CN109015640B - Grabbing method, grabbing system, computer device and readable storage medium - Google Patents
Grabbing method, grabbing system, computer device and readable storage medium Download PDFInfo
- Publication number
- CN109015640B CN109015640B CN201810929707.9A CN201810929707A CN109015640B CN 109015640 B CN109015640 B CN 109015640B CN 201810929707 A CN201810929707 A CN 201810929707A CN 109015640 B CN109015640 B CN 109015640B
- Authority
- CN
- China
- Prior art keywords
- grabbing
- force
- feasible
- points
- target object
- 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.)
- Active
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J9/00—Programme-controlled manipulators
- B25J9/16—Programme controls
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J9/00—Programme-controlled manipulators
- B25J9/16—Programme controls
- B25J9/1602—Programme controls characterised by the control system, structure, architecture
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J9/00—Programme-controlled manipulators
- B25J9/16—Programme controls
- B25J9/1694—Programme controls characterised by use of sensors other than normal servo-feedback from position, speed or acceleration sensors, perception control, multi-sensor controlled systems, sensor fusion
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J9/00—Programme-controlled manipulators
- B25J9/16—Programme controls
- B25J9/1694—Programme controls characterised by use of sensors other than normal servo-feedback from position, speed or acceleration sensors, perception control, multi-sensor controlled systems, sensor fusion
- B25J9/1697—Vision controlled systems
Landscapes
- Engineering & Computer Science (AREA)
- Robotics (AREA)
- Mechanical Engineering (AREA)
- Automation & Control Theory (AREA)
- Manipulator (AREA)
Abstract
The invention provides a grabbing method and a device, a computer device and a readable storage medium, which comprises the steps of obtaining one or more feasible grabbing point pairs corresponding to a target object according to constraint conditions, wherein the constraint conditions comprise the steps of judging whether two grabbing points meet a force sealing condition or not, and taking the two grabbing points as the feasible grabbing point pairs when the two grabbing points meet the force sealing condition; and calculating the grabbing probability of each feasible grabbing point pair, and selecting the feasible grabbing point pair corresponding to the maximum grabbing probability. The invention provides an algorithm for screening the feasible grabbing points with the maximum grabbing probability based on the force closure principle and the friction constraint condition, so that a pair of contact points with the maximum successful grabbing probability is screened out from a plurality of feasible grabbing point combinations which accord with the constraint condition as data points for actual grabbing execution, the success rate of one-time grabbing can be obviously improved, and the cost is reduced.
Description
Technical Field
The invention relates to the field of robots, in particular to a robot-based grabbing method, a robot-based grabbing system, a computer device and a readable storage medium.
Background
This section is intended to provide a background or context to the embodiments of the invention that are recited in the claims and the detailed description. The description herein is not admitted to be prior art by inclusion in this section.
With the increasing popularity of the internet and the mobile internet, electronic commerce has become a field closely related to the life of people. The modern logistics industry is a basic condition for successful operation of electronic commerce, so that commodities can flow rapidly and effectively between online stores and consumption, the sales condition of enterprise products is improved, and the shopping cost and the consumption cost of consumers are saved. The logistics sorting link is an essential part of the whole logistics industry, and needs to be sorted accurately in time in the face of thousands of commodities needing to be distributed every day, so that each article can be sent to a correct assembly line, a packaging point and the like. And a robotic arm grasping system equipped with a flexible robot is an effective tool for performing sorting functions. But the more the manipulator with more freedom degree, the more the price is multiplied, in addition, the more difficult the accurate control is, and the sorting cost of the logistics is increased invisibly. It goes without saying that a robot arm capable of performing the gripping task requires at least two fingers. However, the two-finger robot has relatively low gripping stability compared with a three-finger robot or other more-finger robot. Therefore, how to use the minimum manipulator (i.e. two-finger manipulator) to realize stable mechanical arm grabbing plays a crucial role in saving production cost and improving the industry competitiveness of logistics enterprises.
Generally, the three-finger manipulator is widely used, and three contact points exist when the three-finger manipulator grabs a target object, so that a stable grabbing task can be realized as long as the center of mass of the grabbed target object is located inside a triangle to meet the force closing principle. However, the universal three-finger manipulator is generally in the cost of hundreds of thousands of RMB, and the production cost is greatly increased. The three-finger manipulator made of non-metal materials and manufactured by using a 3D printing technology is also available in the market, but the manipulator is low in manufacturing precision and insufficient in hardness, and is difficult to adapt to a large-density grabbing object.
Generally, in industrial production, after information of a target object to be grabbed is acquired through a camera, various constraint conditions are provided to screen out data points which meet constraints and serve as contact points of specific grabbing operation, then inverse kinematics calculation is performed to obtain coordinates of a tail end position and a corresponding control instruction, the most common mechanical arm control method is PID control, and grabbing tasks are completed after the expected position is reached. The core of PID control is to regulate the parameters of proportional, integral and differential terms of the controller.
However, for a certain target object to be grabbed, under the relevant constraint conditions, there are many groups of data that satisfy the proposed constraint conditions. At this time, a problem is brought to the controller, how to select a group of feasible grabbing points and ensure the grabbing success rate as much as possible is a key problem, and the problems are not well treated in the prior art.
Disclosure of Invention
In view of the foregoing, the present invention provides a capture method, system, computer device and readable storage medium, so as to improve the success rate of capture.
A grabbing method is applied to a robot and comprises the following steps:
acquiring one or more feasible grabbing point pairs corresponding to a target object according to constraint conditions, wherein the constraint conditions comprise that whether two grabbing points meet a force sealing condition or not is judged, and when the two grabbing points meet the force sealing condition, the two grabbing points are used as the feasible grabbing point pairs;
and calculating the grabbing probability of each feasible grabbing point pair, and selecting the feasible grabbing point pair corresponding to the maximum grabbing probability.
Further, in the grasping method, the calculating the grasping probability of each feasible grasping point pair includes:
acquiring the grabbing quality corresponding to the grabbing points contained in each feasible grabbing point pair;
obtaining normal distribution probability of a normal vector corresponding to the grabbing point;
and obtaining the grabbing probability according to the grabbing quality and the normal distribution probability.
Further, in the grasping method, the robot includes a two-finger robot, and the acquiring one or more pairs of feasible grasping points corresponding to the target object according to the constraint condition further includes:
judging whether the distance between the two grabbing points is not less than the maximum opening distance of the two fingers of the manipulator;
when the distance between the two grabbing points is smaller than the maximum opening distance of the two mechanical hands, the two grabbing points are used as feasible grabbing point pairs; or/and
judging whether the two fingers grab the target object obliquely or not;
when the two mechanical hands do not obliquely grab the target object, the two grabbing points are taken as feasible grabbing point pairs; or/and
judging whether the two fingers of the mechanical hand can contact an object to be grabbed in advance in the process of approaching the target object;
and when the two fingers do not contact the object to be grabbed in advance in the process of approaching the target object, the two grabbing points are taken as a feasible grabbing point pair.
Further, in the grasping method, the calculating the grasping probability of each feasible grasping point pair, and selecting the feasible grasping point pair corresponding to the largest grasping probability includes:
acquiring a centroid position corresponding to the target object;
obtaining a contact force and a contact torque of a corresponding grabbing point according to the six-dimensional force sensor, and obtaining a corresponding six-dimensional force rotation according to the contact force and the contact torque;
judging whether a force rotation space corresponding to the six-dimensional force rotation meets corresponding friction constraint;
when the force momentum space corresponding to the six-dimensional force momentum meets the corresponding friction constraint, calculating the volume of a convex hull corresponding to a convex polyhedron according to the convex polyhedron formed by the force momentum space;
calculating a minimum distance from a centroid position of the target object to the surface of the convex polyhedron, and taking the minimum distance as a maximum inscribed sphere radius of the convex hull, wherein the maximum inscribed sphere radius is the grabbing quality;
judging whether the origin of the force rotation volume space is in the convex hull;
and when the origin of the force momentum space is in the convex hull, calculating the product of the normal distribution probability of the contact point corresponding to the grabbing mass and the normal vector of the contact point to obtain the grabbing probability.
Further, in the grasping method, the determining whether a momentum space corresponding to the six-dimensional momentum satisfies a corresponding friction constraint includes:
when the contact model is a point contact model with friction, judging whether the force momentum space meets a first friction constraint condition;
and when the contact model is a soft finger model, judging whether the force rotation space meets a second friction constraint condition.
Further, the grabbing method further comprises the following steps:
generating a motion instruction according to the feasible grabbing point pair corresponding to the maximum grabbing probability, wherein the motion instruction comprises a first control instruction and a second control instruction;
executing the first control instruction to control the action of a mechanical arm of the robot;
and executing the second control instruction to control the action of the manipulator of the robot.
Further, in the grasping method, before the obtaining one or more pairs of feasible grasping points corresponding to the target object according to the constraint condition, the method further includes:
calibrating and setting the 3D depth camera;
and performing filtering operation on a shot image of the target object shot by the 3D depth camera to filter out background interference information in the shot image.
A grasping system, comprising:
the device comprises an acquisition unit, a processing unit and a control unit, wherein the acquisition unit is used for acquiring one or more feasible grabbing point pairs corresponding to a target object according to constraint conditions, the constraint conditions comprise that whether two grabbing points meet a force sealing condition or not is judged, and when the two grabbing points meet the force sealing condition, the two grabbing points are used as the feasible grabbing point pairs;
and the grabbing probability unit is used for calculating the grabbing probability of each feasible grabbing point pair and selecting the feasible grabbing point pair corresponding to the maximum grabbing probability.
A computer arrangement comprising a processor for implementing the steps of the fetching method when executing a computer program stored in a memory.
A readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the fetching method.
The grabbing method, the grabbing system, the computer device and the readable storage medium provide an algorithm for screening the feasible grabbing points with the maximum grabbing probability based on the force closure principle and the friction constraint condition, so that a pair of contact points with the maximum successful grabbing probability is screened out from a plurality of feasible grabbing point combinations meeting the constraint condition to serve as data points for actual grabbing execution, the success rate of one-time grabbing can be remarkably improved, and the cost is reduced. In addition, the grabbing method, the grabbing system, the computer device and the readable storage medium further generate a time-sequential control instruction based on the screened feasible grabbing point pair with the maximum grabbing probability so as to control the actions of the mechanical arm and the mechanical arm in sequence, and thus, the precise control of the mechanical arm is facilitated.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
FIG. 1 is a flow chart of a first preferred embodiment of the grasping method according to the present invention;
FIG. 2 is a flow chart of a preferred embodiment of step 102 of FIG. 1;
FIG. 3 is a flow chart of a second preferred embodiment of the grabbing method of the present invention;
FIG. 4 is a schematic illustration of an environment in which the grabbing method of the present invention is applied;
FIG. 5 is a diagram of a preferred embodiment of the capture system of the present invention applied to a computing device;
FIG. 6 is a block diagram of a preferred embodiment of the grasping system according to the present invention.
Description of the main elements
Six-dimensional force sensor 2
3D depth camera 3
Robot 5
Target object 6
Mechanical arm 7
Storage platform 8
Input/output interface 407
Feasible grasp point pair unit 606
Grab probability unit 608
Detailed Description
So that the manner in which the above recited objects, features and advantages of the present invention can be understood in detail, a more particular description of the invention, briefly summarized above, may be had by reference to the embodiments thereof which are illustrated in the appended drawings. In addition, the embodiments and features of the embodiments of the present application may be combined with each other without conflict.
In the following description, numerous specific details are set forth to provide a thorough understanding of the present invention, and the described embodiments are merely a subset of the embodiments of the present invention, rather than a complete embodiment. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, are within the scope of the present invention.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.
Fig. 1 is a flowchart of a capturing method according to a first embodiment of the present invention, and it should be noted that the capturing method according to this embodiment is not limited to the steps and the sequence in the flowchart shown in fig. 1. Steps in the illustrated flowcharts may be added, removed, or changed in order according to various needs.
The grasping method may be applied to a robot, wherein the robot may include a robot arm, a manipulator, and a force sensor, and the robot arm may include a first control portion, and the first control portion may be configured to configure basic parameters of the robot arm, such as a relative positional relationship between rotational joints of the robot arm, how to travel to a specified position (e.g., a process in which the robot arm approaches a target object to be grasped from an initial position), and the like; the manipulator may include a second control portion that may control the posture of the manipulator, the magnitude of gripping force, and the like.
Preferably, the robot may be a six-degree-of-freedom robot, and the robot may be a two-finger robot, and the robot may have a maximum opening distance. It is understood that the two-finger robot may have a first grasping point and a second grasping point, and the maximum opening distance may be a maximum distance of a line connecting the first grasping point and the second grasping point. Therefore, when the manipulator grasps the target object by the first grasping point and the second grasping point, two grasping points can be formed on the target object. The force sensor may be disposed on the manipulator and may be configured to monitor the force (i.e., the amount of contact force) of the manipulator when closed.
It is understood that the manipulator may perform a grabbing action on the target object, and two grabbing points, such as the first grabbing point and the second grabbing point, exist when the manipulator contacts the target object every time the grabbing action is performed. Therefore, the first grasping point and the second grasping point of the manipulator can be simulated through the two grasping points, whether the two grasping points meet the constraint condition or not can be judged, and when the two grasping points meet the constraint condition, the grasping action can grasp the target object, so that the two grasping points can be used as the grasping point pair. It is understood that the two grabbed points can be extracted from the point cloud corresponding to the target object by a random sampling algorithm, wherein the point cloud of the target object can include several data points, and the point cloud of the target object can be reconstructed from the image captured by the 3D depth camera. In other embodiments, the point cloud data of the target object may be obtained by other methods, such as obtaining the point cloud data through data acquisition by a three-dimensional laser scanner, or calculating the point cloud data through a three-dimensional model.
It will be appreciated that the constraint comprises determining whether the two grasp points satisfy a force closure condition. The judgment condition of the force closure may include: based on the force momentum space, based on the contact force space, or based on the dual space of the force momentum space and the contact force space.
Preferably, for a space based on contact force, the force closure condition requires that all external forces and external moments can be balanced when grabbing, i.e. all external forces and external moment vectors sum to zero. That is, as long as the sum of the external force and the external moment is zero when the grabbing task is performed, it can be determined that the grabbing is successful at this time. In this embodiment, the grabbing mapping relationship between the grabbing force at the contact point corresponding to the two grabbing points and the corresponding contact force can be represented as a grabbing matrix. Therefore, when the two grabbing point pairs meet the force sealing condition, whether the grabbing matrixes corresponding to the two grabbing points are the row full rank matrixes can be judged. And when the grabbing matrixes corresponding to the two grabbing point pairs are row full-rank matrixes, judging that the two grabbing points meet a force sealing condition, and thus indicating that the two grabbing points can successfully grab the target object. At this time, the two grasping points may be taken as the pair of possible grasping points. When the grabbing matrixes corresponding to the two grabbing point pairs do not meet the row full-rank matrix, the target object which the two grabbing points may not successfully grab is represented, and at the moment, the two grabbing points can be removed.
It is to be understood that the constraints may further include: whether the distance between the two grabbing points is smaller than the maximum opening distance of the two fingers, whether the two fingers grab obliquely, and whether the two fingers can contact one or more of the target objects in advance in the process of approaching the target objects.
Preferably, when the distance between the two grabbing points is not less than the maximum opening distance of the two-finger manipulator, failure in grabbing the target object may be caused, and at this time, the two grabbing point pairs may also be removed; when the distance between the two grabbing points is smaller than the maximum opening distance of the two-finger manipulator, the target object can be successfully grabbed, and at the moment, the two grabbing points can be reserved as a feasible grabbing point pair.
Preferably, when the two-finger manipulator obliquely grabs the target object, if the grabbing direction and the normal directions of the two contact points are not located in the same plane, the target object may be grabbed unsuccessfully, and at this time, the two grabbing points may also be removed; when the two mechanical hands do not obliquely grab the target object, if the grabbing direction and the normal directions of the two contact points are located in the same plane, the target object can be successfully grabbed, and at the moment, the two grabbing points can also be reserved as feasible grabbing point pairs.
Preferably, before closing, when the two fingers contact the object to be grabbed in advance in the process of approaching the target object, the target object may be turned and rolled, so that the target point changes, and grabbing failure is caused, and at this time, the two grabbing points can be removed; when the two fingers do not contact the object to be grabbed in advance in the process of approaching the target object, the two grabbing points can be reserved as a feasible grabbing point pair.
And 102, calculating the grabbing probability of each feasible grabbing point pair, and selecting the feasible grabbing point pair corresponding to the maximum grabbing probability.
In this embodiment, the capture probability may be obtained according to the capture quality of the captured point and the normal distribution probability of the corresponding normal vector, and then the feasible capture point pair corresponding to the largest capture point may be selected. It can be understood that, when obtaining the grasping probabilities of the plurality of grasping points, the grasping probabilities of the plurality of grasping points may be sorted, for example, sorted from large to small according to the grasping probability, and for example, the grasping probability of the first position is greater than the grasping probability of the second position, so that a feasible grasping point pair corresponding to the grasping point located on the first position may be obtained.
Referring also to fig. 2, a preferred implementation of the step 102 may include:
and 200, acquiring the centroid position corresponding to the target object.
In this embodiment, the target object may be photographed by a 3D depth camera to obtain a photographed image including RBG information and depth information. The centroid position corresponding to the target object can then be obtained by processing the captured image.
It will be appreciated that the robot can be applied to different contact models, including a friction-bearing point contact model in which contact forces can be resolved into a normal force and two tangential force components, and a soft finger model in which contact forces can be resolved into a normal force, two tangential force components, and a torque about the contact normal vector. Therefore, the contact force and the contact torque corresponding to the grabbing point can be acquired through the six-dimensional force sensor arranged on the manipulator, and the corresponding six-dimensional force rotation can be acquired according to the contact force and the contact torque.
In this embodiment, the friction constraints include friction constraints corresponding to different contact models. For example, the point contact model with friction may correspond to a first friction constraint condition, and the soft finger model may correspond to a second friction constraint condition, wherein the first friction constraint condition may indicate that the sum of the squares and root signs of the two tangential force components is not greater than the product of the normal force and the static friction coefficient, and the normal force is greater than zero; the second friction constraint condition may be expressed as that the sum of the squares of the two tangential force components is not greater than the product of the normal force and the static friction coefficient, the normal force is greater than zero, and the absolute value of the torque of the normal vector is not greater than the product of the normal force and the moment friction coefficient. In this embodiment, the point contact model with friction may be calculated by using a polygonal pyramid to approximately replace a friction cone.
It can be understood that, it can be determined whether the force rotation space corresponding to the six-dimensional force rotation corresponds to the contact model satisfying the friction constraint. For example, when the contact model is a point contact model with friction, it may be determined whether the torque space satisfies the first friction constraint condition; when the contact model is a soft finger model, it may be determined whether the force momentum space satisfies the second friction constraint condition.
In step 206, the grabbing quality Q is set to-1 and step 216 is performed.
When the force curl space does not satisfy the corresponding friction constraint, the grasping mass of the corresponding contact point may be set to-1.
And 208, calculating the volume of a convex hull corresponding to the convex polyhedron according to the convex polyhedron formed by the force momentum space.
It will be appreciated that the calculation of the volume of the convex hull may comprise: the components of the force momentum (two forces in XY direction of the two-dimensional plane and a contact moment) are connected in sequence to form a convex polyhedron (a tetrahedron) formed by the force momentum space, and then the component can be obtained according to the volume calculation formula of the convex polyhedron.
It will be appreciated that the reference coordinate system origin may be the centroid position of the target object. Therefore, the calculated ratio of the volume of the convex hull to the maximum inscribed sphere radius corresponding to the convex hull can be used for quantifying the grabbing capacity of the grabbing point pair.
It is understood that the normal distribution probability of the normal vector of the contact point is established on the premise that the point cloud data of the target object is normally distributed. The method comprises the steps of setting a normal distribution function with an expected covariance and a mean value of zero, and adding random normal distribution error correction values to the measured data point coordinates and normal vectors to compensate the deviation between the real condition and the actual condition caused by various reasons, wherein the relative distance of each data point in the surface point cloud of the target object relative to the position of a centroid is calculated, and then the relative distance can be used as the covariance size of the normal distribution of each data point. The corresponding probability value is the normal distribution probability of the normal vector of the contact point.
It is understood that, in the above steps, the values of the grasping quality may include three types: -1, -Q, and Q, wherein Q is the minimum distance from the origin of the momentum space to the convex polyhedral surface. Therefore, the grabbing probability can be obtained by calculating the product of the grabbing quality and the normal distribution probability of the normal vector of the contact point, so that the grabbing probability of each feasible grabbing point pair can be obtained, and the feasible grabbing point pair corresponding to the maximum grabbing probability can be screened out.
The grabbing method provides an algorithm for screening the feasible grabbing points with the maximum grabbing probability based on a force closing principle and a friction constraint condition, so that a pair of contact points with the maximum successful grabbing probability is screened out from a plurality of feasible grabbing point combinations which accord with the constraint condition as data points for actual grabbing execution, the success rate of one-time grabbing can be obviously improved, and the cost is reduced.
Referring to fig. 3, a second preferred embodiment of the grabbing method of the present invention may include:
and step 300, calibrating and setting the 3D depth camera.
It can be understood that the 3D depth camera can photograph a target object placed on the storage platform to obtain a corresponding photographed image. In order to achieve more accurate positioning, the 3D depth camera may be calibrated, for example, a Zhang Yongyou chessboard calibration method may be used to obtain a transformation matrix from a world coordinate system to a pixel coordinate system.
It can be understood that there may be background interference in the actual identification process of the target object, so that the background interference factor needs to be filtered out first. Because the target object is arranged on the storage platform, namely the distance between the target object and the 3D depth camera is within a certain range, when the distance information in the shot image is not within the range, the shot image can be judged to correspond to background interference information. Therefore, a maximum threshold and a minimum threshold regarding the target object depth information may be set to achieve depth segmentation of the target object from the background interference factors. Preferably, it may be determined whether a pixel whose depth information is not within the maximum threshold and the minimum threshold exists in the captured image, and when the pixel whose depth information is not within the maximum threshold and the minimum threshold exists, the corresponding pixel may be filtered. In this way, a shot image containing only the object to be grasped can be obtained after the filtering operation.
Preferably, in step 200, the centroid position of the target object can be calculated according to the transformation matrix. For example, for a target object with uniformly distributed mass, three-dimensional coordinates of the target object may be calculated according to the transformation matrix, and an average coordinate of all three-dimensional points of the target object may be calculated, where the average coordinate may be a centroid position of the target object.
And 304, acquiring one or more feasible grabbing point pairs according to the shot image, and screening out the feasible grabbing point pairs corresponding to the maximum grabbing probability.
It is understood that step 100 and step 102 have already implemented the function of step 304, and therefore are not described herein.
It is understood that when the feasible grabbing point pairs are screened out, corresponding action commands can be generated so as to control the actions of the mechanical arm and the mechanical arm.
Referring to fig. 4, the target object 6 may be disposed on the storage platform 8, and the 3D depth camera 3 may capture the target object 6 to obtain the captured image, wherein the 3D depth camera 3 may be disposed on the bracket 4; the shot images can be transmitted into a main control computer 1, and the shot images are processed by the main control computer 1. The main control computer 1 can be connected to a robot, the robot can comprise a mechanical arm 7 and a mechanical arm 5, one end of the mechanical arm 7 can be fixed on one position, the other end of the mechanical arm can be connected to the mechanical arm 5, a six-dimensional force sensor 2 can be arranged on the mechanical arm 5, and the six-dimensional force sensor 2 can output corresponding contact force. For example, the main control computer may further generate the first control instruction and the second control instruction according to the pair of feasible capture points obtained by screening.
And 308, executing the first control instruction to control the motion of the mechanical arm, and executing the second control instruction to control the motion of the mechanical arm.
It is understood that the master control computer 1, upon generating the first control command, may be executed by the robot arm 7 to control the actions of the robot arm 7, such as the robot arm 7 being movable to a first position; the master control computer 1 can be executed by the manipulator 5 after generating the second control instruction, so as to control the actions of the manipulator 5, such as controlling the closing action of the manipulator 5. In this embodiment, the first control instruction and the second control instruction may be timing instructions, that is, the second control instruction is executed after the first control instruction is executed. For example, the controller of the robot arm receives a first motion command sent by the main control computer (where the first motion command is obtained by subtracting the measured dimension between the two fingers at the end of the robot arm from the spatial position of the target object), and then the spatial position of the end of the robot arm is obtained through inverse kinematics calculation, so that a command that the end of the robot arm needs to be moved to the spatial position and the angle at which each corresponding joint should rotate is generated to complete the grabbing task), and the robot arm moves to the specified position. When the six-degree-of-freedom mechanical arm moves to a specified position, the main control computer sends an instruction to the mechanical arm, so that the mechanical arm can complete grabbing through opening and closing actions after moving to the specified position.
The grabbing method generates a time-sequential control instruction based on the screened feasible grabbing point pair with the maximum grabbing probability so as to control the actions of the mechanical arm and the mechanical arm in sequence, thereby being beneficial to the accurate control of the mechanical arm.
Fig. 5 is an exemplary structural diagram of a computer device according to an embodiment of the present invention. The computer device 1 provided in this embodiment includes a memory 405, an input/output interface 407, a display 403, a network interface 409, and a processor 401 that exchanges data with the memory 405, the input/output interface 407, the network interface 409, and the display 403 through a bus 411. The input/output interface 407 can be connected to a mouse and/or a keyboard (not shown). The modules referred to in this application are program segments that perform a certain function and are better suited than programs for describing the execution of software on a processor.
The Processor 401 may be a Central Processing Unit (CPU), other general purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf Programmable Gate Array (FPGA) or other Programmable logic device, discrete Gate or transistor logic, discrete hardware components, etc. The general purpose processor may be a microprocessor or the processor may be any conventional processor or the like, and the processor 401 is a control center of the computer apparatus 1 and connects various parts of the whole computer apparatus 1 by various interfaces and lines.
The memory 405 may be used for storing the computer programs and/or modules, and the processor 401 may implement various functions of the computer device 1 by running or executing the computer programs and/or modules stored in the memory 405 and calling the data stored in the memory 405. The memory 405 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application program (such as a graphical interface display function) required by at least one function, and the like; the storage data area may store data (such as audio data, video data, etc.) created according to the use of the computer device, etc. Further, the memory 405 may include high speed random access memory, and may also include non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), at least one magnetic disk storage device, a Flash memory device, or other volatile solid state storage device.
In this embodiment, the display screen 403 may be a display screen with a touch function, so as to facilitate operations of a user. The memory 405 may store program code for execution by the processor 401 to implement the functionality of the grab system 417.
Referring also to fig. 6, a grab system 417 of the present invention may include one or more modules/units that may be stored in a memory of the mobile terminal and configured to be executed by one or more processors (a processor in this embodiment) to accomplish the present invention. For example, as shown in fig. 6, the grab system 417 may include an obtaining unit 600, a filtering unit 602, a calculating unit 604, a feasible grab point pair unit 606, a grab probability unit 608, and a control unit 610. The modules referred to in this application may be program segments that perform particular functions, or may be more specialized than programs that describe the execution of software on a processor.
It should be noted that, corresponding to the above embodiments of the grasping method, the grasping system 417 may include some or all of the functional modules shown in fig. 6, and the functions of the modules will be described in detail below. The same noun and its specific explanation in the above embodiments of the capturing method can also be applied to the following functional description of each module. For brevity and to avoid repetition, further description is omitted.
The acquiring unit 600 is configured to acquire a captured image output by a 3D depth 3 camera, where the captured image is captured by the 3D depth camera 3 and includes RBG information and depth information. In this embodiment, the 3D depth camera 3 may capture the target object 6 placed on the storage platform 8 to obtain a corresponding captured image. For more precise positioning, the 3D depth camera 3 may be calibrated, for example, a zhangnyou chessboard calibration method may be used to obtain a transformation matrix from a world coordinate system to a pixel coordinate system.
The filtering unit 602 is configured to perform a filtering operation on the captured image to filter out background interference information in the captured image. It can be understood that there may be background interference in the actual identification process of the target object, so that the background interference factor needs to be filtered out first. Because the target object is arranged on the storage platform, namely the distance between the target object and the 3D depth camera is within a certain range, when the distance information in the shot image is not within the range, the shot image can be judged to correspond to background interference information. Therefore, the filtering unit 602 may also set a maximum threshold and a minimum threshold for the depth information of the target object to achieve depth segmentation of the target object from the background interference factors. Preferably, the filtering unit 602 may determine whether there is a pixel in the captured image whose depth information is not within the maximum threshold and the minimum threshold, and when there is a pixel in the captured image whose depth information is not within the maximum threshold and the minimum threshold, the filtering unit 602 may filter out the corresponding pixel. In this way, a shot image containing only the object to be grasped can be obtained after the filtering operation.
The calculating unit 604 is configured to calculate a centroid position of the target object according to the transformation matrix. For example, for a target object with uniformly distributed mass, three-dimensional coordinates of the target object may be calculated according to the transformation matrix, and an average coordinate of all three-dimensional points of the target object may be calculated, where the average coordinate may be a centroid position of the target object.
The pair-of-possible-grasp-points unit 606 is configured to obtain one or more pairs of possible-grasp-points corresponding to the target object according to a constraint condition, where the constraint condition includes determining whether the two grasp points satisfy a force closure condition; if yes, the pair-of-possible-grasp-points unit 606 regards the two grasp points as a pair of possible grasp points.
It is understood that the manipulator may perform a grabbing action on the target object, and two grabbing points, such as the first grabbing point and the second grabbing point, exist when the manipulator contacts the target object every time the grabbing action is performed. Therefore, the pair-of-possible-grasp-points unit 606 may simulate the first grasp point and the second grasp point of the manipulator by two grasp points, determine whether the two grasp points satisfy the constraint condition, and when the two grasp points satisfy the constraint condition, indicate that the target object can be grasped by the grasping action, and thus, the two grasp points may be used as the pair of grasp points. It is understood that the two grabbed points can be extracted from the point cloud corresponding to the target object by a random sampling algorithm, wherein the point cloud of the target object can include several data points, and the point cloud of the target object can be reconstructed from the image captured by the 3D depth camera. In other embodiments, the point cloud data of the target object may be obtained by other methods, such as obtaining the point cloud data through data acquisition by a three-dimensional laser scanner, or calculating the point cloud data through a three-dimensional model.
It will be appreciated that the constraint comprises determining whether the two grasp points satisfy a force closure condition. The judgment condition of the force closure may include: based on the force momentum space, based on the contact force space, or based on the dual space of the force momentum space and the contact force space.
Preferably, for a space based on contact force, the force closure condition requires that all external forces and external moments can be balanced when grabbing, i.e. all external forces and external moment vectors sum to zero. That is, as long as the sum of the external force and the external moment is zero when the grabbing task is performed, it can be determined that the grabbing is successful at this time. In this embodiment, the grabbing mapping relationship between the grabbing force at the contact point corresponding to the two grabbing points and the corresponding contact force can be represented as a grabbing matrix. Therefore, the pair-of-possible-grabbing-points unit 606 may determine whether the two grabbing points satisfy the force closure condition by determining whether the grabbing matrix corresponding to the two grabbing points is a row full-rank matrix. And when the grabbing matrixes corresponding to the two grabbing point pairs are row full-rank matrixes, judging that the two grabbing points meet a force sealing condition, and thus indicating that the two grabbing points can successfully grab the target object. At this time, the pair-of-graspable-points unit 606 may regard the two graspable points as the pair of graspable-points. When the capture matrices corresponding to the two capture point pairs do not satisfy the row full rank matrix, which indicates that the two capture points may not be able to successfully capture the target object, at this time, the feasible capture point pair unit 606 may remove the two capture points.
It is to be understood that the constraints may further include: whether the distance between the two grabbing points is smaller than the maximum opening distance of the two fingers of the mechanical hand or not; whether the two-finger manipulator grips obliquely or not; whether the two-finger manipulator contacts the target object in advance in the process of approaching the target object or not.
Preferably, when the distance between two grabbing points is not less than the maximum opening distance of the two-finger robot, which may result in failure of grabbing the target object, the feasible grabbing point pair unit 606 may also remove the two grabbing point pairs; when the distance between the two grabbing points is smaller than the maximum opening distance of the two manipulators, it indicates that the target object can be successfully grabbed, and at this time, the feasible grabbing point pair unit 606 may also keep the two grabbing points as a feasible grabbing point pair.
Preferably, when the two fingers grab the target object obliquely, if the grabbing direction and the normal directions of the two contact points are not located in the same plane, which may result in failure to grab the target object, the feasible grabbing point pair unit 606 may also remove the two grabbing points; when the two manipulators do not grasp the target object obliquely, if the grasping direction and the normal directions of the two contact points are both located in the same plane, it indicates that the target object can be successfully grasped, and at this time, the feasible grasping point pair unit 606 may also keep the two grasping points as feasible grasping point pairs.
Preferably, before closing, when the two fingers contact the object to be grabbed in advance during approaching the object, the object may be turned and rolled, so that the target point changes, and the grabbing fails, and at this time, the feasible grabbing point pair unit 606 may also remove the two grabbing points; when the two fingers do not contact the object to be grasped in advance in the process of approaching the target object, it indicates that the target object can be successfully grasped, and at this time, the feasible grasping point pair unit 606 may also reserve the two grasping points as a feasible grasping point pair.
The grabbing probability unit 608 is configured to calculate a grabbing probability of each feasible grabbing point pair, and select a feasible grabbing point pair corresponding to the largest grabbing probability.
It is to be appreciated that the grab probability unit 608 can obtain a centroid location corresponding to the target object.
In this embodiment, the capture probability unit 608 may capture the target object by a 3D depth camera to obtain a captured image including RBG information and depth information. The centroid position corresponding to the target object can then be obtained by processing the captured image. The grabbing probability unit 608 may obtain a transformation matrix from a world coordinate system to a pixel coordinate system according to a Zhang-Yongyou chessboard scaling method, and may calculate the centroid position of the target object according to the transformation matrix. For example, for a target object with uniformly distributed mass, three-dimensional coordinates of the target object may be calculated according to the transformation matrix, and an average coordinate of all three-dimensional points of the target object may be calculated, where the average coordinate may be a centroid position of the target object.
The capture probability unit 608 may further obtain a contact force and a contact torque corresponding to the capture point according to the six-dimensional force sensor, and obtain a corresponding six-dimensional force rotation according to the contact force and the contact torque.
It will be appreciated that the robot can be applied to different contact models, including a friction-bearing point contact model in which contact forces can be resolved into a normal force and two tangential force components, and a soft finger model in which contact forces can be resolved into a normal force, two tangential force components, and a torque about the contact normal vector. Therefore, the grabbing probability unit 608 can obtain the contact force and the contact torque corresponding to the grabbing point through a six-dimensional force sensor disposed on the manipulator, and can also obtain the corresponding six-dimensional force rotation according to the contact force and the contact torque.
The grabbing probability unit 608 may determine whether a torque space corresponding to the six-dimensional torque satisfies a corresponding friction constraint; if so, the capture probability unit 608 may calculate a volume of a convex hull corresponding to the convex polyhedron according to the convex polyhedron formed by the force momentum space; if not, the fetch probability unit 608 may set the fetch quality Q to-1.
In this embodiment, the friction constraints include friction constraints corresponding to different contact models. For example, the point contact model with friction may correspond to a first friction constraint condition, and the soft finger model may correspond to a second friction constraint condition, wherein the first friction constraint condition may indicate that the sum of the squares and root signs of the two tangential force components is not greater than the product of the normal force and the static friction coefficient, and the normal force is greater than zero; the second friction constraint condition may be expressed as that the sum of the squares of the two tangential force components is not greater than the product of the normal force and the static friction coefficient, the normal force is greater than zero, and the absolute value of the torque of the normal vector is not greater than the product of the normal force and the moment friction coefficient. In this embodiment, the point contact model with friction may be calculated by using a polygonal pyramid to approximately replace a friction cone.
It is to be appreciated that the grab probability unit 608 can determine whether the momentum space corresponding to the six-dimensional momentum corresponds to a contact model that satisfies the friction constraint. For example, when the contact model is a point contact model with friction, it may be determined whether the torque space satisfies the first friction constraint condition; when the contact model is a soft finger model, it may be determined whether the force momentum space satisfies the second friction constraint condition.
The grabbing probability unit 608 may further calculate a minimum distance from the origin of the reference coordinate system corresponding to the target object to the surface of the convex polyhedron, and use the minimum distance as a maximum inscribed sphere radius of the convex hull, where the maximum inscribed sphere radius is the grabbing quality. It will be appreciated that the reference coordinate system origin may be the centroid position of the target object. Therefore, the grabbing probability unit 608 may be used to quantify the grabbing capability of the grabbing point pair by calculating the ratio of the convex hull volume to the maximum inscribed sphere radius corresponding to the convex hull. In one embodiment, the pair of grabbing points may be retained when a ratio of a convex hull volume to a maximum inscribed sphere radius corresponding to the convex hull is within a preset range, otherwise, the pair of grabbing points may be removed.
The grab probability unit 608 may also determine whether the origin of the momentum space is inside the convex hull; if so, the capture probability unit 608 may calculate a product of the normal distribution probability of the contact point corresponding to the capture quality Q and the normal vector of the contact point; if not, the fetch probability unit 608 may set the fetch quality Q to-Q. Namely, the grabbing quality is set to be the opposite number of the minimum distance from the origin of the force momentum space to the surface of the convex polyhedron.
It is understood that the normal distribution probability of the normal vector of the contact point is established on the premise that the point cloud data of the target object is normally distributed. The method comprises the steps of setting a normal distribution function with an expected covariance and a mean value of zero, and adding random normal distribution error correction values to the measured data point coordinates and normal vectors to compensate the deviation between the real condition and the actual condition caused by various reasons, wherein the relative distance of each data point in the surface point cloud of the target object relative to the position of a centroid is calculated, and then the relative distance can be used as the covariance size of the normal distribution of each data point. The corresponding probability value is the normal distribution probability of the normal vector of the contact point.
In an embodiment, the grabbing probability unit 608 may rank the grabbing probabilities corresponding to the one or more pairs of feasible grabbing points, for example, rank the grabbing probabilities from large to small, for example, the grabbing probability at the first position is greater than the grabbing probability at the second position, and thus, the grabbing probability unit 608 may obtain the pairs of feasible grabbing points corresponding to the grabbing points located at the first position.
The control unit 610 may generate a motion command according to the selected feasible grabbing point pairs, where the motion command includes a first control command and a second control command, where the first control command is used to control the motion of the robot arm, and the second control command is used to control the motion of the robot arm.
The control unit 610 may be executed by the robot arm 7 after generating the first control command to control the motion of the robot arm 7, such as the robot arm 7 may move to a first position; the control unit 610 may be executed by the manipulator 5 after generating the second control instruction, so as to control the actions of the manipulator 5, such as controlling the closing action of the manipulator 5. In this embodiment, the first control instruction and the second control instruction may be timing instructions, that is, the second control instruction is executed after the first control instruction is executed. For example, the controller of the robot arm receives a first motion command sent by the main control computer (where the first motion command is obtained by subtracting the measured dimension between the two fingers at the end of the robot arm from the spatial position of the target object), and then the spatial position of the end of the robot arm is obtained through inverse kinematics calculation, so that a command that the end of the robot arm needs to be moved to the spatial position and the angle at which each corresponding joint should rotate is generated to complete the grabbing task), and the robot arm moves to the specified position. When the six-degree-of-freedom mechanical arm moves to a specified position, the main control computer sends an instruction to the mechanical arm, so that the mechanical arm can complete grabbing through opening and closing actions after moving to the specified position.
The grabbing system provides a feasible grabbing point algorithm for screening the maximum grabbing probability based on a force closing principle and a friction constraint condition, so that a pair of contact points with the maximum successful grabbing probability are screened out from a plurality of feasible grabbing point combinations meeting the constraint condition to serve as data points for actual grabbing execution, the success rate of one-time grabbing can be remarkably improved, and the cost is reduced. In addition, the grabbing system generates a time-sequential control instruction based on the screened feasible grabbing point pair with the maximum grabbing probability so as to control the actions of the mechanical arm and the mechanical arm in sequence, and thus, the precise control of the mechanical arm is facilitated.
The modules integrated in the computer device 1 according to the present invention may be stored in a computer-readable storage medium if they are implemented in the form of software functional units and sold or used as independent products. Based on such understanding, all or part of the flow in the volume control method according to the above embodiments may be implemented by a computer program, which may be stored in a computer-readable storage medium, and when the computer program is executed by a processor, the steps in the volume control method according to the above embodiments may be implemented. Wherein the computer program comprises computer program code, which may be in the form of source code, object code, an executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, usb disk, removable hard disk, magnetic disk, optical disk, computer Memory, Read-Only Memory (ROM), Random Access Memory (RAM), electrical carrier wave signals, telecommunications signals, software distribution medium, and the like. It should be noted that the computer readable medium may contain content that is subject to appropriate increase or decrease as required by legislation and patent practice in jurisdictions, for example, in some jurisdictions, computer readable media does not include electrical carrier signals and telecommunications signals as is required by legislation and patent practice.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned. Furthermore, it is obvious that the word "comprising" does not exclude other elements or steps, and the singular does not exclude the plural. A plurality of units, modules or means recited in the system, apparatus or mobile terminal apparatus claims may also be implemented by one and the same unit, module or means by software or hardware. The terms first, second, etc. are used to denote names, but not any particular order.
Although the present invention has been described in detail with reference to the preferred embodiments, it will be understood by those skilled in the art that various changes may be made and equivalents may be substituted without departing from the spirit and scope of the invention.
Claims (9)
1. A grabbing method applied to a robot is characterized by comprising the following steps:
acquiring one or more feasible grabbing point pairs corresponding to a target object according to constraint conditions, wherein the constraint conditions comprise that whether two grabbing points meet a force sealing condition or not is judged, and when the two grabbing points meet the force sealing condition, the two grabbing points are used as the feasible grabbing point pairs;
calculating the grabbing probability of each feasible grabbing point pair, and selecting the feasible grabbing point pair corresponding to the maximum grabbing probability, wherein the calculating the grabbing probability of each feasible grabbing point pair, and the selecting the feasible grabbing point pair corresponding to the maximum grabbing probability comprises:
acquiring a centroid position corresponding to the target object;
obtaining a contact force and a contact torque of a corresponding grabbing point according to the six-dimensional force sensor, and obtaining a corresponding six-dimensional force rotation according to the contact force and the contact torque;
judging whether a force rotation space corresponding to the six-dimensional force rotation meets corresponding friction constraint;
when the force momentum space corresponding to the six-dimensional force momentum meets the corresponding friction constraint, calculating the volume of a convex hull corresponding to a convex polyhedron according to the convex polyhedron formed by the force momentum space;
calculating the minimum distance from the centroid position of the target object to the surface of the convex polyhedron, and taking the minimum distance as the maximum inscribed sphere radius of the convex hull, wherein the maximum inscribed sphere radius is the grabbing quality;
judging whether the origin of the force rotation volume space is in the convex hull;
and when the origin of the force momentum space is in the convex hull, calculating the product of the normal distribution probability of the contact point corresponding to the grabbing mass and the normal vector of the contact point to obtain the grabbing probability.
2. The grasping method according to claim 1, wherein the calculating the grasping probability of each feasible grasping point pair includes:
acquiring the grabbing quality corresponding to the grabbing points contained in each feasible grabbing point pair;
obtaining normal distribution probability of a normal vector corresponding to the grabbing point;
and obtaining the grabbing probability according to the grabbing quality and the normal distribution probability.
3. The method of claim 1, wherein the robot comprises a two-finger robot, and wherein the obtaining one or more pairs of feasible grasping points corresponding to the target object according to the constraint further comprises:
judging whether the distance between the two grabbing points is not less than the maximum opening distance of the two fingers of the manipulator;
when the distance between the two grabbing points is smaller than the maximum opening distance of the two mechanical hands, the two grabbing points are used as feasible grabbing point pairs; or/and
judging whether the two fingers grab the target object obliquely or not;
when the two mechanical hands do not obliquely grab the target object, the two grabbing points are taken as feasible grabbing point pairs; or/and
judging whether the two fingers of the mechanical hand can contact an object to be grabbed in advance in the process of approaching the target object;
and when the two fingers do not contact the object to be grabbed in advance in the process of approaching the target object, the two grabbing points are taken as a feasible grabbing point pair.
4. The grasping method according to claim 1, wherein the determining whether the momentum space corresponding to the six-dimensional momentum satisfies the corresponding friction constraint comprises:
when the contact model is a point contact model with friction, judging whether the force momentum space meets a first friction constraint condition;
and when the contact model is a soft finger model, judging whether the force rotation space meets a second friction constraint condition.
5. The grasping method according to claim 1, characterized in that the grasping method further includes:
generating a motion instruction according to the feasible grabbing point pair corresponding to the maximum grabbing probability, wherein the motion instruction comprises a first control instruction and a second control instruction;
executing the first control instruction to control the action of a mechanical arm of the robot;
and executing the second control instruction to control the action of the manipulator of the robot.
6. The grasping method according to claim 5, wherein the obtaining one or more pairs of feasible grasping points corresponding to the target object according to the constraint condition further includes:
calibrating and setting the 3D depth camera;
and performing filtering operation on a shot image of the target object shot by the 3D depth camera to filter out background interference information in the shot image.
7. A grasping system, characterized in that the system comprises:
the device comprises an acquisition unit, a processing unit and a control unit, wherein the acquisition unit is used for acquiring one or more feasible grabbing point pairs corresponding to a target object according to constraint conditions, the constraint conditions comprise that whether two grabbing points meet a force sealing condition or not is judged, and when the two grabbing points meet the force sealing condition, the two grabbing points are used as the feasible grabbing point pairs;
the grabbing probability unit is used for acquiring the centroid position corresponding to the target object;
obtaining a contact force and a contact torque of a corresponding grabbing point according to the six-dimensional force sensor, and obtaining a corresponding six-dimensional force rotation according to the contact force and the contact torque;
judging whether a force rotation space corresponding to the six-dimensional force rotation meets corresponding friction constraint;
when the force momentum space corresponding to the six-dimensional force momentum meets the corresponding friction constraint, calculating the volume of a convex hull corresponding to a convex polyhedron according to the convex polyhedron formed by the force momentum space;
calculating the minimum distance from the centroid position of the target object to the surface of the convex polyhedron, and taking the minimum distance as the maximum inscribed sphere radius of the convex hull, wherein the maximum inscribed sphere radius is the grabbing quality;
judging whether the origin of the force rotation volume space is in the convex hull;
and when the origin of the force momentum space is in the convex hull, calculating the product of the normal distribution probability of the contact point corresponding to the grabbing mass and the normal vector of the contact point to obtain the grabbing probability.
8. A computer arrangement, characterized in that the computer arrangement comprises a processor for implementing the steps of the fetching method according to any of claims 1-6 when executing a computer program stored in a memory.
9. A readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the grabbing method according to any one of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810929707.9A CN109015640B (en) | 2018-08-15 | 2018-08-15 | Grabbing method, grabbing system, computer device and readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810929707.9A CN109015640B (en) | 2018-08-15 | 2018-08-15 | Grabbing method, grabbing system, computer device and readable storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109015640A CN109015640A (en) | 2018-12-18 |
| CN109015640B true CN109015640B (en) | 2020-07-14 |
Family
ID=64631613
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810929707.9A Active CN109015640B (en) | 2018-08-15 | 2018-08-15 | Grabbing method, grabbing system, computer device and readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109015640B (en) |
Families Citing this family (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109508707B (en) * | 2019-01-08 | 2021-02-12 | 中国科学院自动化研究所 | Monocular vision-based grabbing point acquisition method for stably grabbing object by robot |
| CN109986560B (en) * | 2019-03-19 | 2023-02-14 | 埃夫特智能装备股份有限公司 | Mechanical arm self-adaptive grabbing method for multiple target types |
| CN110000785B (en) * | 2019-04-11 | 2021-12-14 | 上海交通大学 | Agricultural scene calibration-free robot motion vision cooperative servo control method and equipment |
| CN110232710B (en) * | 2019-05-31 | 2021-06-11 | 深圳市皕像科技有限公司 | Article positioning method, system and equipment based on three-dimensional camera |
| CN110271000B (en) * | 2019-06-18 | 2020-09-22 | 清华大学深圳研究生院 | Object grabbing method based on elliptical surface contact |
| CN110653820B (en) * | 2019-09-29 | 2022-11-01 | 东北大学 | Robot grabbing pose estimation method combined with geometric constraint |
| CN110883773B (en) * | 2019-11-20 | 2020-10-27 | 清华大学 | Self-adaptive control method for grabbing operation of two-finger manipulator |
| CN112775959A (en) * | 2020-07-24 | 2021-05-11 | 上海芯翌智能科技有限公司 | Method and system for determining grabbing pose of manipulator and storage medium |
| CN112697553A (en) * | 2020-12-14 | 2021-04-23 | 安徽中青检验检测有限公司 | Sample processing method and device for solid food detection |
| CN112809679B (en) * | 2021-01-25 | 2023-04-07 | 清华大学深圳国际研究生院 | Method and device for grabbing deformable object and computer readable storage medium |
| CN113674348B (en) * | 2021-05-28 | 2024-03-15 | 中国科学院自动化研究所 | Object grabbing method, device and system |
| CN115139325B (en) * | 2022-09-02 | 2022-12-23 | 星猿哲科技(深圳)有限公司 | Object grasping system |
| CN117021122B (en) * | 2023-10-09 | 2024-01-26 | 知行机器人科技（苏州）有限公司 | Grabbing robot control method and system |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5084139B2 (en) * | 2005-12-12 | 2012-11-28 | オリンパスメディカルシステムズ株式会社 | Endoscope device |
| US8267869B2 (en) * | 2008-12-31 | 2012-09-18 | Manua Kea Technologies | Multi-purpose biopsy forceps |
| CN103846922B (en) * | 2013-11-27 | 2016-02-24 | 清华大学 | Tooth bar non-return self adaptation under-actuated robot finger device |
| CN105598965B (en) * | 2015-11-26 | 2018-03-16 | 哈尔滨工业大学 | The autonomous grasping means of robot drive lacking hand based on stereoscopic vision |
| CN108058172A (en) * | 2017-11-30 | 2018-05-22 | 深圳市唯特视科技有限公司 | A kind of manipulator grasping means based on autoregression model |
| CN108280856B (en) * | 2018-02-09 | 2021-05-07 | 哈尔滨工业大学 | Unknown object grabbing pose estimation method based on mixed information input network model |
-
2018
- 2018-08-15 CN CN201810929707.9A patent/CN109015640B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN109015640A (en) | 2018-12-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109015640B (en) | Grabbing method, grabbing system, computer device and readable storage medium | |
| JP6921151B2 (en) | Deep machine learning methods and equipment for robot grip | |
| Schmidt et al. | Grasping of unknown objects using deep convolutional neural networks based on depth images | |
| CN107571260B (en) | Method and device for controlling robot to grab object | |
| RU2700246C1 (en) | Method and system for capturing an object using a robot device | |
| CN112060087A (en) | Point cloud collision detection method for robot to grab scene | |
| CN110271000B (en) | Object grabbing method based on elliptical surface contact | |
| CN107953329A (en) | Object identification and Attitude estimation method, apparatus and mechanical arm grasping system | |
| Jiang et al. | Learning hardware agnostic grasps for a universal jamming gripper | |
| CN115070781B (en) | Object grabbing method and two-mechanical-arm cooperation system | |
| CN113269723A (en) | Unordered grasping system for three-dimensional visual positioning and mechanical arm cooperative work parts | |
| CN115713547A (en) | Motion trail generation method and device and processing equipment | |
| CN116000966A (en) | Workpiece grabbing method, device, equipment and storage medium | |
| JP2021013996A5 (en) | Information processing methods, robot systems, product manufacturing methods, programs, recording media, and information processing equipment | |
| Arents et al. | Construction of a smart vision-guided robot system for manipulation in a dynamic environment | |
| Sahu et al. | Shape features for image-based servo-control using image moments | |
| CN113021333A (en) | Object grabbing method and system and terminal equipment | |
| Zhang et al. | Big-Net: Deep learning for grasping with a bio-inspired soft gripper | |
| CN112894794A (en) | Human body arm action simulation method and device, terminal equipment and storage medium | |
| WO2018161305A1 (en) | Grasp quality detection method, and method and system employing same | |
| Ogas et al. | Object grasping with a robot arm using a convolutional network | |
| Guo et al. | A system of robotic grasping with experience acquisition | |
| Ehlers et al. | Self-scaling Kinematic Hand Skeleton for Real-time 3D Hand-finger Pose Estimation. | |
| CN113561172A (en) | Dexterous hand control method and device based on binocular vision acquisition | |
| Su et al. | Caging a novel object using multi-task learning method |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |