Distributed POS multi-node multi-parameter instant synchronous calibration method
Technical Field
The invention relates to the field of distributed POS measurement, in particular to a distributed POS multi-node multi-parameter instant synchronous calibration method which is suitable for precision calibration of a distributed POS system under the condition that a measurement carrier is elastically deformed, namely under the dynamic condition.
Background
The Position and Orientation System (POS) is composed of an Inertial Measurement Unit (IMU), a navigation Computer System (PCS), and a gps (global Positioning System). The POS can provide high-precision motion navigation information for a high-resolution aerial remote sensing system, improves the imaging quality of a remote sensing load through motion error compensation, and is the key for realizing high-resolution imaging.
China achieves good results in the aspect of single POS imaging and realizes high-resolution two-dimensional imaging. With the continuous development of an airborne earth observation system, the two-dimensional imaging requirement is gradually changed to three-dimensional imaging, so that an airborne distributed array antenna SAR, a flexible multi-baseline interference SAR and a plurality of or a plurality of loads are required to be installed at different positions of an airplane, and the high-precision position and attitude measurement of each load and the time unification of each load data cannot be realized by adopting the traditional single POS. Distributed POS plays a crucial role in multi-load, three-dimensional imaging. The distributed POS includes a master POS, which is mounted in the pod under the belly and acts as a master node, and a plurality of sub-IMUs, which are mounted on the wing at each imaging load and act as sub-nodes. The main node and the sub-nodes respectively measure position and attitude data of each node, high-precision navigation data of the main node is transmitted to the sub-nodes by utilizing a transmission alignment technology, high-precision navigation information of the sub-nodes is realized, and then high-precision motion compensation is completed on a plurality of imaging loads. The distributed POS measurement accuracy is a key index for determining the performance of the distributed POS, and effective accuracy calibration of the distributed POS is one of key technologies to be solved urgently in the field.
At present, the single POS in the published documents has more calibration methods, such as an aerial triangulation method and a turntable calibration method. Calibration methods for distributed POS are relatively few.
The accuracy calibration method and device of the distributed POS comprise the following steps: the content of CN. (application No. CN201810153914. X; publication No. CN108106637A) is that two cameras without a common view field are used for respectively measuring targets pasted on the surface of the distributed POS to perform calibration work of the distributed POS, the relationship calibration between the two cameras uses two fixedly connected plane targets, the attitude relationship between the two cameras is calibrated by using the condition that the included angle of two unit vectors on two target planes before and after the movement of the targets does not become the constraint condition, and the position relationship between the two cameras is calibrated by using the condition that the distance between two points on the targets does not become the constraint condition before and after the movement of the targets.
Disclosure of Invention
The invention solves the problems: the distributed POS multi-node multi-parameter instantaneous synchronous calibration method is simple to operate and easy to realize, only the camera needs to be moved for multiple times, and the whole calculation process can be completed by using hand-eye calibration; the distributed POS multi-node multi-parameter instantaneous synchronous calibration method has the advantages of being high in precision, non-contact in measurement and high in anti-interference capacity, and the distributed POS multi-node multi-parameter instantaneous synchronous calibration can be achieved by expanding the plurality of sub-nodes and the plurality of cameras.
The technical scheme of the invention is as follows: a distributed POS multi-node multi-parameter instant synchronous calibration method comprises the following steps:
calibrating the relative pose relationship of a first camera and a second camera without a common view field by utilizing two calibration plates;
placing the other two targets at the main node and the sub-node respectively, wherein the other two targets are called as the main node target and the sub-node target respectively, and are placed in the visual field of each camera respectively, and adjusting the angle between the cameras to enable the cameras to shoot the targets;
the first camera and the second camera respectively shoot the main node target and the sub-node target, and the poses of the main node and the sub-node in a camera coordinate system are measured;
calculating the relative position and posture between the main sub-node and the sub-node;
and checking the precision of the distributed POS system.
Preferably, the calibrating the relative pose relationship of the first camera and the second camera without a common view field specifically comprises:
rigidly connecting a first camera and a second camera through a tripod;
the first camera and the second camera respectively shoot the corresponding calibration plates, each image is obtained, the position of the camera is changed for multiple times, and the position and posture relation between the first camera and the second camera is not changed into a constraint condition to calibrate the position and posture relation between the first camera and the second camera before and after the camera moves.
Preferably, before and after the camera moves, the pose relationship between the first camera and the second camera does not become a constraint condition, and the pose relationship between the first camera and the second camera is calibrated, specifically including the steps of:
for the first camera, the pose relation between the first camera and the first calibration board at the initial moment is A
0The first camera moves K times, and the pose between the first camera and the first calibration plate is marked as A after each movement
k(K1 … K), the pose of the first camera is T with respect to the pose of the first camera at the initial time
1 k(K1 … K). For the second camera, the pose between the second camera and the second calibration plate at the initial moment is B
0The second camera moves K times, and the pose between the second camera and the second calibration plate is moved each time and is marked as B
k(K1 … K), the pose of the second camera is set to the pose of the second camera with respect to the pose of the second camera at the initial time
Wherein A is
kInvolving rotation R
kAnd translation t
kThe relationship is such that,
before and after the camera moves, the pose relationship between the first camera and the second camera does not become a constraint condition, and the following formula is provided:
wherein, T3The pose relationship between the first camera and the second camera is represented. The equation is a classical hand-eye calibration model.
Preferably, before the first camera and the second camera respectively shoot the main node target and the sub-node target, the method further comprises the following steps:
respectively installing a main POS (point of sale) and a sub IMU (inertial measurement Unit) at two ends of a cantilever beam, respectively defining the main POS and the sub IMU as a main node and a sub node, respectively sticking two targets on the surfaces of one sides of the main POS and the sub IMU, and defining the target on the surface of the main POS as a main node target; the target of the sub-IMU surface is a sub-node target.
Compared with the prior art, the invention can achieve the following technical effects:
aiming at the problem of distributed POS accuracy calibration, the invention firstly calibrates the relative pose relationship of two cameras without a common view field, then measures the poses of a main node and a sub node in a first camera coordinate system and a second camera coordinate system respectively, and finally calculates the relative poses of the main node and the sub node to calibrate the distributed POS measurement accuracy. The method has the advantages that the vision measurement is non-contact, high in frequency, high in precision and capable of measuring in real time, two cameras which are fixedly connected are moved for many times through two calibration plates, the pose relation of the multiple cameras is calibrated based on a hand-eye calibration model, and the defects that the operation of overall calibration on the cameras by utilizing a theodolite or total station three-dimensional coordinate measuring system is complicated and the working efficiency is low are overcome. By expanding the number of the cameras, the pose accuracy of the distributed POS of more sub-nodes can be checked and corrected, and the method can be used for measuring deformation, vibration and the like of large structural members in actual engineering.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without creative efforts.
FIG. 1 is a schematic diagram of a hardware system according to an embodiment of the accuracy calibration method for distributed POS of the present invention;
FIG. 2 is a schematic view of a camera calibration according to an embodiment of the present invention;
fig. 3 is a flowchart illustrating an embodiment of a method for precision calibration of a distributed POS according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention discloses a distributed POS multi-node multi-parameter instant synchronous calibration method, which comprises the following steps: calibrating the relative pose relationship of a first camera and a second camera without a public view field; the first camera and the second camera respectively shoot the main node target and the sub-node target, and the poses of the main node target and the sub-node target in a camera coordinate system are measured; and unifying the pose data measured by the first camera and the second camera to a measurement reference coordinate system, and calculating the relative position and posture of the main node and the sub-node so as to finish the calibration work of the distributed POS system.
As shown in fig. 1 and 2, as an embodiment, the basic components of the system of the present invention include a first camera, a second camera, a first calibration plate, a second calibration plate, a link, a camera tripod, a main node target, a sub node target, a main POS, a sub IMU, and a distributed system installation structure. As shown in fig. 1, the main IMU is installed in the middle of the cantilever as a main node; the sub IMU is installed at a node to be detected on the structure and used as a sub node; the main node target and the sub node target are respectively stuck to the surfaces of the main IMU and the sub IMU; the two cameras respectively shoot the main target and the sub target. Fig. 2 is a schematic view of camera calibration without overlapping fields of view, in which a first calibration plate and a second calibration plate are fixedly connected by a connecting rod, the first and second calibration plates are made of ceramic materials and are engraved with checkerboard patterns or dot patterns, and the distances between the feature points are known.
As shown in fig. 3, an embodiment of the precision calibration method of distributed POS according to the present invention is listed below, and the specific implementation steps are as follows:
step S11: and respectively installing the main POS and the two sub IMUs on corresponding installation nodes of the flexible structure frame, wherein the visual targets (the main node target and the sub node target) are respectively and rigidly connected with one surface of each IMU. Each camera captures a target.
Step S12: and the first calibration plate and the second calibration plate are rigidly connected through a connecting rod, the first camera and the second camera respectively shoot the first calibration plate and the second calibration plate in advance to respectively obtain an image, and the position and posture relation between the first camera and the second camera is calibrated by using the condition that the position and posture relation between the first camera and the second camera does not become a constraint condition before and after the camera moves.
For the first camera, the pose relation between the first camera and the first calibration board at the initial moment is A
0The first camera moves K times, and the pose between the first camera and the first calibration plate is marked as A after each movement
k(K1 … K), the pose of the first camera is T with respect to the pose of the first camera at the initial time
1 k(K1 … K). For the second camera, the pose between the second camera and the second calibration plate at the initial moment is B
0The second camera moves K times, and the pose between the second camera and the second calibration plate is moved each time and is marked as B
k(K1 … K), the pose of the second camera is set to the pose of the second camera with respect to the pose of the second camera at the initial time
Wherein A is
kInvolving rotation R
kAnd translation t
kThe relationship is such that,
before and after the camera moves, the pose relationship between the first camera and the second camera does not become a constraint condition, and the following formula is provided:
wherein, T3The pose relationship between the first camera and the second camera is represented. The above formula is developed:
the above equation set can solve R by utilizing the theory of lie group lie algebra
3And t
3,
Wherein R is
3Representing a rotation matrix between the first camera and the second camera, t
3Representing a translation vector between the first camera and the second camera.
Step S13: and the first camera and the second camera which are calibrated to have a relative relationship shoot the main node target and the sub-node target respectively. The pose measurement of a single target is completed through monocular vision, then a target coordinate system is converted into an IMU coordinate system through a machining structure, the pose relation between a camera and the IMU is calculated, all the pose of the sub-nodes is converted into a reference coordinate system, the reference coordinate system is the coordinate system where the main POS is located, after the coordinate systems are unified, the relative pose between the main and sub-nodes is obtained, the relative pose measurement between the main and sub-nodes is realized, and the calibration of the distributed POS system is completed.
In summary, the invention discloses a distributed POS multi-node multi-parameter instantaneous synchronous calibration method, which comprises the steps of firstly calibrating the relation between two cameras without a common view field, completing the pose measurement of a single target through monocular vision, then converting a target coordinate system into an IMU coordinate system through the known dimension information of a first calibration plate, a second calibration plate and an IMU, finally obtaining the relative pose between main and sub nodes, and finally calibrating the precision of the distributed POS. The pose measurement is realized by using a visual measurement means, and the pose of the two cameras is calibrated by using the constraint condition that the pose relation between the cameras before and after the movement of the cameras is unchanged, so that the defects of complicated operation and large workload of global calibration of the cameras by using other three-dimensional coordinate measurement systems such as a theodolite or a total station and the like in the traditional method are overcome; the calibration method has the characteristics of high precision and strong anti-interference capability, and can complete multi-node multi-parameter instantaneous synchronous calibration on the distributed POS on the ground by expanding the number of the sub-nodes and the cameras.