CN107633536A - A kind of camera calibration method and system based on two-dimensional planar template - Google Patents
A kind of camera calibration method and system based on two-dimensional planar template Download PDFInfo
- Publication number
- CN107633536A CN107633536A CN201710674279.5A CN201710674279A CN107633536A CN 107633536 A CN107633536 A CN 107633536A CN 201710674279 A CN201710674279 A CN 201710674279A CN 107633536 A CN107633536 A CN 107633536A
- Authority
- CN
- China
- Prior art keywords
- homography matrix
- target
- angle point
- camera
- angle
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Abstract
The invention discloses a kind of camera calibration method and system based on two-dimensional planar template, method therein includes：Gridiron pattern uncalibrated image is obtained, the uncalibrated image includes multiple angle points；The error amount of angle point first in the multiple angle point is rejected using RANSAC algorithm and, more than the angle point of the first preset value, obtains Corner；According to the coordinate of the Corner and corresponding world coordinates, homography matrix is obtained；Homography matrix of second error amount more than the second preset value is rejected using RANSAC algorithm, obtains target homography matrix；According to the target homography matrix, target conic model is obtained, and the initial parameter of camera is obtained according to the target conic model；The initial parameter is estimated using the method for maximal possibility estimation, obtains target component, camera is demarcated using the target component.The present invention solves the not high technical problem of accuracy that Zhang Zhengyou scaling methods have camera calibration.
Description
Technical field
The present invention relates to technical field of visual measurement, more particularly to a kind of camera calibration method based on two-dimensional planar template
And system.
Background technology
In vision measurement, the demarcation of camera parameter is the stabilization of unusual the key link, stated accuracy and calibration algorithm
Property directly affect camera work produce result accuracy.The purpose of camera calibration is estimation camera lens and imaging sensor
Parameter, including internal reference, outer ginseng and distortion coefficients of camera lens, these parameters are widely applied to computer vision field, for example correct
Lens distortions, the actual size for measuring object, determine the position of camera in the scene etc..Camera calibration is also widely used for machine
People, navigation system and three-dimensional reconstruction etc..
At present, the method for camera calibration is roughly divided into two classes：Traditional scaling method and Camera Self-Calibration method.Traditional
Scaling method is the calibrating block using known dimensions, establishes the corresponding relation of object coordinates and image coordinate, then by corresponding
Algorithm obtains camera inside and outside parameter, and stated accuracy is higher but the processing of three-dimensional scaling block and difficult in maintenance, and cost is very high；Camera is certainly
Demarcation is mainly using camera motion constraint or the restrained split-flow parameter of scene, and flexibility is stronger, but self-calibrating method is to phase
Machine moves and context restrictions condition is too strong, and in actual use, robustness is poor, and precision is relatively low, and these factors also limit
The use range of self-calibration.
In order to solve the above-mentioned technical problem, Zhang Zhengyou proposes a kind of flexible plane reference method, and this method is between biography
Between the demarcation of system and Camera Self-Calibration method, it is not necessary to specifically demarcate thing, it is only necessary to print a gridiron pattern, can avoid
The problem of traditional scaling method equipment requirement is high, and self-calibrating method precision is low.But the use of Zhang Zhengyou standardization is more
Width chessboard table images, using the corresponding relation of angle point, homography matrix corresponding to each image is calculated, then asked using closing solution
Go out camera internal reference, outer ginseng, finally consider lens distortion, by the use of the closing solution being previously obtained as initial value, calculated using non-linear search
Method estimates all parameters, including internal reference, outer ginseng and distortion factor.But Nonlinear Search Algorithm is to the precision of given initial value
It is required that very strict, because camera parameter is coupling, provided that initial value precision it is not high, then nonlinear optimization performance can very
Difference, locally optimal solution can be also absorbed in, thus cause the accuracy and robustness of solution camera parameter and distortion factor inadequate.
It can be seen that there is the not high technical problem of accuracy of camera calibration in existing Zhang Zhengyou scaling methods.
The content of the invention
The embodiment of the present invention provides a kind of camera calibration method and system based on two-dimensional planar template, existing to solve
The not high technical problem of accuracy of camera calibration be present in Zhang Zhengyou scaling methods.
The invention discloses a kind of camera calibration method based on two-dimensional planar template, methods described includes：
Gridiron pattern uncalibrated image is obtained, the uncalibrated image includes multiple angle points；
The error amount of angle point first in the multiple angle point is rejected using RANSAC algorithm and is more than the first preset value
Angle point, obtain Corner；
According to the coordinate of the Corner and corresponding world coordinates, the generation where the gridiron pattern uncalibrated image is obtained
Homography matrix of boundary's coordinate system to pixel coordinate system；
Homography matrix of second error amount more than the second preset value is rejected using RANSAC algorithm, obtains mesh
Mark homography matrix；
According to the target homography matrix, target conic model is obtained, and according to the target conic section mould
Type obtains the initial parameter of camera, and the initial parameter includes the first internal reference, first ginseng and the first distortion factor outside；
The initial parameter is estimated using the method for maximal possibility estimation, obtains target component, the target ginseng
Number includes the second internal reference, second ginseng and the second distortion factor outside, and camera is demarcated using the target component.
It is described that angle point the is rejected in the multiple angle point using RANSAC algorithm in method provided by the invention
One error amount is more than the angle point of the first preset value, obtains Corner, including：
Obtain the re-projection coordinate of angle point；
Obtain the first distance between the re-projection coordinate of the angle point and the initial coordinate of the angle point；
First distance is rejected more than the angle point corresponding to the first preset value using RANSAC algorithm, so that
Obtain Corner.
In method provided by the invention, the coordinate according to the Corner and corresponding world coordinates, institute is obtained
The world coordinate system where gridiron pattern uncalibrated image is stated to the homography matrix of pixel coordinate system, including：
Four angle points are randomly selected from the gridiron pattern uncalibrated image, and obtain the coordinate of four angle points；
According to world coordinates corresponding to the coordinate and four angle points of four angle points, the gridiron pattern uncalibrated image is obtained
Homography matrix of the world coordinate system at place to pixel coordinate system.
It is described pre- more than second using RANSAC algorithm the second error amount of rejecting in method provided by the invention
If the homography matrix of value, target homography matrix is obtained, including：
According to homography matrix, default conic model is obtained；
The second distance of the homography matrix and the conic model is obtained, using the second distance as poor
Value；
If the difference is more than second preset value, for ineligible homography matrix；
The ineligible homography matrix is rejected from original homography matrix, obtains target homography matrix.
In method provided by the invention, the gridiron pattern uncalibrated image includes multiple, and different uncalibrated images and demarcation
The angle of camera differs.
Based on same inventive concept, second aspect of the present invention provides a kind of camera calibration based on two-dimensional planar template
System, the system include：Acquisition module, for obtaining gridiron pattern uncalibrated image, the uncalibrated image includes multiple angle points；
First obtains module, for rejecting the error of angle point first in the multiple angle point using RANSAC algorithm
Value obtains Corner more than the angle point of the first preset value；
Second obtains module, for the coordinate according to the Corner and corresponding world coordinates, obtains the chessboard
The homography matrix of world coordinate system where lattice uncalibrated image to pixel coordinate system；
3rd obtains module, is more than the second preset value for rejecting the second error amount using RANSAC algorithm
Homography matrix, obtain target homography matrix；
4th obtains module, for according to the target homography matrix, obtaining target conic model, and according to institute
The initial parameter that target conic model obtains camera is stated, the initial parameter includes the ginseng and first outside of the first internal reference, first
Distortion factor；
Demarcating module, for estimating using the method for maximal possibility estimation the initial parameter, obtain target ginseng
Number, the target component include the second internal reference, second ginseng and the second distortion factor outside, and camera is carried out using the target component
Demarcation.
In system provided by the invention, described first obtains module, is additionally operable to：
Obtain the re-projection coordinate of angle point；
Obtain the first distance between the re-projection coordinate of the angle point and the initial coordinate of the angle point；
First distance is rejected more than the angle point corresponding to the first preset value using RANSAC algorithm, so that
Obtain Corner.
In system provided by the invention, described second obtains module, is additionally operable to：
Four angle points are randomly selected from the gridiron pattern uncalibrated image, and obtain the coordinate of four angle points；
According to world coordinates corresponding to the coordinate and four angle points of four angle points, the gridiron pattern uncalibrated image is obtained
Homography matrix of the world coordinate system at place to pixel coordinate system.
In system provided by the invention, the described 3rd obtains module, is additionally operable to：
According to homography matrix, default conic model is obtained；
The second distance of the homography matrix and the conic model is obtained, using the second distance as poor
Value；
If the difference is more than second preset value, for ineligible homography matrix；
The ineligible homography matrix is rejected from original homography matrix, obtains target homography matrix
In system provided by the invention, the gridiron pattern uncalibrated image includes multiple, and different uncalibrated images and demarcation
The angle of camera differs.
Based on same inventive concept, third aspect present invention additionally provides a kind of computer-readable recording medium, thereon
Computer program is stored with, the program realizes following steps when being executed by processor：
Gridiron pattern uncalibrated image is obtained, the uncalibrated image includes multiple angle points；
The error amount of angle point first in the multiple angle point is rejected using RANSAC algorithm and is more than the first preset value
Angle point, obtain Corner；
According to the coordinate of the Corner and corresponding world coordinates, the generation where the gridiron pattern uncalibrated image is obtained
Homography matrix of boundary's coordinate system to pixel coordinate system；
Homography matrix of second error amount more than the second preset value is rejected using RANSAC algorithm, obtains mesh
Mark homography matrix；
According to the target homography matrix, target conic model is obtained, and according to the target conic section mould
Type obtains the initial parameter of camera, and the initial parameter includes the first internal reference, first ginseng and the first distortion factor outside；
The initial parameter is estimated using the method for maximal possibility estimation, obtains target component, the target ginseng
Number includes the second internal reference, second ginseng and the second distortion factor outside, and camera is demarcated using the target component.
Based on same inventive concept, fourth aspect present invention additionally provides a kind of computer equipment, including memory, place
Reason device and storage on a memory and the computer program that can run on a processor, reality during the computing device described program
Existing following steps：
Gridiron pattern uncalibrated image is obtained, the uncalibrated image includes multiple angle points；
The error amount of angle point first in the multiple angle point is rejected using RANSAC algorithm and is more than the first preset value
Angle point, obtain Corner；
According to the coordinate of the Corner and corresponding world coordinates, the generation where the gridiron pattern uncalibrated image is obtained
Homography matrix of boundary's coordinate system to pixel coordinate system；
Homography matrix of second error amount more than the second preset value is rejected using RANSAC algorithm, obtains mesh
Mark homography matrix；
According to the target homography matrix, target conic model is obtained, and according to the target conic section mould
Type obtains the initial parameter of camera, and the initial parameter includes the first internal reference, first ginseng and the first distortion factor outside；
The initial parameter is estimated using the method for maximal possibility estimation, obtains target component, the target ginseng
Number includes the second internal reference, second ginseng and the second distortion factor outside, and camera is demarcated using the target component.
The one or more technical schemes provided in the embodiment of the present invention, have at least the following technical effects or advantages：
The embodiment of the present application provides a kind of camera calibration method based on two-dimensional planar template, and methods described includes：Obtain
Gridiron pattern uncalibrated image is taken, the uncalibrated image includes multiple angle points；Rejected using RANSAC algorithm the multiple
The error amount of angle point first is more than the angle point of the first preset value in angle point, obtains Corner；According to the coordinate of the Corner
With corresponding world coordinates, the world coordinate system where the gridiron pattern uncalibrated image is obtained to the homography square of pixel coordinate system
Battle array；Homography matrix of second error amount more than the second preset value is rejected using RANSAC algorithm, obtains target list
Answering property matrix；According to the target homography matrix, target conic model is obtained, and according to the target conic section mould
Type obtains the initial parameter of camera, and the initial parameter includes the first internal reference, first ginseng and the first distortion factor outside；Using maximum
The method of possibility predication estimated the initial parameter, obtains target component, and the target component includes the second internal reference, the
Ginseng and the second distortion factor, are demarcated using the target component to camera outside two.In the above method provided by the invention, one
Aspect, when obtaining initial value, it is pre- more than first that angle point error amount in multiple angle points is eliminated using RANSAC algorithm
If value angle point, so as to obtain Corner, then further according to Corner calculate homography matrix, due to eliminate precision compared with
Low angle point, so as to improve homography matrix calculating accuracy, so as to improve stated accuracy, on the other hand, using with
Machine sampling consistency algorithm eliminates the homography matrix that error amount is more than the second preset value, so as to obtain target homography square
Then calibrating parameters are solved by battle array further according to the target homography matrix, due to eliminating low-quality uncalibrated image,
So as to further improve the precision of demarcation, solve accuracy and Shandong that existing Zhang Zhengyou scaling methods have camera calibration
The not high technical problem of rod.
Described above is only the general introduction of technical solution of the present invention, in order to better understand the technological means of the present invention,
And can be practiced according to the content of specification, and in order to allow above and other objects of the present invention, feature and advantage can
Become apparent, below especially exemplified by the embodiment of the present invention.
Brief description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
There is the required accompanying drawing used in technology description to be briefly described, it should be apparent that, drawings in the following description are this hairs
Some bright embodiments, for those of ordinary skill in the art, on the premise of not paying creative work, can be with root
Other accompanying drawings are obtained according to these accompanying drawings.
Fig. 1 is a kind of flow chart of the camera calibration method based on two-dimensional planar template in the embodiment of the present invention；
Fig. 2 is a kind of structure chart of the camera calibration system based on two-dimensional planar template in the embodiment of the present invention；
Fig. 3 is a kind of structure chart of computer-readable medium in the embodiment of the present invention；
Fig. 4 is a kind of structure chart of computer equipment in the embodiment of the present invention.
Embodiment
It is existing to solve the embodiments of the invention provide a kind of camera calibration method and system based on two-dimensional planar template
With the presence of the not high technical problem of the accuracy of Zhang Zhengyou scaling method camera calibrations.
Technical scheme in the embodiment of the present application, general thought are as follows：
A kind of camera calibration method based on two-dimensional planar template, methods described include：Obtain gridiron pattern uncalibrated image, institute
State uncalibrated image and include multiple angle points；The error amount of angle point first in the multiple angle point is rejected using RANSAC algorithm
More than the angle point of the first preset value, Corner is obtained；According to the coordinate of the Corner and corresponding world coordinates, obtain
The homography matrix of world coordinate system where the gridiron pattern uncalibrated image to pixel coordinate system；Using random sampling uniformity
Algorithm rejects the homography matrix that the second error amount is more than the second preset value, obtains target homography matrix；According to the target
Homography matrix, target conic model is obtained, and the initial parameter of camera is obtained according to the target conic model,
The initial parameter includes the first internal reference, first ginseng and the first distortion factor outside；Using the method for maximal possibility estimation to described
Initial parameter is estimated, obtains target component, and the target component includes the second internal reference, second ginseng and the second distortion system outside
Number, is demarcated using the target component to camera.
In the above-mentioned methods, on the one hand, when obtaining initial value, multiple angle points are eliminated using RANSAC algorithm
Middle angle point error amount is more than the angle point of the first preset value, so as to obtain Corner, then calculates single answer further according to Corner
Property matrix, the angle point relatively low due to eliminating precision, so as to improve homography matrix calculating accuracy, so as to improve mark
Determine precision, on the other hand, homography matrix of the error amount more than the second preset value eliminated using RANSAC algorithm,
So as to obtain target homography matrix, then calibrating parameters are solved further according to the target homography matrix, due to picking
Except low-quality uncalibrated image, so as to further improve the precision of demarcation, solve existing Zhang Zhengyou scaling methods and deposit
In the not high technical problem of the accuracy and robustness of camera calibration.
To make the purpose, technical scheme and advantage of the embodiment of the present invention clearer, below in conjunction with the embodiment of the present invention
In accompanying drawing, the technical scheme in the embodiment of the present invention is clearly and completely described, it is clear that described embodiment is
Part of the embodiment of the present invention, rather than whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art
The every other embodiment obtained under the premise of creative work is not made, belongs to the scope of protection of the invention.
Embodiment one
A kind of camera calibration method based on two-dimensional planar template is present embodiments provided, refers to Fig. 1, methods described bag
Include：
Step S101：Gridiron pattern uncalibrated image is obtained, the uncalibrated image includes multiple angle points；
Step S102：The error amount of angle point first in the multiple angle point is rejected using RANSAC algorithm and is more than the
The angle point of one preset value, obtain Corner.
Step S103：According to the coordinate of the Corner and corresponding world coordinates, the gridiron pattern calibration maps are obtained
As place world coordinate system to pixel coordinate system homography matrix.
Step S104：Homography square of second error amount more than the second preset value is rejected using RANSAC algorithm
Battle array, obtain target homography matrix.
Step S105：According to the target homography matrix, target conic model is obtained, and according to the target two
Secondary curve model obtains the initial parameter of camera, and the initial parameter includes the first internal reference, first ginseng and the first distortion factor outside；
Step S106：The initial parameter is estimated using the method for maximal possibility estimation, obtains target component, institute
Stating target component includes the second internal reference, second ginseng and the second distortion factor outside, and camera is demarcated using the target component.
It should be noted that present invention applicant is had found by long-term practice, the Zhang Zhengyou side based on plane reference
In method, it is mainly relevant with lower two aspects to solve the precision of initial value, and first, the extraction accuracy of X-comers, second, collection
The quality of uncalibrated image.And in actual applications, camera calibration process can by noise, image is fuzzy, illumination variation and chessboard
Lattice visual angle etc. influences, and causes the extraction accuracy of X-comers relatively low.In addition, meet the uncalibrated image ability of the high quality of specification
Accurate calibration is carried out to camera parameter, such as：The angle of gridiron pattern plane and camera image plane, which is more than 45 degree, can reduce demarcation essence
Degree, gridiron pattern ratio in the picture and gridiron pattern all can produce shadow apart from the far and near of camera photocentre to the precision for solving initial value
Ring.The present processes are based on understanding above, it is proposed that a kind of camera calibration method based on two-dimensional planar template, the present invention
Using double-RANSAC algorithm (D-RANSAC), start with terms of two of initial value solving precision are influenceed, pass through rejecting
The angle point of low precision and the low-quality image for not conforming to specification, so as to improve the solving precision of initial value.Specially：Utilize every width chessboard
The re-projection error size of lattice angle point, and RANSAC algorithm is used, the larger angle point of error is rejected, is then counted again
Calculate the homography matrix of each image.And based on default conic model theory constant in single camera, to every width figure
The homography matrix as corresponding to uses RANSAC algorithm again, rejects the low quality uncalibrated image for not meeting specification.
The initial inside and outside ginseng (i.e. the first internal reference and the second internal reference) of accurate initial value, i.e. camera is finally tried to achieve using closing solution, and
Lens distortion factor is attached to, the second internal reference of accurate camera is tried to achieve using the algorithm of non-linear search, ginseng and the outside second
Two distortion factors, and camera is demarcated, so as to improve the accuracy of demarcation and precision.
Below, a kind of camera calibration method based on two-dimensional planar template provided with reference to Fig. 1 the application carries out detailed
Introduce：
Step S101 is first carried out：Gridiron pattern uncalibrated image is obtained, the uncalibrated image includes multiple angle points；
In specific implementation process, multiple gridiron pattern uncalibrated images can be shot using camera, wherein every gridiron pattern
Uncalibrated image includes multiple angle points.Preferably, the angle of different uncalibrated images and calibration for cameras differs, such as uncalibrated image
It can with multi-angle shoot, ensure that uncalibrated image has enough angle changes with respect to camera, so as to ensure the precision of demarcation, demarcation
The quantity of image can be shot according to actual conditions, such as can be 20,25,30 etc..
Then step S102 is performed：The error of angle point first in the multiple angle point is rejected using RANSAC algorithm
Value obtains Corner more than the angle point of the first preset value.
In specific implementation process, because the precision of X-comers influences final stated accuracy, it is necessary to reject low
The angle point of precision, it is big that the error amount of angle point first in the multiple angle point rejected using RANSAC algorithm in the present embodiment
In the angle point of the first preset value, so as to obtain Corner.
Rejecting the method for low precision angle point can be realized by following step：
Obtain the re-projection coordinate of angle point；
Obtain the first distance between the re-projection coordinate of the angle point and the initial coordinate of the angle point；
First distance is rejected more than the angle point corresponding to the first preset value using RANSAC algorithm, so that
Obtain Corner.
In specific implementation process, camera model is introduced first, camera model is divided into linear model and nonlinear model,
Wherein the linear model based on two-dimensional planar template is as follows：
Wherein
In above-mentioned formula, s represents scale factor, and (u, v) is the pixel of camera imaging point corresponding to gridiron pattern image angle point
Coordinate (i.e. angular coordinate), (X, Y) be gridiron pattern image angle point world coordinates, K be camera intrinsic parameter, r1,r2Outside for camera
Parameter spin matrix R3×3Column vector, t represent Camera extrinsic number translation matrix, R3×3Where t expression gridiron pattern scaling boards
Rotation and translation matrix of the world coordinate system to camera coordinates system.
Next introduce in homography matrix, such as known every width uncalibrated image and camera imaging four pairs of mapping points (u,
V) and (X, Y), then the world coordinates where gridiron pattern scaling board can be obtained by formula (2) to the homography square of pixel coordinate
Battle array H, a homography matrix can be obtained for every width uncalibrated image, as follows：
Wherein
Wherein, homography matrix H has 8 frees degree.
Its angular coordinate (u, v) can be extracted to the chessboard table images of acquisition, then using its corresponding world coordinates (X,
Y) and formula (2) calculates the homography matrix H of the chessboard table images, due to influence meetings such as noise, the fuzzy, illumination variations of image
Cause the angular coordinate and inaccurate of extraction, thus the matrix of the above-mentioned calculating drawnAlso it is inaccurate, therefore public affairs can be passed through
Formula (2) obtains the coordinate of angle point re-projectionAnd be compared with the initial coordinate (u, v) of the angle point, between the two
Euclidean distance is re-projection errord1Value it is smaller, show that the extraction accuracy of angle point is got over
Height, if the error amount is more than the first preset value, show that the precision of the angle point is not high, so as to reject.For example, by first
Preset value is arranged to δ1If d1<δ1, then the angle point meet model needs, be referred to as in angle point, and record and meet the angle of calibration request
Point quantity t1；Otherwise the angle point does not meet model needs, is referred to as outer angle point.In order to ensure accuracy, repetition said process is multiple,
The number repeated is determined by following formula：
N=log (1-p)/log (1-ws)(4)
Above formula represents in n times iteration that it is s to randomly choose smallest sample number every time, then at least once without exterior point
Probability is p, and p values 0.99, the optional sample of w expressions is the probability of interior point.Then t is retained1It is corresponding when maximum to meet
All angle points of model, homography matrix is then reevaluated by least square method using these angle pointsSo as to obtain most
Excellent model.
Next step S103 is performed：According to the coordinate of the Corner and corresponding world coordinates, the chess is obtained
The homography matrix of world coordinate system where disk lattice uncalibrated image to pixel coordinate system.
In specific implementation process, the world where the gridiron pattern uncalibrated image can be obtained by formula (2) is sat
Homography matrix of the mark system to pixel coordinate system.
Specifically, the coordinate according to the Corner and corresponding world coordinates, the gridiron pattern demarcation is obtained
World coordinate system where image to pixel coordinate system homography matrix, including：
Four angle points are randomly selected from the gridiron pattern uncalibrated image, and obtain the coordinate of four angle points；
According to world coordinates corresponding to the coordinate and four angle points of four angle points, the gridiron pattern uncalibrated image is obtained
Homography matrix of the world coordinate system at place to pixel coordinate system.
In specific implementation process, 4 angular coordinates (u, v) are randomly choosed in every width uncalibrated image, pass through formula
(2) and corresponding world coordinates (X, Y), so as to draw homography matrix corresponding with uncalibrated image
Then step S104 is performed：Second error amount is rejected using RANSAC algorithm and is more than the second preset value
Homography matrix, obtain target homography matrix.
Specifically, the list that the second preset value is more than using RANSAC algorithm the second error amount of rejecting should
Property matrix, obtain target homography matrix, including：
According to homography matrix, default conic model is obtained；
The second distance of the homography matrix and the conic model is obtained, using the second distance as poor
Value；
If the difference is more than second preset value, for ineligible homography matrix；
The ineligible homography matrix is rejected from original homography matrix, obtains target homography matrix.
In specific implementation process, internal reference matrix K and default conic section curve model B, Zhang Zhengyou marks are introduced first
Determine in method, utilize r1,r2This orthogonal known conditions of unit, can obtain following two constraints：
h1 TK-TK-1h2=0
h1 TK-TK-1h1=h2 TK-TK-1h2(5)
Wherein, h1,h2For homography matrix H column vector, K is camera internal reference matrix, B=K-TK-1To preset conic section
Curve model (abbreviation ICA), from B expression formula, B only it is relevant with the internal reference matrix K of camera, and with the direction and position of camera
Put unrelated.
Obtain the method for target homography matrix first, can randomly select two homography first by following step
Matrix, ICA, i.e. B=K are calculated by formula (5)-TK-1, each homography matrix is then calculated respectivelyWith ICA distance
d2, wherein d2=(h1 TK-TK-1h2)2+(h1 TK-TK-1h1-1)2+(h2 TK-TK-1h2-1)2(6)；Second preset value is δ2If d2<
δ2, then the homography matrix meet model needs, be referred to as in homography matrix, and write down the homography matrix number for meeting model
t2；Otherwise the homography matrix does not meet model, is referred to as outer homography matrix.In order to ensure accuracy, said process N is repeated2It is secondary
(N2Determined by formula 4), retain t2Correspondingly meet all homography matrixes of model when maximumThese homography matrixes are made
For target homography matrix.
Next step S105 is performed：According to the target homography matrix, target conic model is obtained, and according to
The target conic model obtains the initial parameter of camera, and the initial parameter includes joining outside the first internal reference and first.
In specific implementation process, the target homography matrix that is obtained using abovementioned steps, and use least square method
Reevaluate model and obtain accurate ICA models, the initial parameter of camera is then linearly calculated.
Step S106 is most performed later：The initial parameter is estimated using the method for maximal possibility estimation, obtained
Target component, the target component includes the second internal reference, second ginseng and the second distortion factor outside, using the target component to phase
Machine is demarcated.
In specific implementation process, solved by the use of the initial parameter that step S105 is obtained as closing, and these are closed
Solution is used as initial value, and then all initial parameters are estimated by the method for maximal possibility estimation, including the first internal reference, first
Outer ginseng and the first distortion factor.And by the first distortion factor k1,k2Initial value is arranged to zero.Obtained using formula (7) optimization accurate
Internal reference A, outer ginseng k1,k2, the second distortion factor Ri,ti
Wherein, eventually for the amount of images of demarcation, m after n expressions rejecting low quality uncalibrated imageiRepresent that the i-th width is demarcated
Rejected in image and be used for the quantity for calculating homography matrix angle point, m after low precision angle pointijRepresent j-th of angle point in the i-th width image
Coordinate, MjRepresent mijCorresponding the known world coordinate,Represent MjRe-projection coordinate.The closing obtained using abovementioned steps
Solve A, RiAnd k1=0, k2=0 is used as initial value, and formula (7) is asked by Levenberg-Marquarat iteration optimization algorithms
Solution, so as to obtain accurate target component, the target component includes A, Ri,ti,k1,k2, i.e. the second internal reference of camera, second
Outer ginseng and the second distortion factor, then camera is demarcated by above-mentioned target component, complete the high-precision calibrating of camera.
Based on same inventive concept, the embodiment of the present invention additionally provides a kind of and camera mark based on two-dimensional planar template
The corresponding system of method is determined, referring specifically to embodiment two.
Embodiment two
The embodiment of the present invention two provides a kind of camera calibration system based on two-dimensional planar template, refers to Fig. 2, described
System includes：
Acquisition module 201, for obtaining gridiron pattern uncalibrated image, the uncalibrated image includes multiple angle points；
First obtains module 202, for rejecting angle point first in the multiple angle point using RANSAC algorithm
Error amount is more than the angle point of the first preset value, obtains Corner；
Second obtains module 203, for the coordinate according to the Corner and corresponding world coordinates, obtains the chess
The homography matrix of world coordinate system where disk lattice uncalibrated image to pixel coordinate system；
3rd obtains module 204, default more than second for rejecting the second error amount using RANSAC algorithm
The homography matrix of value, obtain target homography matrix；
4th obtains module 205, for according to the target homography matrix, obtaining target conic model, and root
Obtain the initial parameter of camera according to the target conic model, the initial parameter includes the first internal reference, outside first ginseng and
First distortion factor；
Demarcating module 206, for estimating using the method for maximal possibility estimation the initial parameter, obtain target
Parameter, the target component are included the second internal reference, second ginseng and the second distortion factor outside, camera are entered using the target component
Rower is determined.
In the system that the present embodiment provides, described first obtains mould 202, is additionally operable to：
Obtain the re-projection coordinate of angle point；
Obtain the first distance between the re-projection coordinate of the angle point and the initial coordinate of the angle point；
First distance is rejected more than the angle point corresponding to the first preset value using RANSAC algorithm, so that
Obtain Corner.
In the system that the present embodiment provides, described second obtains module 203, is additionally operable to：
Four angle points are randomly selected from the gridiron pattern uncalibrated image, and obtain the coordinate of four angle points；
According to world coordinates corresponding to the coordinate and four angle points of four angle points, the gridiron pattern uncalibrated image is obtained
Homography matrix of the world coordinate system at place to pixel coordinate system.
In the system that the present embodiment provides, the 3rd acquisition module 204 is additionally operable to：
According to homography matrix, default conic model is obtained；
The second distance of the homography matrix and the conic model is obtained, using the second distance as poor
Value；
If the difference is more than second preset value, for ineligible homography matrix；
The ineligible homography matrix is rejected from original homography matrix, obtains target homography matrix.
The present embodiment provide system in, the gridiron pattern uncalibrated image includes multiple, and different uncalibrated images and
The angle of calibration for cameras differs.
In embodiment one based on various change mode and the instantiation system that is equally applicable to the present embodiment, pass through
Foregoing pair of detailed description, those skilled in the art are clear that in the present embodiment, so the letter for specification
It is clean, it will not be described in detail herein.
Based on same inventive concept, the embodiment of the present invention additionally provides a kind of and camera mark based on two-dimensional planar template
The corresponding computer-readable medium of method is determined, referring specifically to embodiment three.
Embodiment three
The embodiment of the present invention three provides a kind of computer-readable recording medium 300, refers to Fig. 3, is stored thereon with meter
Calculation machine program 301, the program realizes following steps when being executed by processor：
Gridiron pattern uncalibrated image is obtained, the uncalibrated image includes multiple angle points；
The error amount of angle point first in the multiple angle point is rejected using RANSAC algorithm and is more than the first preset value
Angle point, obtain Corner；
According to the coordinate of the Corner and corresponding world coordinates, the generation where the gridiron pattern uncalibrated image is obtained
Homography matrix of boundary's coordinate system to pixel coordinate system；
Homography matrix of second error amount more than the second preset value is rejected using RANSAC algorithm, obtains mesh
Mark homography matrix；
According to the target homography matrix, target conic model is obtained, and according to the target conic section mould
Type obtains the initial parameter of camera, and the initial parameter includes the first internal reference, first ginseng and the first distortion factor outside；
The initial parameter is estimated using the method for maximal possibility estimation, obtains target component, the target ginseng
Number includes the second internal reference, second ginseng and the second distortion factor outside, and camera is demarcated using the target component.
The various change mode and instantiation of scaling method in embodiment one are equally applicable to the calculating of the present embodiment
Machine computer-readable recording medium, by the foregoing detailed description to camera calibration method, those skilled in the art are clear that this reality
The computer-readable medium in example is applied, thus it is succinct for specification, it will not be described in detail herein.
Based on same inventive concept, the embodiment of the present invention additionally provides a kind of and camera mark based on two-dimensional planar template
The corresponding computer equipment of method is determined, referring specifically to example IV.
Example IV
The embodiment of the present invention three provides a kind of computer equipment, refers to Fig. 4, including memory 401, processor and deposits
Storage realizes following step on a memory and the computer program that can run on a processor, during the computing device described program
Suddenly：
Gridiron pattern uncalibrated image is obtained, the uncalibrated image includes multiple angle points；
The error amount of angle point first in the multiple angle point is rejected using RANSAC algorithm and is more than the first preset value
Angle point, obtain Corner；
According to the coordinate of the Corner and corresponding world coordinates, the generation where the gridiron pattern uncalibrated image is obtained
Homography matrix of boundary's coordinate system to pixel coordinate system；
Homography matrix of second error amount more than the second preset value is rejected using RANSAC algorithm, obtains mesh
Mark homography matrix；
According to the target homography matrix, target conic model is obtained, and according to the target conic section mould
Type obtains the initial parameter of camera, and the initial parameter includes the first internal reference, first ginseng and the first distortion factor outside；
The initial parameter is estimated using the method for maximal possibility estimation, obtains target component, the target ginseng
Number includes the second internal reference, second ginseng and the second distortion factor outside, and camera is demarcated using the target component.
For convenience of description, Fig. 4 illustrate only the part related to the embodiment of the present invention, and particular technique details does not disclose
, it refer to present invention method part.Wherein, memory 401 can be used for storage software program and module, processor
402 perform the software program and module that are stored in memory 401 by running, should so as to perform the various functions of mobile terminal
With and data processing.
Memory 401 can mainly include storing program area and storage data field, wherein, storing program area can store operation system
Application program needed for system, at least one function etc.；Storage data field can store uses what is created according to computer equipment
Data etc..The control centre of the mobile communication terminal of processor 402, utilize various interfaces and the whole mobile communication terminal of connection
Various pieces, by running or performing the software program and/or module that are stored in memory 401, and call and be stored in
Data in memory 401, the various functions and processing data of mobile phone are performed, it is overall so as to be carried out to mobile phone
Monitoring.Optionally, processor 402 may include one or more processing units.
The various change mode and instantiation of scaling method in embodiment one are equally applicable to the calculating of the present embodiment
Machine equipment, by the foregoing detailed description to camera calibration method, those skilled in the art are clear that the present embodiment
In computer equipment, so succinct for specification, will not be described in detail herein.
The one or more technical schemes provided in the embodiment of the present invention, have at least the following technical effects or advantages：
The embodiment of the present application provides a kind of camera calibration method based on two-dimensional planar template, and methods described includes：Obtain
Gridiron pattern uncalibrated image is taken, the uncalibrated image includes multiple angle points；Rejected using RANSAC algorithm the multiple
The error amount of angle point first is more than the angle point of the first preset value in angle point, obtains Corner；According to the coordinate of the Corner
With corresponding world coordinates, the world coordinate system where the gridiron pattern uncalibrated image is obtained to the homography square of pixel coordinate system
Battle array；Homography matrix of second error amount more than the second preset value is rejected using RANSAC algorithm, obtains target list
Answering property matrix；According to the target homography matrix, target conic model is obtained, and according to the target conic section mould
Type obtains the initial parameter of camera, and the initial parameter includes joining outside the first internal reference and first；Using the side of maximal possibility estimation
Method is estimated the initial parameter, obtains target component, and the target component includes the second internal reference, second and joins and distort outside
Coefficient, camera is demarcated using the target component.In the above method provided by the invention, on the one hand, obtaining initial value
When, the angle point that angle point error amount in multiple angle points is more than the first preset value is eliminated using RANSAC algorithm, so as to
Obtain Corner, then further according to Corner calculate homography matrix, the angle point relatively low due to eliminating precision, so as to
To improve the accuracy of homography matrix calculating, so as to improve stated accuracy, on the other hand, using RANSAC algorithm
The homography matrix that error amount is more than the second preset value is eliminated, so as to obtain target homography matrix, then further according to described
Target homography matrix solves to calibrating parameters, due to eliminating low-quality uncalibrated image, so as to further carry
The precision of height demarcation, solves that existing Zhang Zhengyou scaling methods have the accuracy of camera calibration and the not high technology of robustness is asked
Topic.
Although preferred embodiments of the present invention have been described, but those skilled in the art once know basic creation
Property concept, then can make other change and modification to these embodiments.So appended claims be intended to be construed to include it is excellent
Select embodiment and fall into having altered and changing for the scope of the invention.
Obviously, those skilled in the art can carry out various changes and modification without departing from this hair to the embodiment of the present invention
The spirit and scope of bright embodiment.So, if these modifications and variations of the embodiment of the present invention belong to the claims in the present invention
And its within the scope of equivalent technologies, then the present invention is also intended to comprising including these changes and modification.
Claims (10)
- A kind of 1. camera calibration method based on two-dimensional planar template, it is characterised in that methods described includes：Gridiron pattern uncalibrated image is obtained, the uncalibrated image includes multiple angle points；It is more than the angle of the first preset value using the error amount of angle point first in the multiple angle point of RANSAC algorithm rejecting Point, obtain Corner；According to the coordinate of the Corner and corresponding world coordinates, the world where obtaining the gridiron pattern uncalibrated image is sat Homography matrix of the mark system to pixel coordinate system；Homography matrix of second error amount more than the second preset value is rejected using RANSAC algorithm, obtains target list Answering property matrix；According to the target homography matrix, target conic model is obtained, and obtain according to the target conic model The initial parameter of camera is obtained, the initial parameter includes the first internal reference, first ginseng and the first distortion factor outside；The initial parameter is estimated using the method for maximal possibility estimation, obtains target component, the target component bag The second internal reference, second ginseng and the second distortion factor outside are included, camera is demarcated using the target component.
- 2. the method as described in claim 1, it is characterised in that described the multiple using RANSAC algorithm rejecting The error amount of angle point first is more than the angle point of the first preset value in angle point, obtains Corner, including：Obtain the re-projection coordinate of angle point；Obtain the first distance between the re-projection coordinate of the angle point and the initial coordinate of the angle point；First distance is rejected more than the angle point corresponding to the first preset value using RANSAC algorithm, so as to obtain Corner.
- 3. the method as described in claim 1, it is characterised in that the coordinate according to the Corner and the corresponding world Coordinate, the world coordinate system where the gridiron pattern uncalibrated image is obtained to the homography matrix of pixel coordinate system, including：Four angle points are randomly selected from the gridiron pattern uncalibrated image, and obtain the coordinate of four angle points；According to world coordinates corresponding to the coordinate and four angle points of four angle points, the gridiron pattern uncalibrated image place is obtained World coordinate system to pixel coordinate system homography matrix.
- 4. the method as described in claim 1, it is characterised in that described that second error is rejected using RANSAC algorithm Value is more than the homography matrix of the second preset value, obtains target homography matrix, including：According to homography matrix, default conic model is obtained；The second distance of the homography matrix and the conic model is obtained, using the second distance as difference；If the difference is more than second preset value, for ineligible homography matrix；The ineligible homography matrix is rejected from original homography matrix, obtains target homography matrix.
- 5. the method as described in any one of claim 1-4 claim, it is characterised in that the gridiron pattern uncalibrated image includes Multiple, and the angle of different uncalibrated images and calibration for cameras differs.
- 6. a kind of camera calibration system based on two-dimensional planar template, it is characterised in that the system includes：Acquisition module, for obtaining gridiron pattern uncalibrated image, the uncalibrated image includes multiple angle points；First obtains module, big for rejecting the error amount of angle point first in the multiple angle point using RANSAC algorithm In the angle point of the first preset value, Corner is obtained；Second obtains module, for the coordinate according to the Corner and corresponding world coordinates, obtains the chessboard case marker The world coordinate system where image is determined to the homography matrix of pixel coordinate system；3rd obtains module, should more than the list of the second preset value for rejecting the second error amount using RANSAC algorithm Property matrix, obtain target homography matrix；4th obtains module, for according to the target homography matrix, obtaining target conic model, and according to the mesh The initial parameter that conic model obtains camera is marked, the initial parameter includes the first internal reference, first ginseng and the first distortion outside Coefficient；Demarcating module, for estimating using the method for maximal possibility estimation the initial parameter, obtain target component, institute Stating target component includes the second internal reference, second ginseng and the second distortion factor outside, and camera is demarcated using the target component.
- 7. system as claimed in claim 6, it is characterised in that described first obtains module, is additionally operable to：Obtain the re-projection coordinate of angle point；Obtain the first distance between the re-projection coordinate of the angle point and the initial coordinate of the angle point；First distance is rejected more than the angle point corresponding to the first preset value using RANSAC algorithm, so as to obtain Corner.
- 8. method as claimed in claim 6, it is characterised in that described second obtains module, is additionally operable to：Four angle points are randomly selected from the gridiron pattern uncalibrated image, and obtain the coordinate of four angle points；According to world coordinates corresponding to the coordinate and four angle points of four angle points, the gridiron pattern uncalibrated image place is obtained World coordinate system to pixel coordinate system homography matrix.
- 9. a kind of computer-readable recording medium, is stored thereon with computer program, it is characterised in that the program is held by processor Following steps are realized during row：Gridiron pattern uncalibrated image is obtained, the uncalibrated image includes multiple angle points；It is more than the angle of the first preset value using the error amount of angle point first in the multiple angle point of RANSAC algorithm rejecting Point, obtain Corner；According to the coordinate of the Corner and corresponding world coordinates, the world where obtaining the gridiron pattern uncalibrated image is sat Homography matrix of the mark system to pixel coordinate system；Homography matrix of second error amount more than the second preset value is rejected using RANSAC algorithm, obtains target list Answering property matrix；According to the target homography matrix, target conic model is obtained, and obtain according to the target conic model The initial parameter of camera is obtained, the initial parameter includes the first internal reference, first ginseng and the first distortion factor outside；The initial parameter is estimated using the method for maximal possibility estimation, obtains target component, the target component bag The second internal reference, second ginseng and the second distortion factor outside are included, camera is demarcated using the target component.
- 10. a kind of computer equipment, including memory, processor and storage are on a memory and the meter that can run on a processor Calculation machine program, it is characterised in that realize following steps during the computing device described program：Gridiron pattern uncalibrated image is obtained, the uncalibrated image includes multiple angle points；It is more than the angle of the first preset value using the error amount of angle point first in the multiple angle point of RANSAC algorithm rejecting Point, obtain Corner；According to the coordinate of the Corner and corresponding world coordinates, the world where obtaining the gridiron pattern uncalibrated image is sat Homography matrix of the mark system to pixel coordinate system；Homography matrix of second error amount more than the second preset value is rejected using RANSAC algorithm, obtains target list Answering property matrix；According to the target homography matrix, target conic model is obtained, and obtain according to the target conic model The initial parameter of camera is obtained, the initial parameter includes the first internal reference, first ginseng and the first distortion factor outside；The initial parameter is estimated using the method for maximal possibility estimation, obtains target component, the target component bag The second internal reference, second ginseng and the second distortion factor outside are included, camera is demarcated using the target component.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710674279.5A CN107633536B (en) | 2017-08-09 | 2017-08-09 | Camera calibration method and system based on two-dimensional plane template |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710674279.5A CN107633536B (en) | 2017-08-09 | 2017-08-09 | Camera calibration method and system based on two-dimensional plane template |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107633536A true CN107633536A (en) | 2018-01-26 |
| CN107633536B CN107633536B (en) | 2020-04-17 |
Family
ID=61099131
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710674279.5A Expired - Fee Related CN107633536B (en) | 2017-08-09 | 2017-08-09 | Camera calibration method and system based on two-dimensional plane template |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107633536B (en) |
Cited By (33)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108447095A (en) * | 2018-01-31 | 2018-08-24 | 潍坊歌尔电子有限公司 | A kind of fisheye camera scaling method and device |
| CN108510551A (en) * | 2018-04-25 | 2018-09-07 | 上海大学 | Method and system for calibrating camera parameters under long-distance large-field-of-view condition |
| CN108682037A (en) * | 2018-04-27 | 2018-10-19 | 腾讯科技（深圳）有限公司 | Method for relocating, device, equipment and the storage medium of camera posture tracing process |
| CN108734744A (en) * | 2018-04-28 | 2018-11-02 | 国网山西省电力公司电力科学研究院 | A kind of remote big field-of-view binocular scaling method based on total powerstation |
| CN108876854A (en) * | 2018-04-27 | 2018-11-23 | 腾讯科技（深圳）有限公司 | Method for relocating, device, equipment and the storage medium of camera posture tracing process |
| CN108895959A (en) * | 2018-04-27 | 2018-11-27 | 电子科技大学 | A kind of camera calibration plate angle point calculating method based on sub-pix |
| CN109003312A (en) * | 2018-08-24 | 2018-12-14 | 重庆邮电大学 | A kind of camera calibration method based on nonlinear optimization |
| CN109146932A (en) * | 2018-07-17 | 2019-01-04 | 北京旷视科技有限公司 | Determine the methods, devices and systems of the world coordinates of target point in image |
| CN109242910A (en) * | 2018-08-21 | 2019-01-18 | 电子科技大学 | A kind of monocular camera self-calibrating method based on any known flat shape |
| CN109521781A (en) * | 2018-10-30 | 2019-03-26 | 普宙飞行器科技（深圳）有限公司 | Unmanned plane positioning system, unmanned plane and unmanned plane localization method |
| CN109658465A (en) * | 2018-12-07 | 2019-04-19 | 广州华端科技有限公司 | Data processing, image rebuilding method and device in image reconstruction process |
| CN109753930A (en) * | 2019-01-03 | 2019-05-14 | 京东方科技集团股份有限公司 | Method for detecting human face and face detection system |
| CN110298892A (en) * | 2019-07-05 | 2019-10-01 | 河南科技大学 | A kind of single line array camera inside and outside parameter scaling method |
| CN110345875A (en) * | 2018-04-04 | 2019-10-18 | 灵动科技（北京）有限公司 | Calibration and distance measuring method, device, electronic equipment and computer readable storage medium |
| CN110728714A (en) * | 2018-07-16 | 2020-01-24 | Oppo广东移动通信有限公司 | Image processing method and device, storage medium and electronic equipment |
| CN110866956A (en) * | 2019-10-28 | 2020-03-06 | 中国科学院深圳先进技术研究院 | Robot calibration method and terminal |
| CN111047614A (en) * | 2019-10-10 | 2020-04-21 | 南昌市微轲联信息技术有限公司 | Feature extraction-based method for extracting target corner of complex scene image |
| CN111311491A (en) * | 2020-01-20 | 2020-06-19 | 当家移动绿色互联网技术集团有限公司 | Image processing method and device, storage medium and electronic equipment |
| CN111429532A (en) * | 2020-04-30 | 2020-07-17 | 南京大学 | Method for improving camera calibration accuracy by utilizing multi-plane calibration plate |
| CN111566701A (en) * | 2020-04-02 | 2020-08-21 | 深圳市瑞立视多媒体科技有限公司 | Method, device and equipment for calibrating scanning field edge under large-space environment and storage medium |
| CN111652937A (en) * | 2019-03-04 | 2020-09-11 | 广州汽车集团股份有限公司 | Vehicle-mounted camera calibration method and device |
| CN111833404A (en) * | 2019-04-23 | 2020-10-27 | 鸿富锦精密电子(天津)有限公司 | Camera calibration system and camera calibration method |
| CN111951335A (en) * | 2020-08-13 | 2020-11-17 | 珠海格力电器股份有限公司 | Method, device, processor and image acquisition system for determining camera calibration parameters |
| CN112634375A (en) * | 2020-12-21 | 2021-04-09 | 杭州东信北邮信息技术有限公司 | Plane calibration and three-dimensional reconstruction method in AI intelligent detection |
| CN112907674A (en) * | 2019-11-19 | 2021-06-04 | 初速度（苏州）科技有限公司 | Camera calibration method and device for unmanned vehicle |
| CN112907677A (en) * | 2019-12-04 | 2021-06-04 | 杭州海康威视数字技术股份有限公司 | Camera calibration method and device for single-frame image and storage medium |
| CN113192145A (en) * | 2021-05-08 | 2021-07-30 | 深圳市商汤科技有限公司 | Equipment calibration method and device, electronic equipment and storage medium |
| CN113256725A (en) * | 2020-02-10 | 2021-08-13 | 武汉Tcl集团工业研究院有限公司 | Camera calibration method and device and storage medium |
| CN113284190A (en) * | 2021-06-09 | 2021-08-20 | 上海商汤临港智能科技有限公司 | Calibration method, calibration device, calibration equipment, storage medium and product |
| CN113516719A (en) * | 2021-04-21 | 2021-10-19 | 深圳臻像科技有限公司 | Camera calibration method, system and storage medium based on multiple homography matrixes |
| CN113706632A (en) * | 2021-08-31 | 2021-11-26 | 上海景吾智能科技有限公司 | Calibration method and system based on three-dimensional visual calibration plate |
| CN117226853A (en) * | 2023-11-13 | 2023-12-15 | 之江实验室 | Robot kinematics calibration method, device, storage medium and equipment |
| WO2024011764A1 (en) * | 2022-07-11 | 2024-01-18 | 深圳思谋信息科技有限公司 | Calibration parameter determination method and apparatus, hybrid calibration board, device, and medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090245690A1 (en) * | 2008-03-26 | 2009-10-01 | City University Of Hong Kong | Auto-calibration method for a projector-camera system |
| CN101673397A (en) * | 2009-09-30 | 2010-03-17 | 青岛大学 | Digital camera nonlinear calibration method based on LCDs |
| CN104851104A (en) * | 2015-05-29 | 2015-08-19 | 大连理工大学 | Flexible-target-based close-range large-field-of-view calibrate method of high-speed camera |
| CN105100640A (en) * | 2015-01-23 | 2015-11-25 | 武汉智源泉信息科技有限公司 | Local registration parallel video stitching method and local registration parallel video stitching system |
| CN105488810A (en) * | 2016-01-20 | 2016-04-13 | 东南大学 | Focused light field camera internal and external parameter calibration method |
-
2017
- 2017-08-09 CN CN201710674279.5A patent/CN107633536B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090245690A1 (en) * | 2008-03-26 | 2009-10-01 | City University Of Hong Kong | Auto-calibration method for a projector-camera system |
| CN101673397A (en) * | 2009-09-30 | 2010-03-17 | 青岛大学 | Digital camera nonlinear calibration method based on LCDs |
| CN105100640A (en) * | 2015-01-23 | 2015-11-25 | 武汉智源泉信息科技有限公司 | Local registration parallel video stitching method and local registration parallel video stitching system |
| CN104851104A (en) * | 2015-05-29 | 2015-08-19 | 大连理工大学 | Flexible-target-based close-range large-field-of-view calibrate method of high-speed camera |
| CN105488810A (en) * | 2016-01-20 | 2016-04-13 | 东南大学 | Focused light field camera internal and external parameter calibration method |
Non-Patent Citations (4)
| Title |
|---|
| ZUWHAN KIM等: "Camera calibration from orthogonally projected coordinates with noisy-RANSAC", 《IEEE》 * |
| 周剑军等: "基于RANSAC的图像拼接方法", 《计算机工程与设计》 * |
| 杨超等: "基于Harris角点和改进SURF描述子的图像快速拼接", 《现代电子技术》 * |
| 黄春燕: "双目立体相机标定算法研究与实现", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (56)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108447095A (en) * | 2018-01-31 | 2018-08-24 | 潍坊歌尔电子有限公司 | A kind of fisheye camera scaling method and device |
| CN110345875B (en) * | 2018-04-04 | 2021-04-27 | 灵动科技（北京）有限公司 | Calibration and ranging method, device, electronic equipment and computer readable storage medium |
| CN110345875A (en) * | 2018-04-04 | 2019-10-18 | 灵动科技（北京）有限公司 | Calibration and distance measuring method, device, electronic equipment and computer readable storage medium |
| CN108510551A (en) * | 2018-04-25 | 2018-09-07 | 上海大学 | Method and system for calibrating camera parameters under long-distance large-field-of-view condition |
| CN108876854A (en) * | 2018-04-27 | 2018-11-23 | 腾讯科技（深圳）有限公司 | Method for relocating, device, equipment and the storage medium of camera posture tracing process |
| CN108895959A (en) * | 2018-04-27 | 2018-11-27 | 电子科技大学 | A kind of camera calibration plate angle point calculating method based on sub-pix |
| CN108682037B (en) * | 2018-04-27 | 2022-03-08 | 腾讯科技（深圳）有限公司 | Method, device and equipment for relocating camera attitude tracking process and storage medium |
| CN108876854B (en) * | 2018-04-27 | 2022-03-08 | 腾讯科技（深圳）有限公司 | Method, device and equipment for relocating camera attitude tracking process and storage medium |
| US11276183B2 (en) | 2018-04-27 | 2022-03-15 | Tencent Technology (Shenzhen) Company Limited | Relocalization method and apparatus in camera pose tracking process, device, and storage medium |
| CN108895959B (en) * | 2018-04-27 | 2020-03-27 | 电子科技大学 | Camera calibration plate corner point calculation method based on sub-pixels |
| CN108682037A (en) * | 2018-04-27 | 2018-10-19 | 腾讯科技（深圳）有限公司 | Method for relocating, device, equipment and the storage medium of camera posture tracing process |
| CN108734744B (en) * | 2018-04-28 | 2022-02-18 | 国网山西省电力公司电力科学研究院 | Long-distance large-view-field binocular calibration method based on total station |
| CN108734744A (en) * | 2018-04-28 | 2018-11-02 | 国网山西省电力公司电力科学研究院 | A kind of remote big field-of-view binocular scaling method based on total powerstation |
| CN110728714A (en) * | 2018-07-16 | 2020-01-24 | Oppo广东移动通信有限公司 | Image processing method and device, storage medium and electronic equipment |
| CN110728714B (en) * | 2018-07-16 | 2023-06-20 | Oppo广东移动通信有限公司 | Image processing method and device, storage medium and electronic equipment |
| CN109146932A (en) * | 2018-07-17 | 2019-01-04 | 北京旷视科技有限公司 | Determine the methods, devices and systems of the world coordinates of target point in image |
| CN109242910A (en) * | 2018-08-21 | 2019-01-18 | 电子科技大学 | A kind of monocular camera self-calibrating method based on any known flat shape |
| CN109242910B (en) * | 2018-08-21 | 2021-10-08 | 电子科技大学 | Monocular camera self-calibration method based on any known plane shape |
| CN109003312B (en) * | 2018-08-24 | 2022-01-28 | 重庆邮电大学 | Camera calibration method based on nonlinear optimization |
| CN109003312A (en) * | 2018-08-24 | 2018-12-14 | 重庆邮电大学 | A kind of camera calibration method based on nonlinear optimization |
| CN109521781A (en) * | 2018-10-30 | 2019-03-26 | 普宙飞行器科技（深圳）有限公司 | Unmanned plane positioning system, unmanned plane and unmanned plane localization method |
| CN109658465A (en) * | 2018-12-07 | 2019-04-19 | 广州华端科技有限公司 | Data processing, image rebuilding method and device in image reconstruction process |
| CN109753930A (en) * | 2019-01-03 | 2019-05-14 | 京东方科技集团股份有限公司 | Method for detecting human face and face detection system |
| CN111652937B (en) * | 2019-03-04 | 2023-11-03 | 广州汽车集团股份有限公司 | Vehicle-mounted camera calibration method and device |
| CN111652937A (en) * | 2019-03-04 | 2020-09-11 | 广州汽车集团股份有限公司 | Vehicle-mounted camera calibration method and device |
| CN111833404B (en) * | 2019-04-23 | 2023-10-31 | 富联精密电子(天津)有限公司 | Camera correction system and camera correction method |
| CN111833404A (en) * | 2019-04-23 | 2020-10-27 | 鸿富锦精密电子(天津)有限公司 | Camera calibration system and camera calibration method |
| CN110298892B (en) * | 2019-07-05 | 2022-10-11 | 河南科技大学 | Method for calibrating internal and external parameters of single-line-array camera |
| CN110298892A (en) * | 2019-07-05 | 2019-10-01 | 河南科技大学 | A kind of single line array camera inside and outside parameter scaling method |
| CN111047614B (en) * | 2019-10-10 | 2023-09-29 | 南昌市微轲联信息技术有限公司 | Feature extraction-based method for extracting target corner of complex scene image |
| CN111047614A (en) * | 2019-10-10 | 2020-04-21 | 南昌市微轲联信息技术有限公司 | Feature extraction-based method for extracting target corner of complex scene image |
| CN110866956A (en) * | 2019-10-28 | 2020-03-06 | 中国科学院深圳先进技术研究院 | Robot calibration method and terminal |
| CN112907674B (en) * | 2019-11-19 | 2022-06-07 | 魔门塔(苏州)科技有限公司 | Camera calibration method and device for unmanned vehicle |
| CN112907674A (en) * | 2019-11-19 | 2021-06-04 | 初速度（苏州）科技有限公司 | Camera calibration method and device for unmanned vehicle |
| CN112907677A (en) * | 2019-12-04 | 2021-06-04 | 杭州海康威视数字技术股份有限公司 | Camera calibration method and device for single-frame image and storage medium |
| CN111311491A (en) * | 2020-01-20 | 2020-06-19 | 当家移动绿色互联网技术集团有限公司 | Image processing method and device, storage medium and electronic equipment |
| CN111311491B (en) * | 2020-01-20 | 2020-12-04 | 当家移动绿色互联网技术集团有限公司 | Image processing method and device, storage medium and electronic equipment |
| CN113256725B (en) * | 2020-02-10 | 2023-06-20 | 武汉Tcl集团工业研究院有限公司 | Camera calibration method, device and storage medium |
| CN113256725A (en) * | 2020-02-10 | 2021-08-13 | 武汉Tcl集团工业研究院有限公司 | Camera calibration method and device and storage medium |
| CN111566701A (en) * | 2020-04-02 | 2020-08-21 | 深圳市瑞立视多媒体科技有限公司 | Method, device and equipment for calibrating scanning field edge under large-space environment and storage medium |
| CN111566701B (en) * | 2020-04-02 | 2021-10-15 | 深圳市瑞立视多媒体科技有限公司 | Method, device and equipment for calibrating scanning field edge under large-space environment and storage medium |
| CN111429532A (en) * | 2020-04-30 | 2020-07-17 | 南京大学 | Method for improving camera calibration accuracy by utilizing multi-plane calibration plate |
| CN111951335A (en) * | 2020-08-13 | 2020-11-17 | 珠海格力电器股份有限公司 | Method, device, processor and image acquisition system for determining camera calibration parameters |
| CN112634375A (en) * | 2020-12-21 | 2021-04-09 | 杭州东信北邮信息技术有限公司 | Plane calibration and three-dimensional reconstruction method in AI intelligent detection |
| CN112634375B (en) * | 2020-12-21 | 2022-08-05 | 杭州东信北邮信息技术有限公司 | Plane calibration and three-dimensional reconstruction method in AI intelligent detection |
| CN113516719B (en) * | 2021-04-21 | 2023-08-15 | 深圳臻像科技有限公司 | Camera calibration method, system and storage medium based on multiple homography matrixes |
| CN113516719A (en) * | 2021-04-21 | 2021-10-19 | 深圳臻像科技有限公司 | Camera calibration method, system and storage medium based on multiple homography matrixes |
| CN113192145B (en) * | 2021-05-08 | 2023-02-28 | 深圳市商汤科技有限公司 | Equipment calibration method and device, electronic equipment and storage medium |
| CN113192145A (en) * | 2021-05-08 | 2021-07-30 | 深圳市商汤科技有限公司 | Equipment calibration method and device, electronic equipment and storage medium |
| CN113284190B (en) * | 2021-06-09 | 2023-04-25 | 上海临港绝影智能科技有限公司 | Calibration method, device, equipment, storage medium and product |
| CN113284190A (en) * | 2021-06-09 | 2021-08-20 | 上海商汤临港智能科技有限公司 | Calibration method, calibration device, calibration equipment, storage medium and product |
| CN113706632A (en) * | 2021-08-31 | 2021-11-26 | 上海景吾智能科技有限公司 | Calibration method and system based on three-dimensional visual calibration plate |
| CN113706632B (en) * | 2021-08-31 | 2024-01-16 | 上海景吾智能科技有限公司 | Calibration method and system based on three-dimensional vision calibration plate |
| WO2024011764A1 (en) * | 2022-07-11 | 2024-01-18 | 深圳思谋信息科技有限公司 | Calibration parameter determination method and apparatus, hybrid calibration board, device, and medium |
| CN117226853A (en) * | 2023-11-13 | 2023-12-15 | 之江实验室 | Robot kinematics calibration method, device, storage medium and equipment |
| CN117226853B (en) * | 2023-11-13 | 2024-02-06 | 之江实验室 | Robot kinematics calibration method, device, storage medium and equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107633536B (en) | 2020-04-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107633536A (en) | A kind of camera calibration method and system based on two-dimensional planar template | |
| CN111429532B (en) | Method for improving camera calibration accuracy by utilizing multi-plane calibration plate | |
| CN106803273B (en) | A kind of panoramic camera scaling method | |
| CN108765328B (en) | High-precision multi-feature plane template and distortion optimization and calibration method thereof | |
| CN106651942B (en) | Three-dimensional rotating detection and rotary shaft localization method based on characteristic point | |
| CN107705333B (en) | Space positioning method and device based on binocular camera | |
| CN104333675B (en) | A kind of panorama electronic image stabilization method based on spherical projection | |
| WO2021004416A1 (en) | Method and apparatus for establishing beacon map on basis of visual beacons | |
| CN107886547B (en) | Fisheye camera calibration method and system | |
| CN110443879B (en) | Perspective error compensation method based on neural network | |
| CN107507246A (en) | A kind of camera marking method based on improvement distortion model | |
| CN106971408B (en) | A kind of camera marking method based on space-time conversion thought | |
| CN108805987B (en) | Hybrid tracking method and device based on deep learning | |
| CN111383194B (en) | Polar coordinate-based camera distortion image correction method | |
| CN113012234B (en) | High-precision camera calibration method based on plane transformation | |
| CN106127261B (en) | A kind of fast multiresolution gray level image template matching method | |
| CN112212788A (en) | Visual space point three-dimensional coordinate measuring method based on multiple mobile phones | |
| CN109003312A (en) | A kind of camera calibration method based on nonlinear optimization | |
| CN101577004A (en) | Rectification method for polar lines, appliance and system thereof | |
| CN109974618A (en) | The overall calibration method of multisensor vision measurement system | |
| Chan et al. | An improved method for fisheye camera calibration and distortion correction | |
| CN115187612A (en) | Plane area measuring method, device and system based on machine vision | |
| CN109145929A (en) | One kind being based on SIFT scale space characteristics information extraction method | |
| CN107067441A (en) | Camera marking method and device | |
| CN112017259B (en) | Indoor positioning and image building method based on depth camera and thermal imager |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200417 Termination date: 20200809 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |