AFM thermal drift image correction method, system, medium and equipment based on vector machine

Abstract

The present invention discloses a method, system, medium and device for correcting thermal drift images of AFM based on vector machines, the method comprising: extracting feature points in a scanned image using Douglas-Peuker algorithm; establishing a sample image database; establishing an offset vector model based on p-value estimation; establishing an image verification vector machine using SVM algorithm; and correcting the thermal drift images of AFM using Lucy-Richardson algorithm. The present invention can complete the verification and correction of scanned images, promote the surface morphology detection system of AFM to perform high temperature operation experiments, and improve the nano-operation environment and efficiency of the surface morphology detection system of AFM.

Description

Technical Field

[0001] The present invention relates to the technical field of atomic force microscope imaging, and in particular to a vector machine-based AFM thermal drift image correction method, system, medium and equipment. Background Art

[0002] Atomic Force Microscope (AFM), as an atomic-level microscopic tool, is based on scanning tunneling microscopy. It can observe the morphology and structure of the surface of nano-sized polymer material samples and analyze the surface properties of materials. At the same time, it can measure the interaction between the AFM microcantilever probe and the atoms on the sample surface, such as interatomic repulsion, van der Waals force, friction, etc. Nowadays, AFM has gradually penetrated into all aspects of polymer research. However, due to the thermal drift of the image caused by temperature during the scanning process, the target object is constantly offset during the imaging process, and the resulting AFM image is severely distorted and blurred, which cannot reflect the true morphology of the target object and cannot complete the dynamic measurement of polymer materials in a heated environment.

[0003] Using digital image correlation technology to process AFM images has become a popular technology. Keller et al. proposed to model the probe morphology by rescanning the sample surface that has been accurately calibrated. Villarrubia et al. published several papers describing the method of establishing a probe tip model by scanning a sharply changing reference surface and using a blind modeling algorithm. Dongmo et al. used a blind modeling method to model the morphology of the three-coordinate measuring machine probe tip, and verified the effectiveness of the algorithm by comparing it with the probe tip image obtained by a scanning electron microscope through comparative experiments. Todd et al. analyzed the impact of noise data in the blind modeling algorithm, and based on this, proposed a blind modeling method to improve the impact of noise and eliminate noise, and then used this method to model the nanoscale probe tip.

[0004] The scanning accuracy of AFM is related to whether AFM can play a greater role in practical applications such as nano-device processing and manufacturing. Improving the accuracy of AFM scanning images is helpful to analyze AFM scanning characteristics, analyze and improve the problems encountered in the AFM scanning process, so that AFM can be more widely used in industrial production and manufacturing, and promote the development of nanotechnology. Summary of the invention

[0005] The purpose of the present invention is to overcome the shortcomings of the prior art and provide an AFM thermal drift image correction method, system, medium and equipment based on a vector machine. The Douglas-Peuker algorithm is used to extract feature points in the scanned image, and the SVM algorithm is used to construct an image calibration vector machine at 20 to 35°C for the surface morphology detection system of the AFM. The image calibration vector machine can identify the offset mode through the ambient temperature, complete the calibration and correction of the scanned image, promote the surface morphology detection system of the AFM to perform high-temperature operation experiments, and improve the nano-operation environment and efficiency of the surface morphology detection system of the AFM.

[0006] The first object of the present invention is to provide an AFM thermal drift image correction method based on a vector machine;

[0007] The second object of the present invention is to provide an AFM thermal drift image correction system based on a vector machine;

[0008] The third object of the present invention is to provide a storage medium;

[0009] A fourth object of the present invention is to provide a computer device.

[0010] In order to achieve the above first purpose, the present invention adopts the following technical solutions:

[0011] A method for correcting AFM thermal drift images based on a vector machine comprises the following steps:

[0012] S1. Collect surface feature images of material standard samples to obtain AFM scanning images, and use Douglas-Peuker algorithm to extract feature points in the scanning images;

[0013] S2, based on the feature points extracted in step S1, continuously scan the sample, and select the reference map and the deformation map to obtain a scanned sample image database;

[0014] S3, based on the scanned sample image database obtained in step S2, using p-value estimation to establish an offset vector model of the surface morphology detection system of the AFM;

[0015] S4, based on the offset vector model obtained in step S3, using the SVM algorithm to establish an image verification vector machine for the surface morphology detection system of the AFM;

[0016] S5. Based on the image calibration vector machine obtained in step S4, the Lucy-Richardson algorithm is used to correct the AFM thermal drift image depending on the relative position relationship between the reference and the deformation map.

[0017] Further, in step S1, the following operations are specifically performed:

[0018] S101, collecting surface feature images of material standard samples under set temperature difference conditions to obtain AFM scanning images;

[0019] S102: Using the Douglas-Peuker algorithm to extract features of the scanned image, determine feature points of the sample.

[0020] Further, in step S2, the following operations are specifically performed:

[0021] The sample is scanned continuously within a set time, and the temperature change curve is recorded. The AFM scanning image at any time is selected as a reference, and the AFM scanning images at other times are selected as deformation maps to obtain a scanned sample image database.

[0022] Further, in step S3, the following operations are specifically performed:

[0023] S301, using a first-order vector form to express a mapping relationship between changes in a feature point between a reference and a deformation map, where the mapping relationship is defined as a mapping function of an offset vector p;

[0024] The offset vector p (p0, p1, p2, p3, p4, p5) is a six-dimensional vector, where p0 and p1 represent the horizontal / vertical offsets of the center positions of the reference map and the deformation map, respectively, p2 and p4 represent the horizontal and vertical offsets of the row scan line, respectively, and p3 and p5 represent the horizontal and vertical offsets of the column scan line, respectively;

[0025] The reference and deformation map generate a mapping relationship about the offset vector p:

[0026] u(x,y,p)=p0+p2(x-x0)+p4(y-y0)

[0027] v(x,y,p)=p1+p3(x-x0)+p5(y-y0)

[0028] Where x and y represent the coordinates of the feature points in the reference image; x0 and y0 represent the coordinates of the center point of the reference image; u(x, y, p) and v(x, y, p) are displacement functions, which represent the displacement of the feature points in the deformation map and the reference image between the scanned images, respectively.

[0029] S302, using the maximum cross-correlation coefficient method, combining the reference map and the deformation map, calculating the similarity of the feature area, and establishing the offset vector model of the surface morphology detection system of the AFM:

[0030]

[0031] Where M and N represent the number of pixels in the x and y directions of the image respectively, g(x,y) represents the height value of point (x,y), and the mapping point height value of point (x,y) in the corrected image is h(x+u(x,y,p),y+v(x,y,p)).

[0032] Further, in step S4, the following operations are specifically performed:

[0033] S401, under different temperature conditions, continuously repeat step S2 and step S3 to establish an offset vector model library that accommodates different operating temperatures T i Offset vector model under conditions Where i = 1, 2, ..., n;

[0034] S402, based on the different operating temperatures T obtained in step S401 i Offset vector model under conditions The SVM algorithm is used to construct the image verification vector machine of the surface morphology detection system of AFM, and its framework is as follows:

[0035]

[0036]

[0037]

[0038] d i ≥0,ξ i ≥0

[0039] Among them, C is the penalty parameter, ξ is the slack variable, and d i is the weight of the offset vector at different temperatures, M is the number of temperature conditions, and the sum of the weights of the offset functions that satisfy various temperature conditions is 1; w, b are vector machine parameters;

[0040] S403, the solution is completed in two steps:

[0041] The first step is to solve the optimal value of the target value J(d) of the image calibration vector machine. At this time, the weight of the offset vector is considered to be unchanged, and the parameters w and b of the image calibration vector machine are solved. Then, the second step is to calculate the gradient of J(d) and solve the weight d of the offset vector. i , the two steps are iterated repeatedly until the optimal solution is found, that is, the calibration function f(x, y)| (x，y)∈p ,

[0042]

[0043] Further, in step S5, the following operations are specifically performed:

[0044] Apply the Lucy-Richardson algorithm, and after iterative convergence, find the calibration image with conditional probability that meets the requirements for the original image:

[0045]

[0046] Where r is the number of iterations; W r+1 (x, y) represents the calibration image after iteration r+1 times, W0(x, y) = h(x, y), h(x, y) represents the height value of the original image; f(x, y) represents the calibration image, when (x, y)∈p, the optimal calibration function f(x, y) of the image calibration vector machine obtained in step S3 is used| (x，y)∈p ; S(x, y) and S(-x, -y) both represent calibration functions; ξ is additive noise;

[0047] As the number of iterations increases, the calibration image W r+1 (x, y) converges according to probability, thus achieving the purpose of image correction.

[0048] In order to achieve the above second purpose, the present invention adopts the following technical solutions:

[0049] A vector machine-based AFM thermal drift image correction system, comprising:

[0050] The feature point extraction module collects surface feature images of material standard samples to obtain AFM scanned images, and uses the Douglas-Peuker algorithm to extract feature points in the scanned images;

[0051] The database module continuously scans the samples based on the feature points extracted by the feature point extraction module, and selects the image at the end of the scan as a reference, and the images at the remaining times as deformation maps, to obtain a scanned sample image database;

[0052] The offset vector model module uses p-value estimation to establish an offset vector model of the surface topography detection system of the AFM based on the scanned sample image database obtained by the database module;

[0053] An image verification vector machine module, based on the offset vector model obtained by the offset vector model building module, uses a SVM algorithm to build an image verification vector machine for the surface morphology detection system of the AFM;

[0054] The image correction module obtains the image correction vector machine based on the image correction vector machine module, adopts the Lucy-Richardson algorithm, and corrects the AFM thermal drift image based on the relative position relationship between the reference and the deformation map.

[0055] In order to achieve the third objective, the present invention adopts the following technical solutions:

[0056] A storage medium stores a program, and when the program is executed by a processor, the AFM thermal drift image correction method based on a vector machine is implemented.

[0057] In order to achieve the fourth objective, the present invention adopts the following technical solutions:

[0058] A computing device comprises a processor and a memory for storing a program executable by the processor, wherein when the processor executes the program stored in the memory, the above-mentioned AFM thermal drift image correction method based on a vector machine is implemented.

[0059] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0060] The present invention adopts the Douglas-Peuker algorithm to extract feature points in the scanned image, and adopts the SVM algorithm to construct an image calibration vector machine at 20-35°C for the surface morphology detection system of the AFM. The image calibration vector machine can identify the offset mode through the ambient temperature, complete the scanned image calibration correction, promote the surface morphology detection system of the AFM to perform high-temperature operation experiments, and improve the nanometer operation environment and efficiency of the surface morphology detection system of the AFM. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 Schematic diagram of the process of AFM thermal drift image correction method based on vector machine.

[0062] Figure 2 The scanned image is at 21.4°C.

[0063] Figure 3 The image is scanned at 22.7°C.

[0064] Figure 4 Schematic diagram for selecting feature points.

[0065] Figure 5 It is a continuous scan image within a certain period of time.

[0066] Figure 6 The images are continuously scanned at 20°C.

[0067] Figure 7 The images are continuous scans at 30°C.

[0068] Figure 8 Schematic diagram of the calibration image.

[0069] Fig. 9 This is the structural block diagram of the AFM thermal drift image correction system based on vector machine. DETAILED DESCRIPTION

[0070] The present invention is further described below in conjunction with specific embodiments, but the use of the present invention is not limited thereto.

[0071] Example 1

[0072] like Figure 1 As shown, this embodiment provides an AFM thermal drift image correction method based on a vector machine, comprising the following steps:

[0073] S1. Collect surface feature images of material standard samples to obtain AFM scanning images, and use Douglas-Peuker algorithm to extract feature points in the scanning images;

[0074] S2, based on the feature points extracted in step S1, continuously scan the sample, and select the reference map and the deformation map to obtain a scanned sample image database;

[0075] S3, based on the scanned sample image database obtained in step S2, using p-value estimation to establish an offset vector model of the surface morphology detection system of the AFM;

[0076] S4, based on the offset vector model obtained in step S3, using the SVM algorithm to establish an image verification vector machine for the surface morphology detection system of the AFM;

[0077] S5. Based on the image calibration vector machine obtained in step S4, the Lucy-Richardson algorithm is used to correct the AFM thermal drift image depending on the relative position relationship between the reference and the deformation map.

[0078] Specifically, in step S1, the following operations are performed:

[0079] S101, turn on the AFM console, and collect surface feature images of a material standard sample (such as graphene) under a temperature difference of 2 to 5°C to obtain an AFM scanning image, such as Figure 2 , Figure 3 As shown, they are the 21.4℃ scanning image and the 22.7℃ scanning image, where: Figure 3 Scan the image for the termination moment;

[0080] S102, using Douglas-Peuker algorithm to extract features of the scanned image and determine the feature points of the sample, such as Figure 4 shown.

[0081] Furthermore, in step S2, the following operations are specifically performed:

[0082] Scan the sample continuously within a certain period of time and record the temperature change curve, such as Figure 5As shown, an AFM scanning image at a certain moment is selected as a reference map, and AFM scanning images at other moments are selected as deformation maps to obtain a scanning sample image database.

[0083] Furthermore, in step S3, the following operations are specifically performed:

[0084] S301, using a first-order vector form to express a mapping relationship between changes in a feature point between a reference and a deformation map, where the mapping relationship is defined as a mapping function of an offset vector p;

[0085] The offset vector p (p0, p1, p2, p3, p4, p5) is a six-dimensional vector, where p0 and p1 represent the horizontal / vertical offsets of the center positions of the reference map and the deformation map, respectively, p2 and p4 represent the horizontal and vertical offsets of the row scan line, respectively, and p3 and p5 represent the horizontal and vertical offsets of the column scan line, respectively;

[0086] The reference and deformation map generate a mapping relationship about the offset vector p:

[0087] u(x,y,p)=p0+p2(x-x0)+p4(y-y0)

[0088] v(x,y,p)=p1+p3(x-x0)+p5(y-y0)

[0089] Where x and y represent the coordinates of the feature points in the reference image; x0 and y0 represent the coordinates of the center point of the reference image; u(x, y, p) and v(x, y, p) are displacement functions, which represent the displacement of the feature points in the deformation map and the reference image between the scanned images, respectively.

[0090] S302, using the maximum cross-correlation coefficient method, combining the reference map and the deformation map, calculating the similarity of the feature area, and establishing the offset vector model of the surface morphology detection system of the AFM:

[0091]

[0092] Where M and N represent the number of pixels in the x and y directions of the image respectively, g(x,y) represents the height value of point (x,y), and the mapping point height value of point (x,y) in the corrected image is h(x+u(x,y,p),y+v(x,y,p)).

[0093] Furthermore, in step S4, the following operations are specifically performed:

[0094] S401, under different temperature conditions, continuously repeat step S2 and step S3, such as Figure 6 , Figure 7 As shown, an offset vector model library is established to accommodate different operating temperatures T iOffset vector model under conditions i=1,2,...,n;

[0095] S402, based on the different operating temperatures T obtained in step S401 i Offset vector model under conditions The SVM algorithm is used to construct the image verification vector machine of the surface morphology detection system of AFM, and its framework is as follows:

[0096]

[0097]

[0098]

[0099] d i ≥0,ξ i ≥0

[0100] Among them, C is the penalty parameter, ξ is the slack variable, and d i is the weight of the offset vector at different temperatures, M is the number of temperature conditions, and the sum of the weights of the offset functions that satisfy various temperature conditions is 1; w, b are vector machine parameters;

[0101] S403, the solution is completed in two steps:

[0102] The first step is to solve the optimal value of the target value J(d) of the image calibration vector machine. At this time, the weight of the offset vector is considered to be unchanged, and the parameters w and b of the image calibration vector machine are solved. Then, the second step is to calculate the gradient of J(d) and solve the weight d of the offset vector. i , the two steps are iterated repeatedly until the optimal solution is found, that is, the calibration function f(x, y)| (x，y)∈p ,

[0103]

[0104] Furthermore, in step S5, the following operations are specifically performed:

[0105] Apply the Lucy-Richardson algorithm, and after iterative convergence, find the calibration image with conditional probability that meets the requirements for the original image:

[0106]

[0107] Where r is the number of iterations; W r+1(x, y) represents the calibration image after iteration r+1 times, W0(x, y) = h(x, y), h(x, y) represents the height value of the original image; f(x, y) represents the calibration image, when (x, y)∈p, the optimal calibration function f(x, y) of the image calibration vector machine obtained in step S3 is used| (x，y)∈p ; S(x, y) and S(-x, -y) both represent calibration functions; ξ is additive noise;

[0108] As the number of iterations increases, the calibration image W r+1 (x, y) converges according to probability, thus achieving the purpose of image correction, such as Figure 8 shown.

[0109] Example 2

[0110] like Fig. 9 As shown, this embodiment provides an AFM thermal drift image correction system based on a vector machine, comprising:

[0111] The feature point extraction module collects surface feature images of material standard samples to obtain AFM scanned images, and uses the Douglas-Peuker algorithm to extract feature points in the scanned images;

[0112] The database module continuously scans the samples based on the feature points extracted by the feature point extraction module, and selects the image at the end of the scan as a reference, and the images at the remaining times as deformation maps, to obtain a scanned sample image database;

[0113] The offset vector model module uses p-value estimation to establish an offset vector model of the surface topography detection system of the AFM based on the scanned sample image database obtained by the database module;

[0114] An image verification vector machine module, based on the offset vector model obtained by the offset vector model building module, uses a SVM algorithm to build an image verification vector machine for the surface morphology detection system of the AFM;

[0115] The image correction module obtains the image correction vector machine based on the image correction vector machine module, adopts the Lucy-Richardson algorithm, and corrects the AFM thermal drift image based on the relative position relationship between the reference and the deformation map.

[0116] Example 3

[0117] This embodiment provides a storage medium, which may be a ROM, RAM, disk, CD or other storage medium, and the storage medium stores one or more programs, which, when executed by a processor, implements the AFM thermal drift image correction method based on a vector machine of the above-mentioned embodiment 1.

[0118] Example 4

[0119] This embodiment provides a computing device, which can be a desktop computer, a laptop computer, a smart phone, a PDA handheld terminal, a tablet computer or other terminal device with a display function. The computing device includes a processor and a memory, and the memory stores one or more programs. When the processor executes the program stored in the memory, the AFM thermal drift image correction method based on the vector machine of the above-mentioned embodiment 1 is implemented.

[0120] In summary, the present invention adopts the Douglas-Peuker algorithm to extract feature points in the scanned image, and adopts the SVM algorithm to construct an image calibration vector machine at 20-35°C for the surface morphology detection system of the AFM. The image calibration vector machine can identify the offset mode through the ambient temperature, complete the scanned image calibration correction, promote the surface morphology detection system of the AFM to perform high-temperature operation experiments, and improve the nano-operation environment and efficiency of the surface morphology detection system of the AFM.

[0121] The embodiments described above are only preferred embodiments of the present invention and are not intended to limit the scope of implementation of the present invention. Therefore, all changes made according to the shape and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for correcting AFM thermal drift images based on a vector machine, characterized in that: The following steps are included: S1. Collect surface feature images of standard material samples to obtain AFM scanned images, and use Douglas-Peuker algorithm to extract feature points in the scanned images; S2, based on the feature points extracted in step S1, continuously scan the sample, and select the reference map and the deformation map to obtain a scanned sample image database; S3, based on the scanned sample image database obtained in step S2, using p-value estimation to establish an offset vector model of the surface topography detection system of the AFM; S4, based on the offset vector model obtained in step S3, using the SVM algorithm to establish an image verification vector machine for the surface morphology detection system of the AFM; S5. Based on the image calibration vector machine obtained in step S4, the Lucy-Richardson algorithm is used to correct the AFM thermal drift image depending on the relative position relationship between the reference and the deformation map.

2. The AFM thermal drift image correction method based on vector machine according to claim 1, characterized in that: In step S1, the following operations are specifically performed: S101, collecting surface feature images of material standard samples under set temperature difference conditions to obtain AFM scanning images; S102: Using the Douglas-Peuker algorithm to extract features of the scanned image, determine feature points of the sample.

3. The AFM thermal drift image correction method based on vector machine according to claim 1, characterized in that: In step S2, the following operations are specifically performed: The sample is scanned continuously within a set time, and the temperature change curve is recorded. The AFM scanning image at any time is selected as a reference, and the AFM scanning images at other times are selected as deformation maps to obtain a scanned sample image database.

4. The AFM thermal drift image correction method based on vector machine according to claim 1, characterized in that: In step S3, the following operations are specifically performed: S301, using a first-order vector form to express a mapping relationship between changes in a feature point between a reference and a deformation map, where the mapping relationship is defined as a mapping function of an offset vector p; The offset vector p (p0, p1, p2, p3, p4, p5) is a six-dimensional vector, where p0 and p1 represent the horizontal / vertical offsets of the center positions of the reference map and the deformation map, respectively, p2 and p4 represent the horizontal and vertical offsets of the row scan line, respectively, and p3 and p5 represent the horizontal and vertical offsets of the column scan line, respectively; The reference and deformation map generate a mapping relationship about the offset vector p: u(x,y,p)=p0+p2(x-x0)+p4(y-y0) v(x,y,p)=p1+p3(x-x0)+p5(y-y0) Where x and y represent the coordinates of the feature points in the reference image; x0 and y0 represent the coordinates of the center point of the reference image; u(x, y, p) and v(x, y, p) are displacement functions, which represent the displacement of the feature points in the deformation map and the reference image between the scanned images, respectively. S302, using the maximum cross-correlation coefficient method, combining the reference map and the deformation map, calculating the similarity of the feature area, and establishing the offset vector model of the surface morphology detection system of the AFM: Where M and N represent the number of pixels in the x and y directions of the image respectively, g(x,y) represents the height value of point (x,y), and the mapping point height value of point (x,y) in the corrected image is h(x+u(x,y,p),y+v(x,y,p)).

5. The AFM thermal drift image correction method based on vector machine according to claim 1, characterized in that: In step S4, the following operations are specifically performed: S401, under different temperature conditions, continuously repeat step S2 and step S3 to establish an offset vector model library that accommodates different operating temperatures T i Migration vector model under conditions Where, i = 1, 2, ..., n; S402, based on the different operating temperatures T obtained in step S401 i Offset vector model under conditions The SVM algorithm is used to construct the image verification vector machine of the surface morphology detection system of AFM, and its framework is as follows: d i ≥0.ξ i ≥0 Among them, C is the penalty parameter, ξ is the slack variable, and d i is the weight of the offset vector at different temperatures, M is the number of temperature conditions, and the sum of the weights of the offset functions that satisfy various temperature conditions is 1; w, b are vector machine parameters; S403, the solution is completed in two steps: The first step is to solve the optimal value of the target value J(d) of the image calibration vector machine. At this time, the weight of the offset vector is considered to be unchanged, and the parameters w and b of the image calibration vector machine are solved. Then, the second step is to calculate the gradient of J(d) and solve the weight d of the offset vector. i , the two steps are iterated repeatedly until the optimal solution is found, that is, the calibration function f(x,y)| (x,y)∈p , 6. The AFM thermal drift image correction method based on vector machine according to claim 1, characterized in that: In step S5, the following operations are specifically performed: Apply the Lucy-Richardson algorithm, and after iterative convergence, find the calibration image with conditional probability that meets the requirements for the original image: Where r is the number of iterations; W r+1 (x, y) represents the calibration image after iteration r+1 times, W0(x, y) = h(x, y), h(x, y) represents the height value of the original image; f(x, y) represents the calibration image, when (x, y)∈p, the optimal calibration function f(x, y) of the image calibration vector machine obtained in step S3 is used| (x，y)∈p ; S(x,y) and S(-x,-y) both represent calibration functions; ξ is additive noise; As the number of iterations increases, the calibration image W r+1 (x,y) converges according to probability, thus achieving the purpose of image correction.

7. An AFM thermal drift image correction system based on a vector machine, characterized in that: include, The feature point extraction module collects surface feature images of material standard samples to obtain AFM scanned images, and uses the Douglas-Peuker algorithm to extract feature points in the scanned images; The database module continuously scans the samples based on the feature points extracted by the feature point extraction module, and selects the image at the end of the scan as a reference, and the images at the remaining times as deformation maps, to obtain a scanned sample image database; The offset vector model module uses p-value estimation to establish an offset vector model of the surface topography detection system of the AFM based on the scanned sample image database obtained by the database module; An image verification vector machine module, based on the offset vector model obtained by the offset vector model establishment module, uses a SVM algorithm to establish an image verification vector machine for the surface morphology detection system of the AFM; The image correction module obtains the image correction vector machine based on the image correction vector machine module, adopts the Lucy-Richardson algorithm, and corrects the AFM thermal drift image based on the relative position relationship between the reference and the deformation map.

8. A storage medium storing a program, characterized in that: When the program is executed by a processor, the AFM thermal drift image correction method based on a vector machine as described in any one of claims 1 to 6 is implemented.

9. A computing device comprising a processor and a memory for storing a program executable by the processor, characterized in that: When the processor executes the program stored in the memory, the AFM thermal drift image correction method based on a vector machine as described in any one of claims 1 to 6 is implemented.

Images