CN113554711A - Camera online calibration method and device, computer equipment and storage medium - Google Patents

Camera online calibration method and device, computer equipment and storage medium Download PDF

Info

Publication number
CN113554711A
CN113554711A CN202010338913.XA CN202010338913A CN113554711A CN 113554711 A CN113554711 A CN 113554711A CN 202010338913 A CN202010338913 A CN 202010338913A CN 113554711 A CN113554711 A CN 113554711A
Authority
CN
China
Prior art keywords
frame image
camera
current frame
pose
initial
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
Application number
CN202010338913.XA
Other languages
Chinese (zh)
Inventor
余一徽
金娜
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shanghai Ofilm Intelligent Vehicle Co ltd
Shanghai OFilm Smart Car Technology Co Ltd
Original Assignee
Shanghai Ofilm Intelligent Vehicle Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Shanghai Ofilm Intelligent Vehicle Co ltd filed Critical Shanghai Ofilm Intelligent Vehicle Co ltd
Priority to CN202010338913.XA priority Critical patent/CN113554711A/en
Publication of CN113554711A publication Critical patent/CN113554711A/en
Pending legal-status Critical Current

Links

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
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30244—Camera pose

Landscapes

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

Abstract

The application relates to the technical field of intelligent automobiles, in particular to a camera online calibration method, a camera online calibration device, computer equipment and a storage medium. The method comprises the following steps: acquiring a current frame image, a previous frame image and the driving mileage of a vehicle from the previous frame image to the current frame image, wherein the current frame image and the previous frame image are shot by a camera; calculating to obtain an initial pose value of each camera corresponding to the current frame image according to the current frame image, the previous frame image and the mileage; acquiring initial heights of cameras corresponding to the current frame image; and performing global optimization according to the initial pose value, the initial height and the driving mileage of each camera to minimize a global reprojection error, and obtaining the current pose value of each camera corresponding to the current frame image and using the current pose value as a calibration value. The method can improve the accuracy of the on-line calibration of the camera.

Description

Camera online calibration method and device, computer equipment and storage medium
Technical Field
The present application relates to the field of intelligent vehicle technologies, and in particular, to a method and an apparatus for online calibrating a camera, a computer device, and a storage medium.
Background
The accuracy of the camera calibration data directly determines the performance of the panoramic looking-around system. In the actual use process, the conventional scheme usually adopts an off-line calibration manner to obtain the camera calibration. However, such a scheme requires a high-precision calibration station and complicated manual operation, is high in cost and not easy to popularize, and cannot meet the real-time calibration requirement. And an online calibration scheme is developed, and vehicles are generally required to run on a specific road for calibration. However, such schemes generally rely on a reasonable initial value, and also require calibration on a road that meets the requirements, and when the initial value is incorrect, an error may occur in the calibration.
Disclosure of Invention
In view of the foregoing, it is desirable to provide an online camera calibration method, an online camera calibration device, a computer device, and a storage medium, which can improve the accuracy of online camera calibration.
An online calibration method for a camera, the method comprising:
acquiring a current frame image, a previous frame image and the driving mileage of a vehicle from the previous frame image to the current frame image, wherein the current frame image and the previous frame image are shot by a camera;
calculating to obtain an initial pose value of each camera corresponding to the current frame image according to the current frame image, the previous frame image and the mileage;
acquiring initial heights of cameras corresponding to the current frame image;
and performing global optimization according to the initial pose value, the initial height and the driving mileage of each camera to minimize a global reprojection error, and obtaining the current pose value of each camera corresponding to the current frame image and using the current pose value as a calibration value.
In the above embodiment, the calibration scene does not need to be limited, and only needs to be operated according to specific driving, and calibration can be completed in any smooth environment with abundant textures.
In one embodiment, the calculating, according to the current frame image, the previous frame image, and the mileage, an initial pose value of each camera corresponding to the current frame image includes:
obtaining a camera pose change matrix of the single camera under a world coordinate system according to the current frame image and the previous frame image of the single camera;
obtaining a vehicle position change matrix of the vehicle corresponding to the current frame image under a world coordinate system according to the driving mileage of the vehicle;
and obtaining the initial pose value of the single camera corresponding to the current frame image according to a hand-eye calibration algorithm.
In the embodiment, the camera pose change matrix of the single camera in the world coordinate system is obtained according to the current frame image and the previous frame image of the camera, and the vehicle can only provide plane motion and can not obtain the initial height value of each camera, so that the vehicle position change matrix of the vehicle corresponding to the current frame image in the world coordinate system is obtained by combining the driving mileage of the vehicle, the initial pose value of the single camera is more accurate, and the accuracy of online calibration of the subsequent camera can be ensured under the condition of the reasonable initial value.
In one embodiment, the obtaining a camera pose change matrix of a single camera in a world coordinate system according to the current frame image and the previous frame image of the single camera includes:
when the current frame image is a second frame image shot by the camera, acquiring a matching point pair in the second frame image and the first frame image;
and calculating to obtain a camera pose change matrix by a method for solving the essential matrix according to the acquired matching point pairs.
In the above embodiment, since there is no initial value of the camera pose at first, in this embodiment, an estimated pose is obtained by means of essential matrix decomposition, so as to perform startup, and further obtain a camera pose change matrix, and further calculate the camera pose in the image of each subsequent frame according to the camera pose change matrix, which is more accurate.
In one embodiment, the obtaining a camera pose change matrix of a single camera in a world coordinate system according to the current frame image and the previous frame image of the single camera includes:
when the current frame image is not the first frame image and the second frame image, acquiring a camera pose matrix of the previous frame image according to a hand-eye calibration algorithm and a camera pose change matrix corresponding to the previous frame image;
acquiring matching point pairs in the previous frame image and the current frame image, and obtaining corresponding three-dimensional estimation points according to the camera pose matrix of the previous frame image and the matching point pairs;
and performing minimum re-projection according to the three-dimensional estimation point and the camera pose matrix to obtain a camera pose change matrix corresponding to the current frame image.
In the embodiment, the pose of the single camera corresponding to each frame of image is obtained in a circulating mode, and circulation is performed according to the camera pose change matrix, so that the camera pose of each frame of image is more accurate.
In one embodiment, the calculating, according to the current frame image, the previous frame image, and the mileage, an initial pose value of each camera corresponding to the current frame image further includes:
and calculating the initial pose value of each camera corresponding to the current frame image according to the current frame image, the previous frame image and the driving mileage by a vanishing point method or an optical flow tracking method.
In the above embodiment, the initial pose value of each camera can be obtained by a vanishing point method or an optical flow tracking method, so that a foundation is laid for subsequent global optimization.
In one embodiment, the acquiring the initial heights of the cameras corresponding to the current frame image includes:
acquiring a termination characteristic point from the current frame image and acquiring a corresponding starting characteristic point from the previous frame image;
and calculating the initial height of each camera according to the distance between the starting point and the ending point and the driving distance of the vehicle.
In the embodiment, the vehicle can only provide plane motion, and the initial height value of each camera cannot be obtained, so that the vehicle position change matrix of the vehicle corresponding to the current frame image in the world coordinate system is obtained by combining the driving mileage of the vehicle, the initial pose value of a single camera is more accurate, and the accuracy of subsequent camera online calibration can be ensured under the condition of the reasonable initial value.
In one embodiment, the acquiring the initial heights of the cameras corresponding to the current frame image includes:
and acquiring the initial height of each camera corresponding to the current frame image through a sensor installed on a vehicle.
In the above embodiment, since the vehicle can only provide planar motion and cannot obtain the initial height value of each camera, the vehicle position change matrix of the vehicle corresponding to the current frame image in the world coordinate system is obtained by combining the driving mileage of the vehicle, so that the initial pose value of a single camera is more accurate, and the accuracy of online calibration of subsequent cameras can be ensured under the condition of the reasonable initial value
In one embodiment, the acquiring the initial heights of the cameras corresponding to the current frame image includes:
and acquiring the initial height of each camera corresponding to the current frame image according to a projection transformation method or a point cloud fitting ground mode.
In the embodiment, the vehicle can only provide plane motion, and the initial height value of each camera cannot be obtained, so that the vehicle position change matrix of the vehicle corresponding to the current frame image in the world coordinate system is obtained by combining the driving mileage of the vehicle, the initial pose value of a single camera is more accurate, and the accuracy of subsequent camera online calibration can be ensured under the condition of the reasonable initial value.
In one embodiment, the performing global optimization according to the initial pose value, the initial height, and the driving range of each camera to minimize a global re-projection error, and obtaining a current pose value of each camera corresponding to the current frame image and using the current pose value as a calibration value includes:
and performing global optimization according to the initial pose value, the initial height and the driving mileage of each camera by adopting a light beam adjustment method to minimize a global reprojection error, and obtaining the current pose value of each camera corresponding to the current frame image and using the current pose value as a calibration value.
In the embodiment, the light beam adjustment method is adopted to minimize the error of the coincidence of all the feature points in the world coordinates by re-projecting, so that the optimal solution of the poses of all the cameras is obtained, and the optimal solution is more accurate.
An online camera calibration device, the device comprising:
the data acquisition module is used for acquiring a current frame image, a previous frame image and the driving mileage of the vehicle from the previous frame image to the current frame image, wherein the current frame image is shot by the camera;
the single camera initial pose value acquisition module is used for calculating an initial pose value of each camera corresponding to the current frame image according to the current frame image, the previous frame image and the mileage;
the initial height acquisition module of a single camera is used for acquiring the initial height of each camera corresponding to the current frame image;
and the global optimization module is used for carrying out global optimization according to the initial pose value, the initial height and the driving mileage of each camera so as to minimize a global reprojection error, and obtaining the current pose value of each camera corresponding to the current frame image and using the current pose value as a calibration value.
In the above embodiment, the calibration scene does not need to be limited, and only needs to be operated according to specific driving, and calibration can be completed in any smooth environment with abundant textures.
A computer device comprising a memory storing a computer program and a processor implementing the steps of any of the methods described above when the processor executes the computer program.
In the above embodiment, the calibration scene does not need to be limited, and only needs to be operated according to specific driving, and calibration can be completed in any smooth environment with abundant textures.
A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the method of any of the above.
In the above embodiment, the calibration scene does not need to be limited, and only needs to be operated according to specific driving, and calibration can be completed in any smooth environment with abundant textures.
Drawings
FIG. 1 is a diagram of an exemplary embodiment of an on-line camera calibration method;
FIG. 2 is a schematic flow chart of an on-line camera calibration method according to an embodiment;
fig. 3 is a flowchart of an initial pose value acquisition step of a single camera in one embodiment;
FIG. 4 is a schematic diagram illustrating an embodiment of solving initial pose values of a single camera by a hand-eye calibration algorithm;
FIG. 5 is a diagram illustrating step S204 in the embodiment shown in FIG. 2;
FIG. 6 is a schematic diagram of a camera initial height acquisition process in one embodiment;
FIG. 7 is a block diagram showing the structure of an on-line camera calibration apparatus according to an embodiment;
FIG. 8 is a diagram illustrating an internal structure of a computer device according to an embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
The camera online calibration method provided by the application can be applied to the application environment shown in fig. 1. Therein, the camera 102 may communicate with a controller 104 on the vehicle. The controller 104 may acquire images captured by the cameras and acquire an initial pose value of each corresponding camera in real time according to the images, for example, calculate an initial pose value of each camera corresponding to a current frame image according to the current frame image, a previous frame image and a mileage, and acquire an initial height of each camera corresponding to the current frame image; therefore, global optimization is carried out according to the initial pose value, the initial height and the driving mileage of each camera, so that the global reprojection error is minimized, the current pose value of each camera corresponding to the current frame image is obtained and serves as a calibration value, a calibration scene is not required to be limited, only specific driving operation is required, calibration can be completed in any smooth environment with rich textures, compared with the traditional calibration scheme and most online calibration schemes in the market, the method is low in cost, minimum in personnel operation requirement and wide in applicable scene, and meanwhile the accuracy of the calibration result can be guaranteed under the operation of different operators.
In one embodiment, as shown in fig. 2, an online calibration method for a camera is provided, which is described by taking the method as an example for being applied to the controller in fig. 1, and includes the following steps:
s202: the method comprises the steps of acquiring a current frame image, a previous frame image and the driving mileage of a vehicle from the previous frame image to the current frame image, wherein the current frame image, the previous frame image and the driving mileage are shot by a camera.
Specifically, a plurality of cameras may be installed on one vehicle, and an image captured by each camera may carry a time stamp, so that the controller may determine a corresponding current frame image according to the time stamp. The current frame image is an image captured by the camera at the current time. The previous frame image refers to the previous frame image which is adjacent to the current frame image and is collected by each camera. The driving range of the vehicle refers to a driving distance of the vehicle during the collection period from the previous frame image to the current frame image, and the distance may be calculated by collecting a driving speed of the vehicle through a sensor mounted on the vehicle, then acquiring a time interval from the previous frame image to the current frame image, and calculating the driving speed and the time interval.
Alternatively, the controller may divide the images taken by the respective cameras, for example, the images taken by one camera constitute one image group, and the image frames are sorted by the time taken in the image group to facilitate the subsequent processing.
S204: and calculating to obtain the initial pose value of each camera corresponding to the current frame image according to the current frame image, the previous frame image and the driving mileage.
Specifically, the controller may first acquire the pose change matrix of the camera according to the current frame image and the previous frame image, because surrounding reference objects, such as the ground or trees, do not change during the driving process of the vehicle, the pose change matrix of the vehicle may be obtained according to the stationary reference objects and the driving range of the vehicle, and the position change matrix of the vehicle may be determined according to the driving range of the vehicle, so that an estimated value of the pose of a single camera, that is, an initial pose value of the camera corresponding to the current frame image, may be obtained.
Specifically, the controller may acquire the initial pose value of the camera in a variety of ways, for example, in one embodiment, the calculating the initial pose value of each camera corresponding to the current frame image according to the current frame image, the previous frame image and the mileage includes: the initial pose value of each camera corresponding to the current frame image is calculated according to the current frame image, the previous frame image and the driving mileage by a vanishing point method, a PnP algorithm (the PnP solving algorithm is an algorithm for solving the external parameters of the camera by using a minimized reprojection error under the condition that the internal parameters of the camera are known or unknown through a plurality of pairs of 3D and 2D matching points) or by using an optical flow tracking method. In the embodiment, the initial pose value of each camera can be obtained by a vanishing point method or an optical flow tracking method, so that a foundation is laid for subsequent global optimization.
S206: the initial height of each camera corresponding to the current frame image is acquired.
Specifically, since the vehicle can only provide plane motion, the initial pose value can only represent the initial translation value of the plane, and the initial height value of each camera cannot be obtained. It is therefore desirable to estimate the initial height of each camera, which may be obtained by various methods, such as by combining odometry information of the vehicle, solving the height estimate of each camera using a projection transformation, or by other means, such as point cloud fitting to a ground equation, adding sensors to obtain the camera height directly, etc.
In one embodiment, acquiring the initial height of each camera corresponding to the current frame image includes: the initial heights of the respective cameras corresponding to the current frame image are acquired by sensors mounted on the vehicle. In one embodiment, acquiring the initial height of each camera corresponding to the current frame image includes: and acquiring the initial height of each camera corresponding to the current frame image according to a projection transformation method or a point cloud fitting ground mode. In the embodiment, the vehicle can only provide plane motion, and the initial height value of each camera cannot be obtained, so that a vehicle position change matrix of the vehicle corresponding to the current frame image in a world coordinate system is obtained by combining the driving mileage of the vehicle, the initial pose value of a single camera is more accurate, and the accuracy of subsequent camera online calibration can be ensured under the condition of the reasonable initial value.
S208: and performing global optimization according to the initial pose value, the initial height and the driving mileage of each camera to minimize a global reprojection error, and obtaining the current pose value of each camera corresponding to the current frame image and using the current pose value as a calibration value.
Specifically, after the controller acquires the initial values of the full degrees of freedom of each camera, that is, the initial pose values and the initial heights of the cameras, the controller may perform global optimization to minimize the global re-projection error, and solve the optimal solution of the current poses of all the cameras as the current pose values of each camera and as the calibration value. The driving mileage of the vehicle provides the constraint on the scale at the same time, and the pose optimization is ensured in a reasonable range in the global optimization step.
The objective of global optimization is to minimize the error of coincidence of all feature points re-projected into world coordinates, wherein the error can be measured by projecting the feature points into the world coordinates using the current camera pose, then calculating the distance of the world coordinates of the matched feature points as the error, finding the expression of the mathematical statistics of all errors, such as the sum of all errors, then optimizing to minimize the sum of the errors, etc., and finding the current pose value of each camera as the calibration value.
Optionally, after each global optimization, the controller determines whether the obtained error is the minimum, if so, the calibration is successful, and then the process is ended, otherwise, the above step S202 is continued, the current frame is obtained again, and the optimization is continued. And in the judgment, a new image frame can be collected again, for example, a next image frame is collected, then the next image frames corresponding to the cameras are spliced, and the same spliced marker is obtained, for example, the online calibration of the cameras is carried out in a flat and rich-texture environment, the corresponding texture features can be obtained as the markers, then the dislocation error of the same marker in the spliced image is judged, if the dislocation error meets the requirement, the calibration is successful, the calibration is finished, otherwise, the next image frame is continuously obtained as the current frame, and the calibration of the cameras is carried out again.
In the above embodiment, the calibration scene does not need to be limited, and only needs to be operated according to specific driving, and calibration can be completed in any smooth environment with abundant textures.
In one embodiment, please refer to fig. 3, fig. 3 is a flowchart of an initial pose value acquiring step of a single camera in an embodiment, in which the method mainly includes the following steps:
s302: and obtaining a camera pose change matrix of the single camera under a world coordinate system according to the current frame image and the previous frame image of the single camera.
Specifically, the controller may calculate the camera pose change matrix in the world coordinate system by means of feature matching, for example, corresponding feature points in the previous frame image and corresponding feature points in the current frame image. For example, the controller may assume a camera pose change matrix containing unknowns, and then substitute the unknowns according to the determined feature points, so as to obtain each unknowns, thereby obtaining the camera pose change matrix in the world coordinate system.
S304: and obtaining a vehicle position change matrix of the vehicle corresponding to the current frame image under the world coordinate system according to the driving mileage of the vehicle.
In particular, the vehicle position change matrix is also in the world coordinate system, such as may be represented by (x)t，yt，zt) To indicate the position of the vehicle in the current frame image, by (x)t-1，yt-1，zt-1) To indicate the position of the vehicle in the previous frame image, so that after the position in the previous frame image is known, and the driving range and the driving direction of the vehicle are known, the controller can calculate the position (x) of the vehicle in the current frame imaget，yt，zt) Thus, a vehicle position change matrix can be determined from the two positions.
S306: and obtaining the initial pose value of the single camera corresponding to the current frame image according to a hand-eye calibration algorithm.
Specifically, the basis of the hand-eye calibration algorithm is as follows:
Hcam2veh*Hcam＝Hveh*Hcam2veh (1)
wherein HcamAs a camera pose change matrix, HvehIs a vehicle position change matrix, Hcam2vehIs the initial pose value of a single camera.
According to the formula, after the camera pose change matrix and the vehicle position change matrix are known, the initial pose value of a single camera can be obtained.
In the embodiment, the camera pose change matrix of the single camera in the world coordinate system is obtained according to the current frame image and the previous frame image of the camera, and the vehicle can only provide plane motion and can not obtain the initial height value of each camera, so that the vehicle position change matrix of the vehicle corresponding to the current frame image in the world coordinate system is obtained by combining the driving mileage of the vehicle, the initial pose value of the single camera is more accurate, and the accuracy of online calibration of the subsequent camera can be ensured under the condition of the reasonable initial value.
In one embodiment, obtaining a camera pose change matrix of a single camera in a world coordinate system according to a current frame image and a previous frame image of the single camera comprises: when the current frame image is a second frame image shot by a camera, acquiring a matching point pair in the second frame image and the first frame image; and calculating to obtain a camera pose change matrix by a method for solving the essential matrix according to the acquired matching point pairs.
In the above embodiment, since there is no initial value of the camera pose at first, in this embodiment, an estimated pose is obtained by means of essential matrix decomposition, so as to perform startup, and further obtain a camera pose change matrix, and further calculate the camera pose in the image of each subsequent frame according to the camera pose change matrix, which is more accurate.
In one embodiment, obtaining a camera pose change matrix of a single camera in a world coordinate system according to a current frame image and a previous frame image of the single camera comprises: when the current frame image is not the first frame image and the second frame image, acquiring a camera pose matrix of the previous frame image according to a hand-eye calibration algorithm and a camera pose change matrix corresponding to the previous frame image; acquiring a matching point pair in the previous frame image and the current frame image, and obtaining a corresponding three-dimensional estimation point according to the camera pose matrix of the previous frame image and the matching point pair; and performing minimum re-projection according to the three-dimensional estimation points and the camera pose matrix to obtain a camera pose change matrix corresponding to the current frame image.
In the embodiment, the pose of the single camera corresponding to each frame of image is obtained in a circulating mode, and circulation is performed according to the camera pose change matrix, so that the camera pose of each frame of image is more accurate.
Specifically, please refer to fig. 4 and 5, wherein fig. 4 is a schematic diagram of solving the initial pose value of a single camera by a hand-eye calibration algorithm in an embodiment, and fig. 5 is a schematic diagram of step S204 in the embodiment shown in fig. 2. In this embodiment, the PnP algorithm solves for the evidence of the camera pose change between two frames of images, and at this time, the pose of the camera (relative to the vehicle) needs to be solved through hand-eye calibration. The PnP algorithm needs to calculate 3D points every time, but does not need the pose of a camera relative to a vehicle, and only needs the pose change of the camera between two frames for proving. For the first frame, no initial value exists, so an estimated pose needs to be obtained in an essential matrix decomposition mode and used as an input of the PnP algorithm to start the PnP algorithm.
Specifically, referring to fig. 4 and fig. 5, taking one of the cameras as an example for description, the controller first acquires a first frame image acquired by the camera, where the first frame image is a valid frame, and the first frame image is not necessarily a first frame acquired by the camera, but may be a first valid frame used for online calibration of the camera.
The controller also acquires a second frame acquired by the camera, and because no initial value exists at the moment, the controller acquires the estimated pose of the camera by an essential matrix decomposition method, wherein the essential matrix is a special case of the basic matrix and is the basic matrix in a normalized image coordinate system. The controller may acquire the second frame image and the matching point pairs in the first frame image, for example, may acquire the matching point pairs according to invariant points on the vehicle, such as the position of a vehicle door, the position of a tire, and the like, so as to acquire a plurality of pairs of matching point pairs, and then calculate a camera pose change matrix, that is, H in the graph, by a method of solving an essential matrixcam(1) Then, the camera pose change moment corresponding to the previous frame of image can be obtained through a hand-eye calibration algorithmArray and vehicle position change matrix between the two frames, and camera pose matrix for obtaining previous frame image, namely Hcam2veh(1)。
For the non-first frame image and the non-second frame image, due to the fact that the PnP has an initial value, namely a camera pose change matrix corresponding to the previous frame image, camera pose proof of the previous frame image can be obtained, then matching points of the images are determined according to the images, corresponding three-dimensional estimation points, namely points after re-projection, can be obtained according to the camera pose matrix and the matching point pairs of the previous frame image, then the controller conducts minimum re-projection according to the three-dimensional estimation points and the camera pose matrix, and a camera pose change matrix corresponding to the current frame image is obtained, wherein the minimum re-projection can be used for determining the camera pose change matrix corresponding to the current frame image according to the difference value between the determined three-dimensional estimation points and actual points (determined according to the current frame image) and minimizing the measurement index.
Thus, the subsequent frames are continuously processed circularly to obtain the corresponding camera pose transformation matrix and the initial pose value.
In the above embodiment, since there is no initial value of the camera pose at the beginning, in this embodiment, an estimated pose is obtained by a way of essential matrix decomposition, so as to perform startup, and further, a camera pose change matrix can be obtained, and further, the camera pose in the image of each subsequent frame is calculated according to the camera pose change matrix, which is more accurate.
In one embodiment, acquiring the initial height of each camera corresponding to the current frame image includes: acquiring a termination characteristic point from a current frame image and acquiring a corresponding starting characteristic point from a previous frame image; and calculating the initial height of each camera according to the distance between the starting point and the ending point and the driving mileage of the vehicle.
Specifically, referring to fig. 6, fig. 6 is a schematic diagram of an initial height acquiring process of a camera in an embodiment.
The controller can acquire an end feature point and a corresponding start feature point from a previous frame of image, wherein the end feature point and the start feature point are mutually matched feature points and are positioned on the cameras, so that the controller can calculate the initial height of each camera according to the distance between the start point and the end point and the driving mileage of the vehicle by using a similarity principle.
Wherein in FIG. 6, x(1)Is the starting feature point, X(2)Is the termination feature point. P(1)Is the point of the starting feature point on the ground plane, P(2)Is the point on the ground plane where the termination feature point corresponds to, where the height of the camera may be:
Figure BDA0002467819340000121
in the embodiment, the vehicle can only provide plane motion, and the initial height value of each camera cannot be obtained, so that the vehicle position change matrix of the vehicle corresponding to the current frame image in the world coordinate system is obtained by combining the driving mileage of the vehicle, the initial pose value of a single camera is more accurate, and the accuracy of subsequent camera online calibration can be ensured under the condition of the reasonable initial value.
In one embodiment, performing global optimization according to the initial pose value, the initial height, and the driving range of each camera to minimize a global reprojection error, and obtaining a current pose value of each camera corresponding to a current frame image and using the current pose value as a calibration value, includes:
and performing global optimization according to the initial pose value, the initial height and the driving mileage of each camera by adopting a light beam adjustment method so as to minimize the global reprojection error, and obtaining the current pose value of each camera corresponding to the current frame image and using the current pose value as a calibration value.
Specifically, after the controller acquires the initial value of the full degree of freedom of each camera, namely the initial pose value and the initial height of each camera, global optimization is performed, so that the global reprojection error is minimized, and the optimal solution of the poses of all the cameras at present is obtained.
Wherein camera pose is optimized such that the error of coincidence of all feature points re-projected into world coordinates is minimized. The metrics may project the feature points into world coordinates using the current camera pose, and then calculate the distances of the world coordinates of the feature points on the resulting match. And obtaining the optimal solution of all the current camera poses by minimizing the distance.
In the embodiment, the light beam adjustment method is adopted to minimize the error of the coincidence of all the feature points in the world coordinates by re-projecting, so that the optimal solution of the poses of all the cameras is obtained, and the optimal solution is more accurate.
Referring to fig. 6, fig. 6 is a flowchart of an online calibration method for cameras in another embodiment, in which each initial pose value of each camera is first calculated, and specifically, the controller obtains pose change H of the camera in the world coordinate system according to feature matching between two frames of a single cameracamObtaining the pose change H of the vehicle by combining with a vehicle odometervehAnd further finding the pose estimation value H of the single cameracam2veh. Since the vehicle can only provide plane motion, the above process can only obtain the initial translation value of the plane, and the initial height value of each camera cannot be obtained. Therefore, the initial height of each camera needs to be calculated and estimated, and finally, after the controller obtains the initial value of the full degree of freedom of each camera, global optimization is performed, so that the global reprojection error is minimum, and the optimal solution of the pose of all the cameras at present is obtained. Algorithms for realizing the initial camera pose value are various, for example, the robust initial camera pose value can be obtained by methods such as PnP, vanishing point utilization, optical flow tracking and the like. By combining the odometer information of the vehicle, the height estimate of each camera can be obtained by utilizing projection transformation solution, or the height estimate of each camera can be obtained by other methods, such as point cloud fitting ground equation, adding a sensor to directly obtain the height of each camera, and the like, and the method is not limited. And the odometer information of the vehicle provides the constraint on the scale at the same time, and the pose optimization is ensured to be in a reasonable range in the global optimization step.
In the above embodiment, the calibration scene does not need to be limited, and only needs to be operated according to specific driving, and calibration can be completed in any smooth environment with abundant textures.
It should be understood that although the steps in the flowcharts of fig. 2, 3, 5 and 6 are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least some of the steps in fig. 2, 3, 5 and 6 may include multiple steps or multiple stages, which are not necessarily performed at the same time, but may be performed at different times, which are not necessarily performed in sequence, but may be performed alternately or alternately with other steps or at least some of the other steps.
In one embodiment, as shown in fig. 7, there is provided an online calibration apparatus for a camera, including: a data acquisition module 100, a single-camera initial pose value acquisition module 200, a single-camera initial height acquisition module 300, and a global optimization module 400, wherein:
a data acquisition module 100, configured to acquire a current frame image, a previous frame image and a driving distance of the vehicle from the previous frame image to the current frame image, which are captured by the camera;
the single camera initial pose value acquisition module 200 is configured to calculate an initial pose value of each camera corresponding to a current frame image according to the current frame image, a previous frame image and the mileage;
an initial height obtaining module 300 for a single camera, configured to obtain initial heights of the cameras corresponding to the current frame image;
and the global optimization module 400 is configured to perform global optimization according to the initial pose value, the initial height, and the driving range of each camera, so as to minimize a global re-projection error, and obtain a current pose value of each camera corresponding to the current frame image and use the current pose value as a calibration value.
In one embodiment, the single-camera initial pose value acquisition module 200 described above may include:
the camera pose change matrix acquisition unit is used for acquiring a camera pose change matrix of the single camera in a world coordinate system according to the current frame image and the previous frame image of the single camera;
the vehicle position change matrix obtaining unit is used for obtaining a vehicle position change matrix of the vehicle corresponding to the current frame image under a world coordinate system according to the driving mileage of the vehicle;
and the initial pose value acquisition unit of the single camera is used for obtaining the initial pose value of the single camera corresponding to the current frame image according to the hand-eye calibration algorithm.
In one embodiment, the camera pose change matrix acquisition unit may include:
a matching point pair obtaining subunit, configured to obtain a matching point pair in the second frame image and the first frame image when the current frame image is a second frame image captured by the camera;
and the first camera pose change matrix acquisition subunit is used for calculating to obtain a camera pose change matrix by a method for solving the essential matrix according to the acquired matching point pairs.
In one embodiment, the camera pose change matrix acquisition unit may further include:
the camera pose matrix acquiring subunit is used for acquiring a camera pose matrix of the previous frame of image according to a hand-eye calibration algorithm and a camera pose change matrix corresponding to the previous frame of image when the current frame of image is not the first frame of image and the second frame of image;
the three-dimensional estimation point acquisition subunit is used for acquiring a previous frame image and a matching point pair in the current frame image and acquiring a corresponding three-dimensional estimation point according to the camera pose matrix of the previous frame image and the matching point pair;
and the second camera pose change matrix acquisition subunit is used for performing minimum re-projection according to the three-dimensional estimation points and the camera pose matrix to obtain a camera pose change matrix corresponding to the current frame image.
In one embodiment, the initial pose value acquiring module of the camera may be further configured to calculate, by a vanishing point method or by using an optical flow tracking method, an initial pose value of each camera corresponding to a current frame image according to the current frame image, a previous frame image, and a mileage.
In one embodiment, the initial height acquisition module 300 for a single camera may include:
the feature point acquisition unit is used for acquiring a termination feature point from a current frame image and acquiring a corresponding starting feature point from a previous frame image;
and the initial height acquisition unit is used for calculating the initial height of each camera according to the distance between the starting point and the ending point and the driving mileage of the vehicle.
In one embodiment, the initial height acquiring module 300 of a single camera may be further configured to acquire the initial heights of the respective cameras corresponding to the current frame image through a sensor mounted on the vehicle.
In one embodiment, the initial height acquiring module 300 of a single camera may be further configured to acquire the initial heights of the cameras corresponding to the current frame image according to a projective transformation method or a point cloud fitting ground method.
In one embodiment, the global optimization module 400 may further be configured to perform global optimization according to the initial pose value, the initial height, and the driving range of each camera by using a light beam adjustment method, so as to minimize a global reprojection error, and obtain a current pose value of each camera corresponding to the current frame image and use the current pose value as a calibration value.
For specific limitations of the online camera calibration device, reference may be made to the above limitations of the online camera calibration method, which are not described herein again. The modules in the camera online calibration device can be wholly or partially realized by software, hardware and a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as shown in fig. 8. The computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The communication interface of the computer device is used for carrying out wired or wireless communication with an external terminal, and the wireless communication can be realized through WIFI, an operator network, NFC (near field communication) or other technologies. The computer program is executed by a processor to implement an online camera calibration method. The display screen of the computer equipment can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer equipment can be a touch layer covered on the display screen, a key, a track ball or a touch pad arranged on the shell of the computer equipment, an external keyboard, a touch pad or a mouse and the like.
Those skilled in the art will appreciate that the architecture shown in fig. 8 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
In one embodiment, a computer device is provided, comprising a memory and a processor, the memory having a computer program stored therein, the processor implementing the following steps when executing the computer program: acquiring a current frame image, a previous frame image and the driving mileage of a vehicle from the previous frame image to the current frame image, wherein the current frame image and the previous frame image are shot by a camera; calculating to obtain an initial pose value of each camera corresponding to the current frame image according to the current frame image, the previous frame image and the driving mileage; acquiring initial heights of cameras corresponding to the current frame image; and performing global optimization according to the initial pose value, the initial height and the driving mileage of each camera to minimize a global reprojection error, and obtaining the current pose value of each camera corresponding to the current frame image and using the current pose value as a calibration value.
In one embodiment, the calculation of the initial pose value of each camera corresponding to the current frame image according to the current frame image, the previous frame image and the mileage, which is realized when the processor executes the computer program, includes: obtaining a camera pose change matrix of the single camera under a world coordinate system according to a current frame image and a previous frame image of the single camera; obtaining a vehicle position change matrix of the vehicle corresponding to the current frame image under a world coordinate system according to the driving mileage of the vehicle; and obtaining the initial pose value of the single camera corresponding to the current frame image according to a hand-eye calibration algorithm.
In one embodiment, the obtaining of the camera pose change matrix of the single camera in the world coordinate system according to the current frame image and the last frame image of the single camera, which is realized by the processor when the processor executes the computer program, comprises: when the current frame image is a second frame image shot by a camera, acquiring a matching point pair in the second frame image and the first frame image; and calculating to obtain a camera pose change matrix by a method for solving the essential matrix according to the acquired matching point pairs.
In one embodiment, the obtaining of the camera pose change matrix of the single camera in the world coordinate system according to the current frame image and the last frame image of the single camera, which is realized by the processor when the processor executes the computer program, comprises: when the current frame image is not the first frame image and the second frame image, acquiring a camera pose matrix of the previous frame image according to a hand-eye calibration algorithm and a camera pose change matrix corresponding to the previous frame image; acquiring a matching point pair in the previous frame image and the current frame image, and obtaining a corresponding three-dimensional estimation point according to the camera pose matrix of the previous frame image and the matching point pair; and performing minimum re-projection according to the three-dimensional estimation points and the camera pose matrix to obtain a camera pose change matrix corresponding to the current frame image.
In one embodiment, the calculation of the initial pose value of each camera corresponding to the current frame image according to the current frame image, the previous frame image and the mileage, which is realized when the processor executes the computer program, further includes: and calculating the initial pose value of each camera corresponding to the current frame image according to the current frame image, the previous frame image and the driving mileage by a vanishing point method or an optical flow tracking method.
In one embodiment, the obtaining of the initial height of each camera corresponding to the current frame image, as implemented by the processor when executing the computer program, comprises: acquiring a termination characteristic point from a current frame image and acquiring a corresponding starting characteristic point from a previous frame image; and calculating the initial height of each camera according to the distance between the starting point and the ending point and the driving mileage of the vehicle.
In one embodiment, the obtaining of the initial height of each camera corresponding to the current frame image, as implemented by the processor when executing the computer program, comprises: the initial heights of the respective cameras corresponding to the current frame image are acquired by sensors mounted on the vehicle.
In one embodiment, the obtaining of the initial height of each camera corresponding to the current frame image, as implemented by the processor when executing the computer program, comprises: and acquiring the initial height of each camera corresponding to the current frame image according to a projection transformation method or a point cloud fitting ground mode.
In one embodiment, the global optimization according to the initial pose value, the initial height, and the driving range of each camera, which is implemented when the processor executes the computer program, so as to minimize the global re-projection error, and obtain the current pose value of each camera corresponding to the current frame image and use the current pose value as the calibration value, includes: and performing global optimization according to the initial pose value, the initial height and the driving mileage of each camera by adopting a light beam adjustment method so as to minimize the global reprojection error, and obtaining the current pose value of each camera corresponding to the current frame image and using the current pose value as a calibration value.
In one embodiment, a computer-readable storage medium is provided, having a computer program stored thereon, which when executed by a processor, performs the steps of: acquiring a current frame image, a previous frame image and the driving mileage of a vehicle from the previous frame image to the current frame image, wherein the current frame image and the previous frame image are shot by a camera; calculating to obtain an initial pose value of each camera corresponding to the current frame image according to the current frame image, the previous frame image and the driving mileage; acquiring initial heights of cameras corresponding to the current frame image; and performing global optimization according to the initial pose value, the initial height and the driving mileage of each camera to minimize a global reprojection error, and obtaining the current pose value of each camera corresponding to the current frame image and using the current pose value as a calibration value.
In one embodiment, the calculation of the initial pose value of each camera corresponding to the current frame image according to the current frame image, the previous frame image and the mileage, when the computer program is executed by the processor, includes: obtaining a camera pose change matrix of the single camera under a world coordinate system according to a current frame image and a previous frame image of the single camera; obtaining a vehicle position change matrix of the vehicle corresponding to the current frame image under a world coordinate system according to the driving mileage of the vehicle; and obtaining the initial pose value of the single camera corresponding to the current frame image according to a hand-eye calibration algorithm.
In one embodiment, the obtaining of the camera pose change matrix of the single camera in the world coordinate system according to the current frame image and the last frame image of the single camera, which is realized when the computer program is executed by the processor, comprises: when the current frame image is a second frame image shot by a camera, acquiring a matching point pair in the second frame image and the first frame image; and calculating to obtain a camera pose change matrix by a method for solving the essential matrix according to the acquired matching point pairs.
In one embodiment, the obtaining of the camera pose change matrix of the single camera in the world coordinate system according to the current frame image and the last frame image of the single camera, which is realized when the computer program is executed by the processor, comprises: when the current frame image is not the first frame image and the second frame image, acquiring a camera pose matrix of the previous frame image according to a hand-eye calibration algorithm and a camera pose change matrix corresponding to the previous frame image; acquiring a matching point pair in the previous frame image and the current frame image, and obtaining a corresponding three-dimensional estimation point according to the camera pose matrix of the previous frame image and the matching point pair; and performing minimum re-projection according to the three-dimensional estimation points and the camera pose matrix to obtain a camera pose change matrix corresponding to the current frame image.
In one embodiment, the calculation of the initial pose value of each camera corresponding to the current frame image according to the current frame image, the previous frame image and the mileage, when the computer program is executed by the processor, further includes: and calculating the initial pose value of each camera corresponding to the current frame image according to the current frame image, the previous frame image and the driving mileage by a vanishing point method or an optical flow tracking method.
In one embodiment, the computer program, when executed by a processor, implements obtaining an initial height for each camera corresponding to a current frame image, comprising: acquiring a termination characteristic point from a current frame image and acquiring a corresponding starting characteristic point from a previous frame image; and calculating the initial height of each camera according to the distance between the starting point and the ending point and the driving mileage of the vehicle.
In one embodiment, the computer program, when executed by a processor, implements obtaining an initial height for each camera corresponding to a current frame image, comprising: the initial heights of the respective cameras corresponding to the current frame image are acquired by sensors mounted on the vehicle.
In one embodiment, the computer program, when executed by a processor, implements obtaining an initial height for each camera corresponding to a current frame image, comprising: and acquiring the initial height of each camera corresponding to the current frame image according to a projection transformation method or a point cloud fitting ground mode.
In one embodiment, the global optimization according to the initial pose value, the initial height, and the mileage of each camera, implemented when the computer program is executed by the processor, so as to minimize the global reprojection error, and obtain the current pose value of each camera corresponding to the current frame image and use the current pose value as the calibration value, includes: and performing global optimization according to the initial pose value, the initial height and the driving mileage of each camera by adopting a light beam adjustment method so as to minimize the global reprojection error, and obtaining the current pose value of each camera corresponding to the current frame image and using the current pose value as a calibration value.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, database or other medium used in the embodiments provided herein can include at least one of non-volatile and volatile memory. Non-volatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical storage, or the like. Volatile Memory can include Random Access Memory (RAM) or external cache Memory. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM), among others.
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.

Claims (12)

1. An online calibration method for a camera, the method comprising:
acquiring a current frame image, a previous frame image and the driving mileage of a vehicle from the previous frame image to the current frame image, wherein the current frame image and the previous frame image are shot by a camera;
calculating to obtain an initial pose value of each camera corresponding to the current frame image according to the current frame image, the previous frame image and the mileage;
acquiring initial heights of cameras corresponding to the current frame image;
and performing global optimization according to the initial pose value, the initial height and the driving mileage of each camera to minimize a global reprojection error, and obtaining the current pose value of each camera corresponding to the current frame image and using the current pose value as a calibration value.
2. The method of claim 1, wherein the calculating an initial pose value of each camera corresponding to the current frame image according to the current frame image, a previous frame image and the mileage includes:
obtaining a camera pose change matrix of the single camera under a world coordinate system according to the current frame image and the previous frame image of the single camera;
obtaining a vehicle position change matrix of the vehicle corresponding to the current frame image under a world coordinate system according to the driving mileage of the vehicle;
and obtaining the initial pose value of the single camera corresponding to the current frame image according to a hand-eye calibration algorithm.
3. The method according to claim 2, wherein the deriving a camera pose change matrix of a single camera in a world coordinate system according to the current frame image and a previous frame image of the single camera comprises:
when the current frame image is a second frame image shot by the camera, acquiring a matching point pair in the second frame image and the first frame image;
and calculating to obtain a camera pose change matrix by a method for solving the essential matrix according to the acquired matching point pairs.
4. The method according to claim 2 or 3, wherein the obtaining a camera pose change matrix of the single camera in a world coordinate system according to the current frame image and the last frame image of the single camera comprises:
when the current frame image is not the first frame image and the second frame image, acquiring a camera pose matrix of the previous frame image according to a hand-eye calibration algorithm and a camera pose change matrix corresponding to the previous frame image;
acquiring matching point pairs in the previous frame image and the current frame image, and obtaining corresponding three-dimensional estimation points according to the camera pose matrix of the previous frame image and the matching point pairs;
and performing minimum re-projection according to the three-dimensional estimation point and the camera pose matrix to obtain a camera pose change matrix corresponding to the current frame image.
5. The method according to claim 1, wherein the calculating an initial pose value of each camera corresponding to the current frame image according to the current frame image, a previous frame image and the mileage includes:
and calculating the initial pose value of each camera corresponding to the current frame image according to the current frame image, the previous frame image and the driving mileage by a vanishing point method or an optical flow tracking method.
6. The method according to any one of claims 1 to 5, wherein the obtaining of the initial height of each camera corresponding to the current frame image comprises:
acquiring a termination characteristic point from the current frame image and acquiring a corresponding starting characteristic point from the previous frame image;
and calculating the initial height of each camera according to the distance between the starting point and the ending point and the driving distance of the vehicle.
7. The method according to any one of claims 1 to 5, wherein the obtaining of the initial height of each camera corresponding to the current frame image comprises:
and acquiring the initial height of each camera corresponding to the current frame image through a sensor installed on a vehicle.
8. The method according to any one of claims 1 to 5, wherein the obtaining of the initial height of each camera corresponding to the current frame image comprises:
and acquiring the initial height of each camera corresponding to the current frame image according to a projection transformation method or a point cloud fitting ground mode.
9. The method according to any one of claims 1 to 5, wherein the global optimization according to the initial pose value, the initial height and the driving range of each camera so as to minimize a global reprojection error, and obtaining the current pose value of each camera corresponding to the current frame image as a calibration value, comprises:
and performing global optimization according to the initial pose value, the initial height and the driving mileage of each camera by adopting a light beam adjustment method to minimize a global reprojection error, and obtaining the current pose value of each camera corresponding to the current frame image and using the current pose value as a calibration value.
10. The online calibration device of camera, its characterized in that, the said device includes:
the data acquisition module is used for acquiring a current frame image, a previous frame image and the driving mileage of the vehicle from the previous frame image to the current frame image, wherein the current frame image is shot by the camera;
the single camera initial pose value acquisition module is used for calculating an initial pose value of each camera corresponding to the current frame image according to the current frame image, the previous frame image and the mileage;
the initial height acquisition module of a single camera is used for acquiring the initial height of each camera corresponding to the current frame image;
and the global optimization module is used for carrying out global optimization according to the initial pose value, the initial height and the driving mileage of each camera so as to minimize a global reprojection error, and obtaining the current pose value of each camera corresponding to the current frame image and using the current pose value as a calibration value.
11. A computer device comprising a memory and a processor, the memory storing a computer program, wherein the processor implements the steps of the method of any one of claims 1 to 9 when executing the computer program.
12. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 9.
CN202010338913.XA 2020-04-26 2020-04-26 Camera online calibration method and device, computer equipment and storage medium Pending CN113554711A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010338913.XA CN113554711A (en) 2020-04-26 2020-04-26 Camera online calibration method and device, computer equipment and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010338913.XA CN113554711A (en) 2020-04-26 2020-04-26 Camera online calibration method and device, computer equipment and storage medium

Publications (1)

Publication Number Publication Date
CN113554711A true CN113554711A (en) 2021-10-26

Family

ID=78101516

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010338913.XA Pending CN113554711A (en) 2020-04-26 2020-04-26 Camera online calibration method and device, computer equipment and storage medium

Country Status (1)

Country Link
CN (1) CN113554711A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117523010A (en) * 2024-01-05 2024-02-06 深圳市欧冶半导体有限公司 Method and device for determining camera pose of vehicle, computer equipment and storage medium

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105844624A (en) * 2016-03-18 2016-08-10 上海欧菲智能车联科技有限公司 Dynamic calibration system, and combined optimization method and combined optimization device in dynamic calibration system
CN108648240A (en) * 2018-05-11 2018-10-12 东南大学 Based on a non-overlapping visual field camera posture scaling method for cloud characteristics map registration
CN108765498A (en) * 2018-05-30 2018-11-06 百度在线网络技术（北京）有限公司 Monocular vision tracking, device and storage medium
CN109389650A (en) * 2018-09-30 2019-02-26 京东方科技集团股份有限公司 A kind of scaling method of in-vehicle camera, device, vehicle and storage medium
CN109544629A (en) * 2018-11-29 2019-03-29 南京人工智能高等研究院有限公司 Camera pose determines method and apparatus and electronic equipment
US20190259178A1 (en) * 2018-02-21 2019-08-22 Ficosa Adas, S.L.U. Methods of calibrating a camera of a vehicle and systems
CN110246193A (en) * 2019-06-20 2019-09-17 南京博蓝奇智能科技有限公司 Industrial robot end camera online calibration method
CN110349219A (en) * 2018-04-04 2019-10-18 杭州海康威视数字技术股份有限公司 A kind of Camera extrinsic scaling method and device
CN110675455A (en) * 2019-08-30 2020-01-10 的卢技术有限公司 Self-calibration method and system for car body all-around camera based on natural scene

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105844624A (en) * 2016-03-18 2016-08-10 上海欧菲智能车联科技有限公司 Dynamic calibration system, and combined optimization method and combined optimization device in dynamic calibration system
US20190259178A1 (en) * 2018-02-21 2019-08-22 Ficosa Adas, S.L.U. Methods of calibrating a camera of a vehicle and systems
CN110349219A (en) * 2018-04-04 2019-10-18 杭州海康威视数字技术股份有限公司 A kind of Camera extrinsic scaling method and device
CN108648240A (en) * 2018-05-11 2018-10-12 东南大学 Based on a non-overlapping visual field camera posture scaling method for cloud characteristics map registration
CN108765498A (en) * 2018-05-30 2018-11-06 百度在线网络技术（北京）有限公司 Monocular vision tracking, device and storage medium
CN109389650A (en) * 2018-09-30 2019-02-26 京东方科技集团股份有限公司 A kind of scaling method of in-vehicle camera, device, vehicle and storage medium
CN109544629A (en) * 2018-11-29 2019-03-29 南京人工智能高等研究院有限公司 Camera pose determines method and apparatus and electronic equipment
CN110246193A (en) * 2019-06-20 2019-09-17 南京博蓝奇智能科技有限公司 Industrial robot end camera online calibration method
CN110675455A (en) * 2019-08-30 2020-01-10 的卢技术有限公司 Self-calibration method and system for car body all-around camera based on natural scene

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117523010A (en) * 2024-01-05 2024-02-06 深圳市欧冶半导体有限公司 Method and device for determining camera pose of vehicle, computer equipment and storage medium
CN117523010B (en) * 2024-01-05 2024-04-09 深圳市欧冶半导体有限公司 Method and device for determining camera pose of vehicle, computer equipment and storage medium

Similar Documents

Publication Publication Date Title
CN111445526B (en) Method, device and storage medium for estimating pose of image frame
CN107796397B (en) Robot binocular vision positioning method and device and storage medium
CN110033489B (en) Method, device and equipment for evaluating vehicle positioning accuracy
CN108955718B (en) Visual odometer and positioning method thereof, robot and storage medium
CN113870343B (en) Relative pose calibration method, device, computer equipment and storage medium
CN109472828B (en) Positioning method, positioning device, electronic equipment and computer readable storage medium
CN108519102B (en) Binocular vision mileage calculation method based on secondary projection
CN111127524A (en) Method, system and device for tracking trajectory and reconstructing three-dimensional image
CN114898313B (en) Method, device, equipment and storage medium for generating bird's eye view of driving scene
CN110648363A (en) Camera posture determining method and device, storage medium and electronic equipment
CN112689234B (en) Indoor vehicle positioning method, device, computer equipment and storage medium
JP2018077162A (en) Vehicle position detection device, vehicle position detection method and computer program for vehicle position detection
CN109029442A (en) Based on the matched positioning device of multi-angle of view and method
CN114217665A (en) Camera and laser radar time synchronization method, device and storage medium
WO2023284358A1 (en) Camera calibration method and apparatus, electronic device, and storage medium
Cui et al. Mcsfm: multi-camera-based incremental structure-from-motion
CN113554711A (en) Camera online calibration method and device, computer equipment and storage medium
CN109708655A (en) Air navigation aid, device, vehicle and computer readable storage medium
CN116563352A (en) Single-line laser radar loop detection method and system integrating depth visual information
CN115222815A (en) Obstacle distance detection method, obstacle distance detection device, computer device, and storage medium
CN111462321B (en) Point cloud map processing method, processing device, electronic device and vehicle
CN114092564A (en) External parameter calibration method, system, terminal and medium of non-overlapping view field multi-camera system
CN116762094A (en) Data processing method and device
CN112598736A (en) Map construction based visual positioning method and device
Miljković et al. Vehicle Distance Estimation Based on Stereo Camera System with Implementation on a Real ADAS Board

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