Hand exoskeleton for capturing finger motion and finger configuration reconstruction method

Abstract

The application discloses a hand exoskeleton for capturing finger movement and a finger configuration reconstruction method, and can realize full-space precise movement capturing of human fingers and reconstruction of finger configurations. The hand exoskeleton is composed of a plurality of serial kinematic chains and a common base, each serial kinematic chain comprises one active joint and five passive joints, and a magnetic rotation angle sensor is arranged on the joint to realize movement capturing of the end of the serial kinematic chain. The serial kinematic chain forms a parallel structure with a human finger to realize six-degree-of-freedom precise movement capturing of the end of the human finger. Captured human finger tip movement is reconstructed through a circular arc fitting algorithm, and the finger movement configuration can be displayed in real time in a three-dimensional visual simulation environment. According to different requirements, the number of serial kinematic chains on the base can be increased or reduced, and hand movement can be reproduced in real time in a three-dimensional simulation environment, and experience is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of robot exoskeleton design and production, and particularly relates to a hand exoskeleton for capturing finger movement and a finger configuration reconstruction method. BACKGROUND

[0002] The hand exoskeleton is actually a kind of robot technology worn on the human hand, which combines exoskeleton bionics technology and information technology. The hand is one of the most important movement execution organs of human beings, and most of the daily work of human beings is completed by the dexterous hands, such as dressing, holding a plate, and picking up food. Therefore, it is of great significance to realize the full-space precise movement capture of human fingers and the finger configuration reconstruction based on the hand exoskeleton for researching the biological and mechanical movement mechanism of human hands, realizing the dexterous operation of robots based on remote operation and hand operation, and rehabilitating the patients with hand movement function damage.

[0003] The current implementation of finger movement capture mainly includes two types: the first type is a non-wearable movement capture device based on vision, such as Leap Motion. However, the non-wearable method has the problems of poor tracking accuracy, easy frame loss caused by occlusion, and poor robustness to external disturbances (such as light, background, etc.). The second type is a wearable data glove or hand exoskeleton device. The data glove can provide relatively accurate finger movement capture capability, but usually does not have the ability of force feedback and is difficult to adapt to different human hand sizes. Although the hand exoskeleton device can provide rich force feedback information for the human hand, due to the complexity and delicacy of the human hand joints, the existing devices have the problems of poor movement compatibility, and it is difficult to realize the full-space movement capture of the fingers. In order to realize the high-degree movement capture of the human hand, the full-drive hand exoskeleton designed has the problems of large size, large weight, and complex structure due to the increase of the number of drives, which seriously affects the wearing experience of human beings.

[0004] Due to the differences in sizes and configurations of different human hands, when the captured finger movement is used for finger configuration reconstruction, a simple and effective algorithm is needed to adapt to different human hand sizes, and mechanical alignment between the exoskeleton device and the fingers is not needed in any way. SUMMARY

[0005] The present application aims at the deficiencies of the prior art, and provides a hand exoskeleton for capturing finger movement and a finger configuration reconstruction method.

[0006] The hand exoskeleton for capturing finger movement provided by the present application comprises a plurality of serial kinematic chains and a base, the serial kinematic chains are fixed at the base, the base is used for binding the back of the wearer's hand, the finger tips are fixed at the end of the serial kinematic chain, and the serial kinematic chain and the wearer's finger form a parallel structure.

[0007] Further, the serial kinematic chain is composed of a first pitch link, a second pitch link, a first axial rotation yaw link, a second axial rotation yaw link and a first cross link and a second cross link connected to the finger tip in sequence through passive joints.

[0008] The first pitch link of the serial kinematic chain is fixed at the base through a driving joint.

[0009] Further, the first pitch link and the second pitch link, the second pitch link and the first axial rotation yaw link, the first axial rotation yaw link and the second axial rotation yaw link, the second axial rotation yaw link and the first cross link, the first cross link and the second cross link are connected through a first passive joint, a second passive joint, a third passive joint, a fourth passive joint and a fifth passive joint, which are composed of a micro bearing, a gasket and a bolt.

[0010] Further, the driving joint is composed of a pair of high-precision bevel gears and a joint shaft, the two bevel gears are engaged with each other, one of the bevel gears is fixed with the first pitch link, and the other bevel gear is fixed with a speed reduction motor; the joint shaft is composed of a bolt and is supported in rotation by a rolling bearing and a flange bearing.

[0011] Further, the magnetic rotation angle sensor is arranged at the driving joint and the passive joint to capture the rotation angle of the driving joint and the passive joint.

[0012] Further, the hand exoskeleton further comprises a single-chip microcomputer device for reading the angle value of the magnetic rotation angle sensor.

[0013] Further, the number of the serial kinematic chains is 1-5, which matches the number of the fingers to be captured.

[0014] On the other hand, the present specification also provides a finger configuration reconstruction method, which comprises the following steps:

[0015] S1, the wearer's hand is bound to the base, and the finger tips are bound to the cross link of the finger tip, so that the serial kinematic chain and the wearer's finger form a parallel structure;

[0016] S2, establish a world coordinate system and fix it on the hand exoskeleton, straighten the fingers and make circular motion around the finger base knuckle joint axis;

[0017] S3, collect the six-degree-of-freedom Cartesian pose dataset of the finger tip during the circular motion through the magnetic rotation angle sensor, perform circular fitting on the dataset, and obtain the circular fitting parameters and the transformation matrix of the finger base coordinate system relative to the world coordinate system;

[0018] S4, calculate the center of the circular fitting and the length of the finger through the circular fitting parameters, and obtain the length of each knuckle through the proportional relationship of each finger knuckle;

[0019] S5, substitute the transformation matrix of the finger base coordinate system relative to the world coordinate system and the six-degree-of-freedom Cartesian pose of the finger tip into the kinematics equation to obtain the homogeneous transformation matrix of the finger tip relative to the finger base coordinate system;

[0020] S6, establish a kinematics model through the homogeneous transformation matrix of the finger tip relative to the finger base coordinate system to obtain the finger joint angle, and input the solved finger knuckle length and finger joint angle into the three-dimensional simulation environment to obtain the finger configuration of the hand wearer in real time.

[0021] Further, the circular fitting parameters are obtained by substituting the position coordinates in the six-degree-of-freedom Cartesian pose dataset into the circular fitting equation.

[0022] Further, the kinematics equation in S5 is as follows:

[0023]

[0024] wherein is the six-degree-of-freedom Cartesian pose of the finger tip; is the transformation matrix of the finger base coordinate system relative to the world coordinate system.

[0025] The beneficial effects of the present application are:

[0026] First, the present application has small volume, light weight, simple structure, good wearability, good motion compatibility, and can realize real-time precise motion capture of the finger tip of the wearer.

[0027] Second, the present application can increase or decrease the number of serial kinematic chains on the base according to the needs of capturing different finger motions.

[0028] Third, the present application can adapt to the differences in hand size and configuration of different wearers and reconstruct the finger configuration of the wearer in real time through the proposed finger configuration reconstruction method.

[0029] Fourth, the method of the present invention can reproduce the wearer's hand movements in real time in a three-dimensional simulation environment, thereby improving the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 Schematic diagram of a hand exoskeleton for capturing finger movements proposed by the present invention;

[0031] Figure 2 2. It is a schematic diagram of a finger motion method for collecting data in a finger configuration reconstruction method provided by an embodiment of the present invention;

[0032] Figure 3 is a schematic diagram of kinematic modeling of a finger in the finger configuration reconstruction method provided by an embodiment of the present invention;

[0033] Figure 4 It is the interface rendering of the three-dimensional simulation environment. DETAILED DESCRIPTION

[0034] The specific embodiments of the present invention are further described in detail below with reference to the accompanying drawings.

[0035] Embodiments of the present invention provide a hand exoskeleton for capturing finger motion and a finger configuration reconstruction method, which are used to achieve precise motion capture of human fingers in the entire space.

[0036] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention are clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0037] Please refer to Figure 1 The present invention discloses a hand exoskeleton for capturing finger motion, which is used to realize the full-space precision motion capture of human fingers and the reconstruction of finger configuration, and includes multiple modular serial motion chains 1 and a common base 2.

[0038] The series kinematic chain is composed of a first pitch link 10, a second pitch link 11, a first axial rotation yaw link 12, a second axial rotation yaw link 13 connected to the base, and a first cross link 14, a second cross link 15 connected to the fingertip. The first pitch link 10 is connected to the base 2 by a pair of high-precision bevel gears 16, and the connecting shaft is fixed on the base by a rolling bearing 17, a flange bearing 18 and a bolt 19. The passive joint shaft between the first pitch link 10, the second pitch link 11, the second pitch link 11 and the first axial rotation yaw link 12, the first axial rotation yaw link 12 and the second axial rotation yaw link 13, the second axial rotation yaw link 13 and the first cross link 14, the first cross link 14 and the second cross link 15 is composed of a micro bearing, a gasket and a bolt, which reduces the axial friction and ensures the continuity of rotation. The six joints, i.e. the active joint 3, the first passive joint 8, the second passive joint 7, the third passive joint 6, the fourth passive joint 5 and the fifth passive joint 4, each have one degree of freedom, so the end of the series kinematic chain has six degrees of freedom. The active joint 3 is composed of a pair of high-precision bevel gears 16 and a joint shaft, the two bevel gears are engaged with each other, one of which is fixed to the first pitch link 10, and the other is fixed to the speed reducer motor 20; the joint shaft is composed of a bolt 19, and the joint shaft is supported by a rolling bearing 17 and a flange bearing 18. Since the rest of the series kinematic chain has the same structure, it will not be described again.

[0039] In the above-mentioned underactuated hand exoskeleton, the fingertip link is equipped with a spring to adapt to different sizes of the fingertip.

[0040] In the above-mentioned underactuated hand exoskeleton, the links are all printed by 3D photosensitive resin to reduce the overall weight and the burden on the human hand.

[0041] In this example, a finger configuration reconstruction method is implemented as follows:

[0042] The wearer's hand is tied to the base, and the fingertips of the fingers are tied to the cross link of the fingertips. The base is tied to the back of the hand, and the roots of the three series kinematic chains are fixed to the base, and the ends of the series kinematic chains are tied to the fingertips of different fingers of the human hand, such as the thumb, index finger and middle finger. The series kinematic chain is composed of one active joint and five passive joints, and the motion of the end of the series kinematic chain is captured by the magnetic rotation angle sensor 9 installed at the six joints, so as to realize the motion capture of the end of the human hand. The rotation angle of the magnetic rotation angle sensor 9 installed on the six joints is read by the single-chip microcomputer, and the forward kinematics calculation is performed to obtain the six-degree-of-freedom Cartesian pose of the end of the series kinematic chain. The series kinematic chain and the wearer's fingers form a parallel structure, so the six-degree-of-freedom Cartesian pose of the end of the wearer's fingers can be obtained.

[0043] In this example, in order to achieve the technical effect of finger configuration reconstruction, a finger base coordinate system calibration method based on circular arc fitting is proposed. As shown in the following step one, this method can adapt to different palm sizes of the wearer, effectively calibrate the spatial position of the finger base coordinate system relative to the hand exoskeleton device, and further realize the reconstruction of different finger configurations. The specific method is as follows:

[0044] The world coordinate system {W} is defined as being fixed on the above-mentioned underactuated hand exoskeleton, the wearer's finger root base coordinate system is {B i}, and the finger tip coordinate system is {EE i}, wherein i = 1, 2, 3, 4, 5 respectively represent the thumb, index finger, middle finger, ring finger and little finger. The six-degree-of-freedom Cartesian pose of the finger tip captured by the above-mentioned underactuated hand exoskeleton device is defined as The lengths of the proximal phalanx, the intermediate phalanx and the distal phalanx of the finger are defined as d 1i , d 2i , d 3i , respectively. The joint angles of the finger are defined as θ 1i , θ 2i , θ 3i , θ 4i , wherein θ 1i is the lateral flexion joint angle, θ 2i is the proximal phalanx flexion joint angle, θ 3i is the intermediate phalanx flexion joint angle, θ 4i is the distal phalanx flexion joint angle, and i = 1, 2, 3, 4, 5 respectively represent the thumb, index finger, middle finger, ring finger and little finger.

[0045] Step one: the position P i = [x i B , y i B , z i B ] of the wearer's finger root base coordinate system {B i} relative to the world coordinate system {W} is obtained. Wherein x B i is the distance of the wearer's finger root base coordinate system {B i} relative to the world coordinate system {W} in the x coordinate axis direction, y B i is the distance of the wearer's finger root base coordinate system {B i} relative to the world coordinate system {W} in the y coordinate axis direction, and z B i is the distance of the wearer's finger root base coordinate system {B i}The distance in the z-axis direction relative to the world coordinate system {W}, i=1, 2, 3, 4, 5 represent the thumb, index finger, middle finger, ring finger and little finger respectively.

[0046] From the above content, we can see that the hand exoskeleton device can capture the six-degree-of-freedom Cartesian posture of the finger end The position coordinates are defined as (x traj ,y traj , z traj ), x traj 、y traj 、z traj are the x, y, and z components of the position coordinates in Cartesian space.

[0047] like Figure 2 As shown in the figure, the fingers are straightened and move in an arc around the phalangeal joint axis at the base of the fingers. The data set of fingertip trajectory collected by the hand exoskeleton device is

[0048] {x traj ij ,y traj ij , z traj ij |j=1, 2, 3, ..., m}

[0049] where x traj ij 、y traj ij 、z traj ij The x, y, and z components of the fingertip trajectory coordinates in Cartesian space are represented in the dataset. i = 1, 2, 3, 4, and 5 represent the thumb, index finger, middle finger, ring finger, and pinky finger, respectively. j = 1, 2, 3, ..., m represents the number of collected data points. Substitute the collected dataset into the following arc fitting equation to solve for the first intermediate parameter a, the second intermediate parameter b, and the third intermediate parameter c.

[0050]

[0051] Therefore, the coordinates of the center of the circle (x B i , z B i )for

[0052]

[0053]

[0054] y B i It can be expressed by averaging the trajectory points as

[0055]

[0056] Thus the position P of the wearer's finger base coordinate system {B} relative to the world coordinate system {W} can be obtained i} relative to the world coordinate system {W} i = [x B i , y B i , z B i} relative to the world coordinate system {W} can be obtained i

[0057] Step two: obtain the lengths d 1i , d 2i , d 3i

[0058] Define the length of the straightened finger as D i , i = 1, 2, 3, 4, 5 respectively represent the thumb, index finger, middle finger, ring finger and little finger. According to the above arc fitting equation can be expressed as

[0059]

[0060] Since there is a relationship of D i = d 1i + d 2i + d 3i when the finger is straightened, that is, the length of the straightened finger is the sum of the lengths of the proximal phalanx, the middle phalanx and the distal phalanx, so d 1i , d 2i , d 3i can be obtained by the following proportional relationship between them.

[0061]

[0062] Step three: obtain the finger lateral flexion joint angle θ 1i , the proximal phalanx flexion joint angle θ 2i , the middle phalanx flexion joint angle θ 3i , the distal phalanx flexion joint angle θ 4i .

[0063] Through the kinematic equation

[0064]

[0065] The homogeneous transformation matrix of the finger tip coordinate system relative to the finger base coordinate system can be obtained​ in The hand exoskeleton device can capture the six-degree-of-freedom Cartesian pose of the finger end. The base coordinate system of the wearer's finger root obtained in step 1 {B i The inverse matrix of the homogeneous transformation matrix of} relative to the world coordinate system {W}. contains the position and posture information of the finger end coordinate system relative to the finger root base coordinate system, which can be used to solve θ 1i ,θ 2i ,θ 3i ,θ 4i value.

[0066] according to Figure 3 The kinematic model shown shows that θ 1i ,θ 2i ,θ 3i ,θ 4i It can be written as

[0067]

[0068] θ 2i =β i -γ i

[0069]

[0070] where ε i To solve θ 1i The first intermediate parameter is expressed as {EE i} and x B i In {EE i A right-angled side of the right triangle formed by the projection of} on the horizontal plane, ∈ i To solve θ 1i The second intermediate parameter is expressed as {EE i} and x B i In {EE i The other right-angled side of the right triangle formed by the projection of} on the horizontal plane, β i To solve θ 2i The first intermediate parameter, denoted as x B i The angle between the proximal phalanx joint axis and the distal phalanx joint axis of the finger, γ i To solve θ 2i The second intermediate parameter is expressed as the angle between the line connecting the proximal joint axis of the finger and the middle joint axis of the finger and the line connecting the proximal joint axis of the finger and the distal joint axis of the finger. To solve θ 3iintermediate parameter, denoted as the angle between the line connecting the proximal phalanx joint axis and the intermediate phalanx joint axis and the line connecting the intermediate phalanx joint axis and the distal phalanx joint axis. ε i , ε i may be directly obtained from the position information in . β i may be indirectly calculated from the position information in and the distal phalanx length d 3i . γ i , may be obtained from the trigonometric relationship of the triangle formed by the proximal phalanx joint axis, the intermediate phalanx joint axis and the distal phalanx joint axis. Due to the coupling relationship between the finger joints, θ 4i may be expressed as:

[0071] θ 4i = k θ 3i

[0072] wherein k is the intermediate proportionality coefficient, which is selected as follows

[0073]

[0074] Step four: as shown in Figure 4 , by solving the finger proximal phalanx, intermediate phalanx, distal phalanx length d 1i , d 2i , d 3i and finger lateral flexion movement, proximal phalanx flexion movement, intermediate phalanx flexion movement, distal phalanx flexion movement joint angle θ 1i , θ 2i , θ 3i , θ 4i , input into the three-dimensional simulation environment, the finger configuration of the hand wearer can be obtained in real time.

[0075] The above embodiments are used to explain and illustrate the present application, but not to limit the present application, any modification and change made to the present application within the spirit and protection scope of the claims of the present application, all fall into the protection scope of the present application.

Claims

1. A finger configuration reconstruction method for a hand exoskeleton that captures finger movements, characterized in that: The hand exoskeleton comprises a plurality of serial kinematic chains (1) and a base (2), wherein the roots of the serial kinematic chains (1) are fixed on the base (2), the base is used to bind the back of the wearer's hand, and the fingertips are fixed on the ends of the serial kinematic chains (1), so that the serial kinematic chains (1) and the wearer's fingers form a parallel structure, and the serial kinematic chains are composed of a first pitch link (10) connected to the base, a second pitch link (11), a first axial rotation yaw link (12), a second axial rotation yaw link (13), and a first cross link (14) and a second cross link (15) connected to the fingertips, which are sequentially connected through passive joints; The root of the first pitch link (10) of the serial kinematic chain (1) is fixed to the base (2) via an active joint (3); The method comprises the following steps: S1. Tie the wearer's hands to the base, and tie the fingertips to the cross links at the fingertips, so that the series kinematic chain and the wearer's fingers form a parallel structure; S2. Establish a world coordinate system and attach it to the hand exoskeleton. Straighten the fingers and perform circular motion around the phalangeal joint axis at the base of the fingers. S3. Collect a six-degree-of-freedom Cartesian pose dataset of the finger tip during circular motion using a magnetic rotation angle sensor, perform arc fitting on the dataset, and obtain arc fitting parameters and a transformation matrix of the base coordinate system of the finger root relative to the world coordinate system; S4, calculating the arc fitting center and the finger length using the arc fitting parameters, and obtaining the length of each phalanx based on the proportional relationship of the phalanges of each finger; S5. Substitute the transformation matrix of the base coordinate system of the finger root relative to the world coordinate system and the six-degree-of-freedom Cartesian pose of the finger tip into the kinematic equation to obtain the homogeneous transformation matrix of the finger tip relative to the base coordinate system of the finger root; S6. A kinematic model is established through the homogeneous transformation matrix of the finger tip relative to the base coordinate system of the finger root to obtain the finger joint angle. The solved finger phalangeal length and finger joint angle are input into the three-dimensional simulation environment to obtain the finger configuration of the wearer in real time.

2. The finger configuration reconstruction method of a hand exoskeleton for capturing finger movements according to claim 1, characterized in that: The first passive joint (8), the second passive joint (7), the third passive joint (6), the fourth passive joint (5) and the fifth passive joint (4) between the first pitch link (10) and the second pitch link (11), the second pitch link (11) and the first axial rotation yaw link (12), the first axial rotation yaw link (12) and the second axial rotation yaw link (13), the second axial rotation yaw link (13) and the first cross link (14), and the first cross link (14) and the second cross link (15) are all composed of miniature bearings, washers and bolts.

3. The finger configuration reconstruction method of a hand exoskeleton for capturing finger movements according to claim 1, characterized in that: The active joint (3) is composed of a pair of bevel gears (16) and a joint shaft. The two bevel gears are engaged with each other, one of the bevel gears is fixedly connected to the first pitch link (10), and the other bevel gear is fixedly connected to the reduction motor (20); the joint shaft is composed of bolts (19), and a rolling bearing (17) and a flange bearing (18) are used to support the joint shaft rotation.

4. The finger configuration reconstruction method of a hand exoskeleton for capturing finger movements according to claim 1, characterized in that: Magnetic rotation angle sensors (9) are provided at both the active joint (3) and the passive joint to capture the rotation angles of the active joint and the passive joint.

5. The finger configuration reconstruction method of a hand exoskeleton for capturing finger movements according to claim 4, characterized in that: The hand exoskeleton also includes a single chip microcomputer device for reading the angle value of the magnetic rotation angle sensor (9).

6. The finger configuration reconstruction method of a hand exoskeleton for capturing finger movements according to claim 1, characterized in that: The number of the serial kinematic chains is 1 to 5, matching the number of fingers whose motion needs to be captured.

7. The finger configuration reconstruction method of a hand exoskeleton for capturing finger movements according to claim 1, characterized in that: The arc fitting parameters are obtained by substituting the position coordinates in the six-degree-of-freedom Cartesian pose data set into the arc fitting equation.

8. The finger configuration reconstruction method of a hand exoskeleton for capturing finger movements according to claim 1, characterized in that: The kinematic equations in S5 are as follows: in is the six-degree-of-freedom Cartesian pose of the finger tip; is the transformation matrix of the finger base coordinate system relative to the world coordinate system.

Images