A bright-dark field hybrid enhanced white light interferometry three-dimensional topography measurement system and method

Abstract

The application discloses a bright-dark field mixed enhanced white light interference three-dimensional profile measurement system and method, belongs to the optical precision measurement technical field, and aims to solve the problems that the existing white light interference measurement is prone to interference signal weakening, data void and measurement integrity deficiency on a complex surface with large slope, low reflectivity and weak texture. The system comprises a bright-dark field mixed illumination, interference imaging collection, interference-scattering signal separation and profile calculation and fusion module, the scattering signal of the complex surface is enhanced through the annular dark field illumination, the interference and scattering information is synchronously collected through single scanning, the two signals are calculated after three-dimensional frequency domain separation, and the profile is reconstructed through pixel-by-pixel fusion based on the signal-to-noise ratio quality mask. The application takes into account the advantages of high longitudinal resolution of the interference measurement and the adaptability of the scattering signal to the complex surface, significantly improves the measurement integrity of the complex surface, and has simple structure and is easy to realize in engineering.

Description

Technical Field

[0001] This application relates to the field of optical precision measurement technology, and in particular to a white light interferometric three-dimensional topography measurement system and method with enhanced bright and dark field mixing. Background Technology

[0002] High-precision measurement of surface morphology at the micro- and nano-scale is a critical requirement in fields such as precision manufacturing, semiconductor device fabrication, metal additive manufacturing, and optical component inspection. Among existing non-contact optical measurement methods, coherent scanning interferometry (CSI, also known as white light interferometry) has become the mainstream technology for high-precision measurement due to its nanometer- to sub-nanometer-scale longitudinal resolution. Among these, the Mirau-type interferometric microstructure has been widely used in white light interferometry systems due to its compact structure and ease of integration.

[0003] Existing Mirau-type systems mostly employ Köhler coaxial illumination, which provides a uniform and stable light field to ensure the stability of interference fringes. However, the effective utilization of reflected light is constrained by the receiving cone limited by the numerical aperture (NA) of the objective lens, making it more suitable for samples with strong surface reflection and gentle morphology. When dealing with rough surfaces, low-reflectivity surfaces, or steep slopes and complex microstructures, local specular reflections tend to deviate from the objective lens's NA receiving range, and surface scattered light is difficult to utilize effectively. This can easily lead to a decrease in interference fringe contrast, a reduction in signal-to-noise ratio, and even interference signal interruption or loss of effective data points, affecting measurement stability and reliability. Another commonly used method, focusing microscopy (FVM), while adaptable to complex morphologies and variations in surface material, has limited longitudinal resolution, making it difficult to achieve the high precision levels of interferometry.

[0004] Furthermore, the existing technology includes Chinese invention patent CN114413783B, "Surface Reconstruction Method of Coherent Scanning Interferometer Based on Instrument Three-Dimensional Transfer Function." This method improves the reconstruction accuracy and spatial resolution of coherent scanning interferometry by preprocessing the three-dimensional interference signal, performing a three-dimensional Fourier transform, and combining it with the instrument's three-dimensional transfer function for frequency domain inversion. This reconstructs the surface film model of the measured surface and further calculates the surface height. This method primarily optimizes the three-dimensional imaging and reconstruction process of the interference signal itself, belonging to a single-interference-channel surface reconstruction approach.

[0005] However, for smooth, steeply sloping, low-reflectivity, or weakly textured surfaces, when local reflections deviate from the objective lens's NA receiving range, the interference modulation signal is easily weakened or even fails, leading to data blind spots. Introducing oblique-incident dark-field illumination greater than the objective lens's NA to enhance the scattered signal can improve the effective utilization of light returning from complex surfaces without changing the objective lens's NA, thereby providing supplementary information for complex areas.

[0006] Therefore, it is still necessary to propose a new scheme that differs from the single interferometric channel reconstruction method mentioned above: introducing mixed illumination of bright and dark fields at the hardware level, using a ring-shaped dark field to enhance the scattering information of complex surfaces, simultaneously acquiring interference modulation information and scattering intensity information in a single scan, and separating, calculating and fusing the two types of signals in the three-dimensional frequency domain, so as to take into account both the high precision advantage of interferometric signal reconstruction and the adaptability of scattering signal reconstruction to complex surfaces, thereby realizing high-precision three-dimensional topography measurement of complex micro-nano surfaces. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention proposes a white light interferometric three-dimensional topography measurement system and method that combines bright and dark field hybrid enhancement. By combining innovative hardware illumination with a frequency domain separation algorithm, it achieves synchronous acquisition and efficient fusion of interference and scattering signals in a single scan, thus resolving the contradiction between accuracy and adaptability in complex surface measurement.

[0008] To solve the technical problem, the technical solution of the present invention is as follows:

[0009] A white light interferometric three-dimensional topography measurement system with enhanced bright and dark field mixing includes: The bright-field and dark-field mixed illumination module is used to provide bright-field coaxial illumination and dark-field oblique-incident illumination to the surface under test; The interferometric imaging acquisition module receives reflected and scattered light from the surface under test via its optical path. It is used to simultaneously acquire a three-dimensional image sequence containing interference modulation information and scattering intensity information under mixed bright and dark illumination conditions. ; The host computer data processing module is connected to the interferometric imaging acquisition module and is used to receive the three-dimensional image sequence, complete all data processing and morphology reconstruction, and output the final three-dimensional morphology map. The light and dark field mixed lighting module includes a coaxial Kohler lighting unit and a ring-shaped dark field lighting unit; The coaxial Kohler illumination unit includes a broadband light source, and a field stop and a dichroic prism arranged sequentially along the illumination optical axis; an aperture stop, a piezoelectric ceramic Z-axis scanning stage, and a Mirau-type interference objective are arranged sequentially along the output optical axis of the dichroic prism; the coaxial Kohler illumination unit is configured in Kohler illumination mode to image the broadband light source onto the entrance pupil of the Mirau-type interference objective; The annular dark field illumination unit is an annular LED illumination array integrated around the periphery of the Mirau-type interference objective. The illumination numerical aperture of the annular LED illumination array is larger than the collection numerical aperture of the Mirau-type interference objective.

[0010] Furthermore, the interferometric imaging acquisition module also includes a high-resolution CCD camera, which is positioned in the reflected light path of the dichroic prism; the Mirau-type interferometric objective lens is used to interferometrically image the object beam with the reference beam; the piezoelectric ceramic Z-axis scanning stage is used to drive the objective lens to scan in steps along the optical axis; and the high-resolution CCD camera is used to simultaneously acquire the three-dimensional image sequence during the scanning process. The Mirau-type interference objective, the piezoelectric ceramic Z-axis scanning stage, and the high-resolution CCD camera constitute the interference imaging acquisition module.

[0011] Furthermore, the host computer data processing module has a built-in interference-scattering signal separation unit and a shape calculation and fusion unit; the interference-scattering signal separation unit is used to perform frequency domain transformation and mask filtering on the three-dimensional image sequence to separate the interference signal component and the scattering signal component; the shape calculation and fusion unit is used to reconstruct the shape of the two components respectively, and fuse them pixel by pixel based on the quality mask to output the final three-dimensional shape map.

[0012] A method for measuring three-dimensional topography using white light interferometry with enhanced bright and dark field mixing, the method comprising the following steps: S1: Activate the bright-dark field mixed illumination module to drive the piezoelectric ceramic Z-axis scanning stage to perform Z-axis step-scanning with the Mirau-type interference objective. Simultaneously, the high-resolution CCD camera acquires a sequence of three-dimensional images containing interference fringes and surface scattering textures. And transmit it to the host computer; S2: The host computer performs a three-dimensional fast Fourier transform on the three-dimensional image sequence to the three-dimensional spatial frequency domain, generates a three-dimensional frequency domain mask corresponding to the interference signal and the scattering signal, and multiplies the original three-dimensional spectrum with the two types of masks respectively to extract the corresponding interference frequency domain component and scattering frequency domain component. S3: The host computer performs three-dimensional inverse Fourier transform on the two separated frequency domain signals to obtain the time-domain interference signal sequence and the time-domain scattering intensity signal sequence; S4: The host computer performs topography reconstruction on the two time-domain signal sequences to obtain the first height map and the second height map; S5: The host computer generates a quality mask, and performs pixel-by-pixel fusion of the first height map and the second height map based on the quality mask to output the final three-dimensional topography map.

[0013] Furthermore, the method for generating the three-dimensional frequency domain mask in step S2 is as follows: exist The two-dimensional frequency domain distribution regions of interference and scattered signals in the cross-section are determined based on the numerical aperture of the objective lens and the spectral distribution of the light source, and then... The axis rotation extension generates the corresponding three-dimensional frequency domain mask.

[0014] Furthermore, the method for morphological reconstruction of the time-domain interferometric signal sequence in step S4 is as follows: The contrast curve of each pixel's interference fringes as a function of scanning position was calculated, and the position of the coherent envelope peak was determined using the envelope centroid method to obtain the axial coarse positioning height. Near this position, a method based on... The orthogonal demodulation algorithm of phase-shifting beats calculates the interference phase information, and the coarse positioning result is combined with the phase information to obtain the first height map.

[0015] Furthermore, the morphology reconstruction method for the temporal scattering intensity signal sequence in step S4 is as follows: the gray-level variance operator is used to calculate the degree of change of neighborhood gray-level difference of each pixel on different scanning layers, and a focusing curve is formed along the Z direction; the position of the focusing peak is determined by three-point Gaussian fitting to obtain the second height map.

[0016] Furthermore, the method for generating and fusing the quality mask in step S5 is as follows: Hilbert transform is performed on the pixel-by-pixel axial interference signal sequence to extract the envelope, the envelope peak value represents the effective signal intensity, the average value of the outer region represents the noise floor, and the signal-to-noise ratio is calculated to generate a binary quality mask; a first height map is used in the region where the signal-to-noise ratio is higher than a set threshold, and a smoothed second height map is used for fusing in the region where the signal-to-noise ratio is lower than a set threshold.

[0017] This application has the following advantages: Improving signal availability of complex surfaces: By introducing annular dark field illumination, the scattered light components caused by surface microstructure, edges and roughness are enhanced, thereby obtaining texture information with a high signal-to-noise ratio in smooth, steep and weakly textured areas that are difficult to obtain under traditional coaxial illumination conditions, providing reliable input for morphology reconstruction based on scattered signals; Enhancing the measurement integrity of complex structures: High-precision topography reconstruction is achieved using interferometric signals in flat areas, and in areas where the contrast of the interferometric signals is low or fails, reconstruction is performed by using scattered signals, thereby reducing measurement blind zones and improving the measurement integrity of complex micro-nano structures (such as steep steps, low-reflection surfaces, etc.). The structure is simple and easy to implement in engineering: Based on the traditional Mirau type interferometer system, the above functions can be achieved simply by adding a ring dark field illumination unit. There is no need to make complex modifications to the original interferometric optical path. While maintaining the stability of the system, the implementation difficulty is reduced, which is conducive to engineering application and promotion. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 is a schematic diagram (a) of the structure of the white light interferometric three-dimensional topography measurement system with bright and dark field mixing enhancement according to the present invention and a flowchart (b) of the measurement method; Figure 2 This is a schematic diagram of K-space frequency domain separation; Figure 3 This is a schematic diagram of the metal wrench sample to be tested and a local measurement area. Figure 4 Comparison of the original 3D images for two lighting methods: (a) the original image with only bright field lighting, and (b) the original image with mixed bright and dark field lighting. Figure 5 The spectrum of a three-dimensional image sequence after Fourier transform; Figure 6 The results are shown in the three-dimensional inverse Fourier transform diagram; (a) is a single-layer cross-sectional view of the interference signal in the time domain, and (b) is a single-layer cross-sectional view of the scattered signal in the time domain. Figure 7 (a) is a height recovery map based on two time-domain signals; (b) is a height recovery map based on the interference signal; Figure 8 This is a quality mask generated based on the signal-to-noise ratio of the interference signal; Figure 9 Comparison of the final blended topography of the two lighting methods: (a) is the blended topography of bright field lighting only, and (b) is the blended topography of mixed bright and dark field lighting. Detailed Implementation

[0020] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0021] Example 1: This embodiment provides a specific hardware implementation of a white light interferometric three-dimensional topography measurement system with enhanced bright and dark field mixing, in conjunction with the attached... Figure 1a and 1bThe system structure, hardware connections, and operating process of this invention are described in detail. This embodiment is a preferred embodiment of the invention and is not intended to limit the scope of protection of the invention.

[0022] I. Overall System Structure; The measurement system described in this embodiment consists of three parts: a mixed illumination module for bright and dark fields, an interferometric imaging acquisition module, and a host computer data processing module. All hardware is installed on an optical platform, and all data processing and topography reconstruction tasks are completed by the host computer (computer). There is no embedded signal processing unit.

[0023] The bright-dark field hybrid illumination module comprises a coaxial Kohler illumination unit and a ring-shaped dark-field illumination unit. The coaxial Kohler illumination unit provides uniform broadband parallel light to meet the conditions for Mirau-type interferometric imaging. The ring-shaped dark-field illumination unit is positioned around the Mirau-type interferometer objective lens, illuminating the surface under test at an oblique incidence angle greater than the numerical aperture of the objective lens. This enhances the scattered light components caused by surface microstructure, edges, and roughness, thereby improving the identifiability of scattered signals in complex surface regions and enhancing the reliability of subsequent fusion.

[0024] Interferometric imaging acquisition module: includes a Mirau-type interferometer objective, a piezoelectric ceramic Z-axis scanning stage, and a high-resolution CCD camera. The Mirau-type interferometer objective integrates a beam-splitting structure and a reference reflector, enabling interference imaging of the object beam and reference beam without the need for an additional reference arm assembly; the piezoelectric ceramic Z-axis scanning stage achieves precise step-scanning along the optical axis; the high-resolution CCD camera simultaneously acquires a three-dimensional image sequence containing interference modulation information and scattering intensity information during the scanning process.

[0025] II. Optical Path Connections; 1. Coaxial Kohler lighting optical path (bright field channel); According to the direction of light propagation, the optical elements are arranged coaxially in sequence: broadband light source → field stop → dichroic prism → aperture stop → piezoelectric ceramic Z-axis scanning stage → Mirau type interference objective lens → object to be measured.

[0026] Specific connection and working process: Broadband white light emitted from the broadband light source is limited to the illumination field of view by the field stop and then incident on the incident surface of the dichroic prism; the dichroic prism transmits the incident light downward along the vertical optical axis, passing sequentially through the aperture stop limiting the numerical aperture of the illumination, the central through-hole of the piezoelectric ceramic Z-axis scanning stage, and finally into the Mirau-type interferometer objective; the coaxial Kohler illumination unit is configured in Kohler illumination mode, imaging the broadband light source at the entrance pupil of the Mirau-type interferometer objective, projecting uniform and parallel broadband coaxial light onto the surface of the object under test, satisfying the conditions for white light interferometric imaging.

[0027] 2. Circular dark field illumination path (dark field channel); The ring-shaped LED illumination array is coaxially fixed on the outer mounting base of the Mirau-type interferometer objective, sharing the same optical axis as the Mirau-type interferometer objective. The light emitted by the ring-shaped LED illumination array is obliquely incident on the surface of the object under test at an incident angle greater than the numerical aperture collected by the Mirau-type interferometer objective, exciting the microscopic scattered light from the surface microstructure, edges, and steep slope areas, thereby enhancing the intensity of the scattered signal from complex surfaces.

[0028] 3. Interferometric imaging acquisition optical path (signal return channel); According to the direction of light propagation, the optical elements are arranged coaxially in sequence: the object under test → Mirau type interference objective lens → piezoelectric ceramic Z-axis scanning stage → aperture stop → dichroic prism → high-resolution CCD camera.

[0029] Specific connection and working process: The specular reflected light (generating interference modulation signal) and scattered light (generating texture intensity signal) returned from the surface of the object under test enter the Mirau-type interference objective lens together; the Mirau-type interference objective lens integrates a beam splitting structure and a reference reflective surface, eliminating the need for an additional reference arm assembly. The internal beam splitting structure superimposes the object beam and the internal reference beam to generate interference, and the resulting interference light returns upward along the original optical path; the returned light passes sequentially through the central through-hole of the piezoelectric ceramic Z-axis scanning stage and the aperture stop, and then enters the reflective surface of the dichroic prism; the dichroic prism reflects the returned light along the horizontal optical axis to the photosensitive surface of the high-resolution CCD camera, completing the conversion of optical signal to electrical signal.

[0030] III. Relationship between electrical control and data transmission; All electrical equipment is directly connected to the host computer, forming an architecture of "unified control by the host computer and unidirectional data transmission to the host computer": The host computer is connected to the piezoelectric ceramic Z-axis scanning stage controller via an RS-232 or USB interface. The host computer sends stepping commands, scanning speed and scanning range parameters to the controller, which controls the piezoelectric ceramic Z-axis scanning stage to drive the Mirau type interference lens to perform precise stepping scanning along the optical axis.

[0031] The host computer connects to the high-resolution CCD camera via a GigE Vision or USB 3.0 high-speed data interface. The host computer sends acquisition trigger signals to the CCD camera, synchronously controlling the CCD camera to acquire two-dimensional images at each scanning position; the CCD camera acquires the original three-dimensional image sequence. It is directly transmitted to the host computer's memory without any intermediate hardware preprocessing.

[0032] The host computer is connected to the ring LED lighting driver circuit via a PWM or analog interface. The host computer adjusts the brightness and on / off status of the ring LED lighting array to achieve precise control of the lighting intensity in the dark field.

[0033] The host computer is connected to the broadband light source drive circuit through an analog interface. The host computer adjusts the output power of the broadband light source to ensure that the interference fringes have the best contrast.

[0034] IV. Mechanical installation and connection relationships; The broadband light source, field stop, dichroic prism, and high-resolution CCD camera are fixed sequentially on the same horizontal optical platform. Their positions are adjusted by an optical bracket to ensure that the coaxiality error of the horizontal optical axis is no greater than 0.02mm.

[0035] The aperture stop and the piezoelectric ceramic Z-axis scanning stage are fixed sequentially on the vertical optical support below the dichroic prism, coaxial with the vertical optical axis, with a coaxiality error of no more than 0.02 mm.

[0036] The Mirau-type interference objective is mounted on the front output end of the piezoelectric ceramic Z-axis scanning stage via a standard threaded interface. It moves synchronously along the Z-axis with the scanning stage, with a travel distance of not less than 100 μm and a stepping accuracy of not less than 1 nm.

[0037] The ring-shaped LED illumination array is fixed to the periphery of the Mirau-type interference objective lens using a special clamp, maintaining a fixed relative position with the objective lens to ensure that all LEDs have the same oblique incidence angle and that the illumination numerical aperture is larger than the collection numerical aperture of the Mirau-type interference objective lens.

[0038] The object to be measured is placed on a three-dimensional manual displacement stage below the objective lens to adjust the measurement area so that the surface to be measured is within the working distance range of the objective lens.

[0039] Example 2 This embodiment provides a three-dimensional topography measurement method based on the system described in Embodiment 1, combined with the appendix. Figure 2-9 The algorithm flow, data processing procedure, and experimental verification results of this invention are described in detail. This embodiment first describes the general structure and standard measurement steps of the host computer data processing module, and then conducts specific experimental verification using the surface of a metal wrench as the object to be tested.

[0040] I. Structure of the host computer data processing module; The host computer data processing module has a built-in interference-scattering signal separation unit and a topography calculation and fusion unit, which are responsible for completing all data processing and topography reconstruction tasks. Interference-scattering signal separation unit: used to perform a three-dimensional fast Fourier transform on the three-dimensional image sequence, converting it to K-space; such as Figure 2As shown, the diagram illustrates the "umbrella-shaped" interference signal region and the "bowtie-shaped" scattering signal region. Based on Ewald's sphere theory, a three-dimensional frequency domain mask corresponding to the interference and scattering signal components is constructed. The interference signal component corresponds to the umbrella-shaped high-frequency region, and the scattering signal component corresponds to the bowtie-shaped low-frequency region. By multiplying and filtering the original three-dimensional spectrum with the two types of masks, the high-frequency interference signal and the low-frequency scattering signal can be separated. Then, a three-dimensional inverse Fourier transform is performed to recover the interference fringe sequence and the scattering intensity image sequence, respectively, avoiding mutual interference between the two types of signals.

[0041] The topography calculation and fusion unit is used to reconstruct the topography of the separated interference signal components and scattering intensity signal components, construct a quality mask based on the signal-to-noise ratio, and perform pixel-by-pixel selective fusion of the two reconstructed height maps to output the final 3D topography map. Specifically, it includes: Interference solution channel: Contrast curves are constructed for the interference intensity sequence of each pixel along the scanning direction. The position of the coherent envelope peak is determined using the envelope centroid method to achieve coarse axial localization. This is then combined with... Orthogonal demodulation method for phase-shifted beats to construct orthogonal components and and utilize Local phase information is extracted and combined with coarse positioning results to reconstruct a high-precision height map.

[0042] Scattering calculation channel: Calculate the gray-level variance of the pixel neighborhood layer by layer for the scattering intensity image sequence to form a focus evaluation curve along the Z direction; perform local fitting in the peak neighborhood to determine the position of the focus peak, thereby recovering a second height map applicable to areas with large slopes, low reflection and complex surfaces.

[0043] Quality evaluation and fusion unit: The Hilbert transform is performed on the pixel-by-pixel axial interference signal sequence to extract the envelope. The signal-to-noise ratio (SNR) is calculated based on the envelope peak value and background noise. A binary quality mask is generated according to a preset threshold or threshold range. In areas with high interference signal quality, the interference reconstruction results are retained first. In areas where the interference signal fails or the SNR is low, the smoothed scattering signal reconstruction results are used to supplement the results, thereby achieving full field-of-view three-dimensional topography measurement without blind spots.

[0044] II. Standard Measurement Procedure; S1: System initialization and parameter adjustment. Turn on the broadband light source and the ring LED illumination array. The host computer adjusts the output power of both to ensure that the acquired image has both clear interference fringes and identifiable scattering texture information. Set the scanning range, stepping accuracy of the piezoelectric ceramic Z-axis scanning stage, and the exposure time and gain parameters of the CCD camera.

[0045] S2: Acquisition of the original 3D image sequence. The host computer controls the piezoelectric ceramic Z-axis scanning stage to drive the Mirau-type interference lens to perform step scanning along the Z-axis, while simultaneously triggering the high-resolution CCD camera to acquire 2D images at each scanning position, forming the original 3D image sequence. And transfer it to the host computer's memory.

[0046] S3: Frequency domain signal separation. The host computer performs a three-dimensional fast Fourier transform on the three-dimensional image sequence to the three-dimensional spatial frequency domain; in ( The two-dimensional frequency domain distribution regions of interference and scattered signals in the cross-section are determined based on the numerical aperture of the objective lens and the spectral distribution of the light source, and then... The axis rotation extension generates the corresponding three-dimensional frequency domain mask; the original three-dimensional spectrum is multiplied by the two types of masks respectively to extract the corresponding interference frequency domain component and scattering frequency domain component; the two separated frequency domain signals are subjected to three-dimensional inverse Fourier transform to obtain the time domain interference signal sequence and the time domain scattering intensity signal sequence.

[0047] S4: Dual-path topography reconstruction. For the temporal interferometric signal sequence: calculate the curve of the contrast of each pixel's interference fringes as a function of the scanning position, and use the envelope centroid method to determine the position of the coherent envelope peak to obtain the axial coarse positioning height; near this position, use an orthogonal demodulation algorithm based on π / 2 phase shift beats to calculate the interference phase information, and combine the coarse positioning result with the phase information to obtain the first height map.

[0048] For the temporal scattering intensity signal sequence: the gray-level variance operator is used to calculate the degree of change of neighborhood gray-level difference of each pixel on different scanning layers, and a focusing curve is formed along the Z direction; the position of the focusing peak is determined by three-point Gaussian fitting, and a second height map is obtained.

[0049] S5: Quality Assessment and Fusion. The Hilbert transform is performed on the pixel-by-pixel axial interference signal sequence to extract the envelope. The envelope peak value represents the effective signal intensity, and the average value outside the envelope represents the noise floor. The signal-to-noise ratio (SNR) is calculated to generate a binary quality mask. A first height map is used in regions with an SNR higher than a set threshold, while a smoothed second height map is used in regions with an SNR lower than the set threshold for fusion. The final output is a full-field-of-view, blind-spot-free 3D topography map.

[0050] III. Experimental Verification and Result Analysis; This experiment uses the surface of a metal wrench as the test object. The wrench surface simultaneously possesses flat areas, edge transition areas, and locally irregular undulating areas. Furthermore, due to the reflective properties of the metal surface, the reflected light intensity varies at different locations, making it suitable as a sample for verifying the ability of this invention to measure the three-dimensional morphology of complex surfaces. During the implementation, a region on the wrench surface containing edge transitions and local undulations was selected as the actual measurement area, such as... Figure 3 As shown. The entire measurement and reconstruction process is as follows: Step S1: Acquisition of the original 3D image sequence; The bright-dark field mixed illumination module was activated, and the parameters of the coaxial Kohler illumination and annular dark field illumination were adjusted to ensure that the acquired images simultaneously possessed clear interference fringes and identifiable scattering texture information. Subsequently, the piezoelectric ceramic Z-axis scanning stage was driven to perform axial stepping scans, and the CCD camera simultaneously recorded two-dimensional images at each scanning position, forming the original three-dimensional image sequence. The original 3D image with only bright field illumination and mixed bright-dark field illumination is, for example... Figure 4 As shown in the figure, under mixed lighting conditions of light and dark fields, the brightness of the originally low grayscale and locally dim areas is significantly improved, the texture information and edge details are enhanced, and the surface grayscale distribution is more uniform and continuous.

[0051] Step S2: Three-dimensional Fourier transform; A three-dimensional fast Fourier transform is performed on the acquired three-dimensional image sequence to transform the data from the spatial domain to the three-dimensional spatial frequency domain. The spectrum is shown below. Figure 5 As shown. After the transformation, in Two typical frequency domain structures can be observed in the cross-section: an umbrella-shaped high-frequency region corresponding to the interference signal and a bowtie-shaped low-frequency region corresponding to the scattered signal.

[0052] Step S3: Frequency domain mask separation and time domain signal recovery; For the interference signal components, in In the meridional section, its frequency domain distribution can be approximately characterized by the superposition relationship between the incident wave vector and the scattered wave vector in the same direction, and the corresponding wave vector components satisfy the following: For the scattered signal component, in In the meridional section, its frequency domain distribution can be approximately characterized by the inverse subtraction relationship between the incident wave vector and the scattered wave vector, and the corresponding wave vector components satisfy the following: .in, The angle between the incident wave vector and the optical axis is... Let be the angle between the scattered wave vector and the optical axis. After determining the ideal two-dimensional frequency domain distribution range of the two types of signals, [the following is a partial translation of the original text, which is incomplete and requires further context.] Axial rotation expansion forms the corresponding three-dimensional frequency domain mask. The original three-dimensional spectrum is then multiplied by the corresponding three-dimensional frequency domain mask to extract the interference frequency domain components and the scattering frequency domain components; subsequently, a three-dimensional inverse Fourier transform is performed to recover the interference fringe sequence and scattering intensity image, as shown in the single-layer cross-section. Figure 6 As shown.

[0053] Step S4: Morphology reconstruction based on interference and scattering signals; For the recovered interference fringe sequence, an axial curve showing the variation of interference fringe contrast with scanning position is constructed pixel-by-pixel. The position of the coherent envelope peak is determined using the envelope centroid method, achieving coarse axial localization. Subsequently, orthogonal demodulation is used near the peak to extract local phase information, and the coarse localization result is combined with the fine phase localization result to recover a high-precision height map. A gray-level variance operator is used to calculate the gray-level variance of the pixel neighborhood layer by layer, forming a focus evaluation curve along the Z direction. Then, local fitting is performed on the peak neighborhood to determine the optimal focus position, thereby recovering the second height map. The recovered two time-domain signal images are shown below. Figure 7 As shown.

[0054] Step S5: Quality evaluation and quality mask generation; The Hilbert transform is performed on the pixel-by-pixel axial interference signal sequence to extract the envelope. The envelope peak value represents the effective signal intensity, and the mean value of the outer region represents the noise floor. The signal-to-noise ratio (SNR) is calculated accordingly. Furthermore, a binarized quality mask is generated based on a preset threshold or threshold range to determine the effectiveness of the interference channel at each pixel location. The generated quality mask is shown below. Figure 8 As shown.

[0055] Step S6: Pixel-by-pixel fusion and final shape output; Based on the quality mask, the reconstruction results of the interference channels are preferentially preserved in areas with high interference signal quality, while the reconstruction results of the scattering channels are used to supplement the interference signals in areas with weak or failed interference signals, thereby obtaining the final fused three-dimensional topography map. The final fusion result is as follows: Figure 9 As shown.

[0056] As can be seen from this embodiment, the bright and dark field hybrid enhanced white light interferometry three-dimensional topography measurement method proposed in this invention can effectively supplement and reconstruct areas with large slopes, low reflections and weak textures by using dark field enhanced scattering signals while maintaining the high precision advantage of interferometry, thereby significantly improving the measurement integrity and result continuity of complex surfaces.

[0057] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0058] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A white light interferometric three-dimensional topography measurement system with enhanced bright and dark field mixing, characterized in that, include: The bright-field and dark-field mixed illumination module is used to provide bright-field coaxial illumination and dark-field oblique-incident illumination to the surface under test; The interferometric imaging acquisition module receives reflected and scattered light from the surface under test via its optical path. It is used to simultaneously acquire a three-dimensional image sequence containing interference modulation information and scattering intensity information under mixed bright and dark illumination conditions. ; The host computer data processing module is connected to the interferometric imaging acquisition module and is used to receive the three-dimensional image sequence, complete all data processing and morphology reconstruction, and output the final three-dimensional morphology map. The light and dark field mixed lighting module includes a coaxial Kohler lighting unit and a ring-shaped dark field lighting unit; The coaxial Kohler illumination unit includes a broadband light source, and a field stop and a dichroic prism arranged sequentially along the illumination optical axis; an aperture stop, a piezoelectric ceramic Z-axis scanning stage, and a Mirau-type interference objective are arranged sequentially along the output optical axis of the dichroic prism; the coaxial Kohler illumination unit is configured in Kohler illumination mode to image the broadband light source onto the entrance pupil of the Mirau-type interference objective; The annular dark field illumination unit is an annular LED illumination array integrated around the periphery of the Mirau-type interference objective. The illumination numerical aperture of the annular LED illumination array is larger than the collection numerical aperture of the Mirau-type interference objective.

2. The three-dimensional topography measurement system for white light interferometry with enhanced bright and dark field mixing as described in claim 1, characterized in that, The interferometric imaging acquisition module further includes a high-resolution CCD camera, which is positioned in the reflected light path of the dichroic prism; the Mirau-type interferometric objective lens is used to interferometrically image the object beam with the reference beam; the piezoelectric ceramic Z-axis scanning stage is used to drive the objective lens to scan in a stepping motion along the optical axis; the high-resolution CCD camera is used to simultaneously acquire the three-dimensional image sequence during the scanning process. The Mirau-type interference objective, the piezoelectric ceramic Z-axis scanning stage, and the high-resolution CCD camera constitute the interference imaging acquisition module.

3. The three-dimensional topography measurement system for white light interferometry with enhanced bright and dark field mixing as described in claim 1, characterized in that, The host computer data processing module has a built-in interference-scattering signal separation unit and a shape calculation and fusion unit. The interference-scattering signal separation unit is used to perform frequency domain transformation and mask filtering on the three-dimensional image sequence to separate the interference signal component and the scattering signal component. The shape calculation and fusion unit is used to reconstruct the shape of the two components respectively, and fuse them pixel by pixel based on the quality mask to output the final three-dimensional shape map.

4. A method for measuring three-dimensional topography using white light interferometry with enhanced bright and dark field mixing, characterized in that, The method includes the following steps: S1: Activate the bright-dark field mixed illumination module to drive the piezoelectric ceramic Z-axis scanning stage to perform Z-axis step-scanning with the Mirau-type interference objective. Simultaneously, the high-resolution CCD camera acquires a sequence of three-dimensional images containing interference fringes and surface scattering textures. And transmit it to the host computer; S2: The host computer performs a three-dimensional fast Fourier transform on the three-dimensional image sequence to the three-dimensional spatial frequency domain, generates a three-dimensional frequency domain mask corresponding to the interference signal and the scattering signal, and multiplies the original three-dimensional spectrum with the two types of masks respectively to extract the corresponding interference frequency domain component and scattering frequency domain component. S3: The host computer performs three-dimensional inverse Fourier transform on the two separated frequency domain signals to obtain the time-domain interference signal sequence and the time-domain scattering intensity signal sequence; S4: The host computer performs topography reconstruction on the two time-domain signal sequences to obtain the first height map and the second height map; S5: The host computer generates a quality mask, and performs pixel-by-pixel fusion of the first height map and the second height map based on the quality mask to output the final three-dimensional topography map.

5. The method for measuring three-dimensional topography using white light interferometry with enhanced bright and dark field mixing as described in claim 4, characterized in that, The method for generating the three-dimensional frequency domain mask in step S2 is as follows: exist The two-dimensional frequency domain distribution regions of interference and scattered signals in the cross-section are determined based on the numerical aperture of the objective lens and the spectral distribution of the light source, and then... The axis rotation extension generates the corresponding three-dimensional frequency domain mask.

6. The method for measuring three-dimensional topography using white light interferometry with enhanced bright and dark field mixing according to claim 4, characterized in that, The method for morphological reconstruction of the time-domain interferometric signal sequence in step S4 is as follows: The contrast curve of each pixel's interference fringes as a function of scanning position was calculated, and the position of the coherent envelope peak was determined using the envelope centroid method to obtain the axial coarse positioning height. Near this position, a method based on... The orthogonal demodulation algorithm of phase-shifting beats calculates the interference phase information, and the coarse positioning result is combined with the phase information to obtain the first height map.

7. The method for measuring three-dimensional topography using white light interferometry with enhanced bright and dark field mixing as described in claim 4, characterized in that, The method for reconstructing the morphology of the temporal scattering intensity signal sequence in step S4 is as follows: the gray-level variance operator is used to calculate the degree of change of neighborhood gray-level difference of each pixel on different scanning layers, and a focusing curve is formed along the Z direction; the position of the focusing peak is determined by three-point Gaussian fitting to obtain the second height map.

8. The method for measuring three-dimensional topography using white light interferometry with enhanced bright and dark field mixing according to claim 4, characterized in that, The method for generating and fusing the quality mask in step S5 is as follows: Hilbert transform is performed on the pixel-by-pixel axial interference signal sequence to extract the envelope, the envelope peak value represents the effective signal intensity, the average value of the outer region represents the noise floor, and the signal-to-noise ratio is calculated to generate a binary quality mask. In regions where the signal-to-noise ratio is higher than a set threshold, a first height map is used, while in regions where the signal-to-noise ratio is lower than a set threshold, a smoothed second height map is used for fusion.

Images