CN111739103A - Multi-camera calibration system based on single-point calibration object - Google Patents
Multi-camera calibration system based on single-point calibration object Download PDFInfo
- Publication number
- CN111739103A CN111739103A CN202010570282.4A CN202010570282A CN111739103A CN 111739103 A CN111739103 A CN 111739103A CN 202010570282 A CN202010570282 A CN 202010570282A CN 111739103 A CN111739103 A CN 111739103A
- Authority
- CN
- China
- Prior art keywords
- calibration
- camera
- cameras
- external parameters
- internal
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 claims description 23
- 239000011159 matrix material Substances 0.000 claims description 10
- 238000005457 optimization Methods 0.000 claims description 8
- 238000013519 translation Methods 0.000 claims description 6
- 238000000354 decomposition reaction Methods 0.000 claims description 4
- 238000010606 normalization Methods 0.000 claims description 4
- 238000006243 chemical reaction Methods 0.000 claims description 2
- 238000012545 processing Methods 0.000 claims description 2
- 238000012216 screening Methods 0.000 claims description 2
- 239000003086 colorant Substances 0.000 claims 1
- 238000010586 diagram Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/80—Analysis of captured images to determine intrinsic or extrinsic camera parameters, i.e. camera calibration
- G06T7/85—Stereo camera calibration
Abstract
The invention relates to a multi-camera calibration system based on a single-point calibration object. Specifically, a plurality of infrared cameras fixed at different positions in a scene are used for collecting a single calibration point which moves freely and image points of an L-shaped rigid body used for indicating a world coordinate system, and the image points are uploaded to an upper computer to realize multi-camera internal and external reference calibration by using image point data. Firstly, calibrating two cameras with a common viewpoint pairwise according to epipolar geometric constraint relations among pinhole, distorted camera models and image point pairs, optimizing internal and external parameters by using an L-M algorithm, then establishing an undirected graph of a multi-camera system according to a common view field relation of all the cameras, determining an optimal cascade path of any camera relative to a reference camera, converting and solving internal and external parameters of all the cameras under a unified reference camera coordinate system by combining the calibration parameters pairwise, binding, adjusting and optimizing by using BA, and finally solving the internal and external parameters of all the cameras under a specified unified world coordinate system by using an L-shaped rigid body by using a kneipPnP algorithm.
Description
Technical Field
The invention relates to the field of computer vision, in particular to a multi-camera calibration system based on a single-point calibration object.
Background
The camera calibration technology recovers the internal and external parameters of the camera by using a two-dimensional image of a calibration object, further reconstructs three-dimensional information of a scene, is a key step in the field of computer vision, and has wide application in the fields of industry, aerospace, cultural originality and the like. In recent years, with the vigorous development of virtual reality, the viewing angle of a single camera cannot meet the expectation of people for a large-range three-dimensional space, and the combination and calibration of multiple cameras gradually become a new research hotspot. However, the existing multi-camera calibration method often requires that the calibration object move in the common view field of all cameras at the same time, which increases the difficulty of practical operation, and the calibration precision is low, so how to obtain the high-precision internal and external parameters of the multi-camera through convenient and fast operation is a problem that needs to be solved at present.
The calibration tool used by the method is simple to manufacture, and the calibration object does not need to be limited to move in the common view field of all cameras, so that the operability is strong; the multi-camera cascade path determined by the common view field relation can enable more image points to participate in operation, and the algorithm robustness is better; through multi-step optimization, the calibration parameters can reach the reprojection error of a sub-pixel level, and the high-precision requirement is completely met.
Disclosure of Invention
The present invention is directed to a multi-camera calibration system based on a single-point calibration object to solve the above problems.
The invention realizes the purpose through the following technical scheme:
a multi-camera calibration system based on a single-point calibration object mainly comprises: the method comprises the steps of collecting a single infrared reflection point which freely moves and image points of an L-shaped rigid body which is used for indicating a world coordinate system by using a plurality of infrared cameras which are fixed at different positions in a scene, and calibrating internal and external parameters of a plurality of cameras by using image point data, wherein specifically:
s10: and collecting calibration data, namely collecting the calibration data by using a calibration rod provided with a single calibration point and an L-shaped right-angle rigid body with known side length.
S20: and pairwise calibration, which comprises pairwise calibration of two cameras with a common viewpoint according to epipolar geometric constraint relations among the pinhole, the distortion camera model and the image point pairs, and optimizing internal and external parameters by using an L-M algorithm.
S30: the method comprises the steps of establishing an undirected graph of a multi-camera system according to the common view field relation of all cameras, determining the optimal cascade path of any camera relative to a reference camera, solving the internal and external references of all cameras in the unified reference camera coordinate system by combining two-to-two calibration parameter conversion, and using BA to bind, adjust and optimize.
S40: the calibration of the internal and external parameters of the multiple cameras under the specified world coordinate system comprises the step of solving the internal and external parameters of all the cameras under the specified unified world coordinate system by using an L-shaped rigid body through a kneipPnP algorithm.
Has the advantages that: the calibration tool used by the method is simple to manufacture, and the calibration object does not need to be limited to move in the common view field of all cameras, so that the operability is strong; the multi-camera cascade path determined by the common view field relation can enable more image points to participate in operation, and the algorithm robustness is better; through multi-step optimization, the calibration parameters can reach the reprojection error of a sub-pixel level, and the high-precision requirement is completely met.
Drawings
FIG. 1 is a schematic diagram of an L-shaped right-angle rigid body structure in a multi-camera internal and external reference calibration assembly based on a single-point calibration object according to the present invention;
FIG. 2 is a schematic diagram of a calibration scenario of a multi-camera internal and external reference calibration method based on a single-point calibration object according to the present invention;
fig. 3 is an undirected graph of a multi-camera system of the multi-camera internal and external reference calibration method based on a single-point calibration object of the present invention;
Detailed Description
The invention will be further described with reference to the accompanying drawings in which:
s10: collecting calibration data, specifically comprising:
s11: and freely swinging a calibration rod provided with a single calibration point in a scene, controlling a plurality of cameras to synchronously acquire calibration point data, and uploading the data to an upper computer in real time.
S12: and placing the L-shaped right-angle rigid body with the known side length in the center of the field, controlling a plurality of cameras to synchronously acquire rigid body data, and uploading the rigid body data to an upper computer in real time.
S20: every two calibration specifically comprises:
s21: and the distortion processing comprises the step of calculating the coordinates of the real image point after distortion removal by using a distortion model according to the distortion coefficient in the camera specification and the physical size of the unit pixel.
S22: and solving the effective focal length, namely translating the original points of the two image coordinate systems to a principal point, solving a basic matrix and a pole by using a normalization 8-point method, simplifying the basic matrix according to a Hartley self-calibration method, and solving the pixel unit focal length of the camera by decomposition.
S23: solving external parameters, namely, carrying out normalization on a basic matrix by using the obtained internal parameters to solve an essential matrix, obtaining four groups of external parameters (R, t) through singular value decomposition, firstly, rebuilding calibration points by using a trigonometry method, and screening out a group of correct external parameters (R, t) based on the principle that a calibration object is necessarily positioned in front of a camera (namely, a Z coordinate is constant). At the moment, the translation vector and the real translation have a difference of a scale factor, and the scale factor can be recovered by utilizing the length of the reconstructed L-shaped rigid body.
S24: and optimizing the internal and external parameters, namely performing nonlinear optimization on the internal and external parameters obtained by pairwise calibration by using an L-M algorithm with the minimum sum of image point epipolar distance errors as an optimization target.
S30: the calibration of internal and external reference of multiple cameras under a unified reference camera coordinate system specifically comprises the following steps:
s31: establishing a multi-camera system undirected graph: the vertexes represent cameras, and if m common viewpoints exist between two cameras, edges exist between the two corresponding vertexes, and the weight of the edges is 1/m.
S32: determining an optimal cascade path of a reference camera to an arbitrary camera: and selecting the camera with the most common viewpoints as a No. 0 reference camera, and calculating a path with the minimum sum of edge weights from the No. 0 camera to any No. k camera by using a Dijkstra method, namely the optimal cascade path of the No. k camera.
S33: the optimal cascade path of the No. k camera is assumed to be 0- > i- > k, and the external parameters of the No. 0 camera to the No. i camera obtained by pairwise calibration are (R)0i,t0i) The external parameters of the i th to j th cameras are (R)ik,tik) And the external parameters from No. 0 to No. k are (R)ikR0i,Rikt0i+tik)。
S34: and repeating the third step, solving the external parameters of all the cameras relative to the unified reference camera, reconstructing the calibration point by using a trigonometry method, and integrally optimizing all the internal and external parameters by using a BA binding algorithm with the minimized reprojection error as an optimization target. Further improve the calibration precision.
S40: the multi-camera internal and external reference calibration under the specified world coordinate system specifically comprises the following steps:
s41: determining the matching relation of the L-shaped rigid body imaged by different cameras: and determining a basic matrix of the two views according to the obtained camera external parameters, and obtaining matched image points by utilizing epipolar geometric constraint.
S42: and reconstructing the coordinates of the L-shaped rigid body under a reference camera coordinate system by using a least square triangle method, setting the coordinates of the L-shaped rigid body under a specified world coordinate system, and calculating the rotation and translation of the world coordinates to the reference camera coordinate system by using a kneipPnP algorithm.
S43: and (4) calculating internal and external parameters of all cameras in the world coordinate system according to the same principle as the S33.
The foregoing illustrates and describes the principles, general features, and advantages of the present invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are described in the specification and illustrated only to illustrate the principle of the present invention, but that various changes and modifications may be made therein without departing from the spirit and scope of the present invention, which fall within the scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (9)
1. The utility model provides a polyphaser calibration system based on single point calibration thing which characterized in that: the method comprises the following steps of acquiring a single calibration point which freely moves and an image point of an L-shaped rigid body used for indicating a world coordinate system by using a plurality of infrared cameras fixed at different positions in a scene, and calibrating internal and external parameters of a plurality of cameras by using image point data, and specifically comprises the following steps:
s10: and collecting calibration data, namely collecting the calibration data by using a calibration rod provided with a single calibration point and an L-shaped right-angle rigid body with known side length.
S20: and pairwise calibration, which comprises pairwise calibration of two cameras with a common viewpoint according to epipolar geometric constraint relations among the pinhole, the distortion camera model and the image point pairs, and optimizing internal and external parameters by using an L-M algorithm.
S30: the method comprises the steps of establishing an undirected graph of a multi-camera system according to the common view field relation of all cameras, determining the optimal cascade path of any camera relative to a reference camera, solving the internal and external references of all cameras in the unified reference camera coordinate system by combining two-to-two calibration parameter conversion, and using BA to bind, adjust and optimize.
S40: the calibration of the internal and external parameters of the multiple cameras under the specified world coordinate system comprises the step of solving the internal and external parameters of all the cameras under the specified unified world coordinate system by using an L-shaped rigid body through a kneipPnP algorithm.
2. The method according to claim 1, wherein the acquiring calibration data comprises:
s11: and freely swinging a calibration rod provided with a single calibration point in a scene, controlling a plurality of cameras to synchronously acquire calibration point data, and uploading the data to an upper computer in real time.
S12: and placing the L-shaped right-angle rigid body with the known side length in the center of the field, controlling a plurality of cameras to synchronously acquire rigid body data, and uploading the rigid body data to an upper computer in real time.
3. The method according to claim 1, wherein the two-by-two calibration comprises:
s21: and the distortion processing comprises the step of calculating the coordinates of the real image point after distortion removal by using a distortion model according to the distortion coefficient in the camera specification and the physical size of the unit pixel.
S22: and solving the effective focal length, namely translating the original points of the two image coordinate systems to a principal point, solving a basic matrix and a pole by using a normalization 8-point method, simplifying the basic matrix according to a Hartley self-calibration method, and solving the pixel unit focal length of the camera by decomposition.
S23: solving external parameters, namely, carrying out normalization on a basic matrix by using the obtained internal parameters to solve an essential matrix, obtaining four groups of external parameters (R, t) through singular value decomposition, firstly, rebuilding calibration points by using a trigonometry method, and screening out a group of correct external parameters (R, t) based on the principle that a calibration object is necessarily positioned in front of a camera (namely, a Z coordinate is constant). At the moment, the translation vector and the real translation have a difference of a scale factor, and the scale factor can be recovered by utilizing the length of the reconstructed L-shaped rigid body.
S24: and optimizing the internal and external parameters, namely performing nonlinear optimization on the internal and external parameters obtained by pairwise calibration by using an L-M algorithm with the minimum sum of image point epipolar distance errors as an optimization target.
4. The system of claim 1, wherein the multi-camera internal and external reference calibration under the unified reference camera coordinate system comprises:
s31: establishing a multi-camera system undirected graph: the vertexes represent cameras, and if m common viewpoints exist between two cameras, edges exist between the two corresponding vertexes, and the weight of the edges is 1/m.
S32: determining an optimal cascade path of a reference camera to an arbitrary camera: and selecting the camera with the most common viewpoints as a No. 0 reference camera, and calculating a path with the minimum sum of edge weights from the No. 0 camera to any No. k camera by using a Dijkstra method, namely the optimal cascade path of the No. k camera.
S33: suppose the optimal cascade path of the k-th camera is 0->i->k, the external parameters of No. 0 to No. i cameras obtained by pairwise calibration are (R)0i,t0i) The external parameters of the i th to j th cameras are (R)ik,tik) And the external parameters from No. 0 to No. k are (R)ikR0i,Rikt0i+tik)。
S34: and repeating the third step, solving the external parameters of all the cameras relative to the unified reference camera, reconstructing the calibration point by using a trigonometry method, and integrally optimizing all the internal and external parameters by using a BA binding algorithm with the minimized reprojection error as an optimization target. Further improve the calibration precision.
5. The system of claim 1, wherein the multi-camera internal and external reference calibration under the specified world coordinate system comprises:
s41: determining the matching relation of the L-shaped rigid body imaged by different cameras: and determining a basic matrix of the two views according to the obtained camera external parameters, and obtaining matched image points by utilizing epipolar geometric constraint.
S42: and reconstructing the coordinates of the L-shaped rigid body under a reference camera coordinate system by using a least square triangle method, setting the coordinates of the L-shaped rigid body under a specified world coordinate system, and calculating the rotation and translation of the world coordinates to the reference camera coordinate system by using a kneipPnP algorithm.
S43: and (4) calculating internal and external parameters of all cameras in the world coordinate system according to the same principle as the S33.
6. A calibration assembly comprising a plurality of image sensors, free-moving calibration points, and an L-shaped rigid body for indicating a world coordinate system. The image sensors are arranged at a plurality of spatial positions relative to the positioning body and used for simultaneously photographing the calibration points and the L-shaped rigid body to acquire a plurality of images.
7. A calibration assembly according to claim 6, wherein said freely moving positioning point is capable of reflecting infrared light, or said positioning body is capable of emitting visible light of different colors.
8. A calibration assembly according to claim 6, wherein said freely movable anchor point is on the calibration rod.
9. The calibration assembly according to claim 6, wherein said L-shaped rigid body is formed by fixing two line segments with different lengths in a vertical relationship, and three calibration points are fixed on the L-shaped rigid body and located at outer end points of the two line segments and intersection points of the two rigid bodies respectively.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010570282.4A CN111739103A (en) | 2020-06-18 | 2020-06-18 | Multi-camera calibration system based on single-point calibration object |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010570282.4A CN111739103A (en) | 2020-06-18 | 2020-06-18 | Multi-camera calibration system based on single-point calibration object |
Publications (1)
Publication Number | Publication Date |
---|---|
CN111739103A true CN111739103A (en) | 2020-10-02 |
Family
ID=72651926
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010570282.4A Pending CN111739103A (en) | 2020-06-18 | 2020-06-18 | Multi-camera calibration system based on single-point calibration object |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111739103A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113077519A (en) * | 2021-03-18 | 2021-07-06 | 中国电子科技集团公司第五十四研究所 | Multi-phase external parameter automatic calibration method based on human skeleton extraction |
CN114742905A (en) * | 2022-06-13 | 2022-07-12 | 魔视智能科技(武汉)有限公司 | Multi-camera parameter calibration method, device, equipment and storage medium |
CN116148823A (en) * | 2023-04-12 | 2023-05-23 | 北京集度科技有限公司 | External parameter calibration method, device, vehicle and computer program product |
-
2020
- 2020-06-18 CN CN202010570282.4A patent/CN111739103A/en active Pending
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113077519A (en) * | 2021-03-18 | 2021-07-06 | 中国电子科技集团公司第五十四研究所 | Multi-phase external parameter automatic calibration method based on human skeleton extraction |
CN114742905A (en) * | 2022-06-13 | 2022-07-12 | 魔视智能科技(武汉)有限公司 | Multi-camera parameter calibration method, device, equipment and storage medium |
CN116148823A (en) * | 2023-04-12 | 2023-05-23 | 北京集度科技有限公司 | External parameter calibration method, device, vehicle and computer program product |
CN116148823B (en) * | 2023-04-12 | 2023-09-19 | 北京集度科技有限公司 | External parameter calibration method, device, vehicle and computer program product |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110296691B (en) | IMU calibration-fused binocular stereo vision measurement method and system | |
CN104537707B (en) | Image space type stereoscopic vision moves real-time measurement system online | |
CN109919911B (en) | Mobile three-dimensional reconstruction method based on multi-view photometric stereo | |
CN105716542B (en) | A kind of three-dimensional data joining method based on flexible characteristic point | |
Zhang et al. | A UAV-based panoramic oblique photogrammetry (POP) approach using spherical projection | |
CN111028155B (en) | Parallax image splicing method based on multiple pairs of binocular cameras | |
CN114399554B (en) | Calibration method and system of multi-camera system | |
CN107155341B (en) | Three-dimensional scanning system and frame | |
CN110728715A (en) | Camera angle self-adaptive adjusting method of intelligent inspection robot | |
CN108416812B (en) | Calibration method of single-camera mirror image binocular vision system | |
CN111739103A (en) | Multi-camera calibration system based on single-point calibration object | |
CN108288291A (en) | Polyphaser calibration based on single-point calibration object | |
CN108629829B (en) | Three-dimensional modeling method and system of the one bulb curtain camera in conjunction with depth camera | |
CN102982551B (en) | Method for solving intrinsic parameters of parabolic catadioptric camera linearly by utilizing three unparallel straight lines in space | |
CN103106661B (en) | Two, space intersecting straight lines linear solution parabolic catadioptric camera intrinsic parameter | |
Ahmadabadian et al. | An automatic 3D reconstruction system for texture-less objects | |
CN107038753B (en) | Stereoscopic vision three-dimensional reconstruction system and method | |
CN113205603A (en) | Three-dimensional point cloud splicing reconstruction method based on rotating platform | |
CN103106662A (en) | Solving parabolic catadioptric camera parameters through one straight line in space | |
CN105374067A (en) | Three-dimensional reconstruction method based on PAL cameras and reconstruction system thereof | |
CN116129037B (en) | Visual touch sensor, three-dimensional reconstruction method, system, equipment and storage medium thereof | |
Resch et al. | On-site semi-automatic calibration and registration of a projector-camera system using arbitrary objects with known geometry | |
CN116051747A (en) | House three-dimensional model reconstruction method, device and medium based on missing point cloud data | |
Wenzel et al. | High-resolution surface reconstruction from imagery for close range cultural Heritage applications | |
CN103116888A (en) | Method for solving intrinsic parameters of cameras by plane triangles |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20201002 |