Automatic alignment system and alignment method based on machine vision and Stewart platform

Abstract

The invention discloses an automatic alignment system and method based on machine vision and a Stewart platform, and solves the problems that in an existing alignment method, human errors exist in interpretation of relative positions of a light spot center and a reticle center; the problem that the final intervisibility field of implosion diagnosis cannot be accurately obtained due to the fact that the alignment precision is limited and the alignment efficiency is low and the position deviation between the light spot center and the reticle center cannot be quantified after alignment is completed is solved. In the invention, the laser is used for emitting laser to form light spots on the front and rear differentiation plates, the relative position relation between the optical axis and the axis of the through sight hole is established, then the Stewart platform kinematics model is solved, the target object is directly adjusted to the corresponding pose through the Stewart platform, iterative adjustment is not needed, and the method has the advantage of high alignment efficiency; the center of the light spot is recognized in a machine vision mode, and compared with a traditional alignment method, the method has the advantages of being high in recognition precision and good in consistency.

Description

Technical Field

[0001] The present invention relates to an alignment method for a beam limiter, a collimator and a light pipe, and in particular to an automatic alignment system and an alignment method based on machine vision and a Stewart platform. Background Art

[0002] In the diagnosis of inertial confinement fusion implosions, multiple beam limiters and collimators with viewing apertures (hereinafter referred to as targets) are often required to limit and collimate the implosion products (such as X-rays, neutrons, and protons) to achieve measurement of the implosion products within a specific field of view. The coaxiality of each target's viewing aperture determines the error between the designed and actual fields of view in implosion diagnosis, significantly impacting the accuracy of the measurement results. Currently, in implosion diagnosis, a reference optical axis that simulates the ray propagation path is typically established using a laser. This optical axis forms a light spot on the graticules on either side of the target's viewing aperture. The human eye observes the position of the light spot on the graticule, and the position of the target in the optical path is continuously adjusted manually or electrically to gradually align the centers of the reticles on both sides of the target with the center of the light spot, completing the alignment of each target.

[0003] Current alignment methods have the following shortcomings: First, because the laser spot typically has a certain area, it is difficult for the human eye to accurately locate the center of the spot during the alignment process. As a result, human errors occur in the interpretation of the relative position between the spot center and the reticle center, limiting alignment accuracy. Second, in current alignment methods, whether manual or electric position adjustment is used, multiple iterative adjustments of the reticles on both sides of the target are required, resulting in low alignment efficiency. Third, after alignment is completed, the positional deviation between the spot center and the reticle center cannot be quantified, making it impossible to accurately obtain the final line-of-sight field for implosion diagnosis. Summary of the Invention

[0004] The present invention aims to solve the problems in existing alignment methods, such as human errors in interpreting the relative positions of the center of the light spot and the center of the reticle, limited alignment accuracy, low alignment efficiency, and the inability to quantify the positional deviation between the center of the light spot and the center of the reticle after alignment is completed, thereby failing to accurately obtain the final through-view field for implosion diagnosis. The present invention provides an automatic alignment system and method based on machine vision and a Stewart platform.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] An automatic alignment system based on machine vision and a Stewart platform includes a six-degree-of-freedom Stewart platform, which comprises a static platform, a dynamic platform, and six retractable support rods, each of which is connected to the static platform and the dynamic platform via a hinge at both ends. The system is also unique in that it also includes a laser, a front camera, a rear camera, a linear servo motor, a front differentiation plate, and a rear differentiation plate.

[0007] Each of the support rods is connected to a linear servo motor, which is used to control the extension or shortening of the support rod;

[0008] The movable platform is used to set a target object with a through-view hole;

[0009] The front differentiation plate and the rear differentiation plate are both translucent plates, and are respectively used to connect to the target object on the side surfaces at both ends of the through-view hole; the front differentiation plate and the rear differentiation plate are both provided with cross engraved lines and circular engraved lines, and the origin of the cross engraved lines and the center of the circular engraved lines both coincide with the axis of the through-view hole;

[0010] The laser is located on the front side of the front differentiation plate and is facing the viewing hole of the target object. The laser is used to emit laser light and pass through the viewing hole of the target object; the front camera is located near the front side of the front differentiation plate and is used to collect the light spot on the front differentiation plate. The rear camera is located on the rear side of the rear differentiation plate and is used to collect the light spot on the rear differentiation plate.

[0011] The pre-differentiation plate is slidably connected to the target object via a linear slide rail.

[0012] One end of the linear slide is fixedly connected to a micro linear servo motor, and the working end of the micro linear servo motor is fixedly connected to the front dividing plate. The micro linear servo motor is used to drive the front dividing plate to slide along the linear slide. A limit block is provided on the linear slide to limit the original center of the cross scale and the center of the circular scale on the front dividing plate to coincide with the axis of the through-view hole.

[0013] The laser, the front camera and the rear camera are all fixed on the ground through brackets.

[0014] An automatic alignment method based on machine vision and a Stewart platform is characterized in that an automatic alignment system based on machine vision and a Stewart platform is used, comprising the following steps:

[0015] S1, turn on the laser, emit laser, and form a front spot on the front reticle;

[0016] S2. Capturing a front spot image on the front grating plate using a front camera, identifying the cross-marked lines and circular marks on the front grating plate in the front spot image, establishing a first coordinate system with the cross-marked lines as coordinate axes, and determining the position coordinates of the center of the front spot in the first coordinate system;

[0017] S3. Remove the front differentiation plate, allow the laser to pass through the viewing hole on the target object, and form a rear light spot on the rear differentiation plate. Use the rear camera to capture the rear light spot image on the rear differentiation plate. Identify the crosshairs and circular lines on the rear reticle in the rear light spot image, and establish a second coordinate system with the crosshairs as coordinate axes. Determine the position coordinates of the center of the rear light spot in the second coordinate system.

[0018] S4. Calculate the posture change parameters required to align the through-hole of the target object based on the position coordinates of the center of the front light spot in the first coordinate system in S1 and the position coordinates of the center of the rear light spot in the second coordinate system in S2, and obtain the expected posture that the Stewart platform needs to achieve.

[0019] S5. Calculate the telescopic length of each support rod based on the expected posture that the Stewart platform needs to achieve and the Stewart platform kinematic model, and calculate the number of revolutions of the linear servo motor based on the telescopic length.

[0020] S6, controlling the rotation of the linear servo motor according to the number of revolutions of the linear servo motor to drive the Stewart platform to reach the desired posture;

[0021] S7. Reset the front reticle and execute S1 to S3. Check whether the obtained position coordinates of the center of the front light spot in the first coordinate system and the center of the rear light spot in the second coordinate system coincide with the origin of the crosshairs. If so, the expected position of the current Stewart platform is the final position of the Stewart platform, and the through-hole of the target object is aligned with the laser. Otherwise, return to S4.

[0022] The specific process of S2 is:

[0023] A front camera is used to capture a front spot image of the front differentiation plate, and the cross-marked lines and circular marks on the front differentiation plate in the front spot image are identified. A first coordinate system is established with the intersection of the cross-marked lines as the origin. The circular marks and the cross-marked lines have four intersections. Using machine vision methods, the pixel distance between the two intersections on the X-axis is L1, and the pixel distance between the two intersections on the Y-axis is L2. Given the actual diameter of the circular marks is D, the mapping relationship between pixel size and actual size during image recognition is obtained by comparing L1 with D and L2 with D.

[0024] The pixels occupied by the front light spot in the front light spot image are identified, and the pixel positions are averaged to obtain the position of the center of the front light spot. According to the established first coordinate system, the position coordinates of the center of the front light spot in the first coordinate system are obtained.

[0025] The specific process of S4 is as follows:

[0026] S41. Use the hinges at both ends of the support rod as hinge points and establish a Cartesian coordinate system on the Stewart platform:

[0027] Take the center of the moving platform as O′, the distance between each hinge point of the moving platform and the platform center O′ as rp, the angle between the adjacent hinge points and O′ as φp, and establish the Cartesian rectangular coordinate system O′XpYpZp with O′ as the origin;

[0028] The distance from each hinge point on the static platform to the platform center O is rb, the angle between adjacent hinge points and O is φb, and a Cartesian rectangular coordinate system OXbYbZb is established with O as the origin; the initial height of the moving platform relative to the static platform is h;

[0029] In the first coordinate system, the X axis is parallel to the Yp axis, and the Y axis is parallel to the Zp axis;

[0030] S42, the position coordinates of the center of the front light spot in the first coordinate system are (x1′, y1′); the position coordinates of the center of the rear light spot in the second coordinate system are (x1, y1);

[0031] Move the target along the Yb axis Move along the Zb axis Rotation around the Yb axis Rotation around the Zb axis

[0032] The pose change parameters required to align the through-hole of the target object are expressed as (0, 0,

[0033] Wherein, δ represents the depth of the target through-hole.

[0034] S5 is specifically:

[0035] S51. According to the expected posture that the Stewart platform needs to achieve and the kinematic model of the Stewart platform, the matrix equation for the coordinate transformation between the moving platform and the static platform is:

[0036]

[0037] Where c is the abbreviation of cos, s is the abbreviation of sin, α, β and γ are the attitude angles of the moving platform relative to the static platform in OY b Z b OX b Z b and OX b Y b Components in the plane, (X bO′ ,Y bO′ ,Z bO′ ) is the centroid O′ of the moving platform in the coordinate system OX b Y bZ b The three-dimensional coordinates in the Stewart platform are (α, β, γ, X bO′ ,Y bO′ ,Z bO′ );

[0038] S52, take the initial state of the Stewart platform posture parameters as (0,0,0,0,0,h); adjust the posture change parameters required for the alignment of the through-hole of the target object (0, 0, ), α=0, X bO′ =0,

[0039] S53. The homogeneous matrix of the position coordinates of the hinge points b1 to b6 on the static platform in the static platform coordinate system is:

[0040]

[0041] Where, φ b is the angle between the two adjacent hinge points on the static platform and the center of the circle, r b represents the radius of the static platform;

[0042] S54. The homogeneous matrix of the position coordinates of the hinge points p1 to p6 on the moving platform in the moving platform coordinate system is:

[0043]

[0044] Where, φ p is the angle between the two adjacent hinge points on the static platform and the center of the circle, r p represents the radius of the static platform;

[0045] S55. According to formula (2) and formula (3), the length λ of each support rod is k The relationship between (k=1,2,…6) and the Stewart platform pose parameters is:

[0046]

[0047] Where, (JP) 1k 、(JP) 2k 、(JP) 3k Represent the elements of the 1st, 2nd, and 3rd rows and the kth column of the JP matrix respectively; B 1k 、B 2k 、B 3k Represent the elements of the 1st, 2nd, and 3rd rows and the kth column of the B matrix respectively;

[0048] S56, calculating the number of revolutions of the linear servo motor (9) according to the telescopic length.

[0049] The number of revolutions of the linear servo motor is calculated based on the telescopic length as follows:

[0050] The length of each support rod of the Stewart platform in the initial position is λ k , calculate the length of each support rod after the posture change parameter changes as λ k ′, then the expansion and contraction of each support rod is λ k ′-λ k , according to the pitch m of the support rod, the corresponding number of revolutions of the linear servo motor is

[0051] Beneficial effects of the present invention:

[0052] 1. The present invention provides an automatic alignment system and method based on machine vision and Stewart platform, which identifies the center of the light spot through machine vision. Compared with traditional alignment methods, it has the advantages of high recognition accuracy and good consistency.

[0053] 2. The present invention provides an automatic alignment system and method based on machine vision and a Stewart platform. The system establishes the relative positional relationship between the optical axis and the axis of the viewing aperture by determining the coordinates of the center of the light spot on the front and rear reticles. The kinematic model of the Stewart platform is then solved. The target object is then directly adjusted to the corresponding position using the Stewart platform, eliminating the need for iterative adjustments and offering the advantage of high alignment efficiency.

[0054] 3. The present invention provides an automatic alignment system and method based on machine vision and a Stewart platform. After alignment is completed, the specific position of the center of the light spot on the reticle is identified by machine vision, and the alignment results are quantitatively evaluated to obtain a true through-view field, which is of great significance for the accurate measurement of implosion products. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 This is a first three-dimensional structural diagram of an embodiment of an automatic alignment system based on machine vision and a Stewart platform according to the present invention;

[0056] Figure 2 This is a second schematic diagram of the three-dimensional structure of an embodiment of an automatic alignment system based on machine vision and a Stewart platform according to the present invention;

[0057] Figure 3 This is a schematic diagram of the position coordinates of the center of the front light spot on the front reticle in an embodiment of an automatic alignment system based on machine vision and a Stewart platform of the present invention, where the crosshairs on the front reticle represent the X-axis and Y-axis of a rectangular coordinate system, and x and y represent the position coordinates of the center of the front light spot in this coordinate system;

[0058] Figure 4 Schematic diagram of a Stewart platform mathematical model in an embodiment of an automatic alignment system based on machine vision and a Stewart platform of the present invention;

[0059] Figure 5 Schematic diagram of the relative positions of a target object and a Stewart platform in an embodiment of an automatic alignment system based on machine vision and a Stewart platform of the present invention.

[0060] In the figure, 1- micro linear servo motor, 2- front camera, 3- optical axis, 4- laser, 5- rear camera, 6- target, 7- moving platform, 8- support rod, 9- linear servo motor, 10- static platform, 11- front reticle, 12- rear reticle, 13- linear slide, 14- light spot, 15- cross mark, 16- circular mark. DETAILED DESCRIPTION

[0061] The following will clearly and completely describe the technical solutions of the present invention in conjunction with the accompanying drawings and embodiments. Obviously, the embodiments described 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.

[0062] This embodiment is an automatic alignment system based on machine vision and Stewart platform, such as Figure 1 and Figure 2 As shown, the Stewart platform includes a six-degree-of-freedom Stewart platform, which includes a static platform 10, a dynamic platform 7, and six retractable support rods 8. The two ends of each support rod 8 are connected to the static platform 10 and the dynamic platform 7 through a hinge seat; the platform also includes a laser 4, a front camera 2, a rear camera 5, a linear servo motor 9, a front differentiation plate 11, and a rear differentiation plate 12;

[0063] Each support rod 8 is connected to a linear servo motor 9, which is used to control the extension or shortening of the support rod 8;

[0064] The movable platform 7 is used to place a target object 6 with a viewing hole; the viewing hole has a relatively large aperture.

[0065] The front and rear differentiation plates 11, 12 are both made of translucent material and are respectively connected to the target object 6 on either side of the viewing aperture. The front differentiation plate 11 is slidably connected to the target object 6 via a linear slide 13. One end of the linear slide 13 is fixedly connected to a micro linear servo motor 1, and the working end of the micro linear servo motor 1 is fixedly connected to the front differentiation plate 11. The micro linear servo motor 1 is used to drive the front reticle 11 to slide along the linear slide 13. Limit blocks are provided on the linear slide 13 to ensure that the origin of the crosshairs and the center of the circular lines on the front reticle 11 coincide with the axis of the viewing aperture.

[0066] The front differentiation plate 11 and the rear differentiation plate 12 are both provided with cross engraved lines and circular engraved lines, and the origins of the cross engraved lines and the centers of the circular engraved lines on the front differentiation plate 11 and the rear differentiation plate 12 coincide with the axes of the through-view holes;

[0067] The laser 4 is located on the front side of the front differentiation plate 11 and is facing the viewing hole of the target object 6. The laser 4 is used to emit laser light and pass through the viewing hole of the target object 6. The front camera 2 is located near the front side of the front differentiation plate 11 and is used to collect the light spot on the front differentiation plate 11. The rear camera 5 is located on the rear side of the rear differentiation plate 12 and is used to collect the light spot on the rear differentiation plate 11.

[0068] The laser 4, the front camera 2 and the rear camera 5 are all fixed on the ground through brackets.

[0069] In this embodiment, the linear servo motor 9 and the micro linear servo motor 1 are both servo motors with linear movement at the output end. The difference is that the micro linear servo motor 1 is small in size and has low output power. The end of the servo motor can also be connected to a ball screw, and the ball screw is threadedly connected to a screw nut. The screw nut corresponding to the linear servo motor 9 is fixedly connected to the free end of the support rod 8; the screw nut of the micro linear servo motor 1 is connected to the front differentiation plate 11.

[0070] The viewing hole of the target object 6 has a relatively large aperture. When in use, it is necessary to adjust the optical axis of the emitted laser and the axis of the viewing hole to be on the same straight line.

[0071] This embodiment also provides an automatic alignment method based on machine vision and a Stewart platform, comprising the following steps:

[0072] S1, turn on the laser 4, emit laser, and form a front spot on the front reticle 11, the position of which is as follows Figure 3 Middle spot 14;

[0073] S2, using the front camera 2 to collect the front spot image on the front differentiation plate 11, identifying the cross-marked lines and the circular-marked lines on the front differentiation plate 11 in the front spot image, and establishing a first coordinate system with the intersection of the cross-marked lines as the origin; Figure 3As shown, the circular scale line 16 of the front reticle and the cross scale line 15 of the front reticle have a total of 4 intersection points. The pixel distance between the two intersection points on the X axis is obtained by machine vision method as L1, and the pixel distance between the two intersection points on the Y axis is L2. The actual diameter of the circular scale line is known to be D. By comparing L1 with D and L2 with D, the mapping relationship between pixel size and actual size during image recognition can be obtained;

[0074] Identify the pixels occupied by the front spot in the front spot image, and calculate the average of the pixel positions to obtain the position of the center of the front spot. According to the established first coordinate system, obtain the position coordinates (x1′, y1′) of the center of the front spot in the first coordinate system.

[0075] S3: Remove the front differentiation plate 11. The laser passes through the viewing hole in the target object 6, forming a rear spot on the rear differentiation plate 12. The rear camera 5 captures the rear spot image on the rear differentiation plate 12. The crosshairs and circular lines on the rear reticle 12 in the rear spot image are identified. A second coordinate system is established with the crosshairs as the coordinate axes. The position (x1, y1) of the center of the rear spot in the second coordinate system is determined using the same method as in S2. Due to the large aperture of the viewing hole, the position of the rear spot in the second coordinate system deviates from its position in the first coordinate system. Therefore, it is necessary to adjust the position of the rear spot in the second coordinate system to the same position as in the first coordinate system.

[0076] S4. Determine the posture change parameters required to align the through-hole of the target object 6 based on the position coordinates of the center of the front light spot in the first coordinate system and the position coordinates of the center of the rear light spot in the second coordinate system, and obtain the expected posture that the Stewart platform needs to achieve. The specific process is as follows:

[0077] S41. Use the hinges at both ends of the support rod 8 as hinge points and establish a Cartesian rectangular coordinate system on the Stewart platform:

[0078] like Figure 4 and Figure 5 As shown, the center of the moving platform 7 is taken as O', and the distance from each hinge point of the moving platform to the platform center O' is r p , the angle between the adjacent hinge point and O′ is φ p , establish the Cartesian rectangular coordinate system O′X with O′ as the origin p Y p Z p ;

[0079] The distance between each hinge point on the static platform 10 and the platform center O is r b , the angle between the adjacent hinge points and O is φ b , establish the Cartesian rectangular coordinate system OX with O as the origin b Y b Z b ; The initial height of the moving platform relative to the static platform is h;

[0080] In the first coordinate system, the X-axis and the Y-axis p Axis parallel, Y axis and Z axis p axis parallel;

[0081] S42, the position coordinates of the center of the front light spot in the first coordinate system are (x1′, y1′); the position coordinates of the center of the rear light spot in the second coordinate system are (x1, y1);

[0082] For target 6 along Y b Axis movement Along Z b Axis movement Around Y b Axis rotation Around Z b Axis rotation

[0083] The pose change parameters required to align the through-hole of the target object 6 are expressed as (0, 0,

[0084] Wherein, δ represents the depth of the target through-hole.

[0085] S5. Given the corresponding parameters of the expected posture of the Stewart platform, and in combination with the expected posture that the Stewart platform needs to achieve, the telescopic length of each support rod 8 is calculated using the Stewart platform kinematic model, specifically:

[0086] S51. Given the corresponding parameters of the expected posture of the Stewart platform and the kinematic model of the Stewart platform, the matrix equation for the coordinate transformation between the moving platform and the static platform is:

[0087]

[0088] Where c is the abbreviation of cos, s is the abbreviation of sin, α, β and γ are the attitude angles of the moving platform relative to the static platform in OY b Z b OX b Z b and OX b Y b Component in the plane, X bO′ ,Y bO′ ,Z bO′ The centroid O′ of the moving platform is in the coordinate system OX b Y b Z b The three-dimensional coordinates in the Stewart platform are α, β, γ, X bO′ ,Y bO′ ,Z bO′ ;

[0089] S52, take the initial state of the Stewart platform posture parameter as 0,0,0,0,0,h; adjust the posture change parameter 0 required for the alignment of the through-hole of the target object 6, 0, α=0, X bO′ =0,

[0090] S53. The homogeneous matrix of the position coordinates of the hinge points b1 to b6 on the static platform in the static platform coordinate system is:

[0091]

[0092] Where, φ b is the angle between the two adjacent hinge points on the static platform and the center of the circle, r b represents the radius of the static platform;

[0093] S54. The homogeneous matrix of the position coordinates of the hinge points p1 to p6 on the moving platform in the moving platform coordinate system is:

[0094]

[0095] Where, φ p is the angle between the two adjacent hinge points on the static platform and the center of the circle, r p represents the radius of the static platform;

[0096] S55, the length of each support rod λ k The relationship between k=1,2,…6 and the Stewart platform posture parameters is:

[0097]

[0098] Where, (JP) 1k 、(JP) 2k 、(JP) 3k Represent the elements of the 1st, 2nd, and 3rd rows and the kth column of the JP matrix respectively; B 1k 、B 2k 、B 3k Represent the elements in the 1st, 2nd, and 3rd rows and the kth column of the B matrix respectively.

[0099] The number of revolutions of the linear servo motor 9 is calculated according to the telescopic length; specifically:

[0100] The length of each support rod of the Stewart platform in the initial position is λ k , calculate the length of each support rod after the posture change parameter changes as λ k ′, then the expansion and contraction of each support rod is λk ′-λ k , according to the pitch m of the support rod, the corresponding number of revolutions of the linear servo motor is

[0101] S6. The number of revolutions of each servo motor n is known. k The linear servo motor 9 is controlled to rotate according to the number of revolutions of the linear servo motor 9, driving the Stewart platform to reach the desired posture, thereby realizing alignment between the through-hole of the target object and the laser.

[0102] The process also includes S7, resetting the front reticle 11, executing S1 to S3, and determining whether the obtained position coordinates of the center of the front light spot in the first coordinate system and the center of the rear light spot in the second coordinate system coincide with the origin of the crosshairs. If so, the expected position of the current Stewart platform is the final position of the final Stewart platform; otherwise, the process returns to S4.

[0103] The present invention provides an automatic alignment system and method based on machine vision and a Stewart platform, which are used to align beam limiters, collimators, light pipes, and other components used in inertial confinement fusion implosion diagnostics. The system's main components include a servo-motor-driven six-degree-of-freedom Stewart platform, a laser, a front camera, a front reticle, a micromotor, a rear camera, a rear reticle, and a target to be aligned. Laser light is first emitted by the laser. The front and rear cameras then read the position coordinates of the center of the light spot on the front and rear reticles, respectively, to determine the position of the light spot. The required pose transformation of the target is then determined based on the relative positional relationship between the light spot and the axis of the target's viewing aperture. The corresponding pose transformation is then achieved using the Stewart platform. After the pose transformation is complete, the position coordinates of the light spot on the front and rear reticles are reacquired to quantitatively assess the alignment accuracy.

[0104] The present invention establishes an automated alignment system based on the Stewart platform and machine vision, which realizes the automatic alignment of targets in inertial confinement fusion implosion diagnosis, and has the advantages of high alignment accuracy, high efficiency, and quantitative evaluation of alignment results.

[0105] The above description is merely a specific embodiment of the present invention, and a comparison of the effects of the specific embodiment with the relevant comparative examples. However, the scope of protection of the present invention is not limited thereto. Any changes or substitutions within the technical scope disclosed in the present invention shall be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be based on the scope of protection of the claims.

Claims

1. An automatic alignment system based on machine vision and Stewart platform, comprising a six-degree-of-freedom Stewart platform, the Stewart platform comprising a static platform (10), a dynamic platform (7) and six retractable support rods (8), each support rod (8) having two ends connected to the static platform (10) and the dynamic platform (7) via hinges; characterized in that: It also includes a laser (4), a front camera (2), a rear camera (5), a linear servo motor (9), a front differentiation plate (11) and a rear differentiation plate (12); Each of the support rods (8) is connected to a linear servo motor (9), and the linear servo motor (9) is used to control the extension or shortening of the support rod (8); The moving platform (7) is used to set a target object (6) with a viewing hole; The front differentiation plate (11) and the rear differentiation plate (12) are both made of translucent material plates, and are respectively used to be connected to the target object (6) on the side surfaces at both ends of the viewing hole; the front differentiation plate (11) and the rear differentiation plate (12) are both provided with cross engraved lines and circular engraved lines, and the original center of the cross engraved line and the center of the circular engraved line are both coincident with the axis of the viewing hole; The laser (4) is located on the front side of the front differentiation plate (11) and directly faces the viewing hole of the target object (6). The laser (4) is used to emit laser light and pass through the viewing hole of the target object (6). The front camera (2) is located near the front side of the front differentiation plate (11) and is used to collect the light spot on the front differentiation plate (11). The rear camera (5) is located on the rear side of the rear differentiation plate (12) and is used to collect the light spot on the rear differentiation plate (11).

2. The automatic alignment system based on machine vision and Stewart platform according to claim 1, characterized in that: The pre-differentiation plate (11) is slidably connected to the target object (6) via a linear slide rail (13).

3. The automatic alignment system based on machine vision and Stewart platform according to claim 2, characterized in that: One end of the linear slide rail (13) is fixedly connected to a micro linear servo motor (1), and a working end of the micro linear servo motor (1) is fixedly connected to a front grating plate (11). The micro linear servo motor (1) is used to drive the front grating plate (11) to slide along the linear slide rail (13). A limit block is provided on the linear slide rail (13) to limit the original center of the cross engraved lines and the center of the circle engraved lines on the front grating plate (11) to coincide with the axis of the through-view hole.

4. The automatic alignment system based on machine vision and Stewart platform according to claim 2, characterized in that: The laser (4), the front camera (2) and the rear camera (5) are all fixed on the ground via a bracket.

5. An automatic alignment method based on machine vision and Stewart platform, characterized in that: Using the automatic alignment system based on machine vision and Stewart platform as described in any one of claims 1 to 4, comprising the following steps: S1, turning on the laser (4), emitting laser light, and forming a front light spot on the front grating plate (11); S2, collecting a front light spot image on the front differentiation plate (11) through a front camera (2), identifying the cross-marked lines and circular-marked lines on the front differentiation plate (11) in the front light spot image, establishing a first coordinate system with the cross-marked lines as coordinate axes, and determining the position coordinates of the center of the front light spot in the first coordinate system; S3, the front differentiation plate (11) is removed, the laser passes through the viewing hole on the target object (6), and a rear light spot is formed on the rear differentiation plate (12), and the rear light spot image on the rear differentiation plate (12) is collected by the rear camera (5), the cross-marked lines and the circular-marked lines on the rear graticule plate (12) in the rear light spot image are identified, and a second coordinate system is established with the cross-marked lines as coordinate axes, and the position coordinates of the center of the rear light spot in the second coordinate system are determined; S4, according to the position coordinates of the center of the front light spot in the first coordinate system in S1 and the position coordinates of the center of the rear light spot in the second coordinate system in S2, calculate the posture change parameters required to align the through-view hole of the target object (6), and obtain the expected posture that the Stewart platform needs to achieve; S5. According to the expected posture that the Stewart platform needs to achieve, the telescopic length of each support rod (8) is calculated in combination with the kinematic model of the Stewart platform, and the number of revolutions of the linear servo motor (9) is calculated according to the telescopic length; S6, controlling the linear servo motor (9) to rotate according to the number of revolutions of the linear servo motor (9), so as to drive the Stewart platform to reach a desired posture; S7, reset the front reticle (11), execute S1 to S3, and determine whether the position coordinates of the center of the front light spot in the first coordinate system and the position coordinates of the center of the rear light spot in the second coordinate system coincide with the origin of the crosshairs. If they coincide, the expected position of the current Stewart platform is the final position of the final Stewart platform, and the through-hole of the target object is aligned with the laser. Otherwise, return to S4.

6. The automatic alignment method based on machine vision and Stewart platform according to claim 5, characterized in that: The specific process of S2 is: The front spot image on the front differentiation plate (11) is collected by a front camera (2), the cross-marked lines and the circular-marked lines on the front differentiation plate (11) in the front spot image are identified, and a first coordinate system is established with the intersection of the cross-marked lines as the origin; the circular-marked lines and the cross-marked lines have a total of four intersections, and the pixel distance between the two intersections on the X-axis is obtained by a machine vision method as L1, and the pixel distance between the two intersections on the Y-axis is obtained as L2. It is known that the actual diameter of the circular-marked line is D, and by comparing L1 with D, and L2 with D, the mapping relationship between the pixel size and the actual size during image recognition can be obtained; The pixels occupied by the front light spot in the front light spot image are identified, and the pixel positions are averaged to obtain the position of the center of the front light spot. According to the established first coordinate system, the position coordinates of the center of the front light spot in the first coordinate system are obtained.

7. The automatic alignment method based on machine vision and Stewart platform according to claim 5, characterized in that: The specific process of S4 is as follows: S41. Using the hinges at both ends of the support rod (8) as hinge points, a Cartesian rectangular coordinate system is established on the Stewart platform: The center of the moving platform (7) is taken as O′, and the distance from each hinge point of the moving platform to the center of the platform O′ is r p , the angle between the adjacent hinge point and O′ is φ p , establish the Cartesian rectangular coordinate system O′X with O′ as the origin p Y p Z p ; The distance from each hinge point on the static platform (10) to the platform center O is r b , the angle between the adjacent hinge point and O is φ b , establish the Cartesian rectangular coordinate system OX with O as the origin b Y b Z b ; The initial height of the moving platform relative to the static platform is h; In the first coordinate system, the X-axis and the Y-axis p Axis parallel, Y axis and Z p Axis parallel; S42, the position coordinates of the center of the front light spot in the first coordinate system are (x1′, y1′); the position coordinates of the center of the rear light spot in the second coordinate system are (x1, y1); Move the target (6) along the Y b Axis movement Along Z b Axis movement Around Y b Axis rotation Around Z b Axis rotation The pose change parameters required to align the through-hole of the target object (6) are expressed as Wherein, δ represents the depth of the target through-hole.

8. The automatic alignment method based on machine vision and Stewart platform according to claim 7, characterized in that: S5 is as follows: S51. According to the expected posture that the Stewart platform needs to achieve, combined with the Stewart platform kinematic model, the matrix equation for the coordinate transformation between the dynamic platform and the static platform is: Where c is the abbreviation of cos, s is the abbreviation of sin, α, β and γ are the attitude angles of the moving platform relative to the static platform in OY b Z b OX b Z b and OX b Y b The component in the plane, (X bO′ ,Y bO′ ,Z bO′ ) is the centroid O′ of the moving platform in the coordinate system OX b Y b Z b The three-dimensional coordinates in the Stewart platform are (α, β, γ, X bO′ ,Y bO′ ,Z bO′ ); S52, taking the posture parameters of the Stewart platform in the initial state as (0,0,0,0,0,h); adjusting the posture change parameters required for the alignment of the through-view hole of the target object (6) α=0, X bO′ =0, S53. The homogeneous matrix of the position coordinates of the hinge points b1 to b6 on the static platform in the static platform coordinate system is: In the formula, φ b is the angle between two adjacent hinge points on the static platform and the center of the circle, r b represents the radius of the static platform; S54. The homogeneous matrix of the position coordinates of the hinge points p1 to p6 on the moving platform in the moving platform coordinate system is: In the formula, φ p is the angle between two adjacent hinge points on the static platform and the center of the circle, r p represents the radius of the static platform; S55. According to formula (2) and formula (3), the length λ of each support rod is k The relationship between (k=1,2,…6) and the Stewart platform posture parameters is: In the formula, (JP) 1k 、(JP) 2k 、(JP) 3k Respectively represent the elements of the 1st, 2nd, 3rd row and kth column of the JP matrix; B 1k , B 2k , B 3k Respectively represent the elements of the 1st, 2nd, and 3rd rows and the kth column of the B matrix; S56, calculating the number of revolutions of the linear servo motor (9) according to the telescopic length.

9. The automatic alignment method based on machine vision and Stewart platform according to claim 7 or 8, characterized in that: The number of revolutions of the linear servo motor (9) is calculated according to the telescopic length as follows: The length of each support rod of the Stewart platform in the initial position is λ k , calculate the length of each support rod after the posture change parameter changes to λ k ′, then the expansion and contraction amount of each support rod is λ k ′-λ k , according to the pitch m of the support rod, the corresponding number of revolutions of the linear servo motor is:

Images