Image processing method, program, image processing device, and scanning probe microscope

Abstract

An image processing method according to the present disclosure is a method for processing a target image generated on the basis of measuring a sample by a scanning probe microscope, the method comprising: a step (S12) for estimating, by an algorithm for calculating a regression line by excluding an outlier from data including the outlier, a first background image of a target image, the first background image arising due to inclination between a plane on which the sample is set and a predetermined plane, during measurement; and a step (S14) for generating a first corrected image in which the inclination of the target image is corrected on the basis of the first background image.

Description

Image processing method, program, image processing device, and scanning probe microscope

[0001] The present disclosure relates to an image processing method, a program, an image processing device, and a scanning probe microscope, and more particularly to processing of images obtained by measurement using a scanning probe microscope.

[0002] A scanning probe microscope (SPM) observes the topography of a sample surface with high resolution by bringing a sufficiently sharpened probe close enough to the sample to be observed and scanning the sample surface horizontally with the probe while adjusting the height of the probe so that the physical quantities acting on the tip of the probe and the surface of the sample remain constant. SPM is a general term for microscopes that observe the topography of a sample surface using the above-mentioned operating principle. Representative SPMs include a scanning tunneling microscope (STM), which detects the current flowing between the probe and the sample as an interaction, and an atomic force microscope (AFM), which detects the atomic force acting between the probe and the sample as an interaction.

[0003] In measurements using an SPM, the resolution in the surface height direction is high, and it is difficult to set the sample surface horizontally at that resolution level. Therefore, tilt correction is generally performed on the measurement data acquired by the SPM to correct the inclined surface so that it is horizontal. By providing an image created based on the corrected measurement data, the user can more accurately recognize the surface condition of the sample.

[0004] As a correction process for images obtained by such an SPM, Japanese Patent No. 6627903 (Patent Document 1) discloses a technique in which at least a portion of the area other than the edge in the image data is extracted as a reference plane area, and the height of the measurement data is corrected based on height information of three points belonging to the reference plane area. Japanese Patent Laid-Open Publication No. 2009-58479 (Patent Document 2) discloses a technique in which errors in the measurement data caused by changes in the measurement environment are corrected by equalizing the average height of all main scan lines in the image data. Japanese Patent Laid-Open Publication No. 11-142416 (Patent Document 3) discloses a technique in which measurement data is corrected based on a histogram of the height distribution obtained from the measurement data, thereby obtaining measurement results with less distortion.

[0005] Patent No. 6627903 JP 2009-58479 JP 11-142416

[0006] In conventional image correction processes, the gradation values ​​of pixels in areas corresponding to the unevenness of the sample surface are equalized with the gradation values ​​of pixels in other areas of the sample, which can make it impossible to measure the height of the surface structure of the sample relative to the substrate surface with sufficient accuracy from the corrected image. When sufficient accuracy cannot be obtained through the correction process, the information on the surface structure of the sample obtained from the corrected image can deviate from the actual surface structure of the sample. Therefore, there is a need for a correction process method that can measure the height of the surface structure of the sample relative to the substrate surface with sufficient accuracy in an image of SPM measurement data acquired with the sample tilted.

[0007] The present disclosure has been made in view of the above circumstances, and its purpose is to provide a correction processing technique for improving the accuracy of images generated based on SPM measurements.

[0008] A first aspect of the present disclosure is a method for processing an object image generated based on measurement of a sample using a scanning probe microscope, comprising the steps of: estimating a first background image of the object image generated due to the tilt between a plane on which the sample is placed and a predetermined plane during measurement using an algorithm that calculates a regression line by excluding outliers from data that includes outliers; and generating a first corrected image in which the tilt of the object image has been corrected based on the first background image.

[0009] A second aspect of the present disclosure is a program that, when executed by one or more processors, causes the one or more processors to perform the image processing method described in the first aspect.

[0010] A third aspect of the present disclosure is an image processing device comprising one or more processors and a storage device storing a program that, when executed by the one or more processors, causes the one or more processors to implement the image processing method described in the first aspect.

[0011] a scanning probe microscope including: a sample stage on which a sample is placed; a cantilever having a probe at its tip; a moving device that changes the distance between the sample stage and the probe; an optical system that irradiates the cantilever with laser light and detects the laser light reflected by the cantilever; a measurement unit that measures the height of the sample based on the displacement of the cantilever obtained by a change in the position of the laser light detected by the optical system; an image generation unit that generates an image of the target based on the measurement; an image processing unit that executes a correction process for the image of the target; and a display unit that displays the corrected image obtained by correcting the image of the target corrected by the image processing unit, wherein the correction process includes the steps of: estimating a background image of the target image generated by the tilt between a plane on which the sample is placed and a predetermined plane during measurement using an algorithm that calculates a regression line by removing outliers from data that includes outliers; and generating the corrected image in which the tilt of the target image has been corrected based on the background image.

[0012] According to the present disclosure, correction processing techniques are provided to improve the accuracy of images generated based on SPM measurements.

[0013] FIG. 1 is a schematic configuration diagram of a scanning probe microscope according to an embodiment. FIG. 2 is an example of an image of a sample generated based on measurement by the scanning probe microscope. FIG. 3 is an example of a first background image estimated based on the image shown in FIG. 2. FIG. 4 is an example of a first corrected image obtained by performing a first correction process on the image shown in FIG. 2. FIG. 5 is a flowchart of an example of a first correction process performed in a scanning probe microscope. FIG. 6 is an example of a first corrected image different from the first corrected image shown in FIG. 4. FIG. 7 is an example of a second background image estimated based on the image shown in FIG. 6. FIG. 8 is an example of a second corrected image obtained by performing a second correction process on the image shown in FIG. 6. FIG. 9 is a flowchart of an example of a second correction process performed in a scanning probe microscope.

[0014] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail with reference to the accompanying drawings, in which the same or corresponding parts are designated by the same reference numerals and will not be described repeatedly.

[0015] [Schematic Configuration of Scanning Probe Microscope] Fig. 1 is a schematic configuration diagram of a scanning probe microscope according to an embodiment. An example of a scanning probe microscope is an AFM. However, the scanning probe microscope may be another type of scanning probe microscope (for example, an STM).

[0016] 1, the scanning probe microscope 1 includes a piezoelectric scanner 111, a sample stage 112, a cantilever 113, a displacement detection mechanism 120, a feedback signal generator 131, a computer 132, a scanning signal generator 133, a storage device 134, and a display unit 135. The scanning probe microscope 1 measures the shape of the surface of a sample 110 placed on the sample stage 112. In Fig. 1, the height direction of the sample 110 is defined as the Z-axis direction, and directions perpendicular to the Z-axis direction are defined as the X-axis direction and the Y-axis direction.

[0017] The sample 110 is placed on a sample stage 112. The sample stage 112 is placed on a piezoelectric scanner 111. The piezoelectric scanner 111 is a moving device for changing the relative positional relationship between the sample 110 and a probe 114 provided at the tip of a cantilever 113. The piezoelectric scanner 111 includes an XY scanner 111xy and a Z scanner 111z. The XY scanner 111xy has a piezoelectric element that deforms based on voltage values ​​Vx and Vy applied from a scan signal generating unit 133. The XY scanner 111xy moves the sample stage 112 in the X-axis and Y-axis directions using this piezoelectric element. The Z scanner 111z has a piezoelectric element that deforms based on a voltage value Vz applied from a feedback signal generating unit. The Z scanner 111z moves the sample stage 112 in the Z-axis direction using this piezoelectric element. Note that the XY scanner 111xy and the Z scanner 111z are not limited to having piezoelectric elements.

[0018] The cantilever 113 is formed in the shape of a leaf spring. The free end of the cantilever 113 is arranged to face the sample 110. In the example shown in FIG. 1 , the cantilever 113 is arranged above the sample 110 in the Z-axis direction. The cantilever 113 has a front surface facing the sample 110 and a back surface opposite the front surface. A probe 114 is arranged on the surface of the tip of the free end of the cantilever 113 so as to face the sample 110. The back surface of the tip is configured to reflect light. The tip of the cantilever 113 is displaced in the Z-axis direction due to a physical quantity (e.g., atomic force) acting between the probe 114 and the sample 110.

[0019] Displacement detection mechanism 120 includes laser diode 115 and photodetector 116. Displacement detection mechanism 120 detects the displacement of cantilever 113. In scanning probe microscope 1, when the tip of probe 114 is brought close to sample 110 to observe the surface, laser light emitted from laser diode 115 is reflected by the back surface of cantilever 113, and the reflected light is received by photodetector 116. As probe 114 is brought closer to the surface of sample 110, cantilever 113 bends like a leaf spring, changing the reflection direction of the reflected light. Therefore, the amount of bending of cantilever 113 can be detected by observing the light-receiving position of photodetector 116.

[0020] The feedback signal generating unit 131 receives a detection signal from the photodetector 116 and calculates the amount of deflection of the cantilever 113 based on the detection signal. The feedback signal generating unit 131 controls the Z-direction position of the sample 110 so that the atomic force between the probe 114 and the surface of the sample 110 is always constant. The feedback signal generating unit 131 calculates a voltage value Vz of the control command value in the Z-axis direction for the piezo scanner 111 based on the amount of deflection of the cantilever 113 and outputs it to the Z scanner 111z.

[0021] The scanning signal generating unit 133 calculates voltage values ​​Vx and Vy of control command values ​​in the X-axis and Y-axis directions, respectively, so that the sample 110 moves relative to the probe 114 in the XY plane according to a predetermined scanning pattern, and outputs them to the XY scanner 111xy.

[0022] A signal reflecting the feedback amount in the Z-axis direction (i.e., the voltage value Vz, which is the voltage applied to the scanner, and the deviation signal Sd) is also sent to the computer 132 and stored in the storage device 134. The computer 132 calculates the amount of surface displacement due to the unevenness of the sample 110 from the voltage value Vz based on correlation information previously stored in the storage device 134, which indicates the relationship between the voltage value Vz and the corresponding amount of surface displacement due to the unevenness of the sample 110. The calculated amount of displacement reflects a value indicating the position of the sample 110 in the Z-axis direction. The computer 132 creates measurement data representing the shape of the surface of the sample 110 by calculating the amount of displacement of the sample 110 in the Z-axis direction at each position in the X-axis and Y-axis directions within the scanning range. The measurement data is matrix data of M×N pieces of height information (amount of displacement in the Z-axis direction) for a predetermined rectangular area on the surface of the sample 110. The row and column directions in the recorded matrix data generally correspond to either the main scanning direction or the sub-scanning direction of the probe 114 of the scanning probe microscope 1. Based on the measurement data, the computer 132 displays information such as the shape of the surface of the sample 110 on the display unit 135. For example, the shape of the surface of the sample 110 is displayed by an image in which the amount of displacement in the Z-axis direction is expressed by grayscale values.

[0023] The computer 132 includes at least one processor, and the storage device 134 stores programs executed by the processor in a non-volatile (non-temporary) manner. The computer 132 can read out the measurement data stored in the storage device 134 as needed and display it on the display unit 135. Furthermore, the computer 132 can correct the measurement data as needed and display it on the display unit 135.

[0024] The computer 132 can perform at least two types of corrections, a first correction and a second correction, for the image obtained by SPM measurement. The first correction is a correction for reducing the effect on the gradation values ​​of the image obtained by SPM measurement due to the tilt between a predetermined plane and the plane on which the sample is actually placed during measurement with the SPM. The second correction is a correction for correcting the difference in gradation values ​​between main scan lines in the image data obtained by the first correction.

[0025] In this embodiment, the predetermined plane is assumed to be a plane on which the user desires to place the sample 110 during SPM measurement. For example, it is a plane parallel to the surface of the sample stage 112. Furthermore, if the user intends to measure the surface structure of the sample based on an image obtained by SPM measurement, the predetermined plane may be a plane parallel to the flat substrate surface of the sample.

[0026] The sample 110 to be measured is, for example, a substrate on which a structure is arranged. An example of a substrate is a mica plate. An example of a structure is nanoparticles or nanofibers of a biological sample. Note that these are merely examples, and the samples that the scanning probe microscope 1 targets are not limited to these. For example, the structure is not limited to structures intentionally arranged on a substrate, and the structure may also refer to unevenness on a substrate. Furthermore, the sample may not include a substrate, but may instead be nanoparticles and nanofibers.

[0027] [Correction of Measured Image According to Comparative Example] When a sample having a structure arranged on a horizontal substrate is measured using an SPM, if the substrate surface of the sample is not placed on a predetermined plane but is placed at an angle, the tilt can cause differences in cantilever displacement on the substrate surface that is on the same plane. In an image (hereinafter referred to as the original image) in which the measured values ​​obtained by such a measurement are used as gradation values ​​without correction, the gradation values ​​of the region corresponding to the substrate of the sample that is on the same plane will not be the same.

[0028] In such cases, the displacement in the Z-axis direction caused by the tilt of the sample may be greater than the unevenness of the sample surface, and the user may not be able to correctly recognize the structure of the sample surface from the original image.

[0029] Since SPM measurements have high resolution in the surface height direction, it is difficult to place the sample substrate surface horizontally on the sample stage at that resolution level. Therefore, a method is generally used in which the tilt of the SPM measurement is estimated relative to the original image and the gradation values ​​are corrected.

[0030] As a correction processing method according to a comparative example, for example, a method has been used in which a plane is estimated from three arbitrary points specified on an original image, and the original image is corrected based on the estimated plane. However, in this correction, the estimated plane differs depending on how the three points on the original image are selected, so even if correction is performed using the same original image, the resulting image may differ depending on the correction processing, and the reproducibility of the correction processing may be low.

[0031] On the other hand, to improve the reproducibility of the correction process, a correction process can be performed in which the correction is automatically performed using the average or median of the gradation values ​​of pixels along the X-axis or Y-axis direction of the original image. However, in this correction, the pixels in the area corresponding to the structure on the sample and the area corresponding to the substrate surface are used as data when calculating the above-mentioned average or median without being distinguished, and the gradation values ​​derived from the structure on the sample and the gradation values ​​derived from the substrate surface may be averaged. If the accuracy of the corrected image is not sufficient, the information on the surface structure of the sample obtained from the corrected image may differ from the actual surface structure of the sample.

[0032] For example, in product management using SPM measurements, it is necessary to improve the reproducibility of the correction process for images obtained by measurement in order to control the quality of the same product in the same way. Also, when semiconductors, whose sample surface shape dimensions are on the scale of angstroms or less, are the subject of product management, it is necessary to improve the measurement accuracy of the surface shape in order to maintain product quality, and it is also necessary to improve the accuracy of the corrected images.

[0033] [Correction of Measured Image According to the Embodiment] Therefore, the first correction according to the present embodiment uses an algorithm that calculates a regression equation by excluding outliers from data that includes outliers. This algorithm estimates a background image (hereinafter referred to as a first background image) in the original image, which is generated by the inclination between a predetermined plane and the plane on which the sample is actually placed. A first-corrected image is created based on the estimated first background image and the original image. In the first correction, the same correction process is performed on both original images, improving the reproducibility of the correction process. Furthermore, in the first correction, the background image is estimated by excluding pixels in the original image corresponding to structures on the sample as outliers. This prevents the gradation values ​​derived from the structures on the sample and the substrate surface from being averaged during the correction process, thereby improving the accuracy of the image obtained after the correction process. This allows the user to recognize the surface condition of the sample from the first-corrected image. In this specification, the original image is sometimes referred to as a target image, and the image after the first correction is sometimes referred to as a first-corrected image. The first background image is an image corresponding to the variation in the gradation value of each pixel caused by the tilt between a predetermined plane in the original image and the plane on which the sample is actually placed, and the first background image allows the user to visually recognize how the predetermined plane in the original image is tilted from the plane on which the sample is actually placed.

[0034] [First Correction] In one implementation example, the computer 132 estimates a first background image caused by the tilt between a predetermined plane and the plane on which the sample 110 is actually placed, and generates a first corrected image in which the tilt has been corrected based on the estimated first background image and the original image.

[0035] FIG. 2 is an example of an original image of a sample 110 generated based on measurements by the scanning probe microscope 1. In the sample 110 corresponding to the image IM10 shown in FIG. 2, four structures P1 to P4 are arranged on a flat substrate. The four structures are arranged two-dimensionally, in two rows in the X-axis direction and two rows in the Y-axis direction, i.e., a 2x2 array. An example of the substrate is a mica plate. An example of the structures is nanoparticles or nanofibers. Note that these are merely examples, and the samples (substrates, structures) that the scanning probe microscope 1 targets are not limited to these.

[0036] The gradation value of a pixel in image IM10 represents the height in the Z-axis direction on the sample 110. In other words, if there is a first pixel corresponding to a certain area on the sample and a second pixel different from the first pixel, and the gradation value of the second pixel is larger than the gradation value corresponding to the first pixel, this indicates that the area on the sample corresponding to the second pixel is located relatively lower in the Z-axis direction than the area on the sample corresponding to the first pixel.

[0037] In image IM10, the substrate of sample 110 is flat, so it is desirable that the gradation values ​​of pixels corresponding to the substrate portion in image IM10 are uniform so that the user can recognize the structure of sample 110. However, in image IM10, the gradation values ​​of the region corresponding to the substrate (the region other than the region in image IM10 where structures P1 to P4 are located) are non-uniform. One reason for this is that the measurement was performed in a state where the predetermined plane and the plane on which the sample was actually placed were not parallel to each other but tilted.

[0038] The first correction process will be described using image IM10. In the first correction process, a first background image is estimated by an algorithm used for the first correction, based on the gradation values ​​corresponding to each pixel of image IM10, which is the original image.

[0039] The first correction uses an algorithm that calculates a regression line by excluding outliers from data that includes outliers. Examples of algorithms that calculate a regression line by excluding outliers from data that includes outliers include RANSAC (RANdom Sample Consensus), Quantile regression, and Theil-Sen regression. The algorithm used for the first correction is not limited to the above-mentioned example, and any algorithm that calculates a regression line by excluding outliers from data that includes outliers may be used. In correction using this algorithm, the parameters used for correction are not changed for each correction process, so that all original images are processed in the same way, improving the reproducibility of the correction process. Furthermore, by using an algorithm that calculates a regression line by excluding outliers from data that includes outliers, structures on the sample substrate in the original image are excluded as outliers from the data used to calculate the regression equation. This allows a background image to be estimated from data corresponding to the sample substrate surface, in which the sample tilt is significantly reflected in the gradation value, thereby preventing portions corresponding to the surface structure from being smoothed out with portions other than the sample surface structure. Therefore, the accuracy of measuring the height of a structure on the substrate surface of a sample relative to the substrate surface can be improved. The gradation values ​​may be adjusted by adding, subtracting, multiplying, or dividing a predetermined value to the gradation values ​​of all pixels included in the first background image. The gradation values ​​may be adjusted so that the average or median of all gradation values ​​becomes an appropriate value while maintaining the magnitude relationship of the gradation values ​​of each pixel, or so that the maximum and minimum values ​​of the pixels included in the image become appropriate values. Adjusting the gradation values ​​improves the visibility of the image for the user.

[0040] Fig. 3 is a diagram showing a first background image estimated based on image IM10. Image IM20 shown in Fig. 3 is obtained in the process of performing a first correction process on image IM10.

[0041] In image IM20, the gradation value increases in the positive direction on the X and Y axes. Therefore, it is estimated from image IM20 that, during the SPM measurement, the sample 110 was placed so that the positive directions of the X and Y axes were lower in the height direction relative to a predetermined plane.

[0042] The algorithm used for the first correction can improve the accuracy of the image after the first correction by adjusting variables according to the characteristics of the sample. Examples of the characteristics of the sample include the ratio of the structure to the substrate surface in the sample, the particle diameter, and the height of the structure relative to the substrate surface. Examples of variables of the algorithm include the number of trials, the residual threshold, and the order of the plane. Preferred values ​​of the above variables are determined by theory and / or experiment.

[0043] The number of trials of the algorithm used for the first correction is the number of times random sampling is repeated, and is, for example, 3,000. The number of trials may be more or less than 3,000. In general, increasing the number of trials improves the accuracy of the image after the first correction, but may increase the amount of calculation and slow the processing speed. On the other hand, decreasing the number of trials reduces the amount of calculation and speeds up the processing speed, but may decrease the accuracy of the image after the first correction. Therefore, the user selects the number of trials as appropriate based on the relationship between the image accuracy desired by the user and the processing time required for correction.

[0044] The residual threshold of the algorithm used for the first correction indicates the maximum residual for a data sample to be classified as inlier, with larger values ​​increasing the range of samples that the algorithm will accept as inliers. The residual threshold may be, for example, 10, but may be greater or less than 10.

[0045] The order of the plane of the algorithm used for the first correction can be changed by the user, and is, for example, 2 or 3. The order represents the order of the fitting function, and may be a value greater than 1 or 3.

[0046] Based on the first background image estimated by the algorithm used for the first correction, a first corrected image is generated, which is an image obtained by correcting the original image. Specifically, the gradation value of a pixel in the first background image located at the same coordinates on the matrix as the pixel in the original image is subtracted from the gradation value of the pixel in the original image to obtain the gradation value of the pixel at that coordinate in the first corrected image. Alternatively, the gradation value of the pixel at that coordinate in the first corrected image may be obtained by dividing the gradation value of the pixel in the original image by the gradation value of the pixel in the first background image located at the same coordinates on the matrix as the pixel in the original image, and multiplying the obtained value by a predetermined value.

[0047] The gradation values ​​of all pixels included in the image after the first correction may be adjusted by adding, subtracting, multiplying, or dividing a predetermined value. The gradation values ​​may be adjusted so that the average or median of all gradation values ​​becomes an appropriate value while maintaining the magnitude relationship of the gradation values ​​of each pixel, or so that the maximum and minimum values ​​of the pixels included in the image become appropriate values. Adjusting the gradation values ​​improves the visibility of the image for the user.

[0048] Fig. 4 is a diagram showing an image after the first correction. In the image IM30 shown in Fig. 4, the non-uniformity of the grayscale values ​​of the pixels in the region corresponding to the substrate surface in the image IM10 of Fig. 2 is reduced.

[0049] Specifically, the gradation value of a pixel in image IM30 is obtained by subtracting the gradation value of a pixel in image IM20 corresponding to the coordinates from the gradation value of a pixel in image IM10 at the same coordinates as the pixel in question.

[0050] [Processing Flow of First Correction] Fig. 5 is a flowchart showing an example of processing performed for the first correction of an image obtained by measurement with the scanning probe microscope 1. In one implementation example, the processing in Fig. 5 is performed by the processor of the computer 132 executing a given program. In this sense, the scanning probe microscope 1 is an example of an image processing device.

[0051] When the processor of the computer 132 starts executing a given program, the scanning probe microscope 1 may receive specifications from the user regarding the number of trials, residual threshold, and plane order in the algorithm used for the first correction.

[0052] In step S10, the scanning probe microscope 1 measures the sample 110 and acquires a target image. An example of the image acquired here corresponds to image IM10 in Fig. 2. In this specification, the original image that is the observation result is an example of a "target image."

[0053] In step S12, the scanning probe microscope 1 estimates a first background image based on the target image using an algorithm that calculates a regression line by removing outliers from data that includes outliers. Note that when estimating the first background image in this step, the first background image may be estimated using some of the pixels included in the target image. By reducing the number of pixels used in the process of estimating the first background image, the amount of calculation required for this process can be reduced, and the burden of arithmetic processing by the computer 132 can be reduced.

[0054] In step S14, the scanning probe microscope 1 generates a first corrected image based on the target image acquired in step S10 and the first background image estimated in step S12. An example of the first corrected image generated here corresponds to image IM30 in Fig. 4. In this specification, the image after the first correction is an example of the "first corrected image."

[0055] In step S16, the scanning probe microscope 1 displays the first corrected image generated in step S14 on the display unit 135. Thereafter, the scanning probe microscope 1 ends the processing of FIG.

[0056] In step S14 and / or step S16, the scanning probe microscope 1 may display the estimated first background image on the display unit 135.

[0057] The above-described correction process adjusts the deviation in the gradation values ​​of the pixels of the original image that occurs because the specified plane and the plane on which the sample 110 is actually placed are not parallel, and the accuracy of the image obtained is improved compared to the accuracy of the image after the correction process in the comparative example, and the accuracy of the height information of the structure on the substrate obtained from the corrected image is improved.

[0058] Furthermore, in the correction process described above, there are no parameters that are determined for each process, and therefore the reproducibility of image processing is improved.

[0059] Furthermore, by improving the accuracy of height information for structures on the substrate, the user does not need to make additional image corrections after checking the corrected image, preventing variations in correction processing from user to user.

[0060] [Second Correction] After performing the first correction described above, a second correction, which is different from the first correction, may be performed on the image after the first correction, thereby allowing the user to more accurately recognize the surface state of the sample 110. The second correction is a correction performed so that the representative value (e.g., the average value) of the gradation values ​​of pixels in each row that coincides with the main scanning direction of the probe 114 in the image after the first correction is the same for each row. The second correction can reduce noise along the main scanning direction in the image after the first correction.

[0061] Fig. 6 is a diagram showing an example of an image after the first correction, which is different from the image IM30 shown in Fig. 4. In the image IM40 of Fig. 6, stripes appear along the X-axis direction. The stripes are noise generated along the main scanning direction of the probe 114, and no structure corresponding to the stripes actually exists on the sample 110. Details of the second correction will be explained below using the image IM40 shown in Fig. 6 as an example. Note that in the following explanation, the main scanning direction of the probe 114 during measurement of the sample 110 by the scanning probe microscope 1 corresponds to the X-axis direction, and the sub-scanning direction corresponds to the Y-axis direction.

[0062] In one implementation, computer 132 performs a second correction on image IM40, which is the image after the first correction.

[0063] The image after the first correction is data in which pixels are arranged in the X-axis direction and the Y-axis direction in a matrix, and each pixel has a gradation value as height information. In the second correction, a representative value is calculated for each row that coincides with the main scanning direction of the probe 114. For example, in image IM40, the average value of the gradation values ​​of the pixels included in each row is taken as the representative value along the X-axis direction, which is the scanning direction of the probe 114. The representative value is replaced with the gradation values ​​of all pixels included in each row to form the second background image. Note that the method for calculating the representative value is not limited to the average value, and for example, any one of the median, quartile, and mode of the gradation values ​​may also be used.

[0064] Fig. 7 shows an image IM40, which is a second background image. In image IM50 in Fig. 7, the average value of each row in IM40 is used as a representative value, and the gradation values ​​of all pixels in each row are replaced with the representative value of the pixels in that row.

[0065] When calculating the representative value, all or some of the pixels in each row may be used. When some of the pixels are used, the pixels in each row may be regularly sampled and used for the calculation, or may be randomly sampled and used for the calculation.

[0066] The first corrected image is corrected based on the second background image, thereby generating a second corrected image in which stripe noise occurring along the main scanning direction of the probe 114 has been corrected. Specifically, the gradation value of a pixel at the same coordinate in the second corrected image is obtained by subtracting the gradation value of a pixel in the second background image located at the same coordinate in the matrix as the pixel in the first corrected image from the gradation value of the pixel in the first corrected image. Alternatively, the gradation value of the pixel at the same coordinate in the second background image as the pixel in the first corrected image may be obtained by dividing the gradation value of the pixel in the first corrected image by the gradation value of the pixel in the second background image located at the same coordinate in the matrix as the pixel in the first corrected image, and multiplying the obtained value by a predetermined value. Note that in this specification, the second corrected image may also be referred to as the second corrected image.

[0067] The gradation values ​​of all pixels included in the image after the second correction may be adjusted by adding, subtracting, multiplying, or dividing a predetermined value. The gradation values ​​may be adjusted so that the average or median of all gradation values ​​becomes an appropriate value while maintaining the magnitude relationship of the gradation values ​​of each pixel, or so that the maximum and minimum values ​​of the pixels included in the image become appropriate values. Adjusting the gradation values ​​improves the visibility of the image for the user.

[0068] 8 is a diagram showing an image after the second correction, in which the striped noise in the image IM40 of FIG. 6 has been reduced in the image IM60 shown in FIG.

[0069] Specifically, image IM60 is obtained by, for example, subtracting the gradation value of each pixel of image IM50 corresponding to each pixel in image IM40 from the gradation value of each pixel in image IM40, and performing adjustment.

[0070] [Flow of the Second Correction Process] Figure 9 is a flowchart of an example of a process for performing the second correction on the first corrected image. In one implementation example, the process of Figure 9 is performed by the processor of the computer 132 executing a given program. In this sense, the scanning probe microscope 1 is an example of an image processing device. Note that the second correction is performed on an image that has been subjected to the first correction process. Therefore, the steps in the first correction process in Figure 9 (steps S10 to S16) are the same as the first correction process described above, and therefore a description thereof will be omitted.

[0071] In step S20, the scanning probe microscope 1 calculates a representative value based on the first corrected image generated in step S14, and estimates a second background image.

[0072] In step S22, the scanning probe microscope 1 generates a second corrected image based on the second background image estimated in step S20 and the first corrected image generated in step S14. An example of the image generated here corresponds to image IM50 in Fig. 8. In this specification, the second corrected image is an example of a "second corrected image."

[0073] In step S24, the scanning probe microscope 1 displays the second corrected image generated in step S22 on the display unit 135. Thereafter, the scanning probe microscope 1 ends the processing of FIG.

[0074] Note that the user may determine whether to execute the processes in and after step S20 based on the first corrected image displayed on the display unit 135 in step S16.

[0075] Furthermore, if it has been decided to perform the second correction before the process of FIG. 9 is started, the process may proceed to step S20 without performing the display process of step S16.

[0076] Furthermore, the scanning probe microscope 1 may appropriately display on the display unit 135 the first background image estimated in step S12 and / or the second background image estimated in step S20.

[0077] Aspects It will be understood by those skilled in the art that the exemplary embodiments described above are examples of the following aspects.

[0078] (Item 1) An image processing method according to one aspect is a method for processing an object image generated based on measurement of a sample using a scanning probe microscope, and may include the steps of: estimating a first background image of the object image generated during the measurement due to a tilt between a plane on which the sample is placed and a predetermined plane using an algorithm that calculates a regression line by removing outliers from data that includes outliers; and generating a first corrected image in which the tilt of the object image is corrected based on the first background image.

[0079] According to the image processing method described in paragraph 1, the accuracy of the image obtained by measurement using a scanning probe microscope is improved, and the user can more clearly recognize the structure of the sample surface.

[0080] (Item 2) In the image processing method described in item 1, the algorithm may include at least one of RANSAC (Random sample consensus), Quantile regression, and Theil-Sen regression.

[0081] According to the image processing method described in paragraph 2, the accuracy of the image obtained by measurement using a scanning probe microscope is improved, and the user can more clearly recognize the structure of the sample surface.

[0082] (Item 3) In the image processing method described in item 1 or 2, the degree of the algorithm may be two or three.

[0083] According to the image processing method described in paragraph 3, the accuracy of the image obtained by measurement using a scanning probe microscope is improved, and the user can more clearly recognize the structure of the sample surface.

[0084] (4) The image processing method according to any one of the first to third paragraphs may further include a step of accepting an order in the algorithm from a user.

[0085] According to the image processing method described in item 4, the accuracy of the image obtained by measurement using a scanning probe microscope is improved, and the user can more clearly recognize the structure of the sample surface.

[0086] (Item 5) In the image processing method described in any one of Items 1 to 5, the outlier may be a gradation value of a pixel in the target image that corresponds to at least one of the sample, a structure on the sample, and a particle on the sample.

[0087] According to the image processing method described in paragraph 5, the accuracy of the image obtained by measurement using a scanning probe microscope is improved, and the user can more clearly recognize the structure of the sample surface.

[0088] (Item 6) In the image processing method described in any one of Items 1 to 5, the target image is an image obtained based on a probe movement consisting of main scanning and sub-scanning in the scanning probe microscope, and the method may further include the steps of: estimating a second background image, which is light and dark noise occurring along the main scanning direction of the probe, in the first corrected image; and generating a second corrected image in which the noise has been corrected based on the second background image.

[0089] According to the image processing method described in paragraph 6, it is possible to correct the light and dark noise that occurs along the main scanning direction in an image obtained after correcting the fluctuation in gradation value caused by the tilt of the sample during measurement with a scanning probe microscope, and the user can more clearly recognize the structure of the sample surface.

[0090] (Clause 7) In the image processing method described in any one of clauses 1 to 6, the step of estimating the second background image comprises the steps of calculating a representative value of the gradation values ​​of pixels included in a predetermined main scanning line from the first corrected image, and setting the representative value as the gradation value of pixels included in the predetermined line in the second background image, and the representative value may be any one of the mean, median, quartile, and mode.

[0091] According to the image processing method described in paragraph 7, it is possible to correct the light and dark noise that occurs along the main scanning direction in an image obtained after correcting the fluctuation in gradation value caused by the tilt of the sample during measurement with a scanning probe microscope, and the user can more clearly recognize the structure of the sample surface.

[0092] (Item 8) A program according to one aspect may be executed by one or more processors to cause the one or more processors to perform the image processing method according to any one of items 1 to 7.

[0093] According to the program described in paragraph 8, the accuracy of the image obtained by measurement using a scanning probe microscope is improved, and the user can more clearly recognize the structure of the sample surface.

[0094] (Clause 9) An image processing device according to one aspect may include one or more processors and a storage device that stores a program that, when executed by the one or more processors, causes the one or more processors to implement the image processing method described in any one of clauses 1 to 7.

[0095] According to the image processing device described in paragraph 9, the accuracy of the image obtained by measurement using a scanning probe microscope is improved, and the user can more clearly recognize the structure of the sample surface.

[0096] (Item 10) A scanning probe microscope according to one aspect includes a sample stage on which a sample is placed, a cantilever having a probe at its tip, a moving device that changes the distance between the sample stage and the probe, an optical system that irradiates the cantilever with laser light and detects the laser light reflected by the cantilever, a measurement unit that measures the height of the sample based on a displacement of the cantilever obtained by a change in the position of the laser light detected by the optical system, an image generation unit that generates an image of the target based on the measurement, an image processing unit that executes a correction process for the image of the target, and a display unit that displays a corrected image obtained by correcting the image of the target corrected by the image processing unit, wherein the correction process may include the steps of: estimating a background image of the target image generated by an inclination between a plane on which the sample is placed and a predetermined plane during measurement using an algorithm that calculates a regression line by removing outliers from data including outliers; and generating the corrected image in which the inclination of the target image has been corrected based on the background image.

[0097] According to the scanning probe microscope described in paragraph 10, the accuracy of the image obtained by measurement using the scanning probe microscope is improved, and the user can more clearly recognize the structure of the sample surface.

[0098] The embodiments disclosed herein should be considered to be illustrative in all respects and not restrictive. The scope of the present disclosure is defined by the claims, not by the description of the above-described embodiments, and is intended to include all modifications within the meaning and scope of the claims. Furthermore, it is intended that each technique in the embodiments can be implemented alone or, if necessary, in combination with other techniques in the embodiments to the extent possible.

[0099] 1 scanning probe microscope, 110 sample, 111 piezo scanner, 111xy XY scanner, 111z Z scanner, 112 sample stage, 113 cantilever, 114 probe, 115 laser diode, 116 photodetector, 120 displacement detection mechanism, 131 feedback signal generator, 132 computer, 133 scanning signal generator, 134 storage device, 135 display unit.

Claims

1. A method for processing an object image generated based on measurement of a sample using a scanning probe microscope, comprising the steps of: estimating a first background image of the object image generated due to the inclination between a plane on which the sample is placed and a specified plane during the measurement using an algorithm that calculates a regression line by removing outliers from data including outliers; and generating a first corrected image in which the inclination of the object image is corrected based on the first background image.

2. The image processing method according to claim 1, wherein the algorithm includes at least one of RANSAC (Random sample consensus), Quantile regression, and Theil-Sen regression.

3. The image processing method according to claim 1 or 2, wherein the degree of said algorithm is 2 or 3.

4. The image processing method according to claim 1 or 2, further comprising the step of accepting an order in said algorithm from a user.

5. An image processing method according to claim 1 or claim 2, wherein the outlier is a gradation value of a pixel in the target image that corresponds to at least one of the sample, a structure on the sample, and a particle on the sample.

6. An image processing method as described in claim 1 or claim 2, wherein the target image is an image obtained based on a probe movement consisting of main scanning and sub-scanning in the scanning probe microscope, and further comprising the steps of: estimating a second background image which is light and dark noise occurring along the main scanning direction of the probe in the first corrected image; and generating a second corrected image in which the noise has been corrected based on the second background image.

7. The image processing method of claim 6, wherein the step of estimating the second background image comprises the steps of: calculating a representative value of the gradation values ​​of pixels included in a predetermined main scanning line from the first corrected image; and setting the representative value as the gradation value of pixels included in the predetermined line in the second background image, wherein the representative value is one of the mean, median, quartile, and mode.

8. A program which, when executed by one or more processors, causes the one or more processors to carry out the image processing method according to claim 1 or 2.

9. An image processing device comprising: one or more processors; and a storage device storing a program that, when executed by the one or more processors, causes the one or more processors to implement the image processing method according to claim 1 or 2.

10. A scanning probe microscope comprising: a sample stage on which a sample is placed; a cantilever having a probe at its tip; a moving device which changes the distance between the sample stage and the probe; an optical system which irradiates laser light onto the cantilever and detects the laser light reflected by the cantilever; a measurement unit which measures the height of the sample based on the displacement of the cantilever obtained by a change in the position of the laser light detected by the optical system; an image generation unit which generates an object image based on the measurement; an image processing unit which executes a correction process for the object image; and a display unit which displays a corrected image obtained by correcting the object image by the image processing unit, wherein the correction process includes the steps of: estimating a background image of the object image generated by the inclination between the plane on which the sample is placed and a specified plane during the measurement by an algorithm which calculates a regression line by removing outliers from data including outliers; and generating the corrected image in which the inclination of the object image is corrected based on the background image.

Images