White light interference attitude angle automatic correction method, system and device and storage medium

Abstract

The invention discloses a white light interference attitude angle automatic correction method, system and device and a storage medium. The method comprises the following steps: acquiring an initial white light interference image obtained by measuring a sample by a white light interference measurement system; resolving the initial white light interference image to obtain at least one group of candidate attitude angles of the sample; each group of candidate attitude angles comprises attitude angle components of the sample in at least two different axial directions; according to each group of candidate attitude angles, performing attitude adjustment on the sample by using the carrier, and obtaining stripe change information after each attitude adjustment; and sequentially judging whether the stripe change information after each adjustment meets a preset change condition or not, and determining the candidate attitude angle corresponding to the stripe change information meeting the preset change condition as a target attitude angle to finish correction. According to the invention, through combination of image analysis and adjustment verification, automatic and high-precision correction of the attitude angle of the sample is realized, and the accuracy and stability of white light interference measurement can be significantly improved.

Description

Technical Field

[0001] This invention relates to the field of optical measurement technology, specifically to an automatic correction method, system, device, and storage medium for white light interference attitude angle. Background Technology

[0002] White light interferometry, with its nanometer-level vertical resolution and non-contact measurement advantages, plays an irreplaceable role in surface topography inspection in fields such as integrated circuits, microelectromechanical systems (MEMS), and precision optics. In automated measurement scenarios, to achieve high-efficiency, high-batch continuous inspection, samples typically need to be placed on a mechanical carrier for rapid positioning and scanning. However, due to factors such as mechanical positioning accuracy, sample clamping errors, or changes in carrier motion attitude, unavoidable attitude angle deviations (including pitch and roll angles) often exist in the sample or carrier. These deviations are directly mapped to the interferogram, causing a deflection of the interference fringes and a change in the fringe width (i.e., spatial frequency), which in turn affects the alignment accuracy of the autofocus system and the stability of phase-shifting scanning, ultimately reducing the accuracy of 3D topography reconstruction and measurement repeatability.

[0003] In existing technologies, the extraction and analysis of fringe parameters either focus on the extraction of parameters for specific types of fringes (such as moiré fringes), or on the angle adjustment of specific interference structures (such as reference mirrors or reflectors inside a Michelson interferometer), or determine the tilt direction by identifying the overall direction of the fringes. These approaches generally suffer from drawbacks such as complex coupling between fringe direction identification and attitude angle calculation, insufficient coupling between fringe parameter extraction and attitude angle closed-loop adjustment, and angle calculation relying on complex fringe direction determination.

[0004] Therefore, existing fringe parameter extraction techniques have failed to effectively solve the problem of fringe variation caused by sample / carrier attitude angle deviation in automated white light interferometry, which easily leads to limited measurement accuracy and insufficient robustness, making it difficult to meet the stringent requirements of high-precision and fully automated measurement. Summary of the Invention

[0005] In view of this, the present invention provides a method, system, device and storage medium for automatic correction of white light interferometry attitude angle, in order to solve the problem that the existing white light interferometry measurement technology does not correct the attitude angle of the sample / carrier, which makes it impossible to effectively solve the fringe variation caused by the pitch angle deviation of the sample / carrier, thus resulting in limited measurement accuracy and insufficient robustness.

[0006] This invention provides an automatic white light interferometric attitude angle correction method for use in a white light interferometric measurement system including a carrier, the method comprising: Acquire the initial white light interferometry image obtained from the sample measured by the white light interferometry system; The initial white light interferometric image is solved to obtain at least one set of candidate attitude angles for the sample; each set of candidate attitude angles includes attitude angle components of the sample in at least two different axes; The sample is adjusted using the carrier according to each set of candidate attitude angles, and the stripe change information after each attitude adjustment is obtained. It is then determined whether the stripe change information after each adjustment meets the preset change conditions, and the candidate attitude angle corresponding to the stripe change information that meets the preset change conditions is determined as the target attitude angle to complete the correction.

[0007] Optionally, each set of candidate attitude angles includes the attitude angle components of the sample in the lateral and longitudinal directions; The initial white light interferometric image is solved to obtain at least one set of candidate attitude angles for the sample, including: A reference point is determined in the initial white light interference image, and grayscale profile sequences are obtained along the horizontal and vertical directions, respectively, with the reference point as the center. Effective peak detection is performed on the grayscale profile sequence in each direction to obtain the peak detection information corresponding to the grayscale profile sequence in each direction. When the peak detection information indicates that there is no valid peak or only one valid peak in the gray-scale profile sequence in the corresponding direction, the attitude angle is assigned to the corresponding gray-scale profile sequence according to the preset rules to obtain the attitude angle component of the sample in the corresponding direction. When the peak detection information indicates that there are two or more valid peaks in the grayscale profile sequence in the corresponding direction, the theoretical fringe period of the white light interferometry system is calculated in advance, and the attitude angle component of the sample in the corresponding direction is calculated based on the theoretical fringe period and the peak detection information in the direction where valid peaks exist.

[0008] Optionally, when the peak detection information indicates that there is no valid peak or only one valid peak in the grayscale profile sequence in the corresponding direction, the attitude angle component of the sample in the corresponding direction is assigned a value of 0.

[0009] Optionally, when the peak detection information indicates that there are no valid peaks in the grayscale profile sequence in the corresponding direction, the peak detection information includes the number of valid peaks, and the number of valid peaks is specifically 0; when the peak detection information indicates that there is only one valid peak in the grayscale profile sequence in the corresponding direction, the peak detection information includes the number of valid peaks, and the number of valid peaks is specifically 1. When the peak detection information indicates that there are two or more valid peaks in the grayscale profile sequence in the corresponding direction, the peak detection information includes the number of valid peaks in the corresponding direction, the position of each valid peak, and the stripe spacing based on the position and number of valid peaks, and the number of valid peaks is greater than or equal to 2. Effective peak detection is performed on the grayscale profile sequence in each direction to obtain peak detection information corresponding to the grayscale profile sequence in each direction, including: A grayscale profile sequence in any direction is selected, and the first-order difference sign change method is used to detect the selected grayscale profile sequence to determine multiple peak candidate points in the selected grayscale profile sequence. The validity of all peak candidate points in the selected grayscale profile sequence is judged. When all peak candidate points in the selected grayscale profile sequence are invalid, the number of valid peaks in the selected grayscale profile sequence is marked as 0; when only one peak candidate point in the selected grayscale profile sequence is valid, the number of valid peaks in the selected grayscale profile sequence is marked as 1. When at least two of the candidate peaks are valid, the number of valid peaks in the selected grayscale profile sequence is counted, and the position of each valid peak in the selected grayscale profile sequence is determined. The position of the first peak and the position of the last peak are determined from all the positions of the valid peaks, and the corresponding stripe spacing is calculated based on the position of the first peak, the position of the last peak and the corresponding number of valid peaks. Using the same method, peak detection information was obtained from another grayscale profile sequence.

[0010] Optionally, the specific calculation formulas for the stripe spacing in the transverse and axial directions are as follows: ; Where a and b are the stripe spacing in the transverse and axial directions, respectively, and N and M are the number of effective peaks in the transverse and axial directions, respectively. first and u last These represent the positions of the first and last peaks in the horizontal direction, respectively. first and v last These represent the positions of the first and last peaks in the vertical direction.

[0011] Optionally, the grayscale profile sequence consists of grayscale values ​​at multiple sampling points; The first-order difference sign change method is used to detect the selected grayscale profile sequence, and multiple peak candidate points in the selected grayscale profile sequence are identified, including: In the selected grayscale profile sequence, the grayscale change of each sampling point is calculated, and the grayscale change of each sampling point is defined as the corresponding first-order difference. For the nth sampling point, if the first difference of the nth sampling point is not greater than 0 and the first difference of the (n-1)th sampling point is greater than 0, then the nth sampling point is determined to be a peak candidate point; otherwise, the nth sampling point is determined not to be a peak candidate point. Following the same determination method, the first difference of each sampling point in the selected grayscale profile sequence is traversed in turn to determine multiple peak candidate points in the selected grayscale profile sequence.

[0012] Optionally, the grayscale profile sequence consists of grayscale values ​​at multiple sampling points; The validity of all candidate peak points in the selected grayscale profile sequence is evaluated, including: In the selected grayscale profile sequence, select any peak candidate point. If the difference between the grayscale value of the selected peak candidate point and the global grayscale mean is greater than or equal to the preset significance threshold, the selected peak candidate point will be retained; otherwise, the selected peak candidate point will be removed. Iterate through each peak candidate point in the selected grayscale profile sequence, and filter all peak candidate points in the selected grayscale profile sequence in the same way. The remaining peak candidate points after filtering are sorted in descending order of grayscale value to obtain the peak candidate sequence. A greedy selection method is used to select the peak candidate point with the highest current grayscale value in the peak candidate sequence as the potential peak. With the potential peak as the center, the remaining peak candidate points in the peak candidate sequence within the preset minimum spacing range are removed. When all peak candidate points in the peak candidate sequence have been processed, the validity of the selected grayscale profile sequence is judged, and the corresponding valid peak set is obtained.

[0013] Optionally, peak detection information corresponding to a grayscale profile sequence in any direction can be selected. When the selected peak detection information indicates that there are two or more valid peaks in the grayscale profile sequence in the corresponding direction, the attitude angle component of the sample in the corresponding direction is calculated based on the theoretical fringe period and the peak detection information in the direction where valid peaks exist. This includes: Based on the theoretical fringe period and the fringe spacing in the selected direction, the absolute value of the attitude angle of the sample in the selected direction is calculated. The specific formulas for calculating the absolute values ​​of the attitude angles in the lateral and axial directions are as follows: ; in, and , respectively, are the absolute values ​​of the attitude angles in the lateral and axial directions, a and b are the stripe spacing in the lateral and axial directions, respectively, and l is the theoretical stripe period; If the peak detection information corresponding to the grayscale profile sequence in another direction indicates that there is no valid peak in the corresponding direction, two attitude angle components with opposite signs are generated based on the absolute value of the attitude angle of the sample in the selected direction. If the peak detection information corresponding to the grayscale profile sequence in another direction also indicates that there is a valid peak in the corresponding direction, two diagonals are constructed based on the reference point and the valid peaks closest to the reference point in the two directions; the grayscale variances corresponding to the two diagonals are obtained, and the attitude angle sign of the sample in the selected direction is determined according to the comparison result between the grayscale variances of the two diagonals; the attitude angle components of the sample in the selected direction are generated according to the absolute value and attitude angle sign of the sample in the selected direction.

[0014] Optionally, the theoretical fringe period of the white light interferometry system is calculated in advance, including: The spectral distribution of the white light interferometry system is obtained, and the spectral distribution is transformed to obtain the coherence function of the white light interferometry system; The corresponding interference light intensity signal is determined based on the coherence function of the white light interferometry system. The oscillation waveform of the interference light intensity signal near zero optical path difference is extracted, and the theoretical fringe period is obtained based on the oscillation waveform.

[0015] Optionally, the specific formula for calculating the coherence function of the white light interferometry system is as follows: ; in, Let S(v) be the coherence function of the white light interferometry system, v be the frequency, and S(v) be the spectral density obtained by frequency transformation of the spectral distribution. This represents the time delay, where i is the imaginary unit; The specific formula for determining the interference light intensity signal is: ; in, The interference light intensity signal of the white light interferometry system. This is for the operation of taking the real part.

[0016] Optionally, grayscale profile sequences are obtained along the horizontal and vertical directions, centered on the reference point, including: In the initial white light interference image, with the reference point as the center, lines passing through the reference point are drawn along the horizontal and vertical directions respectively to obtain the horizontal profile line and the vertical profile line; The pixel grayscale values ​​of the horizontal profile line and the vertical profile line are extracted respectively to obtain the initial grayscale profile image of the horizontal profile and the initial grayscale profile image of the vertical profile. The initial horizontal grayscale profile image is resampled and smoothed to obtain a horizontal grayscale profile sequence; the initial vertical grayscale profile image is resampled and smoothed to obtain a vertical grayscale profile sequence.

[0017] Optionally, based on each group of candidate attitude angles, the carrier is used to adjust the attitude of the sample, and the fringe change information after each attitude adjustment is obtained, including: Based on each group of candidate attitude angles, the attitude of the sample is adjusted using the carrier, and the adjusted white light interference image is obtained after each attitude adjustment. Based on the initial white light interference image and the adjusted white light interference image after each attitude adjustment, fringe change information after each attitude adjustment is obtained; wherein, the fringe change information includes fringe width change and fringe direction change.

[0018] Optionally, it is determined sequentially whether the stripe change information after each adjustment meets the preset change conditions, including: For any stripe change information after a posture adjustment, if the stripe width increases after the stripe width change indicator is adjusted, and the stripe direction approaches the preset ideal direction after the stripe direction change indicator is adjusted, then the corresponding stripe change information is determined to meet the preset change conditions; otherwise, the corresponding stripe change information is determined not to meet the preset change conditions, and the judgment of the stripe change information after the next adjustment continues.

[0019] Furthermore, the present invention also provides an automatic white light interferometric attitude angle correction system, applied in the aforementioned automatic white light interferometric attitude angle correction method, the system comprising: The image acquisition module is used to acquire the initial white light interference image obtained by the white light interferometry system from the sample. The attitude calculation module calculates the initial white light interferometric image to obtain at least one set of candidate attitude angles for the sample; each set of candidate attitude angles includes attitude angle components of the sample in at least two different axes; The adjustment and verification module is used to adjust the attitude of the sample using the carrier according to each group of candidate attitude angles, and to obtain the stripe change information after each attitude adjustment; to determine whether the stripe change information after each adjustment meets the preset change conditions, and to determine the candidate attitude angle corresponding to the stripe change information that meets the preset change conditions as the target attitude angle, thereby completing the correction.

[0020] In addition, the present invention provides an automatic white light interferometric attitude angle correction device, including a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed, it implements the method steps in the aforementioned automatic white light interferometric attitude angle correction method.

[0021] In addition, the present invention also provides a computer storage medium comprising: at least one instruction that, when executed by a computer, implements the method steps in the aforementioned automatic white light interference attitude angle correction method.

[0022] The beneficial effects of this invention are as follows: By acquiring the initial white light interferometry image of the entire measurement system when measuring the sample, this image contains information about the fringe formed by the interference between the incident light reflected from the sample surface and the reflected light reflected from the reference surface (i.e., the surface of the reference mirror in the measurement system). The shape, spacing, and direction of these fringes directly reflect the attitude state of the sample and can provide a data basis for subsequent attitude calculation. By performing a series of precise image processing, analysis, and calculation on the initial white light interferometry image, the attitude angular components of the sample in different directions can be extracted from the complex interference fringes, thereby generating candidate attitude angles. Finally, through actual attitude adjustment and feedback of fringe change information, the candidate attitude angles are verified and screened to ensure that the finally determined target attitude angle can adjust the sample to the ideal measurement attitude. The automatic white light interferometric attitude angle correction method, system, device, and storage medium of this invention achieve automated and high-precision correction of sample attitude angles through the combination of image analysis and adjustment verification. This avoids complex fringe direction identification and phase calculation, significantly simplifies the processing flow, and effectively solves the problem of fringe variation caused by sample / carrier pitch angle deviation, which is not effectively addressed in existing white light interferometric measurement technology due to the lack of sample / carrier attitude angle correction. It can significantly improve the accuracy and stability of white light interferometric measurement, improve measurement efficiency and system robustness, and lay a solid foundation for subsequent high-precision three-dimensional topography measurement. It is especially suitable for fields with high measurement accuracy requirements, such as precision machining, semiconductor manufacturing, and optical component inspection. Attached Figure Description

[0023] The features and advantages of the invention will be more clearly understood by referring to the accompanying drawings, which are schematic and should not be construed as limiting the invention in any way. In the drawings: Figure 1 A flowchart of an automatic white light interference attitude angle correction method according to Embodiment 1 of the present invention is shown; Figure 2 A model diagram of the white light interferometry measurement system in Embodiment 1 of the present invention is shown; Figure 3 A schematic diagram of the sample coordinate system in Embodiment 1 of the present invention is shown; Figure 4 A schematic diagram of the initial white light interference image after denoising processing in Embodiment 1 of the present invention is shown; Figure 5A A schematic diagram of the initial grayscale profile image in the horizontal direction in Embodiment 1 of the present invention is shown; Figure 5B A schematic diagram of the initial longitudinal grayscale profile image in Embodiment 1 of the present invention is shown; Figure 6A A schematic diagram of the horizontal grayscale profile sequence in Embodiment 1 of the present invention is shown; Figure 6B A schematic diagram of a vertical grayscale profile sequence is shown in Embodiment 1 of the present invention; Figure 7 This diagram illustrates the effective peak point that is closest to the reference point in both the horizontal and vertical directions in Embodiment 1 of the present invention. Figure 8 This diagram illustrates the adjusted white light interferometric image and its reference point after attitude adjustment in Embodiment 1 of the present invention. Figure 9A A schematic diagram of the grayscale profile sequence in the lateral direction re-acquired after attitude adjustment in Embodiment 1 of the present invention is shown. Figure 9B A schematic diagram of the longitudinal grayscale profile sequence re-acquired after attitude adjustment in Embodiment 1 of the present invention is shown. Figure 10 The diagram shows a structure of an automatic white light interference attitude angle correction system according to Embodiment 2 of the present invention.

[0024] Explanation of reference numerals in the attached figures: 1. Lens, 2. Reference mirror, 3. Beam splitter, 4. Sample. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] In this embodiment of the invention, the term "and / or" describes the relationship between associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. The character " / " generally indicates that the preceding and following associated objects have an "or" relationship.

[0027] It should be noted that the terms "first," "second," etc., in the specification, claims, and drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0028] In this embodiment of the invention, the term "multiple" refers to two or more, and other quantifiers are similar.

[0029] Example 1 This embodiment provides an automatic white light interferometric attitude angle correction method for use in a white light interferometric measurement system including a carrier, such as... Figure 1 As shown, the method includes: S1: Obtain the initial white light interference image obtained by measuring the sample using the white light interferometry system; S2: Solve the initial white light interferometric image to obtain at least one set of candidate attitude angles for the sample; each set of candidate attitude angles includes attitude angle components of the sample in at least two different axes; S3: Adjust the attitude of the sample using the carrier according to each group of candidate attitude angles, and obtain the stripe change information after each attitude adjustment; determine whether the stripe change information after each adjustment meets the preset change conditions, and determine the candidate attitude angle corresponding to the stripe change information that meets the preset change conditions as the target attitude angle to complete the correction.

[0030] In this embodiment, an initial white light interferometric image of the entire measurement system during sample measurement is acquired. This image contains information about the fringe formed by the interference between the incident light reflected from the sample surface and the reflected light reflected from the reference surface (i.e., the surface of the reference mirror in the measurement system). The shape, spacing, and direction of these fringes directly reflect the attitude state of the sample and provide a data basis for subsequent attitude calculation. By performing a series of precise image processing, analysis, and calculations on the initial white light interferometric image, the attitude angular components of the sample in different directions can be extracted from the complex interference fringes, thereby generating candidate attitude angles. Finally, the candidate attitude angles are verified and screened through actual attitude adjustment and feedback of fringe change information to ensure that the final determined target attitude angle can adjust the sample to the ideal measurement attitude.

[0031] The automatic white light interferometric attitude angle correction method in this embodiment achieves automated and high-precision correction of the sample attitude angle through the combination of image analysis and adjustment verification. It avoids complex fringe direction identification and phase calculation, significantly simplifies the processing flow, and effectively solves the problem of fringe variation caused by sample / carrier pitch angle deviation, which is not effectively addressed in existing white light interferometric measurement technology due to the lack of sample / carrier attitude angle correction. It can significantly improve the accuracy and stability of white light interferometric measurement, improve measurement efficiency and system robustness, and lay a solid foundation for subsequent high-precision three-dimensional topography measurement. It is especially suitable for fields with high measurement accuracy requirements, such as precision machining, semiconductor manufacturing, and optical component inspection.

[0032] The following provides a detailed description of each step in the automatic white light interference attitude angle correction method of this embodiment.

[0033] In this embodiment, the white light interferometry system includes, in addition to the carrier, an interferometric objective lens, an imaging camera, a drive mechanism, and other components. The carrier has attitude angle adjustment capabilities, enabling precise adjustment of the sample's pitch and roll angles to change the relative angular relationship between the sample surface and the incident light and reference plane.

[0034] Interference lenses are used to separate incident light into measurement light and reference light, such as... Figure 2 As shown, it includes lens 1, beam splitter 3, and reference mirror 2. When the incident light (measurement light) shines on the surface of sample 4, it is reflected to form a reflected beam. The reference light, after being split by the beam splitter, shines on reference mirror 2 and also forms a reflected beam. The two reflected beams merge and interfere to form interference fringes. When the distance between the interference objective and the sample changes, the interference pattern of the two reflected beams also changes periodically. If an image is acquired on an inclined plane, periodic interference fringes will be obtained, with adjacent fringes representing a certain height difference between corresponding positions. Based on this information, the attitude angle of the sample can be determined.

[0035] The imaging camera is used to capture white light interference images containing interference fringes, providing raw image data for subsequent attitude calculations. The drive mechanism is used to move the carrier or interference objective to adjust the sample attitude or acquire interference images.

[0036] In step S1, when acquiring the initial white light interference image, it is necessary to ensure that the measurement system is in a stable working state, the sample is placed on the carrier and roughly adjusted to the initial measurement position, and the imaging camera acquires the interference image at this time as the basis for subsequent processing.

[0037] For the sample under test, each set of candidate attitude angles includes attitude angle components in the lateral and longitudinal directions. By using attitude angle components in different axes, the attitude tilt of the sample under test in three-dimensional space can be described more comprehensively, providing multi-dimensional parameter basis for subsequent attitude adjustment.

[0038] To determine the two attitude angular components, it is first necessary to define the sample coordinate system, as shown in the figure. Figure 3 As shown, the attitude angle component in the horizontal direction is also called the pitch angle, and the attitude angle component in the axial direction is also called the roll angle. The pitch angle α refers to the rotation angle of the sample about the Y-axis, and the roll angle β is the rotation angle of the sample about the X-axis. By coordinating the adjustment of the attitude angles in these two directions, the attitude of the sample on the horizontal plane can be optimized, ensuring that its surface maintains a preset ideal relative angle with the measuring optical axis, thereby creating conditions for obtaining high-quality interference fringes and accurate measurement results.

[0039] Preferably, step S2 in this embodiment includes: S21: Determine a reference point in the initial white light interference image, and obtain grayscale profile sequences along the horizontal and vertical directions, respectively, with the reference point as the center; S22: Perform effective peak detection on the grayscale profile sequence in each direction to obtain the peak detection information corresponding to the grayscale profile sequence in each direction. S23A: When the peak detection information indicates that there is no valid peak or only one valid peak in the gray-scale profile sequence in the corresponding direction, the attitude angle is assigned to the corresponding gray-scale profile sequence according to the preset rule to obtain the attitude angle component of the sample in the corresponding direction. S23B: When the peak detection information indicates that there are two or more effective peaks in the gray-scale profile sequence in the corresponding direction, the theoretical fringe period of the white light interferometry system is calculated in advance, and the attitude angle component of the sample in the corresponding direction is calculated based on the theoretical fringe period and the peak detection information in the direction where there are effective peaks.

[0040] Preprocessing to determine the reference point provides an accurate center point for subsequent grayscale profile sequence extraction, ensuring the accuracy and reliability of subsequent analysis. Obtaining grayscale profile sequences along both the horizontal and vertical directions captures the grayscale variation characteristics of interference fringes from two mutually perpendicular directions, providing comprehensive data support for subsequent peak detection and attitude angle component calculation. In the horizontal direction, the grayscale profile sequence reflects the grayscale distribution of fringes along the horizontal direction; in the vertical direction, it reflects the grayscale distribution of fringes along the vertical direction. Combining these two aspects more completely outlines the overall shape of the interference fringes, avoiding attitude angle calculation errors caused by missing information in a single direction. Then, effective peak detection is performed based on the grayscale profile sequences from both directions. Since effective peaks reflect the center positions of the bright fringes in the interference fringes, detecting the number and distribution of these peaks allows for the determination of the fringes' clarity and periodicity, thus providing crucial information for the calculation of attitude angle components. When there is no valid peak or only one valid peak in a certain direction, it indicates that the interference fringes in that direction are not obvious or have not formed a periodic structure. In this case, the attitude angle is assigned according to preset rules (e.g., based on the system's default initial value or historical experience value), which ensures that an initial adjustment direction can be provided even when the fringe quality is poor. When there are two or more valid peaks, it indicates that the fringes have good periodicity. In this case, by combining the peak detection information (e.g., the actual distance between adjacent peaks, i.e., the fringe spacing) with the theoretical fringe period of the system (which is related to the parameters of the interference objective, the wavelength of the light source, and other inherent system properties), the attitude angle component of the sample in that direction can be accurately derived. The attitude angle calculation method described in this embodiment can flexibly select an appropriate calculation strategy according to the actual situation of the interference fringes. While ensuring the accuracy of the attitude angle component calculation, it improves the adaptability of the method to different fringe quality scenarios, ensuring that even when the fringes are not obvious or there is interference, the candidate attitude angle can be stably obtained, thus providing a guarantee for the accuracy of subsequent correction.

[0041] Preferably, in step S21, before determining the reference point in the initial white light interferogram, the method further includes: S211: Perform grayscale conversion and noise reduction on the initial white light interference image.

[0042] By converting and denoising the initial white light interference image to grayscale, color images can be converted to grayscale images to simplify data processing, ensuring the effective length of interference fringes in the image and avoiding noise interference, thus ensuring the clarity and accuracy of the image processed subsequently.

[0043] Preferably, in step S21, determining a reference point in the initial white light interferometry image includes: S212: Select the pixel with the highest gray value in the central region of the initial white light interference image after noise reduction as the reference point.

[0044] The central region is usually the area where interference fringes are relatively stable and concentrated. Selecting the pixel with the highest gray value in this region as the reference point can ensure that the reference point is located at the center of the bright fringes of the interference fringes to the greatest extent possible. This provides an accurate starting point for the subsequent extraction of gray-scale profile sequences along the horizontal and vertical directions, avoiding the gray-scale change feature extraction error caused by the reference point deviating from the fringe center. This lays a reliable foundation for subsequent peak detection and attitude angle component calculation.

[0045] For example, if the resolution of the initial white light interference image after denoising is 1024×1024 pixels, the central region can be set as a square region with a side length of 200 pixels centered at the image center (512, 512). All pixels are traversed within this region, and the coordinates of the pixel with the largest gray value are determined as the reference point.

[0046] In one specific implementation, the initial white light interferometric image after denoising is as follows: Figure 4 As shown, the pixel with the highest gray value in its central region is point P, so point P is determined as the reference point.

[0047] Preferably, in step S21, grayscale profile sequences are obtained along the horizontal and vertical directions, centered on the reference point, including: S213: In the initial white light interference image, with the reference point as the center, draw lines passing through the reference point in the horizontal and vertical directions respectively to obtain the horizontal profile line and the vertical profile line; S214: Extract the pixel grayscale values ​​of the horizontal profile line and the vertical profile line respectively to obtain the initial grayscale profile image of the horizontal profile line and the initial grayscale profile image of the vertical profile line; S215: The initial horizontal grayscale profile image is resampled and smoothed to obtain a horizontal grayscale profile sequence; the initial vertical grayscale profile image is resampled and smoothed to obtain a vertical grayscale profile sequence.

[0048] By drawing lines passing through reference points along both the horizontal and vertical axes, the grayscale variation trajectories of interference fringes in these two key directions can be accurately captured using these horizontal and vertical profile lines, providing a clear physical path for subsequent grayscale profile sequence extraction. After extracting the pixel grayscale values ​​of the horizontal and vertical profile lines, the resulting initial horizontal and vertical grayscale profile images can intuitively display the grayscale distribution along the profile line direction. These original grayscale data are the direct basis for subsequent analysis. Since the original image may have uneven sampling intervals or noise interference, resampling the initial grayscale profile image can adjust the pixels on the profile line to an evenly spaced distribution, ensuring the consistency and comparability of data in subsequent processing. Smoothing filtering effectively removes high-frequency noise from the grayscale data, making the grayscale profile sequence curve smoother, highlighting the periodic variation characteristics of the interference fringes, and reducing the impact of noise on the accuracy of peak detection. The grayscale profile sequences obtained after resampling and smoothing filtering have significantly improved data quality in both the horizontal and vertical directions, providing more reliable and clear input for subsequent effective peak detection and attitude angle component calculation.

[0049] In one specific implementation, for Figure 4 The image shown is plotted with reference point P as the center, and lines passing through the reference point are drawn horizontally and vertically to obtain the horizontal profile line AB and the vertical profile line CD. The initial grayscale profile images obtained after extracting pixel grayscale values ​​from these two profile lines are shown below. Figure 5A and Figure 5B As shown.

[0050] The initial horizontal grayscale profile image and the initial vertical grayscale profile image are resampled and filtered respectively. Specifically, the filtering adopts the Hanning window weighted moving average filtering method, which can effectively suppress the interference of random noise while preserving the main change trend of the grayscale profile sequence, making the grayscale curve of the profile sequence smoother and more consistent.

[0051] Let the window length of the Hanning window be w (an odd number not less than 3). Construct the Hanning window weight, and the specific formula for the weight is as follows: ; Where h[k] is the kth weight; Normalize the above weights to make .

[0052] Then, using all weights, a one-dimensional convolution is performed on the initial horizontal grayscale profile image x0[n] to obtain a smooth sequence, which is the horizontal grayscale profile sequence x[n]: ; Where n is the nth sampling point on the initial horizontal grayscale profile image.

[0053] It should be understood that when the profile length of the horizontal initial grayscale profile image is less than the window length w, the original profile can be used directly, that is, the horizontal initial grayscale profile image is used as the grayscale profile sequence.

[0054] The filtering process for the initial vertical grayscale profile image is similar and will not be described in detail here.

[0055] for Figure 5A The horizontal grayscale profile image shown is processed by smoothing filtering to obtain the horizontal grayscale profile sequence as follows: Figure 6A As shown; for Figure 5B The longitudinal initial grayscale profile image shown, after being processed by smoothing filtering, yields the following longitudinal grayscale profile sequence: Figure 6B As shown, these grayscale profile sequences are all composed of grayscale values ​​from multiple sampling points.

[0056] Preferably, in S22 of this embodiment, when the peak detection information indicates that there is no valid peak in the grayscale profile sequence in the corresponding direction, the peak detection information includes the number of valid peaks, and the number of valid peaks is specifically 0; when the peak detection information indicates that there is only one valid peak in the grayscale profile sequence in the corresponding direction, the peak detection information includes the number of valid peaks, and the number of valid peaks is specifically 1. When the peak detection information indicates that there are two or more valid peaks in the grayscale profile sequence in the corresponding direction, the peak detection information includes the number of valid peaks in the corresponding direction, the position of each valid peak, and the stripe spacing based on the position and number of valid peaks, and the number of valid peaks is greater than or equal to 2.

[0057] Step S22 includes: S221: Select a grayscale profile sequence in any direction, and use the first-order difference sign change method to detect the selected grayscale profile sequence to determine multiple peak candidate points in the selected grayscale profile sequence. S222: Determine the validity of all peak candidate points in the selected grayscale profile sequence; S223A: When all peak candidate points in the selected grayscale profile sequence are invalid, the number of valid peaks in the selected grayscale profile sequence is marked as 0; when only one peak candidate point in the selected grayscale profile sequence is valid, the number of valid peaks in the selected grayscale profile sequence is marked as 1. S223B: When at least two of the peak candidate points are valid, count the number of valid peaks in the selected grayscale profile sequence and determine the position of each valid peak in the selected grayscale profile sequence; determine the first peak position and the last peak position from all valid peak positions, and calculate the corresponding stripe spacing based on the first peak position, the last peak position and the corresponding number of valid peaks. S224: Using the same method, obtain peak detection information in another grayscale profile sequence.

[0058] The obtained grayscale profile sequence may contain multiple peaks. These peaks may be the centers of real bright interference fringes, or they may be spurious peaks caused by noise or local grayscale fluctuations. Therefore, finite peak detection is required. For a grayscale profile sequence in any direction, the first-order difference sign change method is first used to detect peaks and obtain candidate peak points. This can quickly locate the potential peak positions in the sequence and provide preliminary screening results for subsequent validity judgment.

[0059] Then, the validity of all peak candidate points is judged. If all peak candidate points fail the validity judgment, the number of valid peaks is marked as 0, indicating that the interference fringes in that direction are extremely indistinct. If only one peak candidate point meets the condition, the number of valid peaks is marked as 1, indicating that the fringe periodicity is insufficient. When there are two or more valid peaks, in addition to counting the number (i.e., the number of valid peaks), the position coordinates of each valid peak in the profile sequence (i.e., the position of the valid peak) need to be accurately recorded. Based on these position coordinates, the position of the first peak and the position of the last peak can be determined, and then the distance between these two peaks can be calculated. Combined with the number of valid peaks in between (including the first peak and the last peak), the fringe spacing can be obtained. This spacing reflects the actual distance between the centers of adjacent bright fringes and is a key parameter for subsequent calculation of attitude angular components. By performing the above peak detection process on the grayscale profile sequences in both the transverse and longitudinal directions, the peak detection information in both directions can be fully obtained, providing a basis for the calculation of attitude angular components.

[0060] Specifically, in step S221, the first-order difference sign change method is used to detect the selected grayscale profile sequence, and multiple peak candidate points in the selected grayscale profile sequence are determined, including: S2211: In the selected grayscale profile sequence, calculate the grayscale change of each sampling point, and define the grayscale change of each sampling point as the corresponding first-order difference; S2212: For the nth sampling point, if the first difference of the nth sampling point is not greater than 0 and the first difference of the (n-1)th sampling point is greater than 0, then the nth sampling point is determined to be a peak candidate point; otherwise, the nth sampling point is determined not to be a peak candidate point. Following the same determination method, the first difference of each sampling point in the selected grayscale profile sequence is traversed in turn to determine multiple peak candidate points in the selected grayscale profile sequence.

[0061] In a grayscale profile sequence, each sampling point exhibits a corresponding grayscale change relative to the previous sampling point, which can be quantified using first-order differences. Specifically, for the nth sampling point in the horizontal grayscale profile sequence, its grayscale value is denoted as x[n], and the first-order difference Δx[n] of this sampling point is defined as Δx[n] = x[n] - x[n-1]. By calculating the first-order difference of each sampling point, the increasing or decreasing trend of the grayscale value in the sequence can be clearly reflected. When the first-order difference Δx[n] of a certain sampling point is ≤0 and the first-order difference Δx[n-1] of its previous sampling point is >0, it indicates that the grayscale value increases after rising at the (n-1)th sampling point and then begins to decrease at the nth sampling point. Therefore, the nth sampling point is a peak candidate point. By traversing the entire grayscale profile sequence and judging whether each sampling point meets this condition, all possible peak candidate points can be systematically screened, laying the foundation for subsequent validity judgment.

[0062] Specifically, step S222 includes: S2221: Select any peak candidate point from the selected grayscale profile sequence. If the difference between the grayscale value of the selected peak candidate point and the global grayscale mean is greater than or equal to the preset significance threshold, the selected peak candidate point will be retained; otherwise, the selected peak candidate point will be removed. S2222: Traverse each peak candidate point in the selected grayscale profile sequence, and filter all peak candidate points in the selected grayscale profile sequence in the same way. S2223: Arrange the remaining peak candidate points after filtering in descending order of gray value to obtain the peak candidate sequence; use a greedy selection method to select the peak candidate point with the highest gray value in the peak candidate sequence as the potential peak, and remove the other peak candidate points in the peak candidate sequence within the preset minimum spacing range with the potential peak as the center; when all peak candidate points in the peak candidate sequence have been processed, complete the validity judgment of the selected gray profile sequence and obtain the corresponding valid peak set.

[0063] Among all peak candidate points, some peaks may be spurious peaks caused by local noise or random fluctuations in grayscale values. These spurious peaks typically have grayscale values ​​that are only slightly different from the global grayscale mean of the entire grayscale profile sequence, lacking the significant characteristics of a true peak. Therefore, for all peak candidate points, a preset significance threshold is first set, and the grayscale value of each peak candidate point is compared with the global grayscale mean. When the difference is greater than or equal to the threshold, the peak candidate point is considered sufficiently significant and is likely a true peak, thus retaining it; otherwise, it is judged as a spurious peak and discarded. This step can initially filter out insignificant peaks caused by noise and other factors, improving the accuracy of subsequent peak detection. After completing the significance screening, some remaining peak candidate points may still be too close together. These closely spaced peaks may correspond to local fluctuations within the same interference fringe or adjacent spurious peaks, rather than independent valid peaks. Therefore, the retained peak candidate points are arranged in descending order of grayscale value to form a peak candidate sequence, prioritizing candidate points with higher grayscale values ​​that are more likely to be true peaks. A greedy selection method is employed, sequentially selecting the peak candidate point with the highest current grayscale value from the peak candidate sequence as the potential peak. Then, with this potential peak as the center, a preset minimum spacing range is set, and all other peak candidate points within this range are eliminated, avoiding the selection of multiple redundant peaks in the same local area. This method effectively removes neighboring spurious or redundant peaks, ensuring that the peaks in the final valid peak set are independent and representative. After traversing and processing all peak candidate points in the peak candidate sequence, the validity judgment of the grayscale profile sequence is completed, yielding the corresponding valid peak set, providing accurate peak information for subsequent fringe spacing calculation and attitude angular component solution.

[0064] The aforementioned preset salience threshold and preset minimum spacing can be set and adjusted according to actual conditions, and there are no restrictions here.

[0065] In step S2221, for the horizontal grayscale profile sequence, its global grayscale mean is Let the preset significance threshold be p0, then when the gray value of a certain peak candidate point satisfies: If the peak candidate point is selected, it is retained; otherwise, it is discarded.

[0066] After the validity judgment is completed according to step S222 above, if all peak candidate points are invalid, then mark the number of valid peaks according to step S223A; if at least two of the peak candidate points are valid, then count the number of valid peaks according to step S223B.

[0067] Furthermore, in step S223B, the specific calculation formulas for the stripe spacing in the transverse and axial directions are as follows: ; Where a and b are the stripe spacing in the transverse and axial directions, respectively, and N and M are the number of effective peaks in the transverse and axial directions, respectively. first and u last These represent the positions of the first and last peaks in the horizontal direction, respectively. first and v last These represent the positions of the first and last peaks in the vertical direction.

[0068] In the above calculation formula, the difference between the first and last peak positions represents the distance span between the first and last effective peaks in the grayscale profile sequence, while the number of effective peaks minus 1 represents the number of fringe periods between the first and last peaks. Dividing the two yields the average distance between adjacent effective peaks, i.e., the fringe spacing. The fringe spacing calculated in this way provides key geometric parameters for subsequent solution of attitude angular components based on the fringe direction.

[0069] Preferably, in step S23A, when the peak detection information indicates that there is no valid peak or only one valid peak in the grayscale profile sequence in the corresponding direction, the attitude angle component of the sample in the corresponding direction is assigned a value of 0.

[0070] If a grayscale profile sequence has no valid peaks or only one valid peak, it indicates that the interference fringe features in that direction are not obvious or lack periodicity, making it impossible to accurately calculate the attitude angle components based on the fringe direction. To avoid calculation errors caused by invalid fringe information, the attitude angle components in that direction are directly assigned a value of 0, as if there is no attitude angle deflection or the deflection is negligible. Subsequent acquisition of white light interferometry images and recalculation of the attitude angles ensures the stability and reliability of the attitude angle correction process.

[0071] Preferably, in step S23B, when the peak detection information indicates that there are two or more valid peaks in the grayscale profile sequence in the corresponding direction, the theoretical fringe period of the white light interferometry system is pre-calculated, including: S23B1: Obtain the spectral distribution of the white light interferometry system, transform the spectral distribution, and obtain the coherence function of the white light interferometry system; S23B2: Determine the corresponding interference light intensity signal based on the coherence function of the white light interferometry system; S23B3: Extract the oscillation waveform of the interference light intensity signal near zero optical path difference, and obtain the theoretical fringe period based on the oscillation waveform.

[0072] The theoretical fringe period refers to the distance between the centers of two adjacent bright (or dark) fringes when the interference fringes exhibit an ideal state of equal spacing and parallel arrangement under ideal white light interference conditions—that is, when the sample surface is absolutely flat and the measurement system has no attitude deviation. It is a theoretical value calculated based on the inherent parameters of the white light interferometry system (such as the spectral characteristics of the light source and the structure of the optical system) and the interference principle under no attitude deviation. It is used to compare with the actually detected fringe spacing, thus providing a reference benchmark for the accurate calculation of attitude angular components.

[0073] Specifically, in step S23B1, the spectral distribution of the white light interferometry system is obtained, which reflects the light intensity distribution of the light source at different wavelengths. The spectral distribution is then transformed to obtain a coherence function, which describes the degree of coherence between two coherent beams at different optical path differences.

[0074] The specific formula for calculating the coherence function of the white light interferometry system is as follows: ; in, Let S(v) be the coherence function of the white light interferometry system, v be the frequency, and S(v) be the spectral density obtained by frequency transformation of the spectral distribution. For time delay, i is the imaginary unit.

[0075] In step S23B2, the corresponding interference light intensity signal can be further determined based on the coherence function. This signal exhibits a periodic oscillation characteristic that varies with the optical path difference.

[0076] The specific formula for determining the interference light intensity signal is: ; in, The interference light intensity signal of the white light interferometry system. This is for the operation of taking the real part.

[0077] In step S23B3, the oscillation waveform of the interference light intensity signal near zero optical path difference is extracted. The waveform in this region is the most stable and obvious. By analyzing the periodic characteristics of this waveform, the theoretical fringe period can be obtained. Specifically, The time interval between adjacent peaks near zero optical path difference This corresponds to one fringe period. The optical path difference is calculated using the formula... , the time interval Converted to optical path difference interval ΔOPD (i.e. This is the theoretical fringe period l, which is typically measured in nanometers or micrometers. For a Milau-type interference objective, the height difference of the sample surface represented by adjacent fringes is l / 2.

[0078] Preferably, in step S23B, when the peak detection information indicates that there are two or more valid peaks in the grayscale profile sequence in the corresponding direction, the attitude angle component of the sample in the corresponding direction is calculated based on the theoretical fringe period and the peak detection information in the direction where valid peaks exist, including: S23B4: Calculate the absolute value of the attitude angle of the sample in the selected direction based on the theoretical fringe period and the fringe spacing in the selected direction; The specific formulas for calculating the absolute values ​​of the attitude angles in the lateral and axial directions are as follows: ; in, and , respectively, are the absolute values ​​of the attitude angles in the lateral and axial directions, a and b are the stripe spacing in the lateral and axial directions, respectively, and l is the theoretical stripe period; S23B5a: If the peak detection information corresponding to the grayscale profile sequence in another direction indicates that there is no valid peak in the corresponding direction, generate two attitude angle components with opposite attitude angle signs based on the absolute value of the attitude angle of the sample in the selected direction. S23B5b: If the peak detection information corresponding to the grayscale profile sequence in another direction also indicates that there is a valid peak in the corresponding direction, two diagonals are constructed based on the reference point and the valid peaks closest to the reference point in the two directions; the grayscale variances corresponding to the two diagonals are obtained, and the attitude angle sign of the sample in the selected direction is determined according to the comparison result between the grayscale variances of the two diagonals; the attitude angle component of the sample in the selected direction is generated according to the absolute value and attitude angle sign of the sample in the selected direction.

[0079] After calculating the theoretical fringe period and fringe spacing, the absolute values ​​of the attitude angles (including the absolute values ​​of pitch and roll angles) can be accurately calculated using the formula in step S23B4 above. At this point, the absolute values ​​of the attitude angles only reflect the degree of deflection of the sample in the corresponding direction. The attitude angle sign must be considered in conjunction with the attitude angle sign to fully determine the attitude angle components. The determination of the attitude angle sign is related to the tilt direction of the interference fringes. Essentially, it involves determining the tilt direction of the sample surface relative to the ideal horizontal state, i.e., whether it is tilted upwards or downwards. When there are no valid peaks in the grayscale profile sequence in another direction, it means that only one direction of interference fringes can be used for attitude angle calculation. In this case, the attitude angle sign cannot be directly determined through the relationship between the fringes in the two directions. Therefore, two attitude angle components with opposite signs are generated as possible candidate values ​​for subsequent attitude angle correction, so that the final sign can be determined based on other auxiliary information or further judgment logic in practical applications.

[0080] For example, in step S23B5a, when there are two or more valid peaks in the lateral direction, the absolute value of the attitude angle in the lateral direction is calculated. When there is no valid peak or only one valid peak in the longitudinal direction (the corresponding attitude angle component is 0, i.e., the roll angle is 0), the attitude angle sign in the lateral direction includes both ±, i.e., the corresponding attitude angle component (specifically the pitch angle) is +. and- Based on a roll angle of 0, the final output consists of two sets of attitude angular components α1=+ β1=0 and α2=- β2=0.

[0081] Similarly, when there are two or more effective peaks in the longitudinal direction, and the absolute value of the attitude angle in the longitudinal direction is calculated... When there is no valid peak or only one valid peak in the lateral direction (the corresponding attitude angle component is 0, i.e., the pitch angle is 0), the attitude angle sign in the longitudinal direction includes both ±, i.e., the corresponding attitude angle component (specifically the roll angle) is +. and- Since the pitch angle is 0, the final output consists of two sets of attitude angular components: α1=0 and β1=+. and α2=0, β2=- .

[0082] When effective peaks exist in both grayscale profile sequences, the sign of the attitude angle can be determined by constructing diagonals and comparing their grayscale variances. Specifically, first, the effective peaks closest to the reference point are found in both directions. Using the reference point as one vertex, the effective peaks closest to the reference point laterally and longitudinally are used as the other two vertices, thus constructing two distinct diagonals. These two diagonals represent the feature lines corresponding to the two possible fringe tilt directions. Next, the variance of the grayscale values ​​of all pixels along these two diagonals is calculated. Grayscale variance reflects the dispersion of grayscale values ​​in the image. For real interference fringe regions, the grayscale values ​​exhibit periodic brightness changes, resulting in a relatively large grayscale variance; while for non-fringe regions or regions that do not conform to the real fringe direction, the grayscale variance is relatively small. By comparing the grayscale variances of the two diagonals, the fringe tilt direction corresponding to the diagonal with the larger grayscale variance is the actual attitude angle deflection direction of the sample, thus determining the sign of the attitude angle. By combining the calculated absolute value of the attitude angle with the determined sign of the attitude angle, the complete attitude angle components (such as pitch and roll angles) of the sample in that direction can be obtained.

[0083] For example, in step S23B5b, when there are two or more valid peaks in both the horizontal and vertical directions, connecting the peaks closest to the reference point in both directions forms two diagonal lines. One line connects the peak at the top left of the image to the peak at the bottom right (called the first diagonal line), and the other two lines connect the peak at the bottom left of the image to the peak at the top right (called the second diagonal lines). The grayscale variances of the two diagonal lines are compared, and the signs of the pitch and roll angles are determined based on the magnitude of the grayscale variances, resulting in two sets of candidate attitude angles.

[0084] Specifically, when the gray-level variance of the first diagonal is greater than that of the second diagonal, it means that the gray-level change in the image is drastic from the upper left to the lower right. This drastic change occurs in a direction perpendicular to the actual fringe direction, which is from the upper right to the lower left of the image. Therefore, the tilt direction of the sample is from the upper left to the lower right, with the pitch and roll angles sharing the same sign. This gives the corresponding two sets of attitude angular components: α1 = + β1=+ and α2=- β2=- Conversely, if the gray-level variance of the first diagonal is less than that of the second diagonal, it means that the gray-level change in the image is drastic from the lower left to the upper right. This drastic change occurs in a direction perpendicular to the actual fringe direction, which is from the upper left to the lower right of the image. Therefore, the tilt direction of the sample is from the lower left to the upper right, with the pitch and roll angles having opposite signs. This gives the corresponding two sets of attitude angular components: α1 = + β1=- and α2=- β2=+ .

[0085] for Figure 4 In the image shown, there are two effective peaks closest to the reference point in the horizontal direction, namely: Figure 7 The wave crests M and N are shown in the diagram; the effective wave crest closest to the reference point in the longitudinal direction corresponds to one wave crest, which is... Figure 7 The peak point Q is selected from these three peak points. These three peak points construct two diagonal lines MQ and NQ. The gray values ​​on these two lines are resampled and smoothed to obtain a set of gray data points. and The variance of the two point sets is calculated using the following formula, which is the gray-level variance of the two diagonals: ; in, and Let A and B be the grayscale variance along the two diagonals, and A and B be the number of pixels in the two grayscale data point sets. and Let m be the mean gray level of two gray-level data points. i grayscale data point set The grayscale value of the i-th pixel in n j grayscale data point set The grayscale value of the j-th pixel.

[0086] After specific calculations, we obtained > This indicates that the stripe direction is from the upper right corner to the lower left corner of the image, therefore two sets of pitch and roll angles are given: α1=+ β1=+ and α2=- β2=- .

[0087] Preferably, in step S3 of this embodiment, the attitude of the sample is adjusted using the carrier according to each group of candidate attitude angles, and the stripe change information after each attitude adjustment is obtained, including: S31: Adjust the attitude of the sample using the carrier according to each group of candidate attitude angles, and obtain the adjusted white light interference image after each attitude adjustment; S32: Based on the initial white light interference image and the adjusted white light interference image after each attitude adjustment, obtain the fringe change information after each attitude adjustment; wherein, the fringe change information includes the fringe width change and the fringe direction change.

[0088] Using each set of candidate attitude angles as a reference, the carrier is controlled to adjust the attitude of the sample, which can specifically verify the influence of different combinations of attitude angle components on the actual attitude of the sample. By acquiring the adjusted white light interference image (i.e., the adjusted white light interference image) and comparing it with the initial image (i.e., the initial white light interference image), the fringe change information (including changes in fringe width and fringe direction) can be analyzed. This allows for a direct assessment of whether the attitude adjustment has brought the fringes closer to an ideal state (e.g., uniform fringe width and consistent direction). The method described in this embodiment, which verifies the attitude angle components through actual adjustment feedback, can effectively compensate for the sign ambiguity that may exist when relying solely on image feature calculations, providing a reliable experimental basis for finally determining the accurate attitude angle.

[0089] Preferably, in step S3 of this embodiment, it is determined sequentially whether the stripe change information after each adjustment meets the preset change conditions, including: For any stripe change information after a posture adjustment, if the stripe width increases after the stripe width change indicator is adjusted, and the stripe direction approaches the preset ideal direction after the stripe direction change indicator is adjusted, then the corresponding stripe change information is determined to meet the preset change conditions; otherwise, the corresponding stripe change information is determined not to meet the preset change conditions, and the judgment of the stripe change information after the next adjustment continues.

[0090] In determining whether the fringe change information meets the preset change conditions, if the fringe width change indicator increases after adjustment, and the fringe direction change indicator approaches the preset ideal direction after adjustment, this indicates that the current attitude adjustment direction is correct, meaning the selected candidate attitude angle component can effectively improve the sample's attitude deviation. Conversely, if the fringe width change indicator decreases after adjustment, or the fringe direction change indicator deviates from the preset ideal direction after adjustment, it indicates that the current candidate attitude angle component may have a sign error, requiring the attempt to adjust another set of candidate attitude angle components. Through this step-by-step judgment and verification method, the most accurate attitude angle that best matches the actual situation can be selected from multiple sets of candidate attitude angle components, thereby achieving precise correction of the sample's attitude angle.

[0091] For example, when using candidate attitude angular components α1=+ β1=+ After adjustment, if the fringe width significantly increases and the fringe direction gradually approaches the preset horizontal or vertical ideal direction, then the set of attitude angular components can be determined to be effective. If the fringe width decreases or the direction becomes more disordered after adjustment, then it is necessary to switch to another set of candidate attitude angular components α2=- β2=- Adjustments and verifications were performed until the accurate attitude angle component that could make the stripe changes meet the preset conditions was found.

[0092] for Figure 4 The image shown illustrates that when attitude angle calculation yields two sets of attitude angle components α1=+ β1=+ and α2=- β2=- Then, by adjusting the vehicle to make the pitch angle rotation angle α' = -α1 and the roll angle rotation angle β' = -β1, images were acquired again, the reference point P was re-found, and the horizontal and vertical profile lines were redrawn, as shown below. Figure 8 As shown. Then re-acquire the grayscale profile sequence, as shown. Figure 9A and Figure 9B As shown, no peaks were observed in either the horizontal or vertical directions. Therefore, it was determined that the adjustment directions for the pitch and roll angles were correct. The original first set of pitch and roll angles (i.e., α1=+) β1=+ If the actual tilt angle is 0, then the original second set of pitch and roll angles are considered as the actual tilt angles.

[0093] Example 2 An automatic white light interferometric attitude angle correction system is applied to the automatic white light interferometric attitude angle correction method in Embodiment 1, such as... Figure 10 As shown, the system includes: The image acquisition module is used to acquire the initial white light interference image obtained by the white light interferometry system from the sample. The attitude calculation module calculates the initial white light interferometric image to obtain at least one set of candidate attitude angles for the sample; each set of candidate attitude angles includes attitude angle components of the sample in at least two different axes; The adjustment and verification module is used to adjust the attitude of the sample using the carrier according to each group of candidate attitude angles, and to obtain the stripe change information after each attitude adjustment; to determine whether the stripe change information after each adjustment meets the preset change conditions, and to determine the candidate attitude angle corresponding to the stripe change information that meets the preset change conditions as the target attitude angle, thereby completing the correction.

[0094] In this embodiment, the image acquisition module acquires the initial white light interference image of the entire measurement system when measuring the sample. This image contains information about the fringe formed by the interference between the incident light reflected from the sample surface and the reflected light reflected from the reference surface (i.e., the surface of the reference mirror in the measurement system). The shape, spacing, and direction of these fringes directly reflect the attitude state of the sample and can provide a data basis for subsequent attitude calculation. The attitude calculation module performs a series of precise image processing, analysis, and calculation on the initial white light interference image, which can extract the attitude angular components of the sample in different directions from the complex interference fringes, thereby generating candidate attitude angles. Finally, the adjustment and verification module verifies and filters the candidate attitude angles through actual attitude adjustment and feedback of fringe change information, ensuring that the finally determined target attitude angle can adjust the sample to the ideal measurement attitude.

[0095] The automatic white light interferometric attitude angle correction system in this embodiment achieves automated and high-precision correction of the sample attitude angle through the combination of image analysis and adjustment verification. It avoids complex fringe direction identification and phase calculation, significantly simplifies the processing flow, and effectively solves the problem of fringe variation caused by sample / carrier pitch angle deviation, which is not effectively addressed in existing white light interferometric measurement technology due to the lack of sample / carrier attitude angle correction. It can significantly improve the accuracy and stability of white light interferometric measurement, improve measurement efficiency and system robustness, and lay a solid foundation for subsequent high-precision three-dimensional topography measurement. It is especially suitable for fields with high measurement accuracy requirements, such as precision machining, semiconductor manufacturing, and optical component inspection.

[0096] The functions of each module in the white light interferometric attitude angle automatic correction system described in this embodiment are the same as the method steps of the white light interferometric attitude angle automatic correction method described in Embodiment 1. Therefore, for details not covered in this embodiment, please refer to Embodiment 1 and... Figures 1 to 9B The specific details will not be repeated here.

[0097] Example 3 This embodiment also provides an automatic white light interferometric attitude angle correction device, including a processor, a memory, and a computer program stored in the memory and run on the processor. When the computer program runs, it implements the method steps in the automatic white light interferometric attitude angle correction method of Embodiment 1.

[0098] By using a computer program stored in memory and running on a processor, and combining image analysis with adjustment verification, automated and high-precision correction of the sample attitude angle is achieved. This avoids complex fringe direction identification and phase calculation, significantly simplifies the processing flow, and effectively solves the problem of fringe variation caused by sample / carrier pitch angle deviation, which is not effectively addressed in existing white light interferometry techniques due to the lack of correction for sample / carrier attitude angle. It can significantly improve the accuracy and stability of white light interferometry, increase measurement efficiency and system robustness, and lay a solid foundation for subsequent high-precision three-dimensional topography measurement. It is especially suitable for fields with high measurement accuracy requirements, such as precision machining, semiconductor manufacturing, and optical component inspection.

[0099] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the computer device, connecting all parts of the computer device through various interfaces and lines.

[0100] Memory can be used to store computer programs and / or models. The processor performs various functions of the computer device by running or executing the computer programs and / or models stored in the memory, and by accessing data stored in the memory. Memory can primarily include a program storage area and a data storage area. The program storage area can store the operating system and at least one application program required for a function (e.g., sound playback, image playback, etc.); the data storage area can store data created based on the use of the mobile phone (e.g., audio data, video data, etc.). Furthermore, memory can include high-speed random access memory, and can also include non-volatile memory, such as hard disks, RAM, plug-in hard disks, SmartMedia Cards (SMC), Secure Digital (SD) cards, Flash Cards, at least one disk storage device, flash memory device, or other volatile solid-state storage devices.

[0101] It should be understood that each block of a flowchart and / or block diagram, and combinations of blocks in a flowchart and / or block diagram, can be implemented by a computer program. These computer programs can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing device to produce a machine, such that instructions executable by the processor of the computer or other programmable data processing device generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0102] These computer programs may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0103] These computer programs may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0104] This embodiment also provides a computer storage medium, which includes at least one instruction that, when executed by a computer, implements the method steps in the automatic white light interferometric attitude angle correction method of Embodiment 1.

[0105] By executing a computer storage medium containing at least one instruction, and through a combination of image analysis and adjustment verification, automated and high-precision correction of the sample attitude angle is achieved. This avoids complex fringe direction identification and phase calculation, significantly simplifies the processing flow, and effectively solves the problem of fringe variation caused by sample / carrier pitch angle deviation, which is not effectively addressed in existing white light interferometry techniques due to the lack of correction for sample / carrier attitude angle. It can significantly improve the accuracy and stability of white light interferometry, enhance measurement efficiency and system robustness, and lay a solid foundation for subsequent high-precision three-dimensional topography measurement. It is particularly suitable for fields with high measurement accuracy requirements, such as precision machining, semiconductor manufacturing, and optical component inspection.

[0106] Similarly, for details not covered in this embodiment, please refer to Embodiment 1, Embodiment 2, and... Figures 1 to 10 The specific details will not be repeated here.

[0107] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for automatic correction of white light interference attitude angle, characterized in that, For use in a white-light interferometry system including a carrier, the method includes: Acquire the initial white light interferometry image obtained from the sample measured by the white light interferometry system; The initial white light interferometric image is solved to obtain at least one set of candidate attitude angles for the sample; each set of candidate attitude angles includes attitude angle components of the sample in at least two different axes; The sample is adjusted using the carrier according to each set of candidate attitude angles, and the stripe change information after each attitude adjustment is obtained. It is then determined whether the stripe change information after each adjustment meets the preset change conditions, and the candidate attitude angle corresponding to the stripe change information that meets the preset change conditions is determined as the target attitude angle to complete the correction.

2. The automatic white light interference attitude angle correction method according to claim 1, characterized in that, Each group of candidate attitude angles includes the attitude angle components of the sample in the lateral and longitudinal directions; The initial white light interferometry image is solved to obtain at least one set of candidate attitude angles for the sample, including: A reference point is determined in the initial white light interference image, and grayscale profile sequences are obtained along the horizontal and vertical directions, respectively, with the reference point as the center. Effective peak detection is performed on the grayscale profile sequence in each direction to obtain the peak detection information corresponding to the grayscale profile sequence in each direction. When the peak detection information indicates that there is no valid peak or only one valid peak in the gray-scale profile sequence in the corresponding direction, the attitude angle is assigned to the corresponding gray-scale profile sequence according to the preset rules to obtain the attitude angle component of the sample in the corresponding direction. When the peak detection information indicates that there are two or more valid peaks in the grayscale profile sequence in the corresponding direction, the theoretical fringe period of the white light interferometry system is calculated in advance, and the attitude angle component of the sample in the corresponding direction is calculated based on the theoretical fringe period and the peak detection information in the direction where valid peaks exist.

3. The automatic white light interference attitude angle correction method according to claim 2, characterized in that, When the peak detection information indicates that there is no valid peak or only one valid peak in the grayscale profile sequence in the corresponding direction, the attitude angle component of the sample in the corresponding direction is assigned a value of 0.

4. The automatic white light interference attitude angle correction method according to claim 2, characterized in that, When the peak detection information indicates that there is no valid peak in the grayscale profile sequence in the corresponding direction, the peak detection information includes the number of valid peaks, and the number of valid peaks is specifically 0; when the peak detection information indicates that there is only one valid peak in the grayscale profile sequence in the corresponding direction, the peak detection information includes the number of valid peaks, and the number of valid peaks is specifically 1. When the peak detection information indicates that there are two or more valid peaks in the grayscale profile sequence in the corresponding direction, the peak detection information includes the number of valid peaks in the corresponding direction, the position of each valid peak, and the stripe spacing based on the position and number of valid peaks, and the number of valid peaks is greater than or equal to 2. Effective peak detection is performed on the grayscale profile sequence in each direction to obtain the peak detection information corresponding to the grayscale profile sequence in each direction, including: A grayscale profile sequence in any direction is selected, and the first-order difference sign change method is used to detect the selected grayscale profile sequence to determine multiple peak candidate points in the selected grayscale profile sequence. The validity of all peak candidate points in the selected grayscale profile sequence is judged. When all peak candidate points in the selected grayscale profile sequence are invalid, the number of valid peaks in the selected grayscale profile sequence is marked as 0; when only one peak candidate point in the selected grayscale profile sequence is valid, the number of valid peaks in the selected grayscale profile sequence is marked as 1. When at least two of the candidate peaks are valid, the number of valid peaks in the selected grayscale profile sequence is counted, and the position of each valid peak in the selected grayscale profile sequence is determined. The position of the first peak and the position of the last peak are determined from all the positions of the valid peaks, and the corresponding stripe spacing is calculated based on the position of the first peak, the position of the last peak and the corresponding number of valid peaks. Using the same method, peak detection information was obtained from another grayscale profile sequence.

5. The automatic white light interference attitude angle correction method according to claim 4, characterized in that, The specific calculation formulas for the stripe spacing in the transverse and axial directions are as follows: ； Where a and b are the stripe spacing in the transverse and axial directions, respectively, and N and M are the number of effective peaks in the transverse and axial directions, respectively. first and u last These represent the positions of the first and last peaks in the horizontal direction, respectively. first and v last These represent the positions of the first and last peaks in the vertical direction.

6. The automatic white light interference attitude angle correction method according to claim 4, characterized in that, A grayscale profile sequence consists of grayscale values ​​from multiple sampling points; The first-order difference sign change method is used to detect the selected grayscale profile sequence, and multiple peak candidate points in the selected grayscale profile sequence are identified, including: In the selected grayscale profile sequence, the grayscale change of each sampling point is calculated, and the grayscale change of each sampling point is defined as the corresponding first-order difference. For the nth sampling point, if the first difference of the nth sampling point is not greater than 0 and the first difference of the (n-1)th sampling point is greater than 0, then the nth sampling point is determined to be a peak candidate point; otherwise, the nth sampling point is determined not to be a peak candidate point. Following the same determination method, the first difference of each sampling point in the selected grayscale profile sequence is traversed in turn to determine multiple peak candidate points in the selected grayscale profile sequence.

7. The automatic white light interference attitude angle correction method according to claim 4, characterized in that, A grayscale profile sequence consists of grayscale values ​​from multiple sampling points; The validity of all candidate peak points in the selected grayscale profile sequence is evaluated, including: In the selected grayscale profile sequence, select any peak candidate point. If the difference between the grayscale value of the selected peak candidate point and the global grayscale mean is greater than or equal to the preset significance threshold, the selected peak candidate point will be retained; otherwise, the selected peak candidate point will be removed. Iterate through each peak candidate point in the selected grayscale profile sequence, and filter all peak candidate points in the selected grayscale profile sequence in the same way. The remaining peak candidate points after filtering are sorted in descending order of grayscale value to obtain the peak candidate sequence. A greedy selection method is used to select the peak candidate point with the highest current grayscale value in the peak candidate sequence as the potential peak. With the potential peak as the center, the remaining peak candidate points in the peak candidate sequence within the preset minimum spacing range are removed. When all peak candidate points in the peak candidate sequence have been processed, the validity of the selected grayscale profile sequence is judged, and the corresponding valid peak set is obtained.

8. The automatic white light interference attitude angle correction method according to claim 4, characterized in that, The peak detection information corresponding to a grayscale profile sequence in any chosen direction is used. When the selected peak detection information indicates that there are two or more valid peaks in the grayscale profile sequence in the corresponding direction, the attitude angle component of the sample in the corresponding direction is calculated based on the theoretical fringe period and the peak detection information in the direction where valid peaks exist. This includes: Based on the theoretical fringe period and the fringe spacing in the selected direction, the absolute value of the attitude angle of the sample in the selected direction is calculated. The specific formulas for calculating the absolute values ​​of the attitude angles in the lateral and axial directions are as follows: ； in, and , respectively, are the absolute values ​​of the attitude angles in the lateral and axial directions, a and b are the stripe spacing in the lateral and axial directions, respectively, and l is the theoretical stripe period; If the peak detection information corresponding to the grayscale profile sequence in another direction indicates that there is no valid peak in the corresponding direction, two attitude angle components with opposite signs are generated based on the absolute value of the attitude angle of the sample in the selected direction. If the peak detection information corresponding to the grayscale profile sequence in another direction also indicates that there is a valid peak in the corresponding direction, two diagonals are constructed based on the reference point and the valid peaks closest to the reference point in the two directions; the grayscale variances corresponding to the two diagonals are obtained, and the attitude angle sign of the sample in the selected direction is determined according to the comparison result between the grayscale variances of the two diagonals; the attitude angle components of the sample in the selected direction are generated according to the absolute value and attitude angle sign of the sample in the selected direction.

9. The automatic white light interference attitude angle correction method according to claim 2, characterized in that, The theoretical fringe period of the white light interferometry system is calculated in advance, including: The spectral distribution of the white light interferometry system is obtained, and the spectral distribution is transformed to obtain the coherence function of the white light interferometry system; The corresponding interference light intensity signal is determined based on the coherence function of the white light interferometry system. The oscillation waveform of the interference light intensity signal near zero optical path difference is extracted, and the theoretical fringe period is obtained based on the oscillation waveform.

10. The automatic white light interference attitude angle correction method according to claim 9, characterized in that, The specific formula for calculating the coherence function of the white light interferometry system is as follows: ； in, Let S(v) be the coherence function of the white light interferometry system, v be the frequency, and S(v) be the spectral density obtained by frequency transformation of the spectral distribution. This represents the time delay, where i is the imaginary unit; The specific formula for determining the interference light intensity signal is: ； in, The interference light intensity signal of the white light interferometry system. This is for the operation of taking the real part.

11. The automatic white light interference attitude angle correction method according to claim 2, characterized in that, Centered on the reference point, grayscale profile sequences are obtained along the horizontal and vertical directions, including: In the initial white light interference image, with the reference point as the center, lines passing through the reference point are drawn along the horizontal and vertical directions respectively to obtain the horizontal profile line and the vertical profile line; The pixel grayscale values ​​of the horizontal profile line and the vertical profile line are extracted respectively to obtain the initial grayscale profile image of the horizontal profile and the initial grayscale profile image of the vertical profile. The initial horizontal grayscale profile image is resampled and smoothed to obtain a horizontal grayscale profile sequence; the initial vertical grayscale profile image is resampled and smoothed to obtain a vertical grayscale profile sequence.

12. The automatic white light interference attitude angle correction method according to claim 1, characterized in that, Based on each group of candidate attitude angles, the sample is attitude-adjusted using the carrier, and the fringe change information after each attitude adjustment is obtained, including: Based on each group of candidate attitude angles, the attitude of the sample is adjusted using the carrier, and the adjusted white light interference image is obtained after each attitude adjustment. Based on the initial white light interference image and the adjusted white light interference image after each attitude adjustment, fringe change information after each attitude adjustment is obtained; wherein, the fringe change information includes fringe width change and fringe direction change.

13. The automatic white light interference attitude angle correction method according to claim 12, characterized in that, The system sequentially determines whether the stripe change information after each adjustment meets the preset change conditions, including: For any stripe change information after a posture adjustment, if the stripe width increases after the stripe width change indicator is adjusted, and the stripe direction approaches the preset ideal direction after the stripe direction change indicator is adjusted, then the corresponding stripe change information is determined to meet the preset change conditions; otherwise, the corresponding stripe change information is determined not to meet the preset change conditions, and the judgment of the stripe change information after the next adjustment continues.

14. An automatic white light interferometric attitude angle correction system, characterized in that, Applied to the automatic white light interferometric attitude angle correction method as described in any one of claims 1 to 13, the system comprises: The image acquisition module is used to acquire the initial white light interference image obtained by the white light interferometry system from the sample. The attitude calculation module calculates the initial white light interferometric image to obtain at least one set of candidate attitude angles for the sample; each set of candidate attitude angles includes attitude angle components of the sample in at least two different axes; The adjustment and verification module is used to adjust the attitude of the sample using a carrier according to each group of candidate attitude angles, and to obtain the stripe change information after each attitude adjustment; it sequentially determines whether the stripe change information after each adjustment meets the preset change conditions, and determines the candidate attitude angle corresponding to the stripe change information that meets the preset change conditions as the target attitude angle, thus completing the correction.

15. An automatic white light interference attitude angle correction device, characterized in that, It includes a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed, implements the method steps of the automatic white light interference attitude angle correction method as described in any one of claims 1 to 13.

16. A computer storage medium, characterized in that, The computer storage medium includes at least one instruction that, when executed by a computer, implements the method steps of the automatic white light interferometric attitude angle correction method as described in any one of claims 1 to 13.

Images