CN114022551A - Method for accurately identifying and estimating pose of fuel filling cover of fuel vehicle - Google Patents
Method for accurately identifying and estimating pose of fuel filling cover of fuel vehicle Download PDFInfo
- Publication number
- CN114022551A CN114022551A CN202111259240.XA CN202111259240A CN114022551A CN 114022551 A CN114022551 A CN 114022551A CN 202111259240 A CN202111259240 A CN 202111259240A CN 114022551 A CN114022551 A CN 114022551A
- Authority
- CN
- China
- Prior art keywords
- coordinate system
- grabbing
- oil filling
- plug
- cover
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
Abstract
The invention discloses a method for accurately identifying and estimating the pose of an oil filling cover of a fuel vehicle, and belongs to the technical field of automatic oil filling of fuel vehicles and the industrial field of related grabbing. The implementation method of the invention comprises the following steps: and for the outer cover of the oil filling cover, a segmentation algorithm is adopted, self-adaptive factors are introduced, and the point cloud of the outer cover of the oil filling cover is segmented. And for the oil filler plug, acquiring oil filler plug point cloud based on prior knowledge such as characteristics of an oil filler plug handle and a surface. Based on the pose estimation of the point cloud data, adding a direction selection factor, establishing a proper grabbing coordinate system, and establishing the relation between the grabbing coordinate system and the base coordinate system of the mechanical arm through off-line calibration of pose transformation of the camera and the mechanical arm, so that the end tool of the mechanical arm can accurately reach a grabbing point and maintain an accurate grabbing direction. The invention can avoid the requirements of model base construction and high computational power, ensure the requirements of explosion prevention and high precision of the automatic oiling robot system and solve the problem of inaccurate camera positioning and orientation. The invention is particularly suitable for the automatic oiling similar capturing scene of the fuel vehicle.
Description
Technical Field
The invention belongs to the technical field of automatic fuel filling of fuel vehicles and the industrial field of related grabbing, and particularly relates to a method for accurately identifying a fuel filling cover of a fuel vehicle and estimating a pose of the fuel filling cover.
Background
The robot technology and the computer vision technology are continuously updated in an iterative mode, and the robot and the computer vision technology are widely applied to the fields of consumption, medical treatment, construction, industry and the like. At present, a gas station basically depends on manual refueling of workers, so that a large amount of labor and material cost is consumed, and certain potential safety hazards exist due to uncertainty of operation of the workers, so that the automatic refueling technology for the fuel vehicle has wide application prospects.
The intelligent environment perception field of the automatic fuel filling system of the present fuel vehicle has the following scheme: (1) the method is easy to be influenced by ambient light and difficult to meet the explosion-proof requirement of a gas station; (2) by using a visual servo scheme, the mechanical arm is guided to move through the relative position of the oil filling cover in the visual field of the camera, so that the real-time performance is low and the precision is low; (3) the vehicle information is input in a man-machine interaction mode to match with the vehicle data in the database, and the full-automatic aim is difficult to achieve.
Good real-time performance and accuracy are needed for recognizing the outer cover of the oil filling cover and the oil filling plug and estimating the pose of the oil filling cover and the oil filling plug. In addition, the automatic oiling robot system has strict requirements on explosion-proof parameters of a camera, and therefore, how to realize a method for identifying an oiling cover and estimating a pose with high precision and high real-time performance so as to meet the needs of an automatic oiling robot scene is a problem to be solved by technical personnel in the field.
Disclosure of Invention
The invention aims to provide a method for accurately identifying and estimating the pose of an oil filling cover of a fuel vehicle. The invention has the advantages of high precision, high processing speed and the like.
In order to achieve the above purpose, the present invention provides the following specific schemes:
the invention discloses a method for accurately identifying and estimating the pose of a fuel truck refueling cover, which comprises the identification and pose estimation of an outer cover and a refueling plug of the fuel truck refueling cover. And for the outer cover of the oil filling cover, a segmentation algorithm is adopted, self-adaptive factors are introduced, and the point cloud of the outer cover of the oil filling cover is segmented. And for the oil filler plug, acquiring oil filler plug point cloud based on prior knowledge such as characteristics of an oil filler plug handle and a surface. Based on the pose estimation of the point cloud data, adding a direction selection factor, establishing a proper grabbing coordinate system, and establishing the relation between the grabbing coordinate system and a base coordinate system of the mechanical arm through pose transformation of an offline calibration camera and the mechanical arm, so that a tool at the tail end of the mechanical arm can accurately reach the grabbing point of the oil filling cover outer cover and the oil filling plug, and the accurate grabbing direction can be kept. The method is particularly suitable for the field of automatic fuel filling of fuel vehicles and similar capture scenes.
The invention discloses a method for accurately identifying an oil filling cover of a fuel vehicle and estimating a pose, which comprises the following steps of:
step 1: and calibrating the pose of the point on the standard calibration plate in two different coordinate systems, and calibrating the transformation relation Tcb of the 3D camera and the mechanical arm base coordinate system in an off-line manner.
Step 11: fixing a chessboard pattern calibration plate, acquiring the positions of n fixed points on the calibration plate under a camera coordinate system, and recording a point set as P; and acquiring the positions of the n corresponding fixed points on the calibration plate under the mechanical arm base coordinate system, and recording a point set as Q.
Step 12: for the point sets P and Q, decentralization is performed, and then a covariance matrix H is constructed:
wherein n is the number of point concentration points, (X)pi,Ypi,Zpi) As the coordinates of the ith point Pi in the point set P,(Xqi,Yqi,Zqi) Is the coordinates of the ith point Qi in the point set Q.
Step 13: carrying out singular value decomposition on the covariance matrix H to obtain matrixes v and u, and then constructing a rotation matrix R:
R＝vuT
further, a translation matrix t is calculated:
t＝-RPi+Qi
and constructing a homogeneous transformation matrix Tcb of the camera coordinate system and the mechanical arm base coordinate system by using the rotation matrix and the translation matrix to obtain a transformation relation Tcb of the off-line calibration 3D camera and the mechanical arm base coordinate system.
Step 2: the 3D camera with the LED light as the active light source is selected, and the anti-explosion requirement of the oiling robot is met. The close-range shooting pose of the 3D camera is determined according to other visual sensors in the automatic oiling robot system, so that the interference of ambient light is effectively avoided, and point cloud data is collected.
And step 3: introducing an adaptation factor
And screening the segmented point cloud according to the point number of the point cloud data, acquiring the point cloud data of the outer cover of the oil filling cover, and calculating the mass center of the outer cover based on the point cloud data.
Step 31: filtering and downsampling the collected scene point cloud, traversing the points in the scene point cloud, clustering by using Euclidean distance between the points, and introducing self-adaptive factors
And screening the segmented point cloud based on the point number of the point cloud data to obtain the point cloud of the outer cover of the oil filling cover.
Step 32: for the filler cap cover point cloud, the centroid is calculated and used as a reference point Opre for selecting a Region of Interest (ROI) on the filler cap cover.
And 4, step 4: and selecting an ROI on the outer cover of the oil filling cover, and accurately calculating the mass center and normal vector of the ROI through a block averaging algorithm to be respectively used as an origin Oo and a z axis of a grabbing coordinate system of the outer cover of the oil filling cover. Calculating the main direction of the ROI, defining the main direction as a y-axis of a grabbing coordinate system, introducing a direction selection factor _ outer, determining the unique positive direction of the y-axis, constructing the grabbing coordinate system of the outer cover of the oil filling cover, obtaining the pose of the constructed grabbing coordinate system under the mechanical arm base coordinate system by utilizing the relation between a camera coordinate system and the mechanical arm base coordinate system, and realizing the acquisition of the accurate grabbing pose of the outer cover of the oil filling cover.
Step 41: based on the calculated reference point Opre, margins Δ x, Δ y, Δ z are set in the x, y, z directions, respectively, and a target region ROI on the filler cap outer cover is selected.
Step 42: dividing the ROI acquired in the step 41 into a plurality of small blocks, respectively calculating the mass center and normal vector of each point cloud, and taking the mean value of different block mass center coordinates as the origin Oo of the grabbing coordinate system. And for the normal vector, ensuring that the included angle between the normal vector and the viewpoint direction is larger than 90 degrees, taking the average value of the normal vectors of a plurality of small blocks as the normal vector of the ROI area, and defining the normal vector as the positive direction of the z axis of the grabbing coordinate system of the outer cover of the oil filling cover
Step 43: calculating the characteristic vector corresponding to the maximum characteristic value of the ROI acquired in the step 41 by using a principal component analysis algorithm, and setting the characteristic vector as the positive direction of the y axis of the grabbing coordinate system of the outer cover of the oil filling cover
Introducing an outer cover direction selection factor _ outer of the oil filling cover, selecting an included angle between the positive direction of the y axis and the direction vector (0,1,0) of the y axis of the camera coordinate system to be an acute angle or an obtuse angle, uniformly grabbing the positive direction of the y axis of the coordinate system by the outer cover of the oil filling cover, and avoiding the situation that a mechanical arm cannot reach.
Step 44: computing vectors using vector cross-multiplication
Defining the positive direction of the x axis of the grabbing coordinate system of the outer cover of the oil filling cover as
Namely obtaining the required grasping coordinate system for constructing the oil filling cover outer coverAn origin and three directional vectors.
Step 45: constructing a homogeneous matrix Tgo using the origin and direction vectors of the grasping coordinate system:
Tgo＝
wherein the content of the first and second substances,
and capturing a column vector formed by the coordinates of the origin of the coordinate system for the defined oil filling cover outer cover.
Using Tgoc Tgo ═ Tc, where Tc is the camera coordinate system and Tgoc is the relationship of the constructed oil filler cover gripping coordinate system to the camera coordinate system.
And calculating to obtain a transformation relation Tgoc between the grabbing coordinate system and the camera coordinate system, wherein the mechanical arm base coordinate system and the camera coordinate system have the following relation:
Tgob＝Tcb*Tgoc
wherein, Tgob is the transformation relation between the grabbing coordinate system and the mechanical arm base coordinate system.
And further, converting the transformation matrix Tgob into Euler angle representation (Xow, Yow, Zow, Aow, Bow and Cow) by using the relation between the rotation matrix and the Euler angle, namely the pose of the outer cover gripping point of the oil filling cover under the mechanical arm base coordinate system.
And (4) transmitting the grabbing pose of the outer cover of the oil filling cover obtained in the step (4) to the mechanical arm through TCP network communication, opening the outer cover by using a sucker, selecting the mechanical arm to move under a sucker tool coordinate system, and accurately arriving and performing the action of opening the outer cover.
And 5: and identifying the oil filler plug by using the prior geometric characteristics of the oil filler plug, and acquiring point cloud data of the oil filler plug. Based on the relative position of the filler plug handle and the filler plug plane, the segmentation of the filler plug handle is realized, and a reference point for selecting the interested region of the filler plug handle is calculated.
Step 51: and (3) removing noise points of scene point cloud data containing the complete oil filling plug through preprocessing such as filtering and down-sampling. Further, setting a radius threshold of a circle to be fitted, fitting a space circle, and determining the circle center space position (Xc, Yc, ZC) and the normal vector direction (Ac, Bc, Cc) of the space circle.
Step 52: the fitted space circle normal vector is aligned with the vector (0,0,1), the rotation represented by the axis angle is converted into a matrix R0 using the rodregs equation, and the filler point cloud is rotated.
Step 53: the relative position relationship between the filler plug handle and the filler plug plane is utilized, and the grabbing point (Xi, Yi, Zi) of the rotated filler plug handle is calculated through the height difference delta and is used as a reference point Oprei of the interested area of the selected filler plug handle.
Step 6: and selecting an ROI on a handle of the refueling plug, and accurately calculating the mass center and normal vector of the ROI through a block averaging algorithm to be respectively used as an origin Oo and a z axis of a grabbing coordinate system of the refueling plug. And calculating the main direction of the ROI, taking the main direction as a y axis of a grabbing coordinate system, introducing a direction selection factor _ inner, unifying the main direction of the y axis, and constructing the grabbing coordinate system of the oil filling plug. And obtaining the pose of the oil filler plug grabbing coordinate system under the mechanical arm base coordinate system by utilizing the relation between the camera coordinate system and the mechanical arm base coordinate system, so as to realize the acquisition of the accurate grabbing pose of the oil filler plug.
Step 61: based on the calculated reference point Oprei, margins Δ xi, Δ yi, Δ zi are set in the x, y, z directions, respectively, and a target region ROI on the filler plug handle is selected.
Step 62: dividing the ROI acquired in the step 61 into a plurality of small blocks, respectively calculating the mass center and normal vector of each point cloud, and taking the mean value of different block mass center coordinates as the origin Oi of the oil plug grabbing coordinate system. And for the normal vector, ensuring that the included angle between the normal vector and the viewpoint direction is larger than 90 degrees, taking the average value of the normal vectors of a plurality of small blocks as the normal vector of the ROI area, and defining the normal vector as the positive direction of the z axis of the grabbing coordinate system of the outer cover of the oil filling cover
And step 63: for ROI on the handle of the refueling plug, a principal component analysis algorithm is used to obtain the maximum characteristic value pairThe corresponding characteristic vector is defined as the positive direction of the y axis of the oil filler plug grabbing coordinate system
And introducing a direction selection factor of the oil filler plug into the clamping jaw, determining that an included angle between the direction vector (0,1,0) and the y axis of the camera coordinate system is an acute angle or an obtuse angle, uniformly grabbing the positive direction of the y axis of the coordinate system, and ensuring that the mechanical arm can reach when the clamping jaw is clamped.
Step 64: defining the positive direction of the x axis of the grabbing coordinate system of the oil filling plug as
And constructing to obtain a fuel filler plug grabbing coordinate system.
Step 65: constructing a homogeneous matrix Tgi by using the direction vector of the constructed oil filling plug grabbing coordinate system:
wherein the content of the first and second substances,
and capturing a column vector formed by the coordinates of the origin of the coordinate system for the defined oil filling cover outer cover.
And utilizing Tgic Tgi ═ Tc, wherein Tc is a camera coordinate system, and Tgic is the relation between the constructed fuel filler plug grabbing coordinate system and the camera coordinate system.
And calculating to obtain a transformation relation Tgic of the grabbing coordinate system and the camera coordinate system, wherein the mechanical arm base coordinate system and the camera coordinate system have the following relation:
Tgib＝Tcb*Tgic
wherein, Tgib is the transformation relation between the grabbing coordinate system and the mechanical arm base coordinate system.
And further, converting the transformation matrix Tgib into Euler angle representation (Xiw, Yiw, Ziw, Aiw, Biw and Ciw) by using the relation between the rotation matrix and the Euler angle, namely the position of the handle grabbing point of the oil filler plug in the mechanical arm base coordinate system.
And (4) accurately grabbing the pose of the oil filling plug obtained in the step (6), transmitting the pose to a mechanical arm through TCP network communication, tightening and unscrewing the oil filling plug by using a clamping jaw, and selecting the mechanical arm to move under a clamping jaw tool coordinate system to accurately reach and screw the oil filling plug.
Has the advantages that:
1. the invention discloses a method for accurately identifying and estimating the pose of an oil filling cover of a fuel vehicle. And (4) constructing a grabbing coordinate system, and effectively acquiring an accurate grabbing pose by utilizing transformation between the coordinate systems. The method can avoid the requirements of model base construction and high computational power, guarantee the explosion prevention and high precision requirements of the automatic oiling robot system, and solve the problem of inaccurate positioning and orientation of a monocular or binocular camera.
2. The invention discloses a method for accurately identifying an oil filling cover of a fuel oil vehicle and estimating a pose, wherein a structured light camera with an LED light source is selected to meet the explosion-proof requirement of a gas station; other sensors in the automatic oiling robot system are used for guiding the 3D camera to acquire data in a short distance, the influence of ambient light on point cloud data is reduced, and good three-dimensional data is acquired.
3. The invention discloses a method for accurately identifying an oil filling cover of a fuel vehicle and estimating the pose of the oil filling cover. The method comprises the steps of calculating a normal vector and a main direction of point cloud, constructing a grabbing coordinate system, introducing a direction selection factor, uniformly grabbing the positive direction of a y axis of the coordinate system, calculating the accurate pose of a grabbing point of an outer cover of an oil filling cover by utilizing the transformation of the coordinate system between a mechanical arm and a camera, effectively avoiding the phenomenon that the mechanical arm exceeds a soft limit when moving due to non-uniform directions, avoiding modeling and being quicker to process.
4. The invention discloses a method for accurately identifying and estimating the pose of an oil filling cover of a fuel vehicle, which realizes the identification and extraction of a handle point cloud of an oil filling plug by utilizing the prior geometric characteristics of the oil filling plug; when a handle grabbing coordinate system is constructed, a direction selection factor is introduced, the positive direction of the y axis of the grabbing coordinate system is ensured to be uniform, and the effectiveness of pose calculation is improved. When the normal vector is calculated, the average value is calculated in small blocks, and the precision of the z axis can be improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
FIG. 1 is an overall flow chart of a method for accurately identifying and estimating the pose of a fuel truck fuel filler cap according to the invention;
FIG. 2 is a view of the fuel filler cap outer cover to be identified;
FIG. 3 is a diagram of a filler plug to be identified;
FIG. 4 is an original scene point cloud after color rendering, wherein a view (a) is a shot scene point cloud of an outer cover of a refueling cap, and a view (b) is a shot scene point cloud of a refueling plug;
FIG. 5 is a schematic view of a constructed fuel filler cap outer cover gripping coordinate system, wherein FIG. (a) is a schematic view of a defined fuel filler cap outer cover gripping coordinate system and FIG. (b) is a schematic view of a camera coordinate system;
FIG. 6 is a schematic view of a constructed fuel filler plug grip coordinate system, wherein (a) is a schematic view of a defined fuel filler plug handle grip coordinate system and (b) is a schematic view of a camera coordinate system;
FIG. 7 is a visualization of a constructed fuel filler cap outer cover gripping coordinate system, wherein graphs (a) and (b) are fuel filler cap outer cover gripping coordinate systems at different viewing angles, wherein the blue arrow is the z-axis, the green arrow is the y-axis, and the red arrow is the x-axis;
fig. 8 is a visualization result of a constructed fuel plug gripping coordinate system, wherein (a) and (b) are fuel plug gripping coordinate systems under different viewing angles, wherein a blue arrow is a z-axis, a green arrow is a y-axis, and a red arrow is an x-axis.
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.
As shown in fig. 1, the overall flowchart of the method for accurately identifying and pose estimating the fuel truck oil filler cap disclosed in this embodiment includes oil filler cap outer cap identification, oil filler cap outer cap positioning orientation, oil filler plug identification, and oil filler plug positioning orientation. As shown in fig. 2 and 3, the fuel truck fuel filler cap outer cover and the fuel filler plug detected in the embodiment are respectively. As shown in fig. 4, the scene point clouds of the oil cap outer cover and the oil plug respectively need to identify the oil cap outer cover and the oil plug and accurately position and orient the oil cap outer cover and the oil plug, and the method comprises the following specific steps:
step 1: and the transformation relation Tcb between the 3D camera and the mechanical arm base coordinate system is calibrated in an off-line manner through the poses of the points on the standard calibration plate in two different coordinate systems.
Step 11: fixing a chessboard pattern calibration plate, acquiring the positions of n fixed points on the calibration plate under a camera coordinate system, and recording a point set as P; and acquiring the positions of the n corresponding fixed points on the calibration plate under the mechanical arm base coordinate system, and recording a point set as Q.
Step 12: for the point sets P and Q, decentralization is performed, and then a covariance matrix H is constructed:
wherein n is the number of point concentration points, (X)pi,Ypi,Zpi) Is the coordinate of the ith point Pi in the point set P, (X)qi,Yqi,Zqi) For the ith point Qi in the set QAnd (4) coordinates.
Step 13: performing singular value decomposition on the matrix H to obtain matrices v and u, and then constructing a rotation matrix R:
R＝vuT
further, a translation matrix t is calculated:
t＝-RPi+Qi
a homogeneous transformation matrix Tcb of a camera coordinate system and a mechanical arm base coordinate system is constructed by a rotation matrix and a translation matrix, and the homogeneous transformation matrix Tcb is as follows:
step 2: the 3D camera with the LED light as the active light source is selected, and the anti-explosion requirement of the oiling robot is met. The close-range shooting pose of the 3D camera is determined according to other visual sensors in the automatic oiling robot system, so that the interference of ambient light is effectively avoided, and point cloud data is collected.
And step 3: introducing an adaptation factor
The method comprises the following steps of screening and dividing point cloud according to the number of the point cloud data, obtaining the point cloud data of the outer cover of the oil filling cover, and calculating the mass center of the outer cover based on the point cloud data, and comprises the following steps:
step 31: filtering and downsampling the collected scene point cloud, traversing the points in the scene point cloud, clustering by using Euclidean distance between the points, and introducing self-adaptive factors
And screening the segmented point cloud based on the point number of the point cloud data to obtain the point cloud of the outer cover of the oil filling cover.
Step 32: for the filler cap cover point cloud, the centroid is calculated as the reference point Opre for selecting the Region of Interest (ROI) on the filler cap cover as follows:
Opre＝(-8.592 4.453 17.791)。
and 4, step 4: and selecting an ROI on the outer cover of the oil filling cover, and accurately calculating the mass center and normal vector of the ROI through a block averaging algorithm to be respectively used as an origin Oo and a z axis of a grabbing coordinate system of the outer cover of the oil filling cover. Calculating the main direction of the ROI, defining the main direction as a y axis of a grabbing coordinate system, introducing a direction selection factor _ outer, determining the unique positive direction of the y axis, constructing the grabbing coordinate system of the outer cover of the oil filling cover, obtaining the pose of the constructed grabbing coordinate system under the mechanical arm base coordinate system by utilizing the relation between a camera coordinate system and the mechanical arm base coordinate system, and realizing the acquisition of the accurate grabbing pose of the outer cover of the oil filling cover, wherein the method comprises the following steps:
step 41: based on the calculated reference point Opre, margins Δ x, Δ y, Δ z are set in the x, y, z directions, respectively, and a target region ROI on the filler cap outer cover is selected.
Step 42: dividing the ROI acquired in the step 41 into a plurality of small blocks, respectively calculating the mass center and normal vector of each point cloud, and taking the mean value of mass center coordinates of different blocks as the origin Oo (-8.4444.34417.879) of the grabbing coordinate system. And for the normal vector, ensuring that the included angle between the normal vector and the viewpoint direction is larger than 90 degrees, taking the average value of the normal vectors of a plurality of small blocks as the normal vector of the ROI area, and defining the normal vector as the positive direction of the z axis of the grabbing coordinate system of the outer cover of the oil filling cover
Step 43: calculating the characteristic vector corresponding to the maximum characteristic value of the ROI acquired in the step 41 by using a principal component analysis algorithm, and setting the characteristic vector as the positive direction of the y axis of the grabbing coordinate system of the outer cover of the oil filling cover
Introducing an outer cover direction selection factor _ outer of the oil filling cover, selecting an included angle between the positive direction of the y axis and the direction vector (0,1,0) of the y axis of the camera coordinate system to be an acute angle or an obtuse angle, uniformly grabbing the positive direction of the y axis of the coordinate system by the outer cover of the oil filling cover, and avoiding the situation that a mechanical arm cannot reach. Referring to FIG. 5, a schematic diagram of a capturing coordinate system of the fuel filler cover is shown in FIG. (a) and a schematic diagram of a camera coordinate system is shown in FIG. (b), in this embodiment, the positive direction of the y-axis is
The included angle of the direction vector (0,1,0) of the y-axis of the camera coordinate system is an obtuse angle.
Step 44: computing vectors using vector cross-multiplication
Defining the positive direction of the x axis of the grabbing coordinate system of the outer cover of the oil filling cover as
A constructed fuel filler cap outer cover grabbing coordinate system is obtained, the visually constructed coordinate system is shown as reference figure 7, and the figures (a) and (b) are fuel filler cap outer cover grabbing coordinate systems under different viewing angles, wherein a blue arrow is a z-axis, a green arrow is a y-axis, and a red arrow is an x-axis.
Step 45: constructing a homogeneous matrix Tgo using the direction vectors of the grasping coordinate system:
wherein the content of the first and second substances,
and capturing a column vector formed by the coordinates of the origin of the coordinate system for the defined oil filling cover outer cover.
The relationship between the coordinate system and the camera coordinate system is captured by using Tgob Tgo ═ Tc, wherein Tc is the camera coordinate system, and Tgoc is the constructed oil filler cover outer cover.
And calculating to obtain a transformation relation between the grabbing coordinate system and the camera coordinate system:
the mechanical arm base coordinate system and the camera coordinate system have the following relationship:
Tgob＝Tcb*Tgoc
wherein, Tgob is the transformation relation between the grabbing coordinate system and the mechanical arm base coordinate system.
Is calculated to obtain
Further, the transformation matrix Tgob is converted into euler angles by using the relation between the rotation matrix and the euler angles. In this embodiment, the euler angle sequence of the mechanical arm is ZYZ, and the rotation matrix is R:
the euler angles for converting R to the sequence ZYZ are:
the transformation matrix Tgob is used to convert into euler angles: (Xow Yow zhow Aow Bow Cow) ═ 1422.03-266.8741500.4-2.78780.905529.8046, the position of the lid gripping point of the filler cap in the arm base coordinate system.
And (4) transmitting the grabbing pose of the outer cover of the oil filling cover obtained in the step (4) to the mechanical arm through TCP network communication, opening the outer cover by using a sucker, selecting the mechanical arm to move under a sucker tool coordinate system, and accurately arriving and performing the action of opening the outer cover.
And 5: and identifying the oil filler plug by using the prior geometric characteristics of the oil filler plug, and acquiring point cloud data of the oil filler plug. Based on the relative position of the oil filler plug handle and the oil filler plug plane, the segmentation of the oil filler plug handle is realized, and a reference point for selecting an interested area of the oil filler plug handle is calculated, and the method comprises the following steps:
step 51: and (3) removing noise points of scene point cloud data containing the complete oil filling plug through preprocessing such as filtering and down-sampling. Further, setting a radius threshold of a circle to be fitted, fitting the space circle, and determining the center space position (Xc Yc Zc) ═ (14.13659.001-25.299) and the normal vector direction (Ac Bc Cc) — 0.0120.161-0.987 of the space circle.
Step 52: the fitted space circle normal vector is aligned with the vector (0,0,1), the rotation represented by the axis angle is converted into a matrix R0 using the rodregs equation, and the filler point cloud is rotated.
Step 53: using the relative positional relationship of the filler plug handle to the filler plug plane, a point on the rotated filler plug handle is calculated by the height difference Δ as a reference point Oprei for selecting a filler plug handle region of interest (14.3546.142-7.731).
Step 6: and selecting an ROI on a handle of the refueling plug, and accurately calculating the mass center and normal vector of the ROI through a block averaging algorithm to be respectively used as an origin Oo and a z axis of a grabbing coordinate system of the refueling plug. And calculating the main direction of the ROI, taking the main direction as a y axis of a grabbing coordinate system, introducing a direction selection factor _ inner, unifying the main direction of the y axis, and constructing the grabbing coordinate system of the oil filling plug. The method comprises the following steps of obtaining the pose of an oil filler plug grabbing coordinate system under a mechanical arm base coordinate system by utilizing the relation between a camera coordinate system and the mechanical arm base coordinate system, and realizing the acquisition of the precise grabbing pose of the oil filler plug, wherein the method comprises the following steps:
step 61: based on the calculated reference point Oprei, margins Δ xi, Δ yi, Δ zi are set in the x, y, z directions, respectively, and a target region ROI on the filler plug handle is selected.
Step 62: and dividing the ROI acquired in the step 61 into a plurality of small blocks, respectively calculating the mass center and normal vector of each point cloud, and taking the mean value of mass center coordinates of different blocks as the origin Oi of the oil plug grabbing coordinate system (14.5896.142-7.613). And for the normal vector, ensuring that the included angle between the normal vector and the viewpoint direction is larger than 90 degrees, taking the average value of the normal vectors of a plurality of small blocks as the normal vector of the ROI area, and defining the normal vector as the positive direction of the z axis of the grabbing coordinate system of the outer cover of the oil filling cover
And step 63: for ROI on the handle of the oil filler plug, a principal component analysis algorithm is used to obtain a feature vector corresponding to the maximum feature value, and the feature vector is defined as the positive direction of the y axis of a grabbing coordinate system of the oil filler plug
The direction selection factor _ inner of the oil filling plug is introduced to determine that the included angle between the positive direction of the y axis and the direction vector (0,1,0) of the y axis of the camera coordinate system is an acute angle or an obtuse angle, so that the positive direction of the y axis of the coordinate system is grabbed uniformly, and the mechanical arm can reach when the clamping jaw is clamped. Referring to FIG. 6, a schematic diagram of a fuel filler plug handle grabbing coordinate system is shown in (a) and a schematic diagram of a camera coordinate system is shown in (b), wherein the positive direction of the y-axis of the fuel filler plug grabbing coordinate system is shown in the embodiment
And the included angle of the direction vector (0,1,0) and the y axis of the camera coordinate system is an acute angle.
Step 64: defining the positive direction of the x axis of the grabbing coordinate system of the oil filling plug as
And constructing to obtain a fuel filler plug grabbing coordinate system. The visually constructed fuel plug-grasping coordinate system is shown in fig. 8, and fig. (a) and (b) are fuel plug-grasping coordinate systems at different viewing angles, wherein a blue arrow is a z-axis, a green arrow is a y-axis, and a red arrow is an x-axis.
Step 65: constructing a homogeneous matrix Tgi by using the direction vector of the constructed oil filling plug grabbing coordinate system:
wherein the content of the first and second substances,
and capturing a column vector formed by the coordinates of the origin of the coordinate system for the defined oil filling cover outer cover.
And utilizing Tgic Tgi ═ Tc, wherein Tc is a camera coordinate system, and Tgic is the relation between the constructed fuel filler plug grabbing coordinate system and the camera coordinate system.
Calculating to obtain the transformation relation between the capturing coordinate system and the camera coordinate system
The mechanical arm base coordinate system and the camera coordinate system have the following relationship:
Tgib＝Tcb*Tgic
wherein, Tgib is the transformation relation between the grabbing coordinate system and the mechanical arm base coordinate system.
Is calculated to obtain
Further, the transformation matrix Tgib is converted into an Euler angle representation (Xiw Yiiw Ziw Aiw Biw Ciw) ═ 1456.26-269.8461502.14-18.12649.980-152.987) by using the relation between the rotation matrix and the Euler angle, namely the position of the handle grabbing point of the fuel filler plug in the mechanical arm base coordinate system.
And 7: the obtained oil filling plug is accurately grabbed and accurately positioned, the position and the posture are transmitted to the mechanical arm through TCP network communication, the oil filling plug is tightened and unscrewed by using the clamping jaw, the mechanical arm is selected to move under a clamping jaw tool coordinate system, and the action of screwing the oil filling plug is accurately achieved.
In conclusion, the method for accurately identifying and estimating the pose of the fuel cap of the fuel vehicle can finish the intelligent perception of an automatic refueling robot system, realize the identification and pose estimation of the outer cap of the fuel cap and the fuel plug, and meet the high-precision requirement of the system on perception. The method comprises the steps of obtaining point cloud of an outer cover of an oil filling cover based on clustering segmentation of self-adaptive factors, obtaining the point cloud of a handle of the oil filling cover based on priori knowledge and characteristics of the oil filling cover, rapidly and accurately completing target identification, and avoiding model base construction and high computational power requirements. By introducing a direction selection factor, a unified grabbing coordinate system is constructed, the positions and postures of the outer cover of the oil filling cover and the handle of the oil filling plug are accurately obtained by utilizing matrix transformation, and accurate positioning and orientation of a target are realized.
The above detailed description is intended to illustrate the objects, aspects and advantages of the present invention, and it should be understood that the above detailed description is only exemplary of the present invention and is not intended to limit the scope of the present invention, and any modifications, equivalents, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (7)
1. A method for accurately identifying and estimating the pose of an oil filling cover of a fuel vehicle is characterized by comprising the following steps: comprises the following steps of (a) carrying out,
step 1: calibrating the pose of a point on a standard calibration plate under two different coordinate systems, and calibrating the transformation relation Tcb of the 3D camera and the mechanical arm base coordinate system in an off-line manner;
step 2: the 3D camera with the LED light as the active light source is selected to meet the explosion-proof requirement of the oiling robot; the close-range shooting pose of the 3D camera is determined according to other visual sensors in the automatic oiling robot system, so that the interference of ambient light is effectively avoided, and point cloud data is collected;
and step 3: introducing an adaptation factor
Screening the segmented point cloud according to the number of the point cloud data, acquiring the point cloud data of the outer cover of the oil filling cover, and calculating the mass center of the outer cover based on the point cloud data;
and 4, step 4: selecting an ROI on the outer cover of the oil filling cover, and accurately calculating the mass center and normal vector of the ROI through a block averaging algorithm to be respectively used as an origin Oo and a z axis of a grabbing coordinate system of the outer cover of the oil filling cover; calculating the main direction of the ROI, defining the main direction as a y-axis of a grabbing coordinate system, introducing a direction selection factor _ outer, determining the unique positive direction of the y-axis, constructing the grabbing coordinate system of the outer cover of the oil filling cover, and obtaining the pose of the constructed grabbing coordinate system under the mechanical arm base coordinate system by utilizing the relation between a camera coordinate system and the mechanical arm base coordinate system to realize the acquisition of the accurate grabbing pose of the outer cover of the oil filling cover;
and 5: identifying the oil filler plug by using the prior geometric characteristics of the oil filler plug, and acquiring point cloud data of the oil filler plug; based on the relative position of the oil filler plug handle and the oil filler plug plane, the segmentation of the oil filler plug handle is realized, and a reference point for selecting an interested region of the oil filler plug handle is calculated;
step 6: selecting an ROI on a handle of the refueling plug, and accurately calculating the mass center and normal vector of the ROI through a block averaging algorithm to be respectively used as an origin Oo and a z axis of a grabbing coordinate system of the refueling plug; calculating the main direction of the ROI, taking the main direction as a y axis of a grabbing coordinate system, introducing a direction selection factor _ inner, unifying the main direction of the y axis, and constructing a grabbing coordinate system of the oil filling plug; and obtaining the pose of the oil filler plug grabbing coordinate system under the mechanical arm base coordinate system by utilizing the relation between the camera coordinate system and the mechanical arm base coordinate system, so as to realize the acquisition of the accurate grabbing pose of the oil filler plug.
2. The method for accurately identifying and estimating the position and the attitude of the fuel truck fuel filler cap as claimed in claim 1, wherein: also comprises a step 7 of carrying out the following steps,
accurately grabbing the pose of the outer cover of the oil filling cover obtained in the step 4, transmitting the pose to the mechanical arm through TCP network communication, opening the outer cover by using a sucking disc, selecting the mechanical arm to move under a sucking disc tool coordinate system, and accurately achieving and performing the action of opening the outer cover;
and (4) accurately grabbing the pose of the oil filling plug obtained in the step (6), transmitting the pose to a mechanical arm through TCP network communication, tightening and unscrewing the oil filling plug by using a clamping jaw, and selecting the mechanical arm to move under a clamping jaw tool coordinate system to accurately reach and screw the oil filling plug.
3. The method for accurately identifying and estimating the position and the attitude of the fuel truck refueling cover as claimed in claim 1 or 2, wherein the method comprises the following steps: the step 1 is realized by the method that,
step 11: fixing a chessboard pattern calibration plate, acquiring the positions of n fixed points on the calibration plate under a camera coordinate system, and recording a point set as P; acquiring the positions of n corresponding fixed points on the calibration plate under a mechanical arm base coordinate system, and recording a point set as Q;
step 12: for the point sets P and Q, decentralization is performed, and then a covariance matrix H is constructed:
wherein n is the number of point concentration points, (X)pi,Ypi,Zpi) Is the coordinate of the ith point Pi in the point set P, (X)qi,Yqi,Zqi) Is a pointCoordinates of the ith point Qi in the set Q;
step 13: carrying out singular value decomposition on the covariance matrix H to obtain matrixes v and u, and then constructing a rotation matrix R:
R＝vuT
further, a translation matrix t is calculated:
t＝-RPi+Qi
and constructing a homogeneous transformation matrix Tcb of the camera coordinate system and the mechanical arm base coordinate system by using the rotation matrix and the translation matrix to obtain a transformation relation Tcb of the off-line calibration 3D camera and the mechanical arm base coordinate system.
4. The method for accurately identifying and estimating the position and the attitude of the fuel truck fuel filler cap as claimed in claim 3, wherein:
the step 3 is realized by the method that,
step 31: filtering and downsampling the collected scene point cloud, traversing the points in the scene point cloud, clustering by using Euclidean distance between the points, and introducing self-adaptive factors
Screening the segmented point cloud based on the point number of the point cloud data to obtain a point cloud of the outer cover of the oil filling cover;
step 32: for the filler cap cover point cloud, the centroid is calculated and used as a reference point Opre for selecting a Region of Interest (ROI) on the filler cap cover.
5. The method for accurately identifying and estimating the position and the attitude of the fuel truck fuel filler cap as claimed in claim 4, wherein the method comprises the following steps: step 4, the method is realized by the following steps,
step 41: setting margins delta x, delta y and delta z in the directions x, y and z respectively based on the calculated reference point Opre, and selecting a target area ROI on the outer cover of the oil filling cover;
step 42: dividing the ROI acquired in the step 41 into a plurality of small blocks, respectively calculating the mass center and normal vector of each point cloud, and taking the mean value of different block mass center coordinates as the origin Oo of a grabbing coordinate system; for normal vector, ensure it and viewpointThe included angle of the directions is larger than 90 degrees, the mean value of the normal vectors of the small blocks is used as the normal vector of the ROI area, and the normal vector is defined as the positive direction of the z axis of the grabbing coordinate system of the outer cover of the oil filling cover
Step 43: calculating the characteristic vector corresponding to the maximum characteristic value of the ROI acquired in the step 41 by using a principal component analysis algorithm, and setting the characteristic vector as the positive direction of the y axis of the grabbing coordinate system of the outer cover of the oil filling cover
Introducing a direction selection factor _ outer of the outer cover of the oil filling cover, selecting an included angle between the positive direction of the y axis and a direction vector (0,1,0) of the y axis of a camera coordinate system to be an acute angle or an obtuse angle, unifying the positive direction of the grabbing coordinate system y axis of the outer cover of the oil filling cover, and avoiding the situation that a mechanical arm cannot reach;
step 44: computing vectors using vector cross-multiplication
Defining the positive direction of the x axis of the grabbing coordinate system of the outer cover of the oil filling cover as
Obtaining an original point and three direction vectors required by constructing an oil filling cover outer cover grabbing coordinate system;
step 45: constructing a homogeneous matrix Tgo using the origin and direction vectors of the grasping coordinate system:
wherein the content of the first and second substances,
capturing a row vector formed by coordinates of an origin of a coordinate system for a defined oil filling cover outer cover;
utilizing Tgoc Tgo ═ Tc, wherein Tc is a camera coordinate system, and Tgoc is the relation between the constructed oil filling cover outer grabbing coordinate system and the camera coordinate system;
and calculating to obtain a transformation relation Tgoc between the grabbing coordinate system and the camera coordinate system, wherein the mechanical arm base coordinate system and the camera coordinate system have the following relation:
Tgob＝Tcb*Tgoc
wherein, Tgob is the transformation relation between a grabbing coordinate system and a mechanical arm base coordinate system;
and further, converting the transformation matrix Tgob into Euler angle representation (Xow, Yow, Zow, Aow, Bow and Cow) by using the relation between the rotation matrix and the Euler angle, namely the pose of the outer cover gripping point of the oil filling cover under the mechanical arm base coordinate system.
6. The method for accurately identifying and estimating the position and the attitude of the fuel truck fuel filler cap as claimed in claim 5, wherein: step 5 the method is realized by the following steps,
step 51: removing noise points of scene point cloud data containing a complete oil filling plug through preprocessing such as filtering and down-sampling; further, setting a radius threshold of a circle to be fitted, fitting a space circle, and determining the circle center space position (Xc, Yc, ZC) and the normal vector direction (Ac, Bc, Cc) of the space circle;
step 52: aligning the fitted space circle normal vector with the vector (0,0,1), converting the rotation represented by the axial angle into a matrix R0 by using a Rodrigues formula, and further rotating the oil filling plug point cloud;
step 53: the relative position relationship between the filler plug handle and the filler plug plane is utilized, and the grabbing point (Xi, Yi, Zi) of the rotated filler plug handle is calculated through the height difference delta and is used as a reference point Oprei of the interested area of the selected filler plug handle.
7. The method for accurately identifying and estimating the position and the attitude of the fuel truck fuel filler cap as claimed in claim 6, wherein: step 6 is realized by the method that,
step 61: setting margins delta xi, delta yi and delta zi in the directions of x, y and z respectively based on the calculated reference point Oprei, and selecting a target area ROI on the handle of the refueling plug;
step 62: dividing the ROI obtained in the step 61 into a plurality of small blocks, respectively calculating the mass center and normal vector of each point cloud, and taking the mean value of different block mass center coordinates as the origin Oi of the oil plug grabbing coordinate system; and for the normal vector, ensuring that the included angle between the normal vector and the viewpoint direction is larger than 90 degrees, taking the average value of the normal vectors of a plurality of small blocks as the normal vector of the ROI area, and defining the normal vector as the positive direction of the z axis of the grabbing coordinate system of the outer cover of the oil filling cover
And step 63: for ROI on the handle of the oil filler plug, a principal component analysis algorithm is used to obtain a feature vector corresponding to the maximum feature value, and the feature vector is defined as the positive direction of the y axis of a grabbing coordinate system of the oil filler plug
Introducing a direction selection factor _ inner of the oil filling plug, determining that an included angle between the direction vector (0,1,0) and the y axis of the camera coordinate system is an acute angle or an obtuse angle, uniformly grabbing the positive direction of the y axis of the coordinate system, and ensuring that a mechanical arm can reach when the clamping jaw is clamped;
step 64: defining the positive direction of the x axis of the grabbing coordinate system of the oil filling plug as
Constructing and obtaining a grabbing coordinate system of the refueling plug;
step 65: constructing a homogeneous matrix Tgi by using the direction vector of the constructed oil filling plug grabbing coordinate system:
wherein the content of the first and second substances,
capturing a row vector formed by coordinates of an origin of a coordinate system for a defined oil filling cover outer cover;
utilizing Tgic Tgi ═ Tc, wherein Tc is a camera coordinate system, and Tgic is the relation between the constructed fuel filler plug grabbing coordinate system and the camera coordinate system;
and calculating to obtain a transformation relation Tgic of the grabbing coordinate system and the camera coordinate system, wherein the mechanical arm base coordinate system and the camera coordinate system have the following relation:
Tgib＝Tcb*Tgic
wherein Tgib is a transformation relation between a grabbing coordinate system and a mechanical arm base coordinate system;
and further, converting the transformation matrix Tgib into Euler angle representation (Xiw, Yiw, Ziw, Aiw, Biw and Ciw) by using the relation between the rotation matrix and the Euler angle, namely the position of the handle grabbing point of the oil filler plug in the mechanical arm base coordinate system.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111259240.XA CN114022551A (en) | 2021-10-28 | 2021-10-28 | Method for accurately identifying and estimating pose of fuel filling cover of fuel vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111259240.XA CN114022551A (en) | 2021-10-28 | 2021-10-28 | Method for accurately identifying and estimating pose of fuel filling cover of fuel vehicle |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114022551A true CN114022551A (en) | 2022-02-08 |
Family
ID=80058285
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111259240.XA Pending CN114022551A (en) | 2021-10-28 | 2021-10-28 | Method for accurately identifying and estimating pose of fuel filling cover of fuel vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114022551A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114644315A (en) * | 2022-03-11 | 2022-06-21 | 江阴市富仁高科股份有限公司 | Accurate positioning device and method for oil tank cover for automatic oiling |
| CN115893292A (en) * | 2023-01-04 | 2023-04-04 | 浙江驿公里智能科技有限公司 | Automatic refueling equipment and control method thereof |
| WO2023207651A1 (en) * | 2022-04-29 | 2023-11-02 | 华为技术有限公司 | Cover pulling method of robot, and robot |
-
2021
- 2021-10-28 CN CN202111259240.XA patent/CN114022551A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114644315A (en) * | 2022-03-11 | 2022-06-21 | 江阴市富仁高科股份有限公司 | Accurate positioning device and method for oil tank cover for automatic oiling |
| WO2023207651A1 (en) * | 2022-04-29 | 2023-11-02 | 华为技术有限公司 | Cover pulling method of robot, and robot |
| CN115893292A (en) * | 2023-01-04 | 2023-04-04 | 浙江驿公里智能科技有限公司 | Automatic refueling equipment and control method thereof |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107767423B (en) | mechanical arm target positioning and grabbing method based on binocular vision | |
| CN114022551A (en) | Method for accurately identifying and estimating pose of fuel filling cover of fuel vehicle | |
| CN111775146B (en) | Visual alignment method under industrial mechanical arm multi-station operation | |
| CN109344882B (en) | Convolutional neural network-based robot control target pose identification method | |
| CN107590836B (en) | Kinect-based charging pile dynamic identification and positioning method and system | |
| CN111476841B (en) | Point cloud and image-based identification and positioning method and system | |
| CN111360821A (en) | Picking control method, device and equipment and computer scale storage medium | |
| CN113379849B (en) | Robot autonomous recognition intelligent grabbing method and system based on depth camera | |
| CN112509063A (en) | Mechanical arm grabbing system and method based on edge feature matching | |
| CN107300382B (en) | Monocular vision positioning method for underwater robot | |
| CN113103235B (en) | Method for vertically operating cabinet surface equipment based on RGB-D image | |
| CN115213896A (en) | Object grabbing method, system and equipment based on mechanical arm and storage medium | |
| CN113284179A (en) | Robot multi-object sorting method based on deep learning | |
| CN111598172B (en) | Dynamic target grabbing gesture rapid detection method based on heterogeneous depth network fusion | |
| CN106815585B (en) | High-precision visual positioning method for hole characteristics of complex dynamic environment | |
| JP2730457B2 (en) | Three-dimensional position and posture recognition method based on vision and three-dimensional position and posture recognition device based on vision | |
| CN109048911B (en) | Robot vision control method based on rectangular features | |
| JP2778430B2 (en) | Three-dimensional position and posture recognition method based on vision and three-dimensional position and posture recognition device based on vision | |
| Sahu et al. | Shape features for image-based servo-control using image moments | |
| CN110570473A (en) | weight self-adaptive posture estimation method based on point-line fusion | |
| WO2020010625A1 (en) | Method and system for optimizing kinematic model of robot, and storage device. | |
| CN115383740A (en) | Mechanical arm target object grabbing method based on binocular vision | |
| Ren et al. | Vision based object grasping of robotic manipulator | |
| CN114939867A (en) | Calibration method and system for mechanical arm external irregular asymmetric tool based on stereoscopic vision | |
| Pyo et al. | Precise pose estimation using landmark feature extraction and blob analysis for bin picking |
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 |