A high-speed atomic force microscope scanning method based on visual guidance

Abstract

The application provides a high-speed atomic force microscope scanning method based on visual guidance, which comprises the following steps: a direct correlation model of regional gray scale variation and the end position of a scanning mechanism is established, and a magnetic hysteresis compensation strategy based on a PI inverse model is given, so that accurate servo control of a complex scanning track can be realized on the basis of low visual delay; and a device for realizing the method comprises an atomic force microscope, a flexible nanometer positioning platform, a general positioning platform, an inverted microscope, a camera, a sliding table, an inverted microscope clamp, a platform support, a computer and a visual servo control system. The atomic force microscope scanning control method and device provided by the application have the advantages of high upper limit of scanning frequency, high scanning precision and high efficiency in improving the scanning performance of an existing atomic force microscope.

Description

Technical Field

[0001] The invention relates to the fields of high-speed atomic force microscope scanning and visual servo research, and in particular to a high-speed atomic force microscope scanning method based on vision guidance. Background Art

[0002] The atomic force microscope is a high-resolution scanning probe microscope that can image and measure sample surfaces at the nanoscale. It has the advantages of high resolution and the ability to perform non-contact measurements of various properties such as mechanics, electricity, and magnetism in a variety of environments such as air and liquid. It has become one of the scanning imaging and detection and analysis tools widely used in the fields of nanoscience and biotechnology.

[0003] In recent years, with the continuous advancement of scientific research in various fields, there have been higher requirements for the fast scanning performance and scanning stability of AFM systems for real-time research on biological cell movement. However, since AFM is a scanning microscope, its scanning speed is slow and it is not suitable for rapid measurement of large-area samples. Moreover, its scanning speed affects the quality of imaging. In addition, the measurement accuracy of AFM is very sensitive to environmental vibrations, so measurements need to be performed in a relatively stable environment. In AFM scanning control, although open-loop control can minimize the control cycle to the greatest extent, this will also lead to greater position errors. Obviously, the mutual constraint between scanning speed and scanning accuracy is an important technical issue in AFM systems, which needs to be solved through reasonable scanning methods and controls to improve its application efficiency and accuracy. This poses new challenges to the design and control of AFM systems.

[0004] Existing AFM systems typically use piezoelectric ceramics as actuators for scanning probes and sample movement. Piezoelectric actuators are widely used in positioning systems due to their fast response time, high resolution, wide bandwidth, and high output force. However, the nonlinear hysteresis of piezoelectric actuators significantly reduces the positioning accuracy of micro / nanosystems. To compensate for this nonlinear hysteresis and achieve high tracking performance, several piezoelectric actuator control methods have been developed. These control methods can be broadly divided into two categories: 1) feedforward control with an inverse model; and 2) closed-loop control. From a control science perspective, appropriate closed-loop control can effectively mitigate the aforementioned interference and achieve high tracking accuracy. Furthermore, most commercial AFM systems currently use a combination of a general-purpose lock-in amplifier and a data acquisition card for control. Sensors such as cantilevers and optical sensors are used to detect probe position and vibration. This sensing approach can measure basic information such as the probe's position and displacement on the sample surface, but cannot measure multiple degrees of freedom (DOF) pose. To solve this problem, visual sensing methods can be used, but this method is currently rarely used in AFM systems because the existing vision-based closed-loop control cannot meet the high-speed scanning requirements of AFM systems. In our previous research, we developed a microscopic vision system that can effectively measure the pose of a multi-degree-of-freedom compliant nanopositioning platform and applied it to the compliant nanopositioning platform for point-to-point feedback control. Although the measurement frame rate of the proposed vision-based sensing method can reach 200 Hz (Li H, Zhu B, et al Pose sensing and servo control of the compliant nanopositioners based on microscopic vision. IEEE Transactions on Industrial Electronics, 2021, 68(4): 3324-3335.

[0005] Li H, Zhang X, et al An improved template-matching-based pose tracking method for planar nanopositioning stages using enhanced correlation coefficient. IEEE Sensors Journal, 2020, PP(99):1-1), but it is still insufficient for direct closed-loop trajectory tracking control of compliant nanopositioning platforms because the time delay of vision may cause vibration of the compliant nanopositioning platform (the damping ratio of the compliance mechanism is very small).

[0006] Traditional atomic force microscopy (AFM) scanning methods are based on a grid scanning trajectory, scanning samples at a constant linear velocity and acquiring data. Although the grid scanning method has advantages such as stable scanning speed, uniform sampling points, and ease of subsequent image reconstruction, its performance is limited to low-speed scanning (Devasia S, Eleftheriou E, Moheimani SO R. Asurvey of control issues in nanopositioning[J]. IEEE Transactions on ControlSystems Technology, 2007, 15(5): 802-823.), and cannot meet the current researchers' needs for fast or even video-level imaging. In our previous research, we invented an AFM scanning method based on the Cassini oval (Liao Yu, Zhang Xianmin, Li Hai, et al. A non-grid scanning method for atomic force microscopy based on the Cassini oval, CN115420910A [P / OL]). This method can effectively increase the upper limit of the AFM scanning frequency. The constant amplitude and frequency of the signal make the drive controller of the scanning platform easier to design. A vision-guided control method capable of tracking a smooth scanning trajectory is proposed for the scanning control of an atomic force microscope (AFM).

[0007] Therefore, the existing AFM scanning system requires a scanning control method that can track the movement according to its own planned trajectory and has stable signal amplitude and frequency, so as to avoid exciting resonance and easily achieve precise control of the scanning platform, ultimately ensuring imaging quality while improving the existing AFM scanning speed. Summary of the Invention

[0008] The purpose of the present invention is to provide a high-speed atomic force microscope scanning method based on vision guidance. This method can effectively solve the problems of low scanning speed, low scanning positioning accuracy, and easy excitation of resonance during the scanning process of the AFM system, and provide new ideas for the subsequent research and engineering application of high-speed atomic force microscope scanning methods. To achieve the above purpose, the present invention application proposes a high-speed atomic force microscope scanning method and device based on vision guidance.

[0009] A high-speed atomic force microscope scanning method based on vision guidance comprises the following steps:

[0010] Step 1: Plan the AFM scanning trajectory, select a time-dependent parameterized trajectory curve, and discretize the desired trajectory to obtain discrete poses with the same time interval;

[0011] Step 2: Determine the correspondence between the local area of ​​the image space and the expected output trajectory to obtain the expected template;

[0012] Step 3: Design a visual tracking controller to ensure that the expected template area follows the discrete trajectory pose and realize vision-guided atomic force microscope scanning trajectory tracking;

[0013] Step 4: Decouple the XY motion of the nanopositioning platform and use the PI inverse model to perform feedforward control on the nanopositioning platform. Then, in combination with the micro-vision system, the PI inverse model parameters are updated according to the end-stage pose error of the nanopositioning platform to achieve closed-loop control that reduces the hysteresis effect of the nanopositioning platform.

[0014] Step 5: Pre-update the desired template position. Then, use the error between the desired template position and the grayscale image position of the nanopositioning platform end-position captured by the visual feedback actual module as input. The desired position at the next moment is obtained through processing by the visual servo controller, which is then inversely solved into the required joint parameters. The end-effector is then controlled to reach the corresponding position based on the calculated joint parameters, thereby achieving full closed-loop control.

[0015] Furthermore, the nanopositioning platform includes a flexible nanopositioning platform and an ordinary positioning platform, which are used to move in the X / Y direction on the sample surface; the flexible nanopositioning platform is located between the atomic force microscope and the ordinary positioning platform, and the sample placed on the flexible nanopositioning platform can be observed from below through the ordinary positioning platform with a hollow middle portion. The inverted microscope captures the image of the observed sample through a camera and transmits it to the visual servo control system on the computer. The XY movement of the flexible nanopositioning platform is controlled according to the actual trajectory image information and the planned trajectory, so as to realize scanning of the atomic force microscope according to the planned trajectory.

[0016] Furthermore, the Cassini oval is adopted and a translation function is added to make the atomic force microscope scanning area cover the sample surface, obtaining a time-dependent parameterized trajectory curve. The parameterized trajectory is planned in the image space, and then the desired trajectory is discretized into discrete poses with the same time interval.

[0017] Furthermore, a corresponding template region is cropped from the initial image. The template region moves along the desired trajectory in the image space. While discretizing the desired trajectory, the position of the template region in the image space is also discretized. The discrete pose of the discretized template region is used to replace the discrete pose of the discretized trajectory as the trajectory expectation template.

[0018] Furthermore, the visual tracking controller uses a template matching algorithm based on regional grayscale features to establish the relationship between the posture of the flexible nanopositioning platform and the grayscale intensity of the image, converting the visual measurement process into a multivariable nonlinear optimization problem. Combined with the relationship between the posture of the positioning platform and the external matrix mapped to the image, the optimization and control of the template matching algorithm is realized.

[0019] Furthermore, the feedback signal used by the visual tracking controller is the difference in dense photometric information between the expected template and the actual image template area.

[0020] Furthermore, the visual feedback actual module collects a grayscale image of the end position of the nanopositioning platform through a camera.

[0021] Furthermore, in the process of achieving full closed-loop control in step 6, inverse kinematics calculation is required to convert the desired posture into the required joint angles.

[0022] Furthermore, inverse kinematics is the process of solving the angles or positions of each corresponding joint based on the desired position and posture of the end effector of the robotic arm, and is solved by numerical calculation methods.

[0023] Furthermore, after obtaining the required joint angles or positions, the position and posture of the end effector of the nanopositioning platform are calculated through kinematic forward solution. Kinematics is the process of solving the position and posture of the end effector of the nanopositioning platform based on the joint angles or positions of the nanopositioning platform.

[0024] Compared with the existing technology, the beneficial effects of the present invention are:

[0025] The high-speed atomic force microscope scanning method based on visual guidance of the present invention realizes efficient control in the AFM system. This method does not need to extract the image features of the visual acquisition, but directly utilizes the posture of the expected template and the dense photometric information of the template image, which effectively simplifies the image processing process. In addition, unlike the typical closed-loop tracking method, this method does not need to change the reference position in each servo cycle, but keeps the reference signal unchanged. This feature makes the variable tracking problem a fixed-value tracking problem, so that the AFM system can be controlled to perform high-speed scanning according to its own planned trajectory, thereby improving the stability and controllability of the control method. Therefore, the visual servoing method of the present invention provides a more efficient and reliable solution for the control of the AFM system, which has broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 The figure is a schematic diagram of the overall structure of a high-speed atomic force microscope scanning method and device based on vision guidance;

[0027] FIG2 is a schematic diagram of the visual servo full closed-loop control adopted by the AFM system of the present invention;

[0028] Figure 3 Schematic diagram of a process of a high-speed atomic force microscope scanning method based on vision guidance in the present invention;

[0029] In the figure: 1- atomic force microscope, 2- AFM support device, 3- flexible nanopositioning platform, 4- ordinary positioning platform, 5- inverted microscope, 6- imaging lens group, 7- spectrometer, 8- coaxial light source, 9- camera, 10- slide stage, 11- inverted microscope fixture, 12- platform support. DETAILED DESCRIPTION

[0030] To help those skilled in the art better understand the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It is apparent that the embodiments described are only a portion of the present invention, not all of the embodiments. All other embodiments derived by those skilled in the art based on the embodiments of the present invention without inventive effort are also within the scope of protection of the present invention.

[0031] like Figure 3 A high-speed atomic force microscope scanning method based on vision guidance is shown, comprising the following steps:

[0032] Step 1: Establish a vision-guided high-speed atomic force microscope control device, including an atomic force microscope system, a nanopositioning platform, a micro-vision system, a computer, and a visual servo control system;

[0033] Step 2: Plan the AFM scanning trajectory of the high-speed AFM control device, select a time-dependent parameterized trajectory curve, and discretize the desired trajectory to obtain discrete poses with the same time interval;

[0034] The AFM scanning trajectory needs to be smooth and the signal amplitude and frequency must be stable. The scanning speed must be uniform and constant. This embodiment uses the Cassini oval line and adds a translation function to make the scanning area cover the sample surface. A time-dependent trajectory curve can be obtained, and the parameterized trajectory is planned in the image space. Then, this desired trajectory is discretized into discrete positions with the same time interval.

[0035] Step 3. Establish a direct correlation model between regional grayscale changes and the terminal posture of the scanning mechanism, determine the correspondence between the local area of ​​the image space and the expected output trajectory, and obtain the expected template to control the terminal posture of the scanning mechanism. The terminal posture must satisfy the posture discretized according to the output expected trajectory. The grayscale area here refers to the local grayscale area of ​​the image space. By determining the correspondence between the local area of ​​the image space and the expected output trajectory, the correspondence between the expected output trajectory change and the grayscale change of the local area of ​​the image space can be determined.

[0036] The discrete poses of the template region in the image space can be used to replace the discrete poses of the trajectory, and can be used as the expected template of the trajectory.

[0037] Step 4, designing a visual tracking controller to ensure that the template region follows the discrete trajectory pose and realizes visual guidance of the atomic force microscope scanning trajectory tracking;

[0038] According to the expected trajectory pose of the trajectory expected template, the fundamental element of the visual tracking controller is a feedback signal. The feedback signal used in the framework of the controller is the difference between the dense intensity information of the expected template in the image and the actual template region in the image.

[0039] As shown in Figure 2 The core function of the visual tracking controller is to convert the dense intensity error between the expected planned trajectory template and the actual template into the pose error of the compliant nanometer positioning platform, and to control the platform by generating a control signal, so as to realize the goal of the visual tracking controller. The template matching algorithm based on the gray feature of the region is used to establish the relationship between the pose of the compliant nanometer positioning platform and the gray intensity of the image, and the visual measurement process is converted into a multivariable nonlinear optimization problem. Combined with the relationship between the pose of the compliant nanometer positioning platform and the external matrix mapped to the image, the optimization and control of the algorithm are realized.

[0040] Step 5, the hysteresis compensation strategy based on the PI inverse model first decouples the XY motion of the nanometer positioning platform, then uses the P-I inverse model for feedforward control of the nanometer positioning platform, and then updates the P-I inverse model parameters according to the end pose error of the nanometer positioning platform in combination with the micro-vision system, to realize the closed-loop control of reducing the hysteresis effect of the nanometer positioning platform.

[0041] In order to reduce the hysteresis effect of the nanometer positioning platform, a P-I inverse model is added in front of the nanometer positioning platform for feedforward control, and the P-I inverse model parameters are updated according to the end pose error of the nanometer positioning platform in combination with the micro-vision system to realize the first closed-loop control. The end pose of the nanometer positioning platform contains XY motion in two directions, which can be decomposed into pose errors in the two directions by decoupling (with a decoupling feedforward controller (DFC)) to realize closed-loop control respectively. The camera collects the image of the end pose of the nanometer positioning platform and transmits it to the computer, compares it with the image of the expected end pose of the nanometer positioning platform, obtains the end pose error of the nanometer positioning platform, and updates the feedback signal of the closed-loop control.

[0042] The P-I inverse model is:

[0043]

[0044] wherein represents input displacement, represents output voltage, n is the number of play operators adopted, is a weighting constant. and are coefficients to be estimated. is a play operator with threshold , and the specific formula is:

[0045]

[0046]

[0047] wherein is a play operator at time , and is a play operator, is a threshold, represents time, and the time interval is divided into segments.

[0048] Step 6, first pre-update the expected template position, then take the error between the expected template position and the gray image of the end pose of the nano-positioning platform collected by the visual feedback actual module (the visual feedback actual module collects the gray image of the end pose of the nano-positioning platform through a camera) as input, process it through a visual servo controller to obtain the expected pose at the next time, inverse solve it into the required joint parameters, and then control the end effector to reach the corresponding position according to the calculated joint parameters, so as to realize full closed-loop control. In the process of realizing full closed-loop control, inverse kinematics calculation is needed to convert the expected pose into the required joint angle. Inverse kinematics refers to the process of solving the angles (or positions) of each joint corresponding to the expected position and attitude of the end effector of the robotic arm. Generally, the inverse kinematics problem is a multivariate nonlinear equation solving problem, which can be solved by numerical calculation methods (such as Newton method, gradient descent method, etc.).

[0049] After obtaining the required joint angle or position, the position and attitude of the end effector of the nano-positioning platform can be calculated through kinematics forward solution. Kinematics forward solution refers to the process of solving the position and attitude of the end effector of the nano-positioning platform according to the joint angle (or position) of the nano-positioning platform. For the forward kinematics problem of the nano-positioning platform, it can usually be solved by forward iteration method, etc.

[0050] ​Specifically, the previous step realizes the closed-loop control to reduce the hysteresis effect of the nanopositioning platform. The purpose of step 6 is to realize the second closed-loop control of the end posture of the nanopositioning platform based on the visual tracking controller designed earlier. The camera collects the image of the end posture of the nanopositioning platform and transmits it to the computer. The expected template position is first pre-updated, and then the error between the expected template position and the grayscale image position of the nanopositioning platform end posture collected by the visual feedback actual module is used as input. The expected posture at the next moment is obtained through processing by the visual servo controller, which is inversely solved into the required joint parameters, and then the end effector is controlled to reach the corresponding position according to the calculated joint parameters.

[0051] This invention utilizes an image-based visual servoing (IBVS) method, which offers two advantages: first, it eliminates the need for image interpretation, and second, it eliminates errors caused by sensor modeling and camera calibration. By directly utilizing image information for pose control, the IBVS method avoids the tedious feature extraction and matching required in traditional methods. By pre-updating the position of the desired template to update the feedback quantity, the pose and dense photometric information of the desired template can be directly used after the update, eliminating the need to extract visually captured image features. This effectively simplifies the image processing process. Furthermore, unlike typical closed-loop tracking methods, the reference signal in this method remains unchanged, eliminating the need to change the reference position at each servo cycle. This transforms the trajectory tracking problem into a measurement control problem, thereby simplifying the design and implementation of the visual servo controller.

[0052] As an example, Figure 1 As shown, the visually guided high-speed atomic force microscope control device based on the visually guided high-speed atomic force microscope scanning method described in this embodiment includes an atomic force microscope system, a nanopositioning platform (3), a micro-vision system, and a visual servo control system.

[0053] The atomic force microscope system includes an atomic force microscope 1 and an AFM support device 2, which are used to scan the sample surface in the Z direction.

[0054] The nanopositioning platform 3 includes a compliant nanopositioning platform 3 and a common positioning platform 4, and is used to move in the X / Y direction on the sample surface.

[0055] The micro-vision system includes an inverted microscope 5, a camera 9, a slide 10, an inverted microscope fixture 11, and a platform support 12. It is used for visual perception, collecting grayscale information and actual trajectory image information at the end of the positioning platform, and transmitting them to the visual servo control system.

[0056] like Figure 2As shown in FIG, the visual servo control system includes computer hardware and visual servo control system software. The visual servo control system software mainly includes a visual tracking controller, a PI inverse model, and the like.

[0057] The AFM support device 2 is fixed on the ordinary positioning platform 4, and the ordinary positioning platform 4 is stably supported and fixed by four platform pillars 12. The AFM support device 2 is used to support the atomic force microscope 1 so that the atomic force microscope 1 can move in the Z direction within the AFM support device 2. The flexible nanopositioning platform 3 is located between the atomic force microscope 1 and the ordinary positioning platform 4. The inverted microscope 5 is fixed on the slide 10 under the action of the inverted microscope clamp 11. The up and down movement of the slide 10 can be used to adjust the focal length of the inverted microscope 5. The inverted microscope 5 is composed of an imaging lens group 6, a beam splitter 7, and a coaxial light source 8. The coaxial light source 8 is refracted by the beam splitter 7 and focused onto the sample surface through the imaging lens group 6, so that the camera 9 can collect sample images with relatively good quality. The inverted microscope 5 can observe the sample placed on the flexible nanopositioning platform 3 from below through the ordinary positioning platform 4 hollowed out in the middle. The inverted microscope 5 collects images of the observed samples through the camera 9 and transmits them to the visual servo control system on the computer. According to the actual trajectory image information and the planned trajectory, the flexible nanopositioning platform 3 is controlled to move in the XY direction to realize the scanning of the atomic force microscope 1 according to the planned trajectory.

[0058] As an embodiment, the AFM support device 2 is composed of a bracket, a support plate, a lifting slide, and a horizontal slide, and is used to support the atomic force microscope 1 so that the atomic force microscope 1 can move in the Z direction within the AFM support device 2.

[0059] The high-speed AFM scanning adopts dual closed-loop control with vision sensing, and can scan at high speed while keeping the Z-axis sampling frequency of the AFM basically unaffected.

[0060] The above-mentioned dual closed-loop control is specifically as follows: in the first closed-loop control, the XY motion of the nanopositioning platform is decoupled, and the PI inverse model is used to perform feedforward control on the nanopositioning platform. Then, the micro-vision system is combined to update the PI inverse model parameters according to the end posture error of the nanopositioning platform, thereby realizing closed-loop control that reduces the hysteresis effect of the nanopositioning platform; in the first closed-loop control, the expected template position is pre-updated first, and then the error between the expected position and the actual position is used as input. The expected posture at the next moment is obtained through processing by the visual servo controller, which is inversely solved into the required joint parameters, and then the end effector is controlled to reach the corresponding position according to the calculated joint parameters, thereby realizing full closed-loop control.

[0061] The preferred embodiments of the present invention disclosed above are intended only to help illustrate the present invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the present invention to the specific embodiments described. Obviously, many modifications and variations are possible based on the content of this specification. These embodiments are selected and described in detail in this specification to better explain the principles and practical applications of the present invention, thereby enabling those skilled in the art to better understand and utilize the present invention. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A high-speed atomic force microscope scanning method based on vision guidance, characterized in that: The following steps are involved: Step 1: Plan the scanning trajectory of the atomic force microscope, express the desired trajectory as a time-dependent parameterized trajectory curve, and discretize the parameterized trajectory curve to obtain discrete poses with the same time interval; Step 2: Determine the correspondence between the local area of ​​the image space and the expected trajectory to obtain the expected template; Step 3: Design a visual tracking controller to ensure that the expected template follows the discrete trajectory pose and realize vision-guided atomic force microscope scanning trajectory tracking; Step 4: Decouple the X and Y motions of the nanopositioning platform. This involves establishing independent motion control models for the X and Y directions, respectively. The PI inverse model is used for feedforward control of the nanopositioning platform. The micro-vision system is then used to update the PI inverse model parameters based on the end-stage pose error of the nanopositioning platform, achieving closed-loop control that minimizes hysteresis. Step 5: Pre-update the desired template position. Then, use the error between the desired template position and the grayscale image position of the nanopositioning platform end-position captured by the visual feedback actual module as input. The desired position at the next moment is obtained through processing by the visual servo controller, which is then inversely solved into the required joint parameters. The end-effector is then controlled to reach the corresponding position based on the calculated joint parameters, thereby achieving full closed-loop control.

2. The method for high-speed atomic force microscopy scanning based on vision guidance according to claim 1, characterized in that: The nanopositioning platform includes a flexible nanopositioning platform and a common positioning platform, which are used to move in the X / Y direction on the sample surface; the flexible nanopositioning platform is located between the atomic force microscope and the common positioning platform, and the sample placed on the flexible nanopositioning platform can be observed from below through the common positioning platform with the middle hollowed out. The inverted microscope captures the image of the observed sample through a camera and transmits it to the visual servo control system on the computer. The XY movement of the flexible nanopositioning platform is controlled according to the actual trajectory image information and the planned trajectory, so that the atomic force microscope can scan according to the planned trajectory.

3. The method for high-speed atomic force microscopy scanning based on vision guidance according to claim 1, characterized in that: The Cassini oval is adopted and a translation function is added to make the atomic force microscope scanning area cover the sample surface. A time-dependent parameterized trajectory curve is obtained, the parameterized trajectory is planned in the image space, and then the time-dependent parameterized trajectory curve is discretized into discrete poses with the same time interval.

4. The method for high-speed atomic force microscopy scanning based on vision guidance according to claim 1, characterized in that: In step 2, the corresponding template area is cropped from the initial image. The template area moves along the expected trajectory in the image space. While discretizing the expected trajectory, the position of the template area in the image space is also discretized. The discrete pose of the discretized template area is used to replace the discrete pose of the discretized trajectory as the expected template.

5. The high-speed atomic force microscope scanning method based on vision guidance according to claim 2, characterized in that: The visual tracking controller uses a template matching algorithm based on regional grayscale features to establish the relationship between the position and posture of the flexible nanopositioning platform and the grayscale intensity of the image, converting the visual measurement process into a multivariable nonlinear optimization problem. The optimization and control of the template matching algorithm are achieved by combining the relationship between the position and posture of the flexible nanopositioning platform and the external matrix mapped to the image.

6. The method for high-speed atomic force microscopy scanning based on vision guidance according to claim 5, characterized in that: The feedback signal used by the visual tracking controller is the difference in dense photometric information between the expected template and the actual image template region.

7. The method for high-speed atomic force microscopy scanning based on vision guidance according to claim 1, characterized in that: The visual feedback actual module collects a grayscale image of the end position of the nanopositioning platform through a camera.

8. The method for high-speed atomic force microscopy scanning based on vision guidance according to claim 1, characterized in that: In the process of achieving full closed-loop control in step 6, inverse kinematics calculation is required to convert the desired posture into the required joint angles.

9. The method for high-speed atomic force microscopy scanning based on vision guidance according to claim 8, characterized in that: Inverse kinematics is the process of solving the angles or positions of each corresponding joint based on the desired position and posture of the end effector of the robotic arm, and is solved through numerical calculation methods.

10. The method for high-speed atomic force microscopy scanning based on vision guidance according to claim 9, characterized in that: After obtaining the required joint angle or position, the position and posture of the end effector of the nanopositioning platform are calculated through kinematics forward solution. Kinematics is the process of solving the position and posture of the end effector of the nanopositioning platform based on the joint angle or position of the nanopositioning platform.

Images