CN104794718B - A kind of method of single image CT center monitoring camera calibration - Google Patents
A kind of method of single image CT center monitoring camera calibration Download PDFInfo
- Publication number
- CN104794718B CN104794718B CN201510204298.2A CN201510204298A CN104794718B CN 104794718 B CN104794718 B CN 104794718B CN 201510204298 A CN201510204298 A CN 201510204298A CN 104794718 B CN104794718 B CN 104794718B
- Authority
- CN
- China
- Prior art keywords
- image
- camera
- dimensional
- world coordinate
- coordinate system
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 85
- 238000012544 monitoring process Methods 0.000 title claims abstract description 19
- 238000005457 optimization Methods 0.000 claims abstract description 45
- 239000011159 matrix material Substances 0.000 claims description 17
- 238000013507 mapping Methods 0.000 claims description 14
- 239000013598 vector Substances 0.000 claims description 13
- 238000012545 processing Methods 0.000 claims description 12
- 238000013519 translation Methods 0.000 claims description 12
- 238000001514 detection method Methods 0.000 claims description 7
- 239000007787 solid Substances 0.000 claims description 7
- 230000011218 segmentation Effects 0.000 claims description 6
- 230000003247 decreasing effect Effects 0.000 claims description 3
- 230000033001 locomotion Effects 0.000 description 6
- 230000009466 transformation Effects 0.000 description 6
- 230000006872 improvement Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 238000003384 imaging method Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- PCTMTFRHKVHKIS-BMFZQQSSSA-N (1s,3r,4e,6e,8e,10e,12e,14e,16e,18s,19r,20r,21s,25r,27r,30r,31r,33s,35r,37s,38r)-3-[(2r,3s,4s,5s,6r)-4-amino-3,5-dihydroxy-6-methyloxan-2-yl]oxy-19,25,27,30,31,33,35,37-octahydroxy-18,20,21-trimethyl-23-oxo-22,39-dioxabicyclo[33.3.1]nonatriaconta-4,6,8,10 Chemical compound C1C=C2C[C@@H](OS(O)(=O)=O)CC[C@]2(C)[C@@H]2[C@@H]1[C@@H]1CC[C@H]([C@H](C)CCCC(C)C)[C@@]1(C)CC2.O[C@H]1[C@@H](N)[C@H](O)[C@@H](C)O[C@H]1O[C@H]1/C=C/C=C/C=C/C=C/C=C/C=C/C=C/[C@H](C)[C@@H](O)[C@@H](C)[C@H](C)OC(=O)C[C@H](O)C[C@H](O)CC[C@@H](O)[C@H](O)C[C@H](O)C[C@](O)(C[C@H](O)[C@H]2C(O)=O)O[C@H]2C1 PCTMTFRHKVHKIS-BMFZQQSSSA-N 0.000 description 2
- 238000003672 processing method Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 229940050561 matrix product Drugs 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012634 optical imaging Methods 0.000 description 1
- 238000011426 transformation method Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Landscapes
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of methods of single image CT center monitoring camera calibration, comprising: is shot using video camera to be calibrated to rule stereoscopic article known to structure, obtains piece image；According to the structure feature of regular stereoscopic article, obtain several angle points three-dimensional world coordinate value and corresponding two dimensional image coordinate value；The two dimensional image coordinate value of each angle point, three-dimensional world coordinate value are substituted into camera model respectively, first iteration optimization solution is carried out using the iteration preferred method of setting initial value, obtains first camera interior and exterior parameter；According to the structure feature of regular stereoscopic article, obtain several equal parts cut off scatterplot three-dimensional world coordinate value and corresponding two dimensional image coordinate value；Each equal part is cut off into the two dimensional image coordinate value of scatterplot, three-dimensional world coordinate value substitutes into camera model respectively, iteration preferred method using first camera interior and exterior parameter as initial value carries out second iteration Optimization Solution, obtains accurate camera interior and exterior parameter.
Description
Technical Field
The invention relates to the technical field of computer vision detection, in particular to a method for calibrating a monitoring camera of a single-image CT machine room.
Background
In computer vision application, camera calibration is an indispensable process for obtaining the corresponding relationship between image pixel points and actual physical space points. And calibrating the camera, and restoring the object in the space by using the image shot by the camera under the assumption of a camera model. Through processing the images, a series of mathematical transformation and calculation methods are used to obtain the geometric and optical characteristics (namely, internal parameters) inside the camera and the position relation (namely, external parameters) of the camera coordinate system relative to the space coordinate system.
1) The internal parameters include f, (Cx, Cy), k1, sx, wherein:
f: focal length, in millimeters;
(Cx, Cy): coordinates of a central point or a principal point of the image, and the unit is a pixel;
k 1: first order coefficients of lens radial distortion;
sx: a non-deterministic scale factor, which is caused by camera lateral scan and sample timing errors.
2) The extrinsic parameters include R and T, wherein:
r, T are the rotation matrix and translation vector between the world coordinate system and the camera coordinate system, respectively if the camera coordinate system is oriented in the world coordinate system at a counterclockwise rotation angle about the X-axis (alpha, α), a counterclockwise rotation angle about the Y-axis (beta, β), a counterclockwise rotation angle about the Z-axis (gamma, γ), then the rotation matrix is:
wherein,
and T ═ Tx, Ty, Tz ]', where Tx, Ty, Tz: translation along three coordinate axes transformed from the world coordinate system to the camera coordinate system.
In computer vision, the camera model addresses the problem of points in a three-dimensional scene corresponding to points on an image plane. The camera model is a simplification of the optical imaging geometry, and the simplest and most commonly used camera model is the pinhole model (pinhole model). As shown in fig. 1, the pinhole camera model is derived from the principle of lens imaging, and is linear, and thus becomes a linear model of the camera. The pinhole camera model does not consider lens distortion, but can provide a good approximation for an actual camera, and a plurality of camera calibration methods are researched on the basis of the pinhole camera model.
Referring to fig. 1-2, we consider the central projection of a spatial point onto a plane, with the projection center at the origin of a euclidean coordinate system, and the plane z ═ f is called the image plane, where f is the focal length of the camera. Under the pinhole camera model, the spatial coordinates are M ═ (X, Y, Z)TIs mapped to a point on the image plane, which is the intersection of the straight line connecting the point M and the projection center with the image plane. From the similar triangles, the point (X, Y, Z) can be calculated quicklyTIs mapped to a point (fX/Z, fY/Z, f) on the image planeT. After the last image coordinate is omitted, the central projection from world coordinates to image coordinates is:
this is from the 3-dimensional Euclidean space IR3To 2D Euclidean space IR2A mapping of (2).
The center of projection is called the camera center, the perpendicular to the image plane from the camera center is called the principal axis of the camera, and the intersection of the principal axis with the image plane is called the principal point. The plane through the center of the camera parallel to the image plane is called the principal plane of the camera.
If the world and image points are represented by homogeneous vectors, the central projection can be very simply represented as a linear mapping between homogeneous coordinate systems. Specifically, the above formula can be written as a matrix product as follows:
suppose we world point M uses a 4-dimensional homogeneous vector (X, Y, Z,1)TRepresents; the image point m is represented in the form of a 3-dimensional homogeneous vector; p denotes a 3 x 4 homogeneous camera projection matrix. Thus, the above formula can be compactly written as m ═ mAnd PM. It defines a camera matrix of a pinhole model with a central projection as:
the common camera calibration methods include the following methods:
1) the traditional camera calibration method:
when the camera calibration is carried out by using the traditional calibration method, a calibration reference object is needed. Conventional calibration methods may include direct linear transformation calibration (DLT), Radial Alignment Constraint (RAC) calibration, active visual calibration, and planar calibration, among others. The following describes the basic principle of the conventional calibration method by taking a direct linear transformation calibration method and a radial alignment constraint calibration method as examples.
Direct linear transformation scaling (DLT): the direct linear transformation calibration method was proposed by Abdal-Aziz and Karara in the early 70 s. The method comprises the steps of firstly establishing a camera imaging model linear equation set, measuring world coordinates of a set of points in a scene and corresponding coordinates of the points on an imaging plane, and substituting the coordinate values into the linear equation set to obtain unknown coefficients of the linear equation set, wherein the coefficients of the linear equation set comprise internal parameters of a camera.
Tsai two-step camera calibration method: the method comprises the first step of establishing a linear equation by utilizing perspective matrix transformation and solving to obtain an accurate solution of most external parameters; and secondly, taking the obtained parameters as initial values, and substituting the rest external parameters, distortion coefficients and the like into nonlinear coefficients to carry out iterative solution. Since the RAC method takes radial distortion into account, it is more accurate than the DLT method.
2) The calibration method based on active vision comprises the following steps:
the camera is controlled to make some special movements, such as rotation around the broad center or pure translation, and the internal parameters are calculated by utilizing the particularity of the movements.
3) The self-calibration method comprises the following steps:
it is generally assumed that when different images are taken, the internal parameters of the camera do not change, and the correspondence between the image points is determined, and the camera is calibrated by the correspondence between the image points according to the special constraint relationship existing between the image points in the plurality of images.
Because the traditional calibration method always needs to use a calibration reference object in the shooting and calibration processes, great inconvenience is brought to the shooting operation and the use of the calibration method. And a relative movement between the calibration object and the camera is required to be generated, so that a plurality of images are shot with large difference, and the operation is very inconvenient. Each time the camera changes angle or moves position, the calibration object needs to be used again for camera calibration.
When the calibration method based on the active vision is used, the camera must be controlled to make some special motions, but in most cases, the motion state of the camera is difficult to know, or the motion state of the camera cannot be known at all, and in these cases, the calibration method based on the active vision cannot be used.
When the self-calibration method is used: since the calibration is performed on the premise that the internal parameters are not changed and the corresponding relationship between the image points is determined, the calibration accuracy is poor, the stability is not good, and the calibration cannot be used when the internal parameters are changed.
Therefore, there is a need to provide an improved camera calibration method to solve the above problems.
Disclosure of Invention
The embodiment of the invention aims to provide a method for calibrating a single-image CT machine room monitoring camera, which utilizes an unobtrusive regular common object as a calibration reference object, can calibrate the camera by shooting once, and has the advantages of simple and rapid whole operation process and high precision.
The embodiment of the invention provides a method for calibrating a monitoring camera of a single-image CT machine room, which comprises the following steps:
shooting a regular three-dimensional object with a known structure by using a camera to be calibrated to obtain an image;
establishing a world coordinate system for the regular three-dimensional object by taking the regular three-dimensional object as a reference so as to obtain three-dimensional world coordinate values of a plurality of corner points of the regular three-dimensional object;
processing the image to obtain two-dimensional image coordinate values corresponding to the angular points;
respectively substituting the two-dimensional image coordinate value and the three-dimensional world coordinate value of each corner point into a camera model describing the mapping relation between an image coordinate system and a world coordinate system, and performing primary iterative optimization solution by using an iterative optimization method for setting an initial value so as to obtain primary camera internal and external parameters;
obtaining three-dimensional world coordinate values of a plurality of equally-divided discrete points on the boundary line of the regular three-dimensional object according to the established world coordinate system;
processing the image to obtain two-dimensional image coordinate values corresponding to the plurality of equally-divided discrete points;
and respectively substituting the two-dimensional image coordinate value and the three-dimensional world coordinate value of each equal-segmentation discrete point into the camera model describing the mapping relation between the image coordinate system and the world coordinate system, and performing secondary iterative optimization solution by using the initial camera internal and external parameters as initial values, thereby obtaining accurate camera internal and external parameters.
As an improvement of the above solution, the camera model describing the mapping relationship between the image coordinate system and the world coordinate system is:
where s is a scaling factor, (X)w，Yw，Zw1) a coordinate value representing any point P in the world coordinate system, (u, v,1) a coordinate value on the image coordinate system corresponding to the point P, R, t being a rotation matrix and a translation vector in the camera external parameters, respectively, P3×4Projecting a matrix for the camera;
k is the camera intrinsic parameter and satisfies:
wherein (x)0，y0) Is the coordinate value of the center point of the image, fuIs a scale factor, f, on the horizontal axis of the image planevIs the scale factor on the longitudinal axis of the image plane, and fu＝f/dx,fvF is the focal length of the camera to be calibrated.
As a modification of the above, the regular solid object includes, but is not limited to, a cylinder, an ellipsoid, a cube, or a sphere.
As an improvement of the scheme, any one corner point or a middle point between the corner points of the regular three-dimensional object is used as a central point of the constructed world coordinate system.
As an improvement of the above scheme, the number of the equally divided discrete points is more than the number of the corner points.
As an improvement of the above scheme, the image processing method includes a SIFT feature point detection method or a color segmentation method.
As a refinement of the above, the iterative preferred method is a least squares method.
As a modification of the above, the setting of the initial value includes: setting the initial value of the focal length as half of the sum of the width and the height of the image; setting the initial value of the rotation matrix to be 0; the initial value of the translation vector is set to 0.
As an improvement of the scheme, the first iteration optimization solution and the second iteration optimization solution are carried out through the iteration optimization in the Zhangzhen chessboard format calibration method, the Tsai chessboard format calibration method or Hartley, and the solution error is minimized.
Compared with the prior art, the method for calibrating the monitoring camera of the single-image CT machine room utilizes the camera to be calibrated to shoot a regular three-dimensional object with a known structure for one time to obtain an image, utilizes the structural characteristics of the regular three-dimensional object with the known structure to obtain the three-dimensional world coordinate value of the angular point of the object and the corresponding two-dimensional image coordinate value, and then carries out first iterative optimization to obtain the initial internal and external parameters (rough parameters) of the camera; and the structural characteristics of regular three-dimensional objects with known structures are utilized to obtain three-dimensional world coordinate values and corresponding two-dimensional image coordinate values of a plurality of equally-divided discrete points (subdivisions) on the boundary line of the object, and the initial internal and external parameters (rough parameters) are used as initial values to carry out the second iteration optimization to obtain the final internal and external parameters (accurate parameters) of the camera.
Drawings
Fig. 1 is a schematic diagram of a prior art pinhole machine model.
Fig. 2 is a schematic diagram of a prior art pinhole machine model.
Fig. 3 is a flowchart of a method for calibrating a monitoring camera of a single-image CT room in an embodiment of the present invention.
Fig. 4 is a schematic diagram of solving an accurate solution by first iterative optimization in the camera calibration method according to the embodiment of the present invention.
Fig. 5 is a schematic diagram of solving an accurate solution by second iterative optimization in the camera calibration method according to the embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The method for calibrating the monitoring camera of the single-image CT machine room mainly aims at calibrating a high-end camera in a medical system, and the camera of the type can usually ignore distortion parameters and non-deterministic scale factors in internal parameters of the camera, so that the method for calibrating the camera does not need to solve the internal parameters of the two cameras.
Fig. 3 is a schematic flow chart of a method for calibrating a monitoring camera of a single-image CT room according to an embodiment of the present invention. As shown in fig. 3, the method for calibrating a camera according to the present embodiment includes the following steps:
step S101: a regular three-dimensional object with a known structure is shot by a camera to be calibrated to obtain an image.
Specifically, the regular solid object includes, but is not limited to, a cylinder, an ellipsoid, a cube, or a sphere. And known structures may be set or measured as the case may be. Then, a regular three-dimensional object with a known structure is shot by using cameras with internal and external parameters to be calibrated, so that a picture (image) is obtained.
Step S102: and constructing a world coordinate system for the regular three-dimensional object by taking the regular three-dimensional object as a reference so as to obtain three-dimensional world coordinate values of a plurality of corner points of the regular three-dimensional object.
Specifically, a world coordinate system may be constructed by using any one of the corner points or a midpoint between the corner points of the regular three-dimensional object as a central point of the world coordinate system. Since the structure of the regular solid object is known, when one of the corner points or the midpoint between the corner points is used as the central point (origin) of the world coordinate system, the three-dimensional world coordinate values of the other corner points can be calculated correspondingly.
Step S103: and carrying out image processing on the image to obtain the coordinate values of the two-dimensional image corresponding to the plurality of corner points.
And calculating to obtain coordinate values of the two-dimensional image corresponding to the plurality of corner points by adopting image processing such as an SIFT feature point detection method or a color segmentation method.
Step S104: and respectively substituting the two-dimensional image coordinate value and the three-dimensional world coordinate value of each corner point into a camera model for describing the mapping relation between an image coordinate system and a world coordinate system, and performing primary iterative optimization solution by using an iterative optimization method for setting initial values so as to obtain primary camera internal and external parameters.
Specifically, the camera model describing the mapping relationship between the image coordinate system and the world coordinate system is:
where s is a scaling factor, (X)w，Yw，Zw1) a coordinate value representing any point P in the world coordinate system, (u, v,1) a coordinate value on the image coordinate system corresponding to the point P, R, t being a rotation matrix and a translation vector in the camera external parameters, respectively, P3×4Projecting a matrix for the camera;
k is the camera intrinsic parameter and satisfies:
wherein (x)0，y0) Is the coordinate value of the center point of the image, fuIs a scale factor, f, on the horizontal axis of the image planevIs the scale factor on the longitudinal axis of the image plane, and fu＝f/dx,fvF is the focal length of the camera to be calibrated.
After the two-dimensional image coordinate value and the three-dimensional world coordinate value of each angular point are respectively substituted into the camera model in sequence, the first iterative optimization can be carried out through the iterative optimization in the Zhangyingyou chessboard format calibration method, the Tsai chessboard format calibration method or the Hartley to solve the overdetermined linear equation set, and the solving error is minimized.
Preferably, the overdetermined linear equation set can be solved by adopting a least square method to obtain initial internal and external parameters of the camera (rough internal and external parameters of the camera).
It can be understood that, when performing the initial iterative optimization, initial values (internal and external parameters of the camera) of the iterative optimization need to be preset, including a focal length, a rotation matrix, and a translation vector. For example, the initial values set may be: setting the initial value of the focal length as half of the sum of the width and the height of the image; setting the initial value of the rotation matrix to be 0; the initial value of the translation vector is set to 0.
Step S105: and obtaining three-dimensional world coordinate values of a plurality of equally-divided discrete points on the boundary line of the regular three-dimensional object according to the established world coordinate system.
Wherein the number of the equal-segmentation discrete points is more than that of the angular points. In addition, the equally divided discrete points may include the corner points described above. Similarly, since the structure of the regular three-dimensional object is known, when one of the corner points or the midpoint between the corner points is used as the central point (origin) of the world coordinate system, the three-dimensional world coordinate values of the divided discrete points can be calculated accordingly.
Step S106: and carrying out image processing on the image to obtain two-dimensional image coordinate values corresponding to the plurality of equally-divided discrete points.
Specifically, an image processing method such as an SIFT feature point detection method or a color segmentation method may be adopted to detect a boundary line on an image, and then a two-dimensional image coordinate value corresponding to the equally-segmented discrete point is obtained by approximating an equally-segmented curve or dividing a curve according to a decreasing ratio.
Step S107: and respectively substituting the two-dimensional image coordinate value and the three-dimensional world coordinate value of each equal-segmentation discrete point into the camera model describing the mapping relation between the image coordinate system and the world coordinate system, and performing secondary iterative optimization solution by using the initial camera internal and external parameters as initial values, thereby obtaining accurate camera internal and external parameters.
Specifically, the camera model describing the mapping relationship between the image coordinate system and the world coordinate system is, as shown in the formula (1), sequentially substituting the two-dimensional image coordinate value and the three-dimensional world coordinate value of each of the equally-divided discrete points into the camera model, and then performing a second iteration optimization solution on the overdetermined linear equation set by using an iteration optimization or least square method in the Zhang Yongyou chessboard lattice calibration method, the Tsai chessboard lattice calibration method or Hartley, and when performing the second iteration optimization solution, using the first camera internal and external parameters (rough camera internal and external parameters) obtained by the first iteration optimization solution as initial values of the iteration optimization solution, thereby obtaining final camera internal and external parameters (precise camera internal and external parameters).
The method for calibrating a monitoring camera of a single-image CT room according to the present invention will be further described with reference to fig. 4 to 5 by an embodiment.
In this embodiment, the method for calibrating a single image camera mainly includes the following main steps:
1. taking an image of a given reference;
the given references are: a cylinder, sphere, regular solid object with broken lines of known structure, which can be set or measured according to specific situations.
As shown in fig. 4 to 5, a regular object (cylinder) with a known structure is photographed by a camera to be calibrated, so as to obtain an object image.
2. Rough estimation: and obtaining two-dimensional and three-dimensional coordinates of the angular points of the object by using the structural characteristics of the object, performing first iterative optimization, and obtaining and storing internal and external parameters of the camera. The method specifically comprises the following steps:
2.1 Camera model pinhole camera model was used.
The camera mapping formula for mapping the three-dimensional world coordinate system to the two-dimensional image coordinate system is as follows:
wherein,
1)the internal parameters can be simplified, and the focal lengths in two axial directions are equal and are both f; and has no radial direction,And (4) tangential distortion.
2) f: focal length, in millimeters;
3) (x0, y 0): the coordinate of the central point or principal point of the image is a pixel, and the central point of the simplified image is the central point of the shot image and is known;
2.2 calculating three-dimensional world coordinate values
Referring to fig. 4, the corner points of the cylinder include four ABCD points, and the step of calculating three-dimensional world coordinate values of the four points includes:
1) the central point of the world coordinate system can be assumed to be any one of the ABCD points of the cylinder, and can also be the midpoint of the AB or the midpoint of the CD;
2) further assume that the length of AB is a unit of measurement, but of course how many millimeters can be measured with the ruler;
3) in units AB, the three-dimensional world coordinates of A, B, C, D are easily deduced because they are known cylinders.
2.3 calculating the corresponding two-dimensional image coordinate values
And calculating A, B, C, D coordinates of image points A ', B', C 'and D' corresponding to the three-dimensional world coordinate points.
For example, it can be obtained by image processing such as a method of SIFT feature point detection. The curved surface can also be solved by a color segmentation method, and then the coordinate values of the two-dimensional images of the four corner points are found.
And 2.4, setting initial values of the initial iteration optimization, such as focal length, rotation matrix and translation vector.
The initial value of the focal length is set to be half of the sum of the width and the height of the image;
setting the initial value of the rotation matrix to be 0;
the initial value of the translation vector is set to 0
2.5 first iteration optimization solution
Solving the minimum error according to a Zhangnyou chessboard lattice calibration method, a Tsai chessboard lattice calibration method or an iterative optimization solving method in Hartley multi-view geometry:
for the above formula, we now know the values (u, v,1) of the two-dimensional coordinate points of the image, and also know the values (Xw, Yw, Zx,1) of the three-dimensional coordinate points corresponding to the two-dimensional coordinate points of the image, and it is unknown the camera internal parameters K and the external parameters R and t. The overdetermined linear equation set can be solved through a least square method, and primary rough internal and external parameters of the camera are solved to serve as initial values of secondary iteration optimization solution.
3. And (3) accurate estimation: generating subdivided three-dimensional coordinate points and corresponding two-dimensional coordinate points by using the structural characteristics of the object; and performing second iterative optimization by using the obtained two-dimensional and three-dimensional coordinates and the roughly estimated internal and external parameters as initial values to obtain accurate internal and external parameters of the camera. The method specifically comprises the following steps:
3.1 calculating three-dimensional world coordinate values
As shown in fig. 5, the boundary line on the cylinder includes parallel AB curve and CD curve, and its corresponding equally divided discrete points are also parallel, and the points on the curve AB and the points on the curve CD are different from each other by a translational relationship of one direction and length AC or BD, so that the three-dimensional world coordinate values of these discrete points can be obtained;
3.2 calculating two-dimensional image coordinate values
1) As shown in fig. 5, a curve a 'B' and a curve C 'D' may be detected by a method of image processing;
2) similarly, a curve is approximated equally, or divided according to a decreasing ratio;
3) thereby obtaining the coordinate values of the two-dimensional image corresponding to the three-dimensional equally-divided discrete points.
3.3 second iterative optimization solution
Similarly, for the first time, the error equation of the first iterative optimization and the error equation of the second time may be merged together to perform the second iterative optimization (i.e., the coarse internal and external parameters of the camera obtained by the first iterative optimization solution are used as the initial values of the second iterative optimization solution).
Also, for the above formula, we now know the values (u, v,1) of the two-dimensional coordinate points of the image, and also know the values (Xw, Yw, Zx,1) of the three-dimensional coordinate points corresponding to the two-dimensional coordinate points of the image, and it is unknown the camera internal parameter K and the external parameters R and t. The overdetermined linear equation set (the initial value is the rough internal and external parameters of the camera obtained by the first iterative optimization solution) can be solved through a least square method, and the accurate internal and external parameters of the camera in the second step are solved.
In summary, the method for calibrating a single-image CT machine room monitoring camera disclosed in the present invention utilizes a camera to be calibrated to capture a regular three-dimensional object with a known structure once to obtain an image, utilizes the structural characteristics of the regular three-dimensional object with the known structure to obtain three-dimensional world coordinate values of the angular points of the object and corresponding two-dimensional image coordinate values, and then performs a first iterative optimization to obtain initial internal and external parameters (coarse parameters) of the camera; and the structural characteristics of regular three-dimensional objects with known structures are utilized to obtain three-dimensional world coordinate values and corresponding two-dimensional image coordinate values of a plurality of equally-divided discrete points (subdivisions) on the boundary line of the object, and the initial internal and external parameters (rough parameters) are used as initial values to carry out the second iteration optimization to obtain the final internal and external parameters (accurate parameters) of the camera.
While the foregoing is directed to the preferred embodiment of the present invention, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention.
Claims (8)
1. A method for calibrating a monitoring camera of a single-image CT machine room is characterized by comprising the following steps:
shooting a regular three-dimensional object with a known structure by using a camera to be calibrated to obtain an image; wherein the regular solid object has an arc-shaped boundary;
establishing a world coordinate system for the regular three-dimensional object by taking the regular three-dimensional object as a reference so as to obtain three-dimensional world coordinate values of a plurality of corner points of the regular three-dimensional object;
processing the image to obtain two-dimensional image coordinate values corresponding to the angular points;
respectively substituting the two-dimensional image coordinate value and the three-dimensional world coordinate value of each corner point into a camera model describing the mapping relation between an image coordinate system and a world coordinate system, and performing primary iterative optimization solution by using an iterative optimization method for setting an initial value so as to obtain primary camera internal and external parameters;
obtaining three-dimensional world coordinate values of a plurality of equally-divided discrete points on the boundary line of the regular three-dimensional object according to the established world coordinate system; wherein the boundary line is an arc boundary line;
processing the image to obtain two-dimensional image coordinate values corresponding to the plurality of equally-divided discrete points; wherein, the coordinate value of the two-dimensional image corresponding to the equally-divided discrete point is obtained by approximating an equally-divided curve or dividing the curve according to a decreasing ratio;
respectively substituting the two-dimensional image coordinate value and the three-dimensional world coordinate value of each equal-segmentation discrete point into the camera model describing the mapping relation between the image coordinate system and the world coordinate system, and performing secondary iterative optimization solution by using the initial camera internal and external parameters as initial values to obtain accurate camera internal and external parameters;
the number of the equally divided discrete points is more than that of the angular points.
2. The method for calibrating a monitoring camera of a single-image CT room according to claim 1, wherein the camera model describing the mapping relationship between the image coordinate system and the world coordinate system is:
where s is a scaling factor, (X)w，Yw，Zw1) coordinate values of an arbitrary point P in the world coordinate system, (u, v,1) coordinate values on the image coordinate system corresponding to the point P, and R, t are camera external parametersOf (1) a rotation matrix and a translation vector, P3×4Projecting a matrix for the camera;
k is the camera intrinsic parameter and satisfies:
wherein (x)0，y0) Is the coordinate value of the center point of the image, fuIs a scale factor, f, on the horizontal axis of the image planevIs the scale factor on the longitudinal axis of the image plane, and fu＝f/dx,fvF is the focal length of the camera to be calibrated.
3. The method of claim 1, wherein the regular solid object includes but is not limited to a cylinder, an ellipsoid or a sphere.
4. The method for calibrating the monitoring camera of the single-image CT machine room as claimed in claim 1, wherein any one of the corner points or the midpoint between the corner points of the regular solid object is taken as the central point of the constructed world coordinate system.
5. The method for calibrating the monitoring camera of the single-image CT machine room according to claim 1, wherein an SIFT feature point detection method or a color segmentation method is adopted to perform image processing on the image to obtain the coordinate values of the two-dimensional image corresponding to the plurality of corner points; and carrying out image processing on the image by adopting an SIFT feature point detection method or a color segmentation method to obtain the coordinate values of the two-dimensional image corresponding to the plurality of equally-segmented discrete points.
6. The method for calibrating the monitoring camera of the single-image CT machine room as recited in claim 1, wherein the iterative optimization method is a least square method.
7. The method for calibrating the monitoring camera of the single-image CT machine room as claimed in claim 1 or 2, wherein the setting the initial value comprises: setting the initial value of the focal length as half of the sum of the width and the height of the image; setting the initial value of the rotation matrix to be 0; the initial value of the translation vector is set to 0.
8. The method for calibrating the monitoring camera of the single-image CT machine room according to claim 1 or 2, wherein the first iterative optimization solution and the second iterative optimization solution are performed by iterative optimization in the zhangnyou checkerboard calibration method, the Tsai checkerboard calibration method or Hartley, and the solution error is minimized.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510204298.2A CN104794718B (en) | 2015-04-27 | 2015-04-27 | A kind of method of single image CT center monitoring camera calibration |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510204298.2A CN104794718B (en) | 2015-04-27 | 2015-04-27 | A kind of method of single image CT center monitoring camera calibration |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104794718A CN104794718A (en) | 2015-07-22 |
| CN104794718B true CN104794718B (en) | 2019-01-15 |
Family
ID=53559498
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510204298.2A Expired - Fee Related CN104794718B (en) | 2015-04-27 | 2015-04-27 | A kind of method of single image CT center monitoring camera calibration |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104794718B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107368014A (en) * | 2017-08-24 | 2017-11-21 | 国网黑龙江省电力有限公司信息通信公司 | Machine room monitoring system and method |
| CN107644444B (en) * | 2017-09-07 | 2020-04-03 | 广东工业大学 | Single-image camera calibration method based on compressed sensing |
| CN107582085B (en) * | 2017-09-14 | 2021-02-05 | 广州七喜医疗设备有限公司 | Intelligent digital X-ray exposure control device and method |
| CN108109179B (en) * | 2017-12-29 | 2021-05-18 | 天津科技大学 | Camera attitude correction method based on pinhole camera model |
| CN113554710A (en) * | 2020-04-24 | 2021-10-26 | 西门子(深圳)磁共振有限公司 | Calibration method, system and storage medium of 3D camera in medical image system |
| CN111686378B (en) * | 2020-07-14 | 2023-02-17 | 上海联影医疗科技股份有限公司 | Bed body movement precision detection method, device, equipment and storage medium |
| CN113066086B (en) * | 2021-04-26 | 2022-08-30 | 深圳市商汤科技有限公司 | Road disease detection method and device, electronic equipment and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101025824A (en) * | 2007-03-30 | 2007-08-29 | 东南大学 | Calibrating method based on fixed parameters and variable parameters for three-dimensional scanning system |
| CN102184563B (en) * | 2011-03-23 | 2013-01-02 | 华中科技大学 | Three-dimensional scanning method, three-dimensional scanning system and three-dimensional scanning device used for plant organ form |
| CN103258329A (en) * | 2013-05-24 | 2013-08-21 | 西安电子科技大学 | Camera calibration method based on one-dimensional feature of balls |
-
2015
- 2015-04-27 CN CN201510204298.2A patent/CN104794718B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101025824A (en) * | 2007-03-30 | 2007-08-29 | 东南大学 | Calibrating method based on fixed parameters and variable parameters for three-dimensional scanning system |
| CN102184563B (en) * | 2011-03-23 | 2013-01-02 | 华中科技大学 | Three-dimensional scanning method, three-dimensional scanning system and three-dimensional scanning device used for plant organ form |
| CN103258329A (en) * | 2013-05-24 | 2013-08-21 | 西安电子科技大学 | Camera calibration method based on one-dimensional feature of balls |
Non-Patent Citations (2)
| Title |
|---|
| Direct Calibration by Fitting of Cuboids to a Single Image Using Differential Evolution;Luis Gerardo de la Frage, Oliver Schutze;《International Journal of Computer Vision》;20090228;第81卷(第2期);第2-3节，图1 |
| 一种改进的摄像机内外参数初值计算方法;康晶 等;《计算机工程与科学》;20071231;第29卷(第1期);第2-3节 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104794718A (en) | 2015-07-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104794718B (en) | A kind of method of single image CT center monitoring camera calibration | |
| CN106595528B (en) | A kind of micro- binocular stereo vision measurement method of telecentricity based on digital speckle | |
| US20200096317A1 (en) | Three-dimensional measurement apparatus, processing method, and non-transitory computer-readable storage medium | |
| Zhang | Camera calibration with one-dimensional objects | |
| US10846885B2 (en) | Methods and computer program products for calibrating stereo imaging systems by using a planar mirror | |
| WO2018076154A1 (en) | Spatial positioning calibration of fisheye camera-based panoramic video generating method | |
| CN103839227B (en) | Fisheye image correcting method and device | |
| JPH10221072A (en) | System and method for photogrammetry | |
| WO2011145285A1 (en) | Image processing device, image processing method and program | |
| JP2011086111A (en) | Imaging apparatus calibration method and image synthesis device | |
| Yang et al. | Flexible and accurate implementation of a binocular structured light system | |
| CN112229323B (en) | Six-degree-of-freedom measurement method of checkerboard cooperative target based on monocular vision of mobile phone and application of six-degree-of-freedom measurement method | |
| CN110490943B (en) | Rapid and accurate calibration method and system of 4D holographic capture system and storage medium | |
| CN113744340A (en) | Calibrating cameras with non-central camera models of axial viewpoint offset and computing point projections | |
| JP2002516443A (en) | Method and apparatus for three-dimensional display | |
| CN115861445B (en) | Hand-eye calibration method based on three-dimensional point cloud of calibration plate | |
| CA3233222A1 (en) | Method, apparatus and device for photogrammetry, and storage medium | |
| CN115272080A (en) | Global deformation measurement method and system based on image stitching | |
| Yamauchi et al. | Calibration of a structured light system by observing planar object from unknown viewpoints | |
| GB2569609A (en) | Method and device for digital 3D reconstruction | |
| CN113483669B (en) | Multi-sensor pose calibration method and device based on three-dimensional target | |
| Chen et al. | Efficient vision-based calibration for visual surveillance systems with multiple PTZ cameras | |
| Fantin et al. | An efficient mesh oriented algorithm for 3d measurement in multiple camera fringe projection | |
| CN110163920B (en) | Method and system for calibrating camera by using pole line property and circular ring point | |
| Dhome | Visual Perception Through Video Imagery |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate 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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20190115 Termination date: 20200427 |