A filter film thickness detection device and a detection method

Abstract

The application relates to the technical field of light filtering film, and discloses a light filtering film thickness detection device and a detection method.The method comprises the following steps: acquiring a first parallel light beam, a second parallel light beam and a third parallel light beam; fixing a to-be-detected light filtering film sample on a sample table and measuring the to-be-detected light filtering film sample to obtain a reflection spectrum and an initial refractive index parameter; irradiating the to-be-detected light filtering film sample with the first parallel light beam, the second parallel light beam and the third parallel light beam, dividing a measurement area into a plurality of subareas according to the initial refractive index parameter, and obtaining an optimal thickness value and a refractive index value of each subarea through least square fitting; and establishing a mapping relationship between spectral reflectivity and thickness based on the optimal thickness value, the refractive index value and the reflection spectrum, and obtaining a target thickness distribution diagram through an iterative algorithm solution and error compensation calculation.The application solves the problem of uneven measurement of large-area light filtering film thickness, and significantly improves the precision, efficiency and reliability of light filtering film thickness detection.

Description

Technical Field

[0001] This invention relates to the field of optical filter technology, and in particular to an optical filter thickness detection device and detection method. Background Technology

[0002] Traditional methods for measuring filter thickness mainly rely on ellipsometers and single-wavelength interferometry. While these methods can meet basic measurement needs, they have significant limitations when dealing with nanometer-level precision requirements and complex filter structures. In particular, when measuring large-area, non-uniform, or multilayer composite filter structures, single-wavelength interferometry is prone to phase ambiguity, leading to inaccurate measurement results. While ellipsometers offer higher precision, they are slow, require high sample surface roughness, and struggle to achieve rapid large-area scanning measurements, failing to meet the demands of efficient production in the modern optical thin film industry.

[0003] Spectral reflectance analysis, as a precise tool for measuring filter thickness, has become the standard testing method in the optical thin film industry. However, due to the need to process spectral data at a large number of wavelengths, traditional reflectance analysis methods suffer from low computational efficiency and are easily affected by the choice of initial values. This is especially true for multilayer composite filters, where the inversion calculation process is complex, has poor convergence, and limited measurement accuracy. Furthermore, traditional methods do not adequately consider environmental factors such as temperature changes, incident angle fluctuations, and the wavelength stability of the measurement system itself, making it difficult to systematically compensate for various error sources. This results in insufficient reliability of measurement results in actual production environments. Summary of the Invention

[0004] The main objective of this invention is to provide a filter film thickness detection device and detection method. This invention solves the problem of uneven measurement of large-area filter film thickness and significantly improves the accuracy, efficiency and reliability of filter film thickness detection.

[0005] To achieve the above objectives, the present invention provides a method for detecting the thickness of a filter film, comprising the following steps: The first wavelength light source, the second wavelength light source, and the third wavelength light source are processed by an optical collimation system and an angle control device to obtain a first parallel beam, a second parallel beam, and a third parallel beam. The filter film sample to be tested is fixed on the sample stage and the incident angle is set. The filter film sample to be tested is measured by a spectrophotometer to obtain the reflectance spectrum and initial refractive index parameters. The sample of the filter film to be tested is illuminated by the first parallel beam, the second parallel beam and the third parallel beam, and the measurement area is divided into multiple sub-regions according to the initial refractive index parameter. The optimal thickness value and refractive index value of each sub-region are obtained by fitting using the least squares method. Based on the optimal thickness value, the refractive index value, and the reflection spectrum, a mapping relationship between spectral reflectance and thickness is established, and the target thickness distribution map is obtained by solving the problem through an iterative algorithm and error compensation calculation.

[0006] The present invention also provides a filter film thickness detection device, comprising: The processing module is used to process the first wavelength light source, the second wavelength light source, and the third wavelength light source through an optical collimation system and an angle control device to obtain a first parallel beam, a second parallel beam, and a third parallel beam. The measurement module is used to fix the filter film sample to be tested on the sample stage and set the incident angle, and measure the filter film sample to be tested by a spectrophotometer to obtain the reflectance spectrum and initial refractive index parameters. The fitting module is used to irradiate the filter film sample under test with the first parallel beam, the second parallel beam and the third parallel beam, divide the measurement area into multiple sub-regions according to the initial refractive index parameter, and obtain the optimal thickness value and refractive index value of each sub-region by fitting with the least squares method. The calculation module is used to establish a mapping relationship between spectral reflectance and thickness based on the optimal thickness value, the refractive index value and the reflection spectrum, and obtain the target thickness distribution map by solving the problem through an iterative algorithm and error compensation calculation.

[0007] In summary, the technical solution provided by this invention effectively expands the measurement range and reduces errors caused by phase ambiguity by employing a collaborative measurement strategy using three different wavelength light sources. Simultaneously, it solves the problem of non-uniform thickness measurement of large-area filter films by combining adaptive region partitioning technology. This method establishes a systematic multi-factor error compensation mechanism, including compensation for refractive index error, incident angle error, temperature error, wavelength error, and phase measurement error, thereby improving the accuracy of thickness measurement. Data processing efficiency is improved through GPU parallel computing technology and adaptive mesh refinement strategies. Precise spectral reflectance-thickness mapping models are established for different types of filter films, enhancing the accuracy and applicability of the inversion algorithm. Relative errors have been reduced through cross-validation and comparative measurements, making it suitable for online monitoring and quality control in industrial production lines, significantly improving the accuracy, efficiency, and reliability of filter film thickness detection. Attached Figure Description

[0008] Figure 1 This is a schematic diagram of the steps of a filter film thickness detection method in one embodiment of the present invention; Figure 2 This is a structural block diagram of a filter film thickness detection device in one embodiment of the present invention.

[0009] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0010] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0011] Reference Figure 1 This embodiment provides a method for detecting the thickness of a filter film, including the following steps: S1 processes the first wavelength light source, the second wavelength light source, and the third wavelength light source through an optical collimation system and an angle control device to obtain a first parallel beam, a second parallel beam, and a third parallel beam. The selected laser source was wavelength-tuned to meet the wavelength requirements of high-precision interferometry. Three lasers with different wavelengths were configured: blue, green, and red lasers. The first wavelength source was a blue laser, whose center wavelength was precisely adjusted to 450±5nm using a wavelength tuning device. This wavelength band has the characteristic of a relatively short wavelength, which is used to improve the spatial resolution of the interference fringes. The second wavelength source was a green laser, whose center wavelength was tuned to 532±3nm, which is a medium wavelength in the visible light spectrum, helping to form composite interference with other wavelengths and expanding the measurement range. The third wavelength source was a red laser, whose center wavelength was stabilized around 650±5nm by controlling its emission band. This longer wavelength band helps to enhance the sensitivity of the composite wavelength to large-scale film thickness variations. Through the above tuning, the first, second, and third center wavelengths were obtained, respectively. To ensure that the three laser beams can form spatially consistent parallel beams stably and consistently in subsequent interferometry measurements, the three beams were collimated sequentially through the same or a cooperating optical collimation system. The collimation system is equipped with a precisely adjustable lens group. Lens parameters, including focal length, position, and spacing, are optimized based on three determined center wavelengths to ensure that lasers of different wavelengths are converted into parallel beams with low divergence angles and flat wavefronts after passing through the lens system. For blue light, due to its shorter wavelength, optical glass with low chromatic aberration is selected to avoid optical path distortion caused by dispersion; green and red light require lenses with low curvature to maintain their wavefront stability. After processing by the collimation system, the three laser beams form initial parallel beams and propagate along the optical axis. An angle control device is introduced to further deflect and adjust these three initial parallel beams to achieve angle calibration. This angle control device is a rotary stage or fine-tuning prism mechanism with sub-angular resolution and a control accuracy better than 0.01°, allowing users to precisely set the laser incident angle within any range of 0° to 85° according to measurement requirements. This device enables precise deflection of three laser beams after collimation, forming angle-calibrated parallel beams with uniform incident angle parameters. This ensures that the beam incident on the filter sample surface is completely consistent with the assumptions in the interference theory model, thereby eliminating phase distortion caused by angle errors. After the three laser beams are collimated and angled, they are split into reference and measurement paths for interference. This process is accomplished by a highly stable beam splitter, which divides each laser beam into two paths proportionally or according to a set ratio. One path serves as the reference beam, which does not interact with the sample and is directly reflected and transmitted to the spectral detector to form the reference phase in the interference signal. The other path is the measurement beam, which, after angle control, illuminates the surface of the filter sample at a specific incident angle. Part of it is reflected by the sample's upper surface and interferes with the reference beam in the interference detection module, forming interference fringes containing film thickness information.The interference pattern was then acquired in real time by a two-dimensional photodetector array and digitally processed. Through the synergistic effect of the above-mentioned light source wavelength modulation, optical collimation, angle adjustment and path beam splitting, high-quality interferometry at three wavelengths was finally achieved on the filter film sample, thereby obtaining the first parallel beam, the second parallel beam and the third parallel beam respectively.

[0012] S2, Fix the filter film sample to be tested on the sample stage and set the incident angle. Measure the filter film sample to be tested using a spectrophotometer to obtain the reflectance spectrum and initial refractive index parameters. Specifically, the filter sample to be tested is mounted on a sample stage equipped with a high-precision three-dimensional micro-adjustment mechanism. This sample stage is equipped with a precision displacement control device, which adjusts the spatial position of the sample within a nanometer range to ensure that the orientation of the filter in the XYZ three-dimensional directions meets the preset requirements. The displacement control device fine-tunes the sample to ensure the surface flatness of the filter while maintaining its stable and fixed position, avoiding systematic errors caused by displacement or vibration during subsequent optical measurements. After the sample is successfully fixed, the filter layer type is identified through reflectance pre-scanning or user-input technical parameters, determining whether the sample is a dielectric multilayer film, a metal film, or a hybrid thin film. Based on the identified layer type, preset standard incident angle parameters are automatically matched; for example, the incident angle is set to approximately five degrees for dielectric multilayer films, fifteen degrees for metal films, and ten degrees for hybrid thin films. Combined with allowable minute adjustment ranges, a precise angle setting value is obtained, which serves as the basis for the incident angle of incident light in subsequent measurements. Based on the set incident angle, a high-resolution spectrophotometer configured in the system is used to perform spectral scanning on the sample. This spectrophotometer continuously scans within the visible light range of 380 nm to 780 nm, recording the reflectance corresponding to each wavelength with a wavelength resolution up to 0.5 nm, thus obtaining a reflectance spectrum dataset. This reflectance spectrum contains interference oscillation information caused by the thickness and optical properties of the filter film, and reflects the intrinsic characteristics of the film material's refractive index as a function of wavelength. Based on the reflectance spectrum, a commonly used optical Cauchy model is constructed, which describes the relationship between refractive index and wavelength in mathematical function form. To obtain more accurate model parameters, the coefficients of the Cauchy model are fitted using the least squares method. By minimizing the squared error between the reflectance curve and the model calculation results, the initial refractive index parameters of the filter film within the measured wavelength range are finally obtained.

[0013] S3, the sample of the filter film to be tested is irradiated with the first parallel beam, the second parallel beam and the third parallel beam, and the measurement area is divided into multiple sub-regions according to the initial refractive index parameter. The optimal thickness value and refractive index value of each sub-region are obtained by fitting with the least squares method. It should be noted that a configured electronic shutter controller sequentially activates three light sources of different wavelengths. This controller possesses high response speed and precise timing control capabilities, ensuring that the three beams of light sequentially illuminate the filter film sample surface within their respective independent time windows, avoiding optical crosstalk caused by simultaneous interference of multiple wavelengths. During the activation process, the blue laser emitted by the first wavelength light source forms the first parallel beam illuminating the filter film. Subsequently, the first light source is turned off and the second wavelength green laser is turned on to form the second parallel beam. Finally, the third wavelength red laser is activated sequentially to complete the third parallel beam illumination process. Each laser irradiation forms a clear interference fringe pattern on the filter film surface. These fringes record information about the optical path difference of the film layers and are directly related to the film thickness distribution. The high-resolution two-dimensional photodetector array in the system performs high-precision acquisition of the original interference fringe images generated during these three irradiations, forming a digital interference image dataset. These original images undergo dark-field correction processing, that is, acquiring the background dark-field image without laser irradiation and performing pixel-level difference with the actual acquired image to remove the background response deviation of the detector system itself, thereby obtaining the corrected interference image. Background noise reduction is performed on the corrected interferometric image. A wavelet transform-based image denoising algorithm, such as the Daubechies wavelet function, is used for multi-level decomposition to remove high-frequency noise components, thus preserving the true spatial structural features of the interference fringes and obtaining the interference image. An improved five-step phase-shifting method is used to extract phase from each set of interferometric images. This method sets a phase-shifting step size and sequentially acquires interferometric images at multiple phase-shifting states, extracting optical phase information related to the fringe structure to obtain wrapper phase maps corresponding to three wavelengths. Since the wrapper phase map exhibits a 2π periodic jump characteristic, it is converted into a continuous phase distribution map using a corresponding phase unpacking algorithm. A quality-guided phase unpacking technique is employed, which prioritizes unpacking regions with stable phase changes based on image quality factors, gradually expanding towards boundaries and abrupt change regions to eliminate jumps and obtain continuous first, second, and third phase distribution maps. Based on the initial refractive index parameters obtained using a spectrophotometer, the entire interferometric measurement region was spatially divided. Regions with similar phase characteristics were further subdivided into multiple sub-regions. This division dynamically determined the region boundaries based on a coarse thickness gradient and phase change trend, ensuring that the optical properties within each sub-region were as consistent as possible. Within each sub-region, a fitting model was established between light intensity, film thickness, and refractive index. The objective function was optimized using the least squares method to minimize the error between the theoretical and measured light intensity, thereby solving for the optimal thickness and optimal refractive index values ​​for each sub-region.

[0014] The first synthesized wavelength is calculated by combining the first and second phase distribution maps. Its physical meaning is to extend the effective measurement depth by utilizing the interference beat effect between two nearest-neighbor wavelengths. Then, the second synthesized wavelength is constructed using the first and third phase distribution maps to obtain a longer-scale optical modulation response. By designing the synthesized wavelength, the 2π ambiguity problem present in traditional single-wavelength interferometry is effectively avoided, and the non-jump range of film thickness measurement is significantly broadened. After the synthesized wavelength calculation is completed, a short-wavelength phase map and a long-wavelength phase map are constructed based on these two synthesized wavelengths. The short-wavelength phase map has higher phase sensitivity, while the long-wavelength phase map has stronger thickness penetration capability. Combining these two types of information enables multi-scale fusion from local details to the overall contour. Based on the long-wavelength phase map and the initial refractive index parameters obtained by fitting the Cauchy model, the coarse film thickness value at each pixel is calculated, generating an original thickness distribution map that approximately reflects the spatial thickness distribution trend of the filter film surface. Because long synthesized wavelengths possess strong resistance to transitions, the original thickness map maintains global continuity while realistically representing film undulations over a large range. A dynamic threshold segmentation operation is performed on the original thickness distribution map to achieve adaptive division of the measurement region. This segmentation process automatically calculates the threshold step based on the minimum and maximum values ​​in the thickness distribution map and determines the number and boundaries of sub-regions based on the changing trend of the film thickness gradient within the region. This ensures that the thickness variation within each sub-region tends to be gradual, facilitating subsequent local modeling and fitting. After completing the sub-region division, local least squares fitting optimization is performed on each sub-region. The optimization objective function is to minimize the sum of squared residuals between the measured light intensity and the theoretically calculated light intensity, with variables including film thickness and refractive index. During the fitting process, an initial value selection strategy guided by gradient information is introduced, and the numerically stable Levenberg-Marquardt algorithm is used for nonlinear optimization, ensuring that each sub-region ultimately obtains a set of optimal thickness and refractive index values ​​with good convergence and high accuracy.

[0015] S4. Based on the optimal thickness value, refractive index value, and reflection spectrum, a mapping relationship between spectral reflectance and thickness is established. The target thickness distribution map is obtained by solving the problem through an iterative algorithm and calculating error compensation.

[0016] Specifically, a suitable optical modeling method is selected to establish a reflectance calculation model based on the specific structural characteristics of the filter film. For single-layer homogeneous film samples, a spectral reflectance calculation model based on the principle of interference optics is constructed. This model accurately simulates the reflectance variation trend at different wavelengths by considering the interrelationships between parameters such as film thickness, refractive index, and incident angle of light, obtaining a set of mapping relationships. This mapping relationship converts any given film thickness and refractive index into the expected reflectance curve. For multilayer composite filter film samples with more complex structures, a feature matrix calculation model is constructed. This model incorporates the optical characteristic parameters of each film material, such as refractive index, thickness, and optical admittance, into a matrix form, and obtains the overall optical behavior of the entire film system through layer-by-layer matrix multiplication, thus obtaining the reflectance mapping relationship of the multilayer film. The optimal thickness and refractive index values ​​of each sub-region obtained in the previous stage through phase inversion and local fitting are substituted into the corresponding reflectance mapping model as initial estimates for thickness inversion calculation. To improve inversion accuracy, the Newton-Raphson iterative method is introduced. By calculating the derivative relationship between reflectivity and thickness, the true thickness and refractive index values ​​are gradually approximated, minimizing the error between the theoretically calculated reflectivity and the actual measured reflectivity, thus obtaining a set of preliminary thickness inversion results. Since performing high-resolution inversion calculations simultaneously on a large-area filter sample requires a huge amount of computation, a parallel computing strategy is implemented to avoid the inefficiency of traditional serial algorithms. The entire filter surface is divided into multiple computational blocks, each independently assigned tasks according to a fixed pixel dimension. GPU acceleration technology is used to simultaneously solve for and iteratively update the reflectivity of all blocks in parallel, thereby significantly improving computational speed while maintaining inversion accuracy. After parallel processing, a detailed analysis is performed on local abrupt change regions in the thickness data. By performing boundary scanning and gradient analysis on the accelerated thickness map, the locations of thickness abrupt change regions are identified. An adaptive mesh refinement strategy is applied to these regions, making the computational units within the abrupt change regions denser to improve local fitting and inversion accuracy. After refinement, a higher resolution initial thickness and refractive index distribution map is reconstructed, and an error compensation process is performed based on this data. Compensation calculations for five typical error sources are sequentially applied to the initial thickness map: refractive index error, incident angle error, temperature error, wavelength error, and phase measurement error. Specifically, refractive index error is compensated by introducing a dispersion function correction term; incident angle error is corrected by fitting a thickness-to-angle sensitivity function; temperature error is corrected using a combination of thermal expansion and thermo-optical effect coefficients; wavelength error is updated using correction values ​​obtained from standard sample calibration; and phase error is smoothed and adjusted through statistical weighting of multiple measurements. After compensating for all the above error sources one by one, the final target thickness distribution map is generated.

[0017] The system identifies and corrects refractive index errors introduced by inaccurate estimation of intrinsic material parameters in the initial thickness distribution map. By introducing multiple auxiliary measurement wavelengths, measured reflectance data at these wavelengths are obtained on the filter sample. Based on this data, a more accurate dispersion function is constructed, and the initial refractive index value is dynamically adjusted using the corrected dispersion relation to obtain a new set of refractive index error compensation data. Differential sensitivity analysis is performed based on the compensated refractive index data. By establishing a response function of thickness to changes in the incident angle (i.e., a partial derivative model of thickness with respect to the incident angle), the trend of film thickness variation under different angular deviations is evaluated. The thickness data is then finely corrected using measured incident angle deviation values, resulting in a dataset including incident angle error compensation. Based on the existing incident angle compensation data, temperature error compensation is performed according to the thermal expansion coefficient and thermo-optic coefficient of the filter material. The system measures the deviation between the current ambient temperature and the reference temperature, introducing temperature response correction factors to the thickness and refractive index values ​​respectively, obtaining temperature error compensation data. This ensures that the film thickness information measured under different experimental environments has good thermal stability and comparability. Based on this, to eliminate systematic errors caused by wavelength shifts in the light source, a single-crystal silicon wafer with known thickness and optical properties is used as a standard reference material. By analyzing the deviation between its interference pattern and the theoretical model, the actual operating wavelengths of each laser source are deduced, and the wavelength parameters used in the inversion model are updated accordingly. Wavelength error correction is applied to the temperature error compensation data to obtain wavelength error compensation data. After compensating for the above four sources of error, the phase measurement error that is unavoidable in the measurement is addressed. By performing multiple repeated measurements on the same area, the average and standard deviation of each phase extraction result are calculated. A weighted average method combined with the confidence coefficients of each measurement is used to construct a final phase estimation map. The thickness data corresponding to this map is then applied to the previously compensated wavelength data to generate statistically robust phase error compensation data. The data results after all compensation steps are integrated to generate a target thickness distribution map.

[0018] In one example, a first wavelength light source, a second wavelength light source, and a third wavelength light source are processed by an optical collimation system and an angle control device to obtain a first parallel beam, a second parallel beam, and a third parallel beam, including: The wavelength of the first wavelength light source is tuned to a blue laser with a center wavelength of 450±5nm to obtain the first center wavelength. The wavelength of the second wavelength light source is tuned to a green laser with a center wavelength of 532±3nm to obtain the second center wavelength. Wavelength modulation is performed on the third wavelength light source, and the center wavelength of the third wavelength light source is adjusted to a red laser with a wavelength of 650±5nm to obtain the third center wavelength; The lens group parameters of the optical collimation system are set based on the first center wavelength, the second center wavelength and the third center wavelength. The three laser beams are collimated by the optical collimation system to obtain an initial parallel beam. An angle control device is used to adjust the deflection of the initial parallel beam to obtain an angle-calibrated parallel beam. The angle-calibrated parallel beam is split by a beam splitter, and each beam is divided into a reference beam and a measurement beam. The reference beam directly enters the spectral detector, and the measurement beam illuminates the surface of the filter film to be tested, resulting in a first parallel beam, a second parallel beam, and a third parallel beam.

[0019] In this example, three lasers with different center wavelengths are selected as the first, second, and third wavelength sources, corresponding to the blue, green, and red light bands, respectively. To ensure that these three lasers meet the predetermined spectral specifications, the wavelength of each laser is tuned. This process relies on the coordinated control of built-in temperature control elements and current modulation mechanisms to achieve fine-tuning of the center wavelength. The first wavelength source is stabilizing its output wavelength in the blue light range of 450±5nm by adjusting the laser's operating temperature and electric drive current, ensuring that it possesses the characteristics of a shorter wavelength, high spatial resolution, and strong interference sensitivity, thus obtaining the first center wavelength that meets the design specifications. The second wavelength source is tuned to the green light range of 532±3nm. This band is located in the middle of the visible spectrum, has good system compatibility and moderate penetration ability, and can provide a stable signal in regions with different film thicknesses, thus obtaining the second center wavelength. The third wavelength source is set in the red light region of 650±5nm through the same thermoelectric modulation and drive current adjustment. The longer red light wavelength has a better interference modulation depth, which helps maintain the clarity of interference fringes in thick film regions, thus obtaining the required third center wavelength. Once the center wavelengths of the three laser beams reach the aforementioned control targets, they are guided into an optical collimation system. This system consists of multiple precisely designed lens assemblies, capable of adjusting focal length, radius of curvature, and lens spacing. Its design principle is based on configuring the optimal lens group combination for different wavelengths according to the relationship between laser wavelength and divergence angle. Since different wavelengths of light will produce varying degrees of chromatic aberration and optical axis shift when propagating through the lenses, the propagation paths of blue, green, and red lasers in the lens system are calculated separately during the design process. The arrangement order of each lens group, optical axis concentricity, and focal length matching are adjusted to ensure that all three beams, after collimation, can be converted into high-quality initial parallel beams with low divergence angles, flat wavefronts, and uniform beam waists. To guarantee the collimation effect, an automatic alignment mechanism and an online beam quality monitoring device are provided. Through real-time feedback, the lens positions are adjusted to precisely convert the three laser beams from point source forms into collimated parallel beams of consistent quality. Although the three collimated beams already possess consistent parallelism in the spatial direction, directional deflection adjustment is still required via an angle control device. This angle control device consists of a high-precision electric rotating platform and a fine-tuning prism or beam deflection module. Its main function is to apply a tiny angular offset to the propagation direction of the light beam, thereby achieving precise setting of the incident angle. The device boasts a control accuracy of up to 0.01°, allowing the system to flexibly adjust the incident angle according to the filter's structure; for example, setting it to 5° for a dielectric film, 15° for a metallic film, and 10° for a hybrid film. Angle adjustment ensures the controllability of the three beams' directions and guarantees that they have incident conditions consistent with the theoretical model in subsequent interference, eliminating phase errors caused by incident angle deviations.After angle adjustment, the three laser beams form angle-calibrated parallel beams with the target incident angle, providing precise light source conditions for interferometric imaging. These three angle-calibrated parallel beams are sequentially introduced into a high-stability beam splitter for path separation. The beam splitter employs a partially transparent, partially reflective optical film structure design, which can rationally allocate each laser beam into two optical paths: one for reference and the other for measurement. The reference beam directly enters the spectral detector after beam splitting to record the standard optical path information in the interference signal that is unaffected by the sample. The measurement beam continues to propagate at its original incident angle and illuminates the surface of the filter film fixed on the sample stage. After reflection from the filter film surface, it superimposes with the reference beam in the interference system to form interference fringes. The phase change carried by the measurement beam after passing through the filter film is caused by the optical path difference resulting from the changes in film thickness and refractive index. By observing the phase difference between the measurement beam and the reference beam, a mapping relationship between film thickness and interference phase is established. Through the above process, the first, second, and third parallel beams are obtained.

[0020] In one example, the filter sample to be tested is fixed on the sample stage and the incident angle is set. The reflectance spectrum and initial refractive index parameters are obtained by measuring the filter sample using a spectrophotometer, including: The filter film sample to be tested is mounted on a sample stage with a three-dimensional fine-tuning mechanism. The position of the sample stage is adjusted by a displacement control device to obtain a fixed filter film sample to be tested. The film type of the filter film sample under test in a fixed state is identified to obtain the film type, and a specific incident angle is set according to the film type to obtain the angle setting value; Based on the angle setting value, the test filter film sample in a fixed state is scanned in the wavelength range of 380nm to 780nm using a spectrophotometer to obtain the reflectance spectrum; A Cauchy model was constructed based on the reflection spectrum, and the parameters of the Cauchy model were fitted by the least squares method to obtain the initial refractive index parameters.

[0021] In this example, the filter film sample to be measured is mounted on a sample stage equipped with a high-resolution three-dimensional fine-tuning mechanism, which enables nano-scale micro-displacement adjustment along three directions X, Y and Z respectively. During installation, the filter film sample is stably placed in the central area of the sample stage through a mechanical limit device or a vacuum adsorption structure, and then precision alignment is performed through the displacement control device matched with the sample stage. The displacement control system is equipped with a closed-loop feedback system and a high-resolution electric platform, which can achieve an adjustment step of 10 nanometers, and monitor the position change of the sample in real time through a high-precision encoder, so that the included angle between the normal direction of the sample and the incident light direction reaches an ideal state on the premise of ensuring the flatness of the sample surface. When the system judges that the sample is stable and accurately positioned, the sample can be regarded as in a fixed state. After the sample is fixed, the system calls its built-in film type identification algorithm to automatically identify the filter film. The identification is carried out by combining multiple methods such as material database, historical spectral data matching and user input information. By analyzing the basic parameters of the sample such as color and luster, initial surface reflectance, preparation process information, etc., it is determined whether the filter film belongs to a dielectric multilayer film, a metal film or a hybrid film structure. The identified film type affects the selection of the film model and more directly determines the basis for setting the subsequent incident angle of the light source. For example, if the system identifies the sample as a dielectric multilayer film, the incident angle is set to a small value, such as 5 degrees, to avoid multiple interference at an excessively large angle; if it is identified as a metal film, it is suitable to be set to 15 degrees because its reflectance characteristics are relatively stable with respect to the angle; if it is identified as a composite hybrid film, an intermediate value, such as 10 degrees, is taken to balance phase stability while retaining reflection sensitivity. The angle setting value of the current sample is calculated by matching the identification result with the incident angle database, and this value is automatically transmitted to the angle control module to complete the subsequent calibration preset of the incident direction of the light source. After the sample is fixed and the incident angle is set, the spectrophotometer is started to perform a comprehensive scan of the reflection spectrum of the filter film sample. The spectrophotometer has the characteristics of high wavelength resolution and low stray light, the scanning range covers the entire visible spectral band from 380 nanometers to 780 nanometers, and performs equidistant sampling with a step of 0.5 nanometers, so that it can accurately capture the small reflectance fluctuation caused by film interference during the entire measurement process. During scanning, the incident light irradiates the surface of the filter film at a preset angle, and the reflected light converges into the detection optical path through an optical sampler, and is finally received by a high-sensitivity detector and converted into a digital signal. The system stores, normalizes and suppresses noise on the collected full-spectrum reflectance data to obtain a group of reflection spectrum curves with high signal-to-noise ratio. The curve contains the information of the film structure and optical path difference, and reflects the refraction characteristics of the film material at different wavelengths. A Cauchy model is constructed based on the obtained reflectance spectrum data. Cauchy model is a classical empirical function that describes the relationship between the refractive index of optically transparent materials and wavelength, and is used in thin film optics.The model is presented in a three-parameter form, where the refractive index is expressed as a wavelength function. By fitting reflectance data across the entire visible light spectrum, the refractive properties of the material at various wavelengths are deduced. To improve fitting accuracy and robustness, the Cauchy model is optimized using the least squares method. The measured reflectance curves are compared with the theoretical curves calculated by the model, and the sum of squared errors is used as the objective function. A nonlinear fitting algorithm is used for iterative solution to minimize the error between the model output and the measured data. During the fitting process, to avoid local optima traps, multiple initial value strategies and boundary conditions are introduced, and a convergence threshold and a maximum number of iterations are set to ensure the convergence and efficiency of the fitting process. Through this least squares fitting process, the initial set of refractive index parameters for the filter film within the measured spectral range is obtained.

[0022] In one example, the sample of the filter film under test is illuminated with a first parallel beam, a second parallel beam, and a third parallel beam. The measurement area is divided into multiple sub-regions based on the initial refractive index parameters. The optimal thickness and refractive index values ​​for each sub-region are obtained by least-squares fitting, including: The first wavelength light source, the second wavelength light source, and the third wavelength light source are activated sequentially by an electronic shutter controller, so that the first parallel beam, the second parallel beam, and the third parallel beam illuminate the filter film sample under test in sequence, and three sets of original interference fringe images are obtained. Three sets of original interference fringe images were acquired using a two-dimensional photodetector array to obtain digitized interference fringe image data. Dark field correction was then performed on the digitized interference fringe image data to obtain corrected interference fringe images. Background noise was eliminated from the corrected interference fringe image to obtain a noise-reduced interference fringe image. The phase of the noise-reduced interference fringe image was then extracted using a five-step phase-shifting method to obtain wrap-around phase maps corresponding to the three wavelengths. The 2π jump elimination is performed on the packaged phase map using the quality-guided phase unpacking algorithm to obtain the first phase distribution map, the second phase distribution map, and the third phase distribution map; Based on the first phase distribution map, the second phase distribution map, the third phase distribution map, and the initial refractive index parameter, the measurement area is divided into multiple sub-regions, and the least squares method is used for fitting to obtain the optimal thickness and refractive index values ​​for each sub-region.

[0023] In this example, an electronic shutter controller sequentially activates three lasers with different center wavelengths. This controller possesses high-precision timing management capabilities, independently scheduling the opening and closing sequence of the three laser beams with millisecond-level control accuracy. This ensures that the blue laser (corresponding to the first wavelength), the green laser (corresponding to the second wavelength), and the red laser (corresponding to the third wavelength) are strictly staggered in time, each forming an independent illumination window. Whenever a laser of a wavelength is activated, the corresponding collimated parallel beam illuminates the surface of the pre-adjusted filter sample, forming an interference fringe image with the beam of the same wavelength returning from the reference light path in the interferometric detection system. Through this step, the original interference fringe images for blue, green, and red wavelengths are obtained respectively. The system's built-in two-dimensional photodetector array acquires the three sets of interference images at high resolution. This detector array has a 2048×2048 pixel structure, with each pixel size controlled at the micrometer level, and is equipped with high dynamic range 16-bit grayscale imaging capability, ensuring high-fidelity data recording under different light intensity distributions. The detector operates in synchronous sampling mode, achieving independent acquisition of images corresponding to each laser illumination through signal linkage with the shutter controller. After acquisition, these three sets of image data are formatted and stored, and the dark-field correction process is initiated. During dark-field correction, a dark-field image is acquired under no-light conditions to record the detector's electrical noise and system background signal. This dark-field image is then subjected to pixel-level interpolation from the actual acquired images, effectively eliminating errors caused by non-optical components to obtain the corrected interference image. Background noise reduction is then performed on the corrected image. To ensure that the fringe information of the interference image is not weakened during processing, a multi-scale wavelet transform method is used for noise reduction. A five-level decomposition using the Daubechies wavelet basis is employed, with noise thresholds set and soft thresholding applied in each wavelet coefficient level. Finally, an inverse wavelet transform is used to reconstruct a smooth interference image that preserves edge features. After noise reduction, a five-step phase-shifting method is used to extract phase information from each set of images. By sequentially acquiring five interference images with a known phase difference (π / 2) and calculating the corresponding wrapped phase map using a formula, the optical phase information in the interference fringes can be accurately restored. This method is highly noise-resistant and has high resolution, making it suitable for high-precision optical interferometry systems. In this step, each wavelength image group yields a corresponding wrapped phase map. The phase values ​​in these images are still in a 2π ambiguity state and cannot be directly used for film thickness calculation; therefore, phase unpacking processing is performed. To restore the wrapped phase map to a true and continuous phase distribution map, a quality-guided phase unpacking algorithm is used. This algorithm evaluates the phase quality factor of each pixel in the entire image, determines its local phase gradient, noise distribution, and edge continuity, and constructs an unpacking path priority map based on this.The algorithm prioritizes phase unfolding in high-quality regions and gradually expands towards edges and complex regions, effectively avoiding the propagation of jump errors. This results in the acquisition of first, second, and third phase distribution maps for blue, green, and red light bands, respectively. These three phase maps provide continuous phase information with high spatial resolution and phase accuracy. After obtaining the phase maps, the measurement area is intelligently divided based on the initial refractive index parameters obtained through Cauchy model fitting. The division method uses local gradient changes, adaptive thickness distribution estimation, and spatial structure characteristics in the phase distribution maps to divide the entire interferometric field of view into multiple sub-regions, ensuring relatively uniform optical properties within each sub-region, which improves the accuracy of local inversion. For each sub-region, a nonlinear relationship model is established, encompassing reflectivity, phase, thickness, and refractive index. The least squares method is then used to fit the parameters of this model. The least squares method iteratively solves for the optimal film thickness and refractive index by minimizing the sum of squares of the difference between the actual measured light intensity and the theoretical model's calculated light intensity. In the solution process, the Levenberg-Marquardt optimization algorithm is introduced as the core of the numerical solution to improve the fitting accuracy and convergence speed. The optimal film thickness and refractive index values ​​in each spatial sub-region are obtained.

[0024] In one example, the measurement area is divided into multiple sub-regions based on the first phase distribution map, the second phase distribution map, the third phase distribution map, and the initial refractive index parameter. Least squares fitting is then performed to obtain the optimal thickness and refractive index values ​​for each sub-region, including: Wavelength synthesis calculation is performed based on the first phase distribution map and the second phase distribution map to calculate the first synthesized wavelength, and the second synthesized wavelength is calculated based on the first phase distribution map and the third phase distribution map. Based on the first and second composite wavelengths, a composite phase is constructed using the first phase distribution map, the second phase distribution map, and the third phase distribution map, respectively, to obtain the short composite wavelength phase map and the long composite wavelength phase map; A rough thickness distribution is calculated based on the phase diagram of the long composite wavelength and the initial refractive index parameters to obtain the original thickness distribution diagram. The original thickness distribution map is dynamically thresholded to obtain multiple sub-regions, and local optimization is performed on each sub-region to obtain the optimal thickness and refractive index values ​​for each sub-region.

[0025] In this example, the first and second phase distribution maps are used to calculate the composite wavelength. These two phase maps correspond to the interferometric measurement results of the first and second wavelengths, respectively. Therefore, their phase difference reflects the difference in the interference period of the two wavelengths under the same optical path. The first composite wavelength is then calculated using the wavelength synthesis formula. The physical significance of this wavelength value is that a composite wavelength is constructed by the difference between different wavelengths, and its period is much larger than the original wavelength, thus effectively expanding the unambiguous measurement range of the interferometric measurement system and reducing the impact of phase jump errors. Using the same method, the second composite wavelength is calculated based on the first and third phase distribution maps, resulting in two composite wavelengths: the first composite wavelength is shorter, and the second composite wavelength is longer. These are used to construct phase distribution maps at different measurement scales. The first, second, and third phase distribution maps are combined, and a composite phase map is generated through point-by-point phase difference calculations. Specifically, the differences between the first and second phase distribution maps are calculated separately, corresponding to the first and second composite wavelengths, respectively, to obtain two composite phase maps. The first synthetic phase map, with its shorter wavelength scale, possesses higher spatial resolution and is suitable for capturing minute local thickness variations. The second synthetic phase map, with its longer wavelength and larger interference fringe spacing, exhibits better resistance to 2π jumps and is suitable for extracting large-scale thickness structure contours. The two synthetic phase maps are named the short-wavelength synthetic phase map and the long-wavelength synthetic phase map, respectively, and are used as the basis for thickness calculation. The long-wavelength synthetic phase map is selected as the initial basis for thickness estimation because its large wavelength characteristic effectively avoids phase jumps in thick film regions, thus providing strong global continuity for the coarse thickness estimation. During thickness calculation, the initial refractive index parameter obtained through the Cauchy model is used as the effective refractive index input. Combined with the synthetic wavelength and phase map data, the thickness-phase inverse formula is used to calculate the film thickness value at each pixel, resulting in a complete original thickness distribution map that macroscopically reflects the thickness variation trend of the filter film sample throughout the entire field of view. Dynamic threshold segmentation is applied to the original thickness distribution map, dividing the measurement area into several sub-regions. Dynamic thresholding segmentation is based on the difference range between the maximum and minimum values ​​in the thickness map. By setting an adaptive threshold stepping strategy, the thickness space is divided into multiple intervals, and the number and boundaries of these intervals are determined according to the gradient trend of thickness change. This segmentation method ensures that the thickness change within each sub-region is relatively gradual, thereby improving the stability and accuracy of the subsequent fitting process. After segmentation, each sub-region is treated as an independent modeling unit for fine-grained optimization calculations of local thickness and refractive index.In the local optimization process, an objective function is constructed that includes the squared difference between the theoretical and measured reflectivities. The least squares method is introduced as an optimization technique, and an iterative solution strategy is employed to continuously adjust the thickness and refractive index parameters to minimize the error term. To improve computational efficiency and convergence performance, the Levenberg-Marquardt algorithm is introduced during the optimization process, combining the advantages of gradient descent and Newton's method, enabling rapid convergence when the initial values ​​are close to the solution. Simultaneously, at the boundary of each sub-region, a bicubic interpolation strategy is introduced to smooth the fitting results, ensuring the continuity and spatial consistency of the final thickness value across sub-regions and avoiding local discontinuities in thickness abrupt changes. The optimal thickness and refractive index values ​​for each sub-region are then obtained.

[0026] In one example, a mapping relationship between spectral reflectance and thickness is established based on the optimal thickness value, refractive index value, and reflection spectrum. This relationship is then solved using an iterative algorithm and error compensation calculations to obtain a target thickness distribution map, including: A spectral reflectance calculation model was established based on a single-layer homogeneous film, and reflectance was calculated based on the spectral reflectance calculation model to obtain the single-layer film reflectance mapping relationship. A feature matrix calculation model is established based on the multilayer composite filter film, and the reflectance is calculated based on the feature matrix calculation model to obtain the reflectance mapping relationship of the multilayer film. The optimal thickness and refractive index values ​​are substituted into the single-layer film reflectivity mapping relationship or the multilayer film reflectivity mapping relationship to perform thickness inversion, and the Newton-Raphson iterative method is used to solve the problem to obtain the preliminary thickness inversion results. A parallel computing strategy was implemented on the preliminary thickness inversion results. The surface of the filter film was divided into multiple computing blocks, and the computing tasks were processed simultaneously using GPU acceleration technology to obtain accelerated thickness data. Adaptive mesh refinement is performed on the thickness abrupt change regions in the accelerated thickness data to obtain refined thickness data, and an initial thickness distribution map and refractive index distribution map are generated based on the refined thickness data; The initial thickness distribution map is compensated for by calculations for refractive index error, incident angle error, temperature error, wavelength error, and phase measurement error, and the target thickness distribution map is obtained.

[0027] In this example, for a single-layer homogeneous film sample, a spectral reflectance calculation model is established based on thin-film interference theory. This model uses incident light wavelength, film thickness, film refractive index, substrate refractive index, and incident angle as input parameters, and derives the reflectance expression of light interference at multiple interfaces between air, film, and substrate by combining electromagnetic boundary conditions. This model reveals the periodic variation of reflectance with film thickness and wavelength, and is used to construct the mapping relationship between film thickness and reflectance. Through extensive parameter simulation calculations, a reflectance mapping map covering different thicknesses and refractive indices is generated. For multilayer composite filter samples with more complex structures, the characteristic matrix method is used for modeling. In this method, each film layer is represented as a 2×2 matrix with specific optical admittance and phase delay. These characteristic matrices are multiplied sequentially according to their hierarchy to form the equivalent propagation matrix of the entire film system. Using this total matrix, the complex reflection coefficient and reflectance at any wavelength are calculated. Because the interference behavior of light in multilayer film systems exhibits strong interlayer coupling characteristics, simple analytical models are insufficient to accurately represent it. Therefore, this matrix model provides a highly versatile and accurate calculation method, making the system applicable to filter film structures with any number of layers and any combination of materials. After modeling is completed and reflectivity mapping relationships are generated, the optimal thickness and refractive index values ​​obtained in the previous sub-regions are substituted into the corresponding models to perform thickness inversion. For known reflectivity measurements and preliminary estimated optical parameters, the Newton-Raphson iterative method is used for inversion solutions. This method uses the reflectivity model as the objective function and gradually corrects the film thickness and refractive index by calculating the first derivative (i.e., partial derivative information) of the function at the current parameter point, gradually approximating the actual material solution as the error decreases. A convergence threshold and a maximum number of iterations are set to ensure accuracy while controlling computational complexity, yielding preliminary thickness inversion results for each calculation point. Considering the large area and high resolution of the entire filter sample, a parallel computing strategy was implemented, dividing the entire filter surface into multiple computational blocks. Each block contains a fixed number of pixel units, such as a 128×128 pixel array. These blocks were allocated to different computing cores of the GPU, simultaneously performing reflectivity calculation, model iteration, and thickness determination. Leveraging the CUDA parallel computing platform, the system scheduled a large number of concurrent threads to maximize the GPU's computing power, significantly improving the overall computational speed and enabling the originally time-consuming inversion calculation to be completed efficiently within an acceptable timeframe. To address the issue of abrupt or drastic changes in the filter surface thickness in local areas, edge detection and difference analysis were performed on the accelerated thickness data. By determining whether the thickness difference between adjacent pixels exceeded a set threshold, regions of abrupt thickness changes were identified.In these regions, to improve model resolution and fitting accuracy, an adaptive mesh refinement strategy is initiated, reducing the original mesh size to one-quarter or smaller, generating denser computational units, and re-performing local optimization calculations within the refined mesh. This approach maintains overall computational efficiency while ensuring detail accuracy, significantly improving film thickness modeling in complex structural regions. After processing data from all regions, the thickness results are reassembled to generate initial thickness distribution maps and corresponding refractive index distribution maps. These images maintain spatial resolution consistent with the initially acquired interferometric images, achieving high-precision mapping from phase data to physical parameters. Considering the influence of various error sources during measurement, such as refractive index model errors, incident angle deviations, ambient temperature fluctuations, laser wavelength drift, and phase extraction uncertainties, multi-factor compensation is applied to the initial thickness results to improve the accuracy of thickness measurements. To address refractive index errors, a dispersion correction strategy is introduced. Reflectivity measurements of the filter film are performed at multiple additional wavelength points, and the refractive index curve is corrected across the entire wavelength range using the Sellmeier equation. For incident angle errors, a response function of thickness to angle changes is established through differential sensitivity analysis, and the thickness data is corrected based on the actual incident angle deviation. For temperature errors, a compensation formula is used to jointly correct for thickness and refractive index based on the thermal expansion coefficient and thermo-optic coefficient of the film material. For wavelength errors, the interference spectrum of a single-crystal silicon wafer standard sample is used for wavelength calibration, and the light source offset is corrected in reverse. For phase measurement errors, multiple measurements and statistical averaging of the same area are performed, and a weighted phase estimation method is used to improve the reliability of the phase data. After all errors are compensated, a fully corrected target thickness distribution map is finally generated.

[0028] In one example, compensation calculations are performed on the initial thickness distribution map to account for refractive index error, incident angle error, temperature error, wavelength error, and phase measurement error, resulting in the target thickness distribution map, including: The refractive index error in the initial thickness distribution map is compensated by calculation to obtain refractive index error compensation data. Based on differential sensitivity analysis, incident angle error compensation is performed on the refractive index error compensation data to obtain incident angle error compensation data; Based on the incident angle error compensation data, temperature error compensation is performed using the thermal expansion coefficient and thermo-optic coefficient of the material to obtain temperature error compensation data. Wavelength error compensation data was obtained by using standard materials from single-crystal silicon wafers to compensate for temperature error compensation data. Phase measurement error compensation is performed on the wavelength error compensation data to obtain phase error compensation data, and a target thickness distribution map is generated based on the phase error compensation data.

[0029] In this example, refractive index error compensation calculations are performed on the initial thickness distribution map. Since the refractive index was obtained based on the Cauchy model or empirical fitting in early modeling, but the actual material's dispersion characteristics deviate due to process fluctuations, material impurities, or interface structure differences, multi-band spectral reflectance measurement data is introduced to correct the refractive index. By adding five auxiliary wavelength points, such as 480 nm, 500 nm, 550 nm, 600 nm, and 630 nm, corresponding reflectance values ​​are obtained. These data are compared with the theoretical model calculation results, and a corrected dispersion model is constructed using the Sellmeier equation to obtain the refractive index deviation at each wavelength. These deviation values ​​are then weighted and substituted back into the original refractive index distribution data in a weighted functional form to obtain the corrected refractive index error compensation data, from which the film thickness value under conditions closer to the true refractive index is derived. After completing the refractive index error compensation, the measurement error caused by the incident angle deviation is identified. In actual measurements, the laser incident angle may deviate slightly from the set value due to the influence of mechanical tolerances, platform drift, or thermal stress deformation of the optical system. This small angular error significantly affects the phase distribution of the interference fringes, thus impacting the thickness inversion results. A partial derivative model of thickness with respect to the incident angle is established using differential sensitivity analysis, i.e., a response function of thickness with respect to angle changes is constructed. By fitting the trend of film thickness variation under different angular conditions, the sensitivity function is obtained. Film thickness is measured at multiple deviation angles such as θ±0.5°, θ±1°, and θ±1.5°, and these measurements are substituted into the sensitivity function for inverse correction to obtain incident angle error compensation data. Temperature-induced errors are addressed. Since the physical dimensions and optical properties of the filter film material change under different temperature conditions, the thermal expansion coefficient and thermo-optical coefficient of the material are combined to perform bidirectional compensation for thickness and refractive index. The system detects the difference between the ambient temperature and the reference temperature during measurement, then performs thermal expansion correction on the physical thickness of the film layer according to the thickness compensation formula, and uses the refractive index compensation formula to perform thermo-optical modulation correction on the optical refractive index value. Temperature correction is applied sequentially to all pixels to generate temperature error compensation data, ensuring good consistency and stability of measurement results under different ambient temperature conditions. After temperature compensation, wavelength error compensation is performed. The deviation between the actual operating wavelength of the laser and its nominal value is identified. Since the laser experiences center wavelength drift after long-term operation, and environmental conditions also affect the laser output wavelength during measurement, a single-crystal silicon wafer with known thickness and stable material properties is used as a standard reference material. Interferometry is performed on this material to obtain the phase data corresponding to the interference fringes, which is then compared with the theoretical calculation value to deduce the current actual operating wavelength. After obtaining the true center wavelength of each light source, the wavelength parameters in the thickness calculation are updated, and the reflectivity model is recalculated to obtain the film thickness data after wavelength compensation, i.e., the wavelength error compensation data.After compensating for the four types of systematic errors mentioned above, the measurement error caused by the uncertainty in phase extraction is addressed. Especially in multi-wavelength interferometry, phase images in different wavelength bands may experience phase shifts or jitter due to uneven light source intensity, detector nonlinear response, or poor interference fringe quality. A statistical averaging method is used to correct the phase. Multiple measurements are performed on each region to obtain multiple sets of phase images. The average phase value and standard deviation of each pixel are calculated, and a weighted averaging algorithm is used to assign greater weight to phases with higher stability, thus constructing a weighted average phase image. This image is used for the final thickness inversion calculation, and the corresponding results are combined with the wavelength error compensation data to obtain the final thickness result after phase measurement error compensation.

[0030] Reference Figure 2 This embodiment provides a filter film thickness detection device, including: Processing module 1 is used to process the first wavelength light source, the second wavelength light source, and the third wavelength light source through an optical collimation system and an angle control device to obtain a first parallel beam, a second parallel beam, and a third parallel beam; Measurement module 2 is used to fix the filter film sample to be tested on the sample stage and set the incident angle. The spectrophotometer is used to measure the filter film sample to be tested to obtain the reflectance spectrum and initial refractive index parameters. Fitting module 3 is used to irradiate the filter film sample under test with the first parallel beam, the second parallel beam and the third parallel beam, and divide the measurement area into multiple sub-regions according to the initial refractive index parameters, and obtain the optimal thickness value and refractive index value of each sub-region by fitting with the least squares method. Calculation module 4 is used to establish the mapping relationship between spectral reflectance and thickness based on the optimal thickness value, refractive index value and reflection spectrum, and obtain the target thickness distribution map by solving through iterative algorithm and error compensation calculation.

[0031] In this embodiment, the specific implementation of each unit in the above device embodiment is described in the above method embodiment, and will not be repeated here.

[0032] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0033] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for detecting the thickness of a light filter film, characterized in that, Includes the following steps: The first, second, and third wavelength light sources are processed by an optical collimation system and an angle control device to obtain a first parallel beam, a second parallel beam, and a third parallel beam. Specifically, this includes: wavelength modulation of a blue laser as the first wavelength light source, adjusting the center wavelength to 450±5nm to obtain a first center wavelength; wavelength modulation of a green laser as the second wavelength light source, adjusting the center wavelength to 532±3nm to obtain a second center wavelength; and wavelength modulation of a red laser as the third wavelength light source, adjusting the center wavelength to 650±5nm to obtain a third parallel beam. Center wavelength; based on the first center wavelength, the second center wavelength, and the third center wavelength, the lens group parameters of the optical collimation system are set, and the three laser beams are collimated by the optical collimation system to obtain an initial parallel beam; the initial parallel beam is deflected and adjusted by an angle control device to obtain an angle-calibrated parallel beam; the angle-calibrated parallel beam is split by a beam splitter to divide each beam into a reference beam and a measurement beam, wherein the reference beam directly enters the spectral detector, and the measurement beam illuminates the surface of the filter film to be tested, to obtain a first parallel beam, a second parallel beam, and a third parallel beam; The filter film sample to be tested is fixed on the sample stage and the incident angle is set. The filter film sample to be tested is measured by a spectrophotometer to obtain the reflectance spectrum and initial refractive index parameters. The test filter sample is illuminated using a first, second, and third parallel light beam. The measurement area is divided into multiple sub-regions based on the initial refractive index parameter. The optimal thickness and refractive index values ​​for each sub-region are obtained through least-squares fitting. Specifically, this includes: sequentially activating the first, second, and third wavelength light sources using an electronic shutter controller, causing the first, second, and third parallel light beams to sequentially illuminate the test filter sample, obtaining three sets of original interference fringe images; acquiring the three sets of original interference fringe images using a two-dimensional photodetector array to obtain digitized interference fringe image data; performing dark-field correction processing on the digitized interference fringe image data to obtain corrected interference fringe images; eliminating background noise from the corrected interference fringe images to obtain denoised interference fringe images; and using a five-step phase-shifting method to further refine the denoised interference fringe images. Phase extraction is performed to obtain wrapper phase maps corresponding to three wavelengths. A quality-guided phase unpacking algorithm is used to eliminate 2π jumps in the wrapper phase maps, resulting in a first phase distribution map, a second phase distribution map, and a third phase distribution map. Wavelength synthesis calculation is performed based on the first and second phase distribution maps to calculate a first synthesized wavelength, and a second synthesized wavelength is calculated based on the first and third phase distribution maps. Synthetic phases are constructed using the first, second, and third phase distribution maps simultaneously, based on the first and second synthesized wavelengths, to obtain short-synthetic wavelength phase maps and long-synthetic wavelength phase maps. A coarse thickness distribution is calculated based on the long-synthetic wavelength phase map and the initial refractive index parameter to obtain an original thickness distribution map. The original thickness distribution map is then dynamically thresholded to obtain multiple sub-regions, and local optimization is performed on each sub-region to obtain the optimal thickness and refractive index values ​​for each sub-region. Based on the optimal thickness value, the refractive index value, and the reflection spectrum, a mapping relationship between spectral reflectance and thickness is established, and the target thickness distribution map is obtained by solving the problem through an iterative algorithm and error compensation calculation.

2. The method for detecting the thickness of a filter film according to claim 1, characterized in that, The process of fixing the filter sample to be tested on the sample stage and setting the incident angle, and measuring the filter sample using a spectrophotometer to obtain the reflectance spectrum and initial refractive index parameters includes: The filter film sample to be tested is mounted on a sample stage with a three-dimensional fine-tuning mechanism. The position of the sample stage is adjusted by a displacement control device to obtain a fixed filter film sample to be tested. The film type of the fixed-state filter film sample is identified to obtain the film type, and a specific incident angle is set according to the film type to obtain the angle setting value. Based on the angle setting value, the test filter film sample in the fixed state is scanned in the wavelength range of 380nm to 780nm using a spectrophotometer to obtain the reflectance spectrum; A Cauchy model is constructed based on the reflection spectrum, and the parameters of the Cauchy model are fitted by the least squares method to obtain the initial refractive index parameters.

3. The method for detecting the thickness of a filter film according to claim 1, characterized in that, The process involves establishing a mapping relationship between spectral reflectance and thickness based on the optimal thickness value, the refractive index value, and the reflection spectrum, and obtaining a target thickness distribution map through iterative algorithm solving and error compensation calculations. This includes: A spectral reflectance calculation model is established based on a single-layer homogeneous film, and reflectance is calculated based on the spectral reflectance calculation model to obtain the single-layer film reflectance mapping relationship. A feature matrix calculation model is established based on the multilayer composite filter film, and the reflectance is calculated based on the feature matrix calculation model to obtain the multilayer film reflectance mapping relationship. The optimal thickness value and the refractive index value are substituted into the single-layer film reflectivity mapping relationship or the multilayer film reflectivity mapping relationship to perform thickness inversion, and the Newton-Raphson iterative method is used to solve the problem to obtain the preliminary thickness inversion result. A parallel computing strategy is implemented on the preliminary thickness inversion results. The surface of the filter film is divided into multiple computing blocks, and the computing tasks are processed simultaneously by GPU acceleration technology to obtain accelerated thickness data. Adaptive mesh refinement is performed on the thickness abrupt change regions in the accelerated thickness data to obtain refined thickness data, and an initial thickness distribution map and refractive index distribution map are generated based on the refined thickness data. The initial thickness distribution map is compensated for errors in refractive index, incident angle, temperature, wavelength, and phase measurement to obtain the target thickness distribution map.

4. The method for detecting the thickness of a filter film according to claim 3, characterized in that, The calculation of compensation for refractive index error, incident angle error, temperature error, wavelength error, and phase measurement error in the initial thickness distribution map to obtain the target thickness distribution map includes: The refractive index error in the initial thickness distribution map is compensated to obtain refractive index error compensation data. Based on differential sensitivity analysis, the incident angle error is compensated for using the refractive index error compensation data to obtain incident angle error compensation data. Based on the incident angle error compensation data, temperature error compensation is performed using the thermal expansion coefficient and thermo-optic coefficient of the material to obtain temperature error compensation data. Wavelength error compensation data is obtained by using standard materials from single-crystal silicon wafers to compensate for the temperature error compensation data. Phase measurement error compensation is performed on the wavelength error compensation data to obtain phase error compensation data, and a target thickness distribution map is generated based on the phase error compensation data.

5. A filter film thickness detection device, characterized in that, For implementing the filter film thickness detection method according to any one of claims 1 to 4, the filter film thickness detection device comprises: The processing module is used to process the first wavelength light source, the second wavelength light source, and the third wavelength light source through an optical collimation system and an angle control device to obtain a first parallel beam, a second parallel beam, and a third parallel beam. The measurement module is used to fix the filter film sample to be tested on the sample stage and set the incident angle, and measure the filter film sample to be tested by a spectrophotometer to obtain the reflectance spectrum and initial refractive index parameters. The fitting module is used to irradiate the filter film sample under test with the first parallel beam, the second parallel beam and the third parallel beam, divide the measurement area into multiple sub-regions according to the initial refractive index parameter, and obtain the optimal thickness value and refractive index value of each sub-region by fitting with the least squares method. The calculation module is used to establish a mapping relationship between spectral reflectance and thickness based on the optimal thickness value, the refractive index value and the reflection spectrum, and obtain the target thickness distribution map by solving the problem through an iterative algorithm and error compensation calculation.

6. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 4.

Images