal length in the height direction; u. u0、v0The coordinates of the center of the camera image plane;
and S52, calculating the rotation posture of the measured object based on the initial position according to the three-dimensional coordinates, wherein the rotation posture of the measured object is obtained by the following formula:
P0i＝Pni×Rn+Tn
wherein, P0iThe method comprises the steps of obtaining mark points of an initial position, wherein the mark points are Euclidean three-dimensional coordinate values of the initial position, n is the number of attitude measurement, i is the number of the mark points, R and T are respectively a rotation matrix and a translation vector of a measured object relative to the initial position during nth measurement, and the rotation matrix and the translation vector represent the rotation attitude of the measured object.
Further, the method for calibrating and calibrating the system parameters of the binocular camera in advance comprises the following steps:
firstly, shooting a target by using a binocular camera, and extracting coordinates of feature points in a shot image to obtain target feature point image coordinates p (u, v);
and secondly, calculating system parameters of the binocular camera according to a visual imaging mathematical model formula by using the physical coordinates P (X, Y, Z) of the target feature points and the image coordinates P (u, v) of the target feature points.
Based on the same conception of the invention, the invention also provides an attitude measurement device based on phase-guided binocular vision dense marking point matching, which comprises at least one processor and a memory in communication connection with the at least one processor; the memory stores instructions executable by the at least one processor to enable the at least one processor to perform any of the methods described above.
Compared with the prior art, the invention has the following beneficial effects:
1. when the layout amount of the marker point sets is large and regular, searching for a homonymy point set to be matched through polar line search, then searching for an unambiguous matching point based on a phase value in the homonymy point set to be matched, optimizing coordinates, and searching for an accurate optimal matching point of the left image marker point, so that the problems that in the prior art, when the layout amount of the marker point sets is large and regular, matching errors occur, and more polar line matching error points are caused by image distortion, perspective projection and image noise are solved;
2. because the method for extracting the centers of the left circle and the right circle based on the ellipse boundary fitting has errors in the extraction of the centers of the left circle and the right circle from the view point of image processing, the centers of the corresponding object surfaces of the left image and the right image which have the same name point are not completely the same, and therefore, a certain error exists in the calculated polar line; on the other hand, with the window phase data search, the amount of calculation is greatly reduced compared to the entire image search.
Drawings
FIG. 1 is a schematic diagram of the principle of ambiguity matching problem generated by homonymous points of binocular visual features based on epipolar lines;
FIG. 2 is a flow chart of an attitude measurement method based on phase-guided binocular vision dense marker point matching;
FIG. 3 is a set of marker extraction and matching points of example 1;
fig. 4 is a schematic diagram of the phase information calculation and characteristic point phase mapping process in embodiment 1;
FIG. 5 is a diagram showing the result of matching the marked points based on the phase equivalence search in example 1;
fig. 6 shows the feature points of the regular circle center photographed by the left and right cameras of the binocular camera of the embodiment 2;
FIG. 7 is a drawing of center coordinates of a left image according to example 2;
FIG. 8 is a left image of the epipolar-line based binocular visual feature homonymous point matching graph of example 2;
FIG. 9 is the right image of the epipolar-line based binocular visual feature homonymous point matching graph of example 2;
FIG. 10 is a phase diagram of the marker points in the left and right images of example 2;
fig. 11 is a distribution diagram of the three-dimensional coordinate calculation results of 660 circle centers in example 2.
Detailed Description
The present invention will be described in further detail with reference to test examples and specific embodiments. It should be understood that the scope of the above-described subject matter is not limited to the following examples, and any techniques implemented based on the disclosure of the present invention are within the scope of the present invention.
Example 1
As the prior art searches for matching points on an epipolar line, ambiguous matching occurs, as shown in fig. 1, a P point and an N point are located on the same polar plane, corresponding epipolar lines of the two points are the same, and a left image point m has m ' and N ' on a corresponding epipolar line l ' of a right image (in addition, more epipolar line matching error points are also caused by image distortion, perspective projection and image noise), the present embodiment provides a posture measurement method based on phase-guided binocular vision dense marking point matching, as shown in fig. 2, specifically including the following steps:
s101, calibrating and calibrating parameters of a binocular camera system, wherein the method comprises the following steps:
calculating the relative pose relationship between the internal parameters of a camera and the left and right cameras in the binocular vision system by adopting a two-dimensional plane target algorithm proposed by Zhang Zhengyou to complete the measurement and storage of system structure parameters;
the calibration plate is a checkerboard target with 20 multiplied by 20 which is 400 angular points, then a binocular camera is used for shooting the target, the coordinates of the checkerboard angular points in each shot image are extracted, the image coordinates of the checkerboard angular points are recorded as P (u, v), the known checkerboard physical coordinates P (X, Y, Z) are combined, the camera internal parameter calibration can be completed by utilizing the following visual imaging mathematical model formula,
the calibrated intrinsic parameter has a focal length f in the image width directionμFocal length in the height direction fvPrincipal point coordinates (u) of the camera0,v0) And λ is an arbitrary constant. R and T are respectively a rotation matrix and a translation vector converted from a world coordinate system to a camera coordinate system, and the rotation matrix and the translation vector jointly form external parameters of the camera; wherein the component of T is (T)x,ty,tz) (ii) a The matrix R is a 3 × 3 orthogonal matrix, and the elements of T are (R)1,…,r9)。
S102, acquiring a left image and a right image of a measured object by using a calibrated binocular camera system, and extracting mark points from the centers of circles of the left image and the right image based on ellipse boundary fitting respectively, so as to obtain image coordinates p (u, v) of each mark point from the shot left image and the shot right image respectively and store the image coordinates p (u, v) of each mark point; the object to be measured in this embodiment is an antenna;
as shown in fig. 3, (a) in fig. 3 is a diagram of marker extraction positions of a binocular vision system left image based on epipolar lines.
S103, calculating to obtain a homonymous point set to be matched of the left image mark point in the right image according to the epipolar geometry principle by using the image coordinates of the left image mark point and the system parameters of the binocular camera;
as shown in fig. 3, (b) in fig. 3 is a diagram of calculating a set of matching points in the corresponding right image for the 5 marking points in the diagram (a) in fig. 3 according to the epipolar geometry principle; wherein, the red solid line is the polar line of the right image corresponding to the 5 marking points in the left image;
it can be seen that the epipolar lines of the right images corresponding to the mark points 3 and 4 in the left image are very close, and the epipolar lines of the right images corresponding to the mark points 5 and 1 in the left image are overlapped, so that the corresponding homonymous point of the mark point in the left image cannot be found by only depending on the epipolar line.
S104, acquiring phase values of all the mark points in the left image and the right image; then calculating the phase difference between each point in the same-name point set to be matched in the right image and the mark point of the left image, wherein the point with the minimum phase difference is an unambiguous matching point;
the method for obtaining the phase value of each mark point in the left image and the right image comprises the steps of obtaining a binocular stripe adjustment image shot by an antenna when the attitude of a measured object, namely the antenna is measured, then carrying out image processing, Fourier transformation, frequency domain window fundamental frequency filtering, inverse Fourier transformation, truncation phase obtaining, phase unfolding and the like on the binocular stripe adjustment image to obtain a truncation phase and an unfolding phase of the surface of the antenna shot by a binocular camera system, and obtaining the phase value of each mark point in the left image and the right image to be recorded as the phase value of each mark point in the left image and the right image as shown in figure 4
S105, carrying out coordinate optimization based on the unambiguous matching points to obtain an optimal matching point equivalent to the phase of the left image mark point; the specific method comprises the following steps:
selecting a window consisting of n multiplied by n points by taking the unambiguous matching point in the right image as a central point, and calculating and searching the best matching point with the phase equivalent to the marking point of the left image in the window; the value of n is selected based on the system parameters of the binocular camera and the density of the projection stripes, and is a positive integer not less than 3, and n is set to be 5 based on the system parameters of the binocular camera and the density of the projection stripes in the embodiment;
the specific method for calculating and searching the best matching point equivalent to the phase of the left image mark point in the window is to traverse the phase values of n multiplied by n points in the window, wherein the point with the phase value equal to the phase value of the left image mark point is the best matching point; or calculating the correlation between the phase value of n multiplied by n points in the window and the phase value of the left image marking point, wherein when the correlation is maximum, the corresponding point is the best matching point;
the matching result is shown in fig. 5.
The method has the technical effects that when the arrangement amount of the marker point sets is large and the marker point sets are regular, the homonymy point set to be matched is found through polar line search, then the homonymy matching point is found in the homonymy point set to be matched based on the phase value and coordinate optimization is carried out, the accurate optimal matching point of the left image marker point is found, and the problems that in the prior art, when the arrangement amount of the marker point sets is large and the rule is regular, matching errors occur, and more polar line matching error points are caused by image distortion, perspective projection and image noise are solved.
Because the method for extracting the centers of the left circle and the right circle based on the ellipse boundary fitting has errors in the extraction of the centers of the left circle and the right circle from the view point of image processing, the centers of the corresponding object surfaces of the homonymous points of the left image and the right image are not completely the same, and certain errors exist in calculated epipolar lines.
S106, calculating to obtain the three-dimensional coordinates and the rotating posture of the measured object based on the optimal matching point, and the specific steps comprise:
calculating the three-dimensional coordinate of the measured object by a collinear equation according to the optimal matching points in the left image marking point and the right image and the system parameters of the binocular camera, wherein the collinear equation is as follows:
wherein r is1～r9As a system of said binocular camerasElements of a rotation matrix R in the system parameters, wherein the rotation matrix R is a 3 multiplied by 3 orthogonal matrix; t is tx、txAnd tzIs an element of a translation vector T, fμIs the focal length in the image width direction, f, in the system parameters of the binocular cameravIs the focal length in the height direction; u. u0、v0Is the center coordinate of the camera image plane;
then, calculating a rotation posture of the measured object based on the initial position according to the three-dimensional coordinates, wherein the rotation posture of the measured object is obtained by the following formula:
P0i＝Pni×Rn+Tn
wherein n is the number of attitude measurement, i is the number of mark points, P0iIs a mark point of an initial position in an European three-dimensional coordinate value, PniAnd the three-dimensional coordinates of the mark point during the nth measurement are corresponding to the mark point, R and T are respectively a rotation matrix and a translation vector of the measured object relative to the initial position during the nth measurement, and the rotation matrix and the translation vector represent the rotation posture of the measured object.
Example 2
In the prior art, error matching is most easily caused by only depending on epipolar line matching on regularly arranged feature points, so in this embodiment, the mark points are pasted into a relatively regular row-column distribution, and then the method described in embodiment 1 is used, and the left and right camera mark points shot by the binocular camera system calibrated in embodiment 1 are shown in fig. 6.
S201, circle center extraction based on ellipse boundary fitting is respectively carried out on the shot left and right images, wherein the circle center extraction comprises binarization, area selection, edge detection, ellipse center positioning and the like, 660 mark points of the shot images are successfully extracted, and the mark point extraction result of the left image is shown in FIG. 7.
S202, randomly selecting 20 marking points in the left image, searching homonymous points in the right image, and matching homonymous points in an epipolar line searching mode inevitably causes a phenomenon of matching failure. As shown in fig. 8, 20 marked points of the left image are selected for illustration, and the matching results are shown in fig. 9, and many error matching results occur.
And S203, calculating to obtain a set of homonymous points to be matched of the left image mark points in the right image according to the epipolar geometry principle by using the extracted image coordinates of the left image mark points and the system parameters of the binocular camera, and then acquiring phase values of the mark points in the left image and the right image, as shown in FIG. 10.
S204, calculating the phase difference between each point in the same-name point set to be matched in the right image and the mark point of the left image, finding the point with the minimum phase difference as an unambiguous matching point, and performing coordinate optimization based on the unambiguous matching point to obtain the best matching point equivalent to the mark point of the left image in phase.
In order to verify the matching correctness of the 660 optimal matching points, the corresponding spatial three-dimensional point coordinates of the optimal matching point sets are calculated through a collinear equation, if the 660 feature point homonymous coordinates are correctly matched, the calculated three-dimensional coordinates are the actual spatial distribution of the object surface circle features, all the feature points are in the same plane and regularly distributed in a net shape, the three-dimensional point coordinates are shown in fig. 11, it can be seen that the optimal matching points are correct matching points, and the matching result accords with the expectation.
Example 3
The attitude measurement device for binocular vision dense marker point matching based on phase guidance comprises at least one processor and a memory which is in communication connection with the at least one processor; the memory stores instructions executable by the at least one processor to enable the at least one processor to perform any of the methods described above.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.
Claims (10)
1. The attitude measurement method based on phase-guided binocular vision dense marking point matching is characterized by comprising the following steps of:
s1, acquiring a left image and a right image of the measured object by using a binocular camera, and extracting mark points from the left image and the right image respectively;
s2, calculating to obtain a homonymy point set to be matched of the left image mark point in the right image according to the epipolar geometry principle by using the image coordinates of the left image mark point and the system parameters of the binocular camera;
s3, acquiring phase values of the marking points in the left image and the right image; then calculating the phase difference between each point in the same-name point set to be matched in the right image and the mark point of the left image, wherein the point with the minimum phase difference is an unambiguous matching point;
s4, carrying out coordinate optimization based on the unambiguous matching points to obtain the best matching point equivalent to the phase of the left image mark point;
and S5, calculating and obtaining the three-dimensional coordinates and the rotation posture of the measured object based on the optimal matching point.
2. The pose measurement method based on phase-guided binocular vision dense marker point matching according to claim 1, wherein in step S3, the acquiring the phase value of each marker point in the left image and the right image specifically comprises the following steps:
s31, acquiring a binocular stripe adjustment image of the measured object;
and S32, performing Fourier transform, frequency domain window fundamental frequency filtering, inverse Fourier transform and phase expansion on the binocular fringe adjustment image to obtain phase values of the mark points in the left image and the right image.
3. The method for measuring pose based on phase-guided binocular vision dense marker matching according to claim 1, wherein in step S4, the specific method for performing coordinate optimization based on the unambiguous matching point is to select a window consisting of n × n points with the unambiguous matching point in the right image as a center point, and calculate and search a best matching point with a phase equivalent to the left image marker in the window.
4. The pose measurement method based on phase-guided binocular vision dense marker point matching according to claim 3, wherein the specific method for calculating and searching the best matching point in the window having the same phase value as the left image marker point is to traverse phase values of n x n points in the window, wherein the point where the phase value is equal to the phase value of the left image marker point is the best matching point.
5. The pose measurement method based on phase-guided binocular vision dense marker point matching according to claim 3, wherein the specific method for calculating and searching the best matching point with the same phase value as the left image marker point in the window is to calculate the correlation between the phase value of n x n points in the window and the phase value of the left image marker point, and when the correlation is the maximum, the corresponding point is the best matching point.
6. The attitude measurement method based on phase-guided binocular vision dense marker point matching according to claim 3, wherein a value of n is an integer greater than or equal to 3.
7. The attitude measurement method based on phase-guided binocular vision dense marker matching according to claim 1, wherein the marker extraction method in step S1 is to perform circle center extraction based on ellipse boundary fitting for marker types to obtain marker image coordinates.
8. The attitude measurement method based on phase-guided binocular vision dense marker point matching according to claim 1, wherein the specific steps of step S5 include:
s51, calculating the three-dimensional coordinate of the measured object by a collinear equation according to the mark point of the left image, the best matching point in the right image and the system parameters of the binocular camera, wherein the collinear equation is as follows:
wherein r is1～r9To the binocular phaseThe method comprises the following steps of rotating elements of a torque matrix in system parameters of the machine, wherein the rotation matrix is a 3 x 3 orthogonal matrix; t is tx、tyAnd tzIs an element of a translation vector, fμIs the focal length in the image width direction, f, among the system parameters of the binocular cameravIs the focal length in the height direction; u. u0、v0Is the center coordinate of the camera image plane;
and S52, calculating the rotation posture of the measured object based on the initial position according to the three-dimensional coordinates, wherein the rotation posture of the measured object is obtained by the following formula:
P0i＝Pni×Rn+Tn
wherein, P0iThe mark points are European three-dimensional coordinate values of initial positions, n is the number of attitude measurement times, and i is the number of the mark points; rnAnd TnAnd respectively representing the rotation matrix and the translation vector of the measured object relative to the initial position during the nth measurement, wherein the rotation matrix and the translation vector represent the rotation posture of the measured object.
9. The attitude measurement method based on phase-guided binocular vision dense marker point matching according to any one of claims 1 to 8, wherein before step S1, system parameters of a binocular camera are calibrated and calibrated in advance, and the method comprises the following steps:
firstly, shooting a target by using a binocular camera, and extracting coordinates of feature points in a shot image to obtain target feature point image coordinates p (u, v);
and secondly, calculating system parameters of the binocular camera according to a visual imaging mathematical model formula by using the physical coordinates P (X, Y, Z) of the target feature points and the image coordinates P (u, v) of the target feature points.
10. The attitude measurement device for binocular vision dense marker point matching based on phase guidance is characterized by comprising at least one processor and a memory which is in communication connection with the at least one processor; the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110491674.6A CN113240740B (en) | 2021-05-06 | 2021-05-06 | Attitude measurement method based on phase-guided binocular vision dense marking point matching |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110491674.6A CN113240740B (en) | 2021-05-06 | 2021-05-06 | Attitude measurement method based on phase-guided binocular vision dense marking point matching |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113240740A CN113240740A (en) | 2021-08-10 |
| CN113240740B true CN113240740B (en) | 2022-07-12 |
Family
ID=77132121
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110491674.6A Active CN113240740B (en) | 2021-05-06 | 2021-05-06 | Attitude measurement method based on phase-guided binocular vision dense marking point matching |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113240740B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113610931A (en) * | 2021-08-13 | 2021-11-05 | 南京航空航天大学 | Machine vision method for measuring parachute-shaped state parameters |
| CN113642681B (en) * | 2021-10-13 | 2022-01-04 | 中国空气动力研究与发展中心低速空气动力研究所 | Matching method of aircraft model surface mark points |
| CN114353780B (en) * | 2021-12-31 | 2024-04-02 | 高德软件有限公司 | Gesture optimization method and device |
| CN115063468B (en) * | 2022-06-17 | 2023-06-27 | 梅卡曼德（北京）机器人科技有限公司 | Binocular stereo matching method, computer storage medium and electronic equipment |
| CN114964316B (en) * | 2022-07-27 | 2022-11-01 | 湖南科天健光电技术有限公司 | Position and attitude calibration method and device, and method and system for measuring target to be measured |
| CN114993243A (en) * | 2022-08-04 | 2022-09-02 | 深圳粤讯通信科技有限公司 | Antenna attitude monitoring and early warning system based on Internet of things |
| CN116309829B (en) * | 2023-02-28 | 2024-03-19 | 无锡赛锐斯医疗器械有限公司 | Cuboid scanning body group decoding and pose measuring method based on multi-view vision |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5684890A (en) * | 1994-02-28 | 1997-11-04 | Nec Corporation | Three-dimensional reference image segmenting method and apparatus |
| US6108435A (en) * | 1994-12-15 | 2000-08-22 | Fuji Electric Co., Ltd. | Method of detecting phase difference between a pair of images |
| CN110378341A (en) * | 2019-07-24 | 2019-10-25 | 西南交通大学 | A kind of binocular vision pedestrian distance detection method |
| CN110567398A (en) * | 2019-09-02 | 2019-12-13 | 武汉光发科技有限公司 | Binocular stereo vision three-dimensional measurement method and system, server and storage medium |
| CN110852979A (en) * | 2019-11-12 | 2020-02-28 | 广东省智能机器人研究院 | Point cloud registration and fusion method based on phase information matching |
| CN111062990A (en) * | 2019-12-13 | 2020-04-24 | 哈尔滨工程大学 | Binocular vision positioning method for underwater robot target grabbing |
| CN111429571A (en) * | 2020-04-15 | 2020-07-17 | 四川大学 | Rapid stereo matching method based on spatio-temporal image information joint correlation |
| CN111649694A (en) * | 2020-06-04 | 2020-09-11 | 四川大学 | Implicit phase-parallax mapping binocular measurement missing point cloud interpolation method |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI635256B (en) * | 2017-10-11 | 2018-09-11 | 宏碁股份有限公司 | Phase detection auto-focus-based positioning method and system thereof |
| CN111028284A (en) * | 2019-10-31 | 2020-04-17 | 浙江未来技术研究院(嘉兴) | Binocular vision stereo matching method and device based on homonymous mark points |
-
2021
- 2021-05-06 CN CN202110491674.6A patent/CN113240740B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5684890A (en) * | 1994-02-28 | 1997-11-04 | Nec Corporation | Three-dimensional reference image segmenting method and apparatus |
| US6108435A (en) * | 1994-12-15 | 2000-08-22 | Fuji Electric Co., Ltd. | Method of detecting phase difference between a pair of images |
| CN110378341A (en) * | 2019-07-24 | 2019-10-25 | 西南交通大学 | A kind of binocular vision pedestrian distance detection method |
| CN110567398A (en) * | 2019-09-02 | 2019-12-13 | 武汉光发科技有限公司 | Binocular stereo vision three-dimensional measurement method and system, server and storage medium |
| CN110852979A (en) * | 2019-11-12 | 2020-02-28 | 广东省智能机器人研究院 | Point cloud registration and fusion method based on phase information matching |
| CN111062990A (en) * | 2019-12-13 | 2020-04-24 | 哈尔滨工程大学 | Binocular vision positioning method for underwater robot target grabbing |
| CN111429571A (en) * | 2020-04-15 | 2020-07-17 | 四川大学 | Rapid stereo matching method based on spatio-temporal image information joint correlation |
| CN111649694A (en) * | 2020-06-04 | 2020-09-11 | 四川大学 | Implicit phase-parallax mapping binocular measurement missing point cloud interpolation method |
Non-Patent Citations (8)
| Title |
|---|
| High-speed Stereo Matching Algorithm for Ultra-high Resolution Binocular Image;Baopeng xu等;《2018 IEEE International Conference on Automation, Electronics and Electrical Engineering (AUTEEE)》;IEEE;20190523;第87-90页 * |
| Tractor path tracking control based on binocular vision;Shuo Zhang等;《Information Processing in Agriculture》;ELSEVIER;20181231;第5卷(第4期);第422-432页 * |
| 基于传像束的小视场双目结构光三维形貌恢复;袁卓凡等;《应用光学》;20210115;第42卷(第1期);第113-118页 * |
| 基于单双目融合的遮挡区域点云获取技术研究;张利萍等;《微型机与应用》;20170225(第04期);第74-77页 * |
| 基于双目视觉的三维重建和拼接技术研究;吕耀文等;《光电子技术》;20161231(第04期);第23-27页 * |
| 基于双目视觉的原木材积检测方法研究;赵亚凤;《中国博士学位论文全文数据库 信息科技辑》;20170215(第2期);I138-109 * |
| 基于极线校正的亚像素相位立体匹配方法;肖志涛等;《红外与激光工程》;20141225;第225-230页 * |
| 基于相位映射的双目视觉缺失点云插补方法;李承杭等;《光学学报》;20190909;第40卷(第1期);第260-269页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113240740A (en) | 2021-08-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113240740B (en) | Attitude measurement method based on phase-guided binocular vision dense marking point matching | |
| KR101533182B1 (en) | 3d streets | |
| JP6421395B2 (en) | 3D topographic map formation method from SAR map | |
| WO2007133620A2 (en) | System and architecture for automatic image registration | |
| US11212511B1 (en) | Residual error mitigation in multiview calibration | |
| Rupnik et al. | Refined satellite image orientation in the free open-source photogrammetric tools Apero/Micmac | |
| Tjahjadi et al. | Precise wide baseline stereo image matching for compact digital cameras | |
| CN116957987A (en) | Multi-eye polar line correction method, device, computer equipment and storage medium | |
| KR102015817B1 (en) | A method of automatic correction of provided rpc of stereo satellite images | |
| Wu | Photogrammetry: 3-D from imagery | |
| Petitpas et al. | Roughness measurement from multi-stereo reconstruction | |
| Pedersini et al. | Automatic monitoring and 3D reconstruction applied to cultural heritage | |
| Zhang et al. | Non-rigid registration of mural images and laser scanning data based on the optimization of the edges of interest | |
| Evans et al. | Surveillance camera calibration from observations of a pedestrian | |
| Oda et al. | Qualification of point clouds measured by SfM software | |
| Kanuki et al. | Automatic compensation of radial distortion by minimizing entropy of histogram of oriented gradients | |
| Zhang et al. | Registration of CBERS-02B satellite imagery in quick GIS updating | |
| CN104331882B (en) | Method for measuring speed of aircraft | |
| CN108764161A (en) | The remote sensing image treating method and apparatus of the ill singularity for cracking Sparse Array initiation based on polar coordinate system | |
| CN116091610B (en) | Combined calibration method of radar and camera based on three-dimensional tower type checkerboard | |
| Li et al. | DEM registration, alignment and evaluation for SAR interferometry | |
| Tommaselli et al. | Camera calibration using straight lines: assessment of a model based on plane equivalence | |
| Deliś et al. | 3D modeling of architectural objects from video data obtained with the fixed focal length lens geometry | |
| Ding et al. | Large scale object's measurement method research based on multi-view Reconstruction | |
| Zhang et al. | Stereo image matching for vehicle-borne mobile mapping system based on digital parallax model |
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 |