Quantitative optical measurement method and device based on optical interference and medium

Abstract

The invention discloses a quantitative optical measurement method and device based on optical interference and a medium, and relates to the technical field of optical precision measurement, and the method comprises the steps: building an interference vision measurement light path, obtaining light source wavelength information, collecting a calibration image, forming a correction alignment rule, collecting a reference fringe image, and carrying out the processing according to the correction alignment rule, generating a reference phase and a reference stripe baseline; a measurement stripe image is collected and processed according to the correction alignment rule, a correction stripe sequence is generated, and observable pixel subsets are screened in the correction stripe sequence according to the reference stripe baseline; and performing phase shift self-calibration and phase inversion in the observable pixel subset to obtain a wrapped phase, a phase fitting residual error and fringe quality characteristics. Phase shift self-calibration and phase inversion are completed in an observable pixel subset, interference of unobservable pixels and phase shift deviation on phase calculation is isolated outside a calculation domain, and stability and consistency of a phase result in a visual measurement view field are improved.

Description

Technical Field

[0001] This invention relates to the field of optical precision measurement technology, and in particular to a quantitative optical measurement method, device and medium based on optical interferometry. Background Technology

[0002] Optical interferometry, as a non-contact measurement method with nanometer-level precision, has been continuously developing in the field of precision metrology since the advent of lasers. White light interferometry solves the phase ambiguity problem through short coherent light sources and is suitable for measuring stepped heights and surfaces with large roughness. In recent years, technologies such as partially coherent interferometry, polarization interferometry, and wavelength scanning interferometry have continuously expanded the measurement dimensions. Combined with the visual measurement acquisition capabilities of high-speed cameras and parallel processing algorithms, three-dimensional topographic measurement of dynamic processes has been realized. At the same time, active interferometry methods such as structured light projection and digital holography have extended the measurement objects from optically smooth surfaces to diffuse reflectors and biological samples, promoting the penetration of this technology into interdisciplinary fields such as industrial inspection and biomedicine.

[0003] Existing methods have shortcomings. Environmental disturbances and imaging noise can cause unobservable pixels in the stripe sequence within the visual measurement field of view, such as weak contrast, saturation underexposure, and local dominance frequency drift. If all pixels in the field of view are still used uniformly for phase inversion and nominal phase shift processing is adopted, it is easy to introduce phase shift bias and phase fitting residual accumulation, resulting in local instability and systematic bias in the wrapped phase. In addition, complex surfaces, especially at steep edges and in regions with large gradients, are more prone to interval periodic misalignment and boundary phase breakage. A single unfolding strategy cannot simultaneously take into account the phase continuity of high-quality regions and the repair of missing parts in low-quality regions, which can easily lead to boundary cracks and loss of details and reduce the usability and consistency of measurement results. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a quantitative optical measurement method based on optical interferometry to solve the problems of phase inversion instability and boundary misalignment and fracture.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a quantitative optical measurement method based on optical interferometry, comprising: establishing an interferometric visual measurement optical path; acquiring light source wavelength information and acquiring calibration images; forming correction alignment rules; acquiring reference fringe images and processing them according to the correction alignment rules to generate a reference phase and a reference fringe baseline; acquiring measurement fringe images and processing them according to the correction alignment rules to generate a correction fringe sequence; selecting an observable pixel subset from the correction fringe sequence based on the reference fringe baseline; and performing phase shift self-calibration and phase inversion in the observable pixel subset to obtain the wrapping phase, phase fitting residual, and fringe quality characteristics. Quality partitions are formed based on the wrapping phase, fringe quality characteristics, and reference fringe baseline, and routing rules are generated. Continuous phase recovery and edge repair are performed on different quality partitions according to the routing rules to obtain continuous phase and consistency markers. The phase change is obtained based on the continuous phase and reference phase, and the optical path difference distribution is generated based on the light source wavelength information. The corresponding quantitative conversion rules are executed according to the measurement object type to obtain quantitative measurement results, and the confidence characterization is generated by the phase fitting residual and consistency markers.

[0007] As a preferred embodiment of the quantitative optical measurement method based on optical interferometry described in this invention, the specific steps for establishing the interferometric visual measurement optical path, acquiring light source wavelength information and collecting calibration images, and forming correction alignment rules are as follows. The interferometric vision measurement optical path components are connected and fixed in optical path sequence to form the interferometric vision measurement optical path, and the wavelength information of the light source is read from the light source instruction manual. Under the condition that the optical path of the interferometric vision measurement remains unchanged, the calibration image is acquired and the dark field calibration image, flat field calibration image and geometric calibration target calibration image are output. The calibration calculations and parameter processing are performed on the dark field calibration image, flat field calibration image, and geometric calibration target calibration image to generate correction alignment rules.

[0008] As a preferred embodiment of the quantitative optical measurement method based on optical interferometry described in this invention, the steps of acquiring a reference fringe image and processing it according to a correction alignment rule to generate a reference phase and a reference fringe baseline are as follows. A reference fringe image is acquired under the interferometric visual measurement optical path and processed according to the correction alignment rules to obtain a reference corrected fringe image; Perform phase solving and fringe quality calculation on the reference corrected fringe image to generate the reference phase and reference fringe baseline.

[0009] As a preferred embodiment of the quantitative optical measurement method based on optical interferometry described in this invention, the specific steps for obtaining the wrapping phase, phase fitting residual, and fringe quality characteristics are as follows: Acquire measurement fringe images to form a measurement fringe image set, perform intensity correction and coordinate alignment on the measurement fringe image set according to the correction alignment rules, and output the correction fringe sequence; Based on the reference fringe baseline, the observability of the corrected fringe sequence is evaluated and a subset of observable pixels is selected. Phase shift self-calibration is then performed on the corrected fringe sequence within the subset of observable pixels to generate actual phase shift information. Phase inversion is performed on the corrected fringe sequence based on actual phase shift information to obtain the wrapping phase, and the phase fitting residual is calculated. Fringe quality features are then extracted based on the corrected fringe sequence and the phase fitting residual.

[0010] As a preferred embodiment of the quantitative optical measurement method based on optical interferometry described in this invention, the specific steps for forming quality partitions and generating routing rules based on the wrapping phase, fringe quality characteristics, and reference fringe baseline are as follows: A quality score map is generated by comparing the stripe quality characteristics against a reference stripe baseline. Based on the quality score map, perform quality partitioning on the pixel positions corresponding to the package phase, and output a quality partition map and a quality partition set; Perform partition classification and action mapping on the quality partition set to generate routing rules.

[0011] As a preferred embodiment of the quantitative optical measurement method based on optical interferometry described in this invention, the specific steps of performing continuous phase recovery and edge repair on different quality partitions according to routing rules to obtain continuous phase and consistency markers are as follows. Based on the quality partition map, the quality partition boundaries are extracted. According to the routing rules, the continuous phase of the package is restored by partition and edge repair is performed at the quality partition boundaries. The continuous phase of the partition is then output. Consistency fusion is performed on the continuous phases of the partition to generate continuous phases, and phase jump test and phase gradient continuity test are performed at the boundary of the quality partition based on the continuous phases to generate consistency markers.

[0012] As a preferred embodiment of the quantitative optical measurement method based on optical interference described in this invention, the specific steps for obtaining the phase change based on the continuous phase and the reference phase, and generating the optical path difference distribution based on the light source wavelength information, are as follows: Given that the continuous phase and the reference phase are at the same pixel coordinate position, the phase change is generated by subtracting the continuous phase and the reference phase pixel by pixel. The phase change is combined with the wavelength information of the light source and converted according to the conversion relationship from phase to optical path difference to generate the optical path difference distribution.

[0013] As a preferred embodiment of the quantitative optical measurement method based on optical interferometry described in this invention, the quantitative measurement result is obtained by performing the corresponding quantitative conversion rule according to the type of measurement object, and a reliability characterization is generated by the phase fitting residual and consistency marker. The specific steps are as follows: Select the corresponding quantitative conversion rule according to the type of measurement object and perform quantitative conversion on the optical path difference distribution to generate quantitative measurement results; The residuals of the phase fitting are classified into residual levels and the consistency labels are classified into consistency levels. The residual levels and consistency levels are combined and mapped to generate a credibility characterization.

[0014] In a second aspect, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein when the computer program is executed by the processor, it implements any step of the quantitative optical measurement method based on optical interferometry as described in the first aspect of the present invention.

[0015] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the quantitative optical measurement method based on optical interference as described in the first aspect of the present invention.

[0016] The beneficial effects of this invention are as follows: By performing phase shift self-calibration and phase inversion within a subset of observable pixels, the interference of unobservable pixels and phase shift deviations on phase calculations is isolated outside the computational domain, thereby improving the stability and consistency of phase results within the visual measurement field of view. This reduces the amplification effect of deviations in subsequent phase changes, optical path differences, and quantitative conversions, and improves the accuracy and repeatability of quantitative measurements at the micro-nano scale. By partitioning the quality and performing continuous phase recovery, boundary repair, and consistency checks on different partitions according to routing rules, targeted suppression of periodic misalignment and phase breakage at complex surface boundaries is achieved. This ensures that the continuous phase within the visual measurement field of view remains coherent even in high-gradient and steep edge regions, reducing boundary cracks and detail loss, and improving the usability of quantitative measurement results and the traceability of quality control. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a quantitative optical measurement method based on optical interferometry.

[0019] Figure 2 A flowchart for generating the reference phase and reference fringe baseline.

[0020] Figure 3 A flowchart for generating routing rules.

[0021] Figure 4 Flowchart for generating quantitative measurement results and confidence characterization Detailed Implementation To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0022] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0023] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0024] Reference Figures 1-4 This is one embodiment of the present invention, which provides a quantitative optical measurement method based on optical interferometry, comprising the following steps: S1. Establish the interferometric visual measurement optical path, acquire the wavelength information of the light source and collect the calibration image, form the correction alignment rule, collect the reference fringe image and process it according to the correction alignment rule to generate the reference phase and reference fringe baseline.

[0025] S1.1 Connect and fix the interferometric visual measurement optical path components in optical path sequence to form the interferometric visual measurement optical path, and read the wavelength information of the light source from the light source instruction manual.

[0026] It should be noted that the interferometric visual measurement optical path assembly includes a light source, a beam splitter, a reference optical path reflector, a measurement optical path reflector, a beam combiner, an imaging lens, and a camera. The optical path sequence refers to the light emitted from the light source being sequentially split into a reference optical path and a measurement optical path by the beam splitter. The reference optical path is reflected back after reaching the reference optical path reflector and then merges into the beam combiner path through the beam splitter. The measurement optical path is reflected back after reaching the measurement optical path reflector and then merges into the beam combiner path through the beam splitter. The beam combiner path is superimposed by the beam combiner to form interference light, which is then imaged onto the camera through the imaging lens.

[0027] According to the optical path sequence, the light source, beam splitter, reference optical path reflector, measurement optical path reflector, beam combiner, imaging lens, and camera are installed, positioned, and fixed in sequence to form the interferometric visual measurement optical path. Under the condition that the interferometric visual measurement optical path remains fixed, the center wavelength of the light source is read from the light source instruction manual and recorded as the light source wavelength information.

[0028] S1.2 Acquire calibration images while keeping the interferometric visual measurement optical path unchanged, output dark field calibration images, flat field calibration images and geometric calibration target calibration images, and perform calibration calculations and parameter processing on the three types of calibration images to generate correction alignment rules.

[0029] It should be noted that, under the condition that the interferometric visual measurement optical path remains unchanged, the camera exposure time, camera gain, camera resolution and camera frame rate are fixed, and the working state of the light source and the position of the camera field of view are kept unchanged. Under the acquisition conditions that the light source is turned off and the incident light is blocked, the dark field calibration image sequence is acquired, and the mean value is calculated at each pixel position of the dark field calibration image sequence to obtain the dark field bias term.

[0030] Under the conditions of light source on and uniform brightness in the camera field of view, a flat field calibration image sequence is acquired. The dark field bias term is first subtracted from the flat field calibration image sequence to obtain the flat field response distribution. The mean of the flat field response distribution is calculated for the entire field of view, and the ratio of the flat field response distribution at each pixel position is normalized by the mean of the entire field of view to obtain the relative response. The reciprocal of the relative response is recorded as the gain compensation term.

[0031] Under the premise that the camera acquisition conditions remain unchanged, a geometric calibration target is placed and a geometric calibration target calibration image sequence is acquired. The geometric calibration target calibration image sequence is subjected to intensity correction pixel by pixel to obtain an intensity-corrected geometric calibration target calibration image sequence. A frame is selected as a reference frame in the intensity-corrected geometric calibration target calibration image sequence. The geometric calibration target area is defined in the reference frame. The corner positions are extracted in the geometric calibration target area and sub-pixel localization is performed on the corner positions. The corner points are numbered according to the row and column scanning order to obtain the pixel coordinates of the geometric calibration target feature points. The pixel coordinates of the geometric calibration target feature points are paired with the geometric calibration target reference coordinates (from the geometric calibration target instruction manual) according to the same number to form a correspondence. Let the matrix elements of the coordinate mapping matrix be the parameters to be solved. The correspondence is input into the solution process of the mapping parameters. The coordinate mapping matrix is ​​obtained by using "the sum of the squares of the coordinate deviations between the coordinates after matrix mapping of all correspondences and the reference coordinates of the geometric calibration target is minimized". The expression for performing pixel-by-pixel intensity correction on the geometric calibration target calibration image sequence is as follows: ; in, The pixel coordinates after intensity correction are: Image pixel values; Indicates pixel coordinates The pixel values ​​of the geometric calibration target calibration image sequence; Indicates pixel coordinates Dark field bias term; Indicates pixel coordinates Gain compensation term; Indicates pixel position coordinates; subscript This is an abbreviation for "corrected" after strength correction.

[0032] The parameters of the dark field bias term, gain compensation term, and coordinate mapping matrix are processed. The dark field bias term and gain compensation term are merged into the intensity correction part, and the coordinate mapping matrix is ​​merged into the coordinate alignment part, forming the correction alignment rule.

[0033] S1.3 Acquire a reference fringe image under the interferometric visual measurement optical path and process it according to the correction alignment rules to obtain a reference corrected fringe image. Perform phase solving and fringe quality calculation on the reference corrected fringe image to generate the reference phase and reference fringe baseline.

[0034] It should be noted that, under the conditions of fixed camera exposure time, camera gain, camera resolution and camera frame rate in the interferometric visual measurement optical path, and to ensure that the camera field of view presents a stable interference fringe state, multiple frames of reference fringe images are continuously acquired and the reference frame index is recorded to form a reference fringe image; intensity correction is performed on the reference fringe image pixel by pixel, and the intensity-corrected reference fringe image is resampled and aligned according to the coordinate alignment part to obtain the reference corrected fringe image.

[0035] The reference correction fringe image is organized into multiple grayscale sequences at the same pixel position according to the reference frame index. A phase-shift interference intensity relationship is established at each pixel position, and the phase shift of each frame is used as the reference phase shift information to be determined. The reference phase shift information corresponding to the reference frame index is obtained using a least-squares fitting method. The multiple grayscale sequences are weighted and accumulated according to the cosine and sine terms of the reference phase shift information of each frame, forming a cosine weighted sum and a sine weighted sum. The wrapping phase at the corresponding pixel position is calculated from the sine weighted sum and the cosine weighted sum. The wrapping phase is then written point-by-point into a phase map of the same size as the reference correction fringe image according to the pixel coordinates. This phase map is the reference phase. The expression for the phase-shift interference intensity relationship is as follows: ; ; in, Represents pixel coordinates First The image pixel values ​​of the frame; Indicates pixel coordinates The average pixel value of the image at that location; Indicates pixel coordinates The combined amplitude of the cosine-weighted sum and the sine-weighted sum at a given point; Indicates pixel coordinates The encapsulation phase at the location; Indicates the first Reference phase shift information of the frame reference correction stripe image; Indicates the reference frame index; Indicates pixel coordinates Sine weighted sum at the point; Indicates pixel coordinates The cosine weighted sum at the point.

[0036] In the reference correction fringe image, pixels whose grayscale value equals the upper boundary of the camera's grayscale value are marked as reference saturated pixels, pixels whose grayscale value equals the lower boundary of the camera's grayscale value are marked as reference underexposed pixels, and the remaining pixels are marked as reference valid pixels. A reference contrast baseline map is formed pixel by pixel on the reference correction fringe image using a sliding window. Specifically, for each window, the maximum and minimum grayscale values ​​within the window are taken, and the intensity of the fringe brightness fluctuation is characterized by calculating the difference between the maximum and minimum grayscale values. The local contrast value of the pixel at the center of the window is obtained by calculating the sum of the maximum and minimum grayscale values ​​and normalizing it, and then organized into a reference contrast baseline map.

[0037] A reference frequency stability baseline map is formed pixel by pixel using a sliding window. Specifically, the grayscale values ​​within the window are de-meaned, a discrete spectrum transformation is performed on the window, and the position of the spectral peak with the largest amplitude in the non-zero frequency part is found. The position of the spectral peak with the largest amplitude is represented by the "spectral discrete point index" and used as the peak position index. Using the number of pixels corresponding to the side length of the sliding window as the normalization scale, the ratio of the peak position index to the number of pixels of the side length of the sliding window is calculated to obtain the local fringe dominant frequency of the center pixel of the window. The local fringe dominant frequency is a dimensionless quantity. Peak location and normalization are repeated in adjacent overlapping windows to obtain the local fringe dominant frequencies of adjacent windows. The variation amplitude of the local fringe dominant frequency in the neighborhood of the same pixel is statistically analyzed, and the statistical results are compiled into a reference frequency stability baseline map.

[0038] A reference noise baseline map is formed pixel by pixel. Specifically, local smoothing is performed on the grayscale within the sliding window (the grayscale of the pixels within the sliding window is smoothed by mean smoothing to obtain the smoothing result). The difference between the original grayscale and the smoothing result is calculated to obtain the smoothing residual. The mean of the absolute values ​​of the smoothing residuals within the sliding window is used to complete the smoothing residual statistics. The smoothing residual statistics are organized into a reference noise baseline map. The reference contrast baseline map, reference frequency stability baseline map, reference noise baseline map and reference effective pixels are aligned according to the pixel position to form a reference stripe baseline.

[0039] S2. Acquire and measure fringe images and process them according to the calibration alignment rules to generate a calibration fringe sequence. Select an observable pixel subset from the calibration fringe sequence based on the reference fringe baseline. Perform phase shift self-calibration and phase inversion on the observable pixel subset to obtain the wrapping phase, phase fitting residual and fringe quality characteristics.

[0040] S2.1 Acquire measurement fringe images to form a measurement fringe image set. Perform intensity correction and coordinate alignment on the measurement fringe image set according to the correction alignment rules, and output the correction fringe sequence.

[0041] It should be noted that in the interferometric vision measurement optical path, the object to be measured is placed at the measurement position, and the reference mirror is controlled to perform step displacement by a piezoelectric ceramic actuator to acquire a set of measurement fringe images containing multiple phase-shifted images. The measurement fringe image set is processed by applying correction alignment rules: the dark field bias term in the correction alignment rules is used to subtract background noise from each measurement fringe image to eliminate the influence of detector dark current; the gain compensation term in the correction alignment rules is used to correct the illumination uniformity of each measurement fringe image to compensate for uneven light field distribution; the coordinate mapping matrix in the correction alignment rules is used to perform resampling alignment on each measurement fringe image to ensure that the pixel coordinates of the measurement fringe image are consistent with the reference coordinates of the geometric calibration target; after intensity correction and coordinate alignment processing, a correction fringe sequence with consistent geometry and intensity is output.

[0042] S2.2. Based on the reference fringe baseline, perform observability evaluation on the corrected fringe sequence and select an observable pixel subset. Perform phase shift self-calibration on the corrected fringe sequence within the observable pixel subset to generate actual phase shift information.

[0043] It should be noted that the measurement fringe image set is written frame by frame into the measurement frame index according to the acquisition order, so that each frame in the calibration fringe sequence corresponds to the measurement frame index. In the calibration fringe sequence, each pixel gray level is checked frame by frame to see if it reaches the upper boundary of the camera gray level and is marked as a measurement saturated pixel, and whether it reaches the lower boundary of the camera gray level and is marked as a measurement underexposed pixel. Pixels that are neither saturated nor underexposed are marked as measurement valid pixels. At the measurement valid pixel position, the difference between the maximum gray level and the minimum gray level is calculated pixel by pixel in the calibration fringe sequence using a sliding window as the local fringe contrast, which is used to characterize the intensity of the fringe brightness fluctuation. The local fringe contrast, the sum of the maximum gray level and the minimum gray level are normalized to obtain the local contrast value of the pixel in the center of the window. The local contrast value is compared with the reference contrast baseline map in the reference fringe baseline at the same pixel position to obtain the contrast deviation.

[0044] Within the same sliding window, pixel grayscale values ​​are mean-reduced and a discrete spectrum transform is performed. The spectral peak with the largest amplitude in the non-zero frequency region is located, and the position of the spectral peak with the largest amplitude is taken as the local fringe dominant frequency of the center pixel of the window. The local fringe dominant frequency is obtained by normalizing the peak position index according to the number of pixels of the sliding window side length, and is a dimensionless quantity. The dominant frequency location is repeated in the corresponding windows of adjacent frames, and the cross-frame dispersion of the local fringe dominant frequency is statistically analyzed to obtain the fringe frequency stability index of the center pixel of the window. The local fringe dominant frequency is compared with the reference frequency stability baseline map at the same pixel position to obtain the frequency deviation. The expression for calculating the fringe frequency stability index is as follows: ; ; in, Represents pixel coordinates The stripe frequency stability index at the location, the larger the value, the more stable the main frequency across frames; This indicates the total number of measurement frame indices contained in the correction stripe sequence; Indicates the measurement frame index; Represents pixel coordinates First The dominant frequency of local stripes in a frame; Represents pixel coordinates The average frequency across frames at that location.

[0045] Within the same sliding window, the gray-level gradient direction is calculated, and the consistency of the gradient directions within the window is statistically analyzed. This consistency is defined as the stripe direction consistency index. The expression for calculating the stripe direction consistency index is as follows: ; in, Represents pixel coordinates First The frame stripe direction consistency index ranges from 0 to 1. The larger the value, the more consistent the direction within the window. Represented in pixel coordinates The set of pixels centered on the sliding window; Represents any pixel coordinate variable within the pixel set of the sliding window; This represents the total number of pixels within the sliding window; Represents pixel coordinates First The grayscale gradient direction angle of the frame; Represents the natural constant; It represents the imaginary unit.

[0046] The contrast deviation, frequency deviation, and fringe direction consistency index are combined into an observability score map according to pixel position, and then combined with saturated pixel markers and underexposed pixel markers to form a set of effective pixels for measurement, which is the observable pixel subset. Within the observable pixel subset, the phase shift interference intensity relationship is established using the correction fringe sequence and the measurement frame index, and the phase shift of each frame is taken as the actual phase shift information to be obtained. The actual phase shift information corresponding to the measurement frame index is obtained by least squares solution.

[0047] S2.3. Based on the actual phase shift information, perform phase inversion on the corrected fringe sequence to obtain the wrapping phase, and calculate the phase fitting residual. Extract fringe quality features based on the corrected fringe sequence and the phase fitting residual.

[0048] It should be noted that in the calibration fringe sequence, multiple frames of gray values ​​at the same pixel position are organized into a gray value sequence according to the measurement frame index. Within the observable pixel subset, the gray value sequence is weighted by cosine terms and sine terms respectively based on the phase shift amount given by the actual phase shift information to obtain the cosine weighted sum and sine weighted sum. The wrapping phase at the corresponding pixel position is then calculated using the cosine weighted sum and sine weighted sum. At the same pixel position, the background term and modulation amplitude are obtained based on the multiple frames of gray value sequence. The background term, modulation amplitude, wrapping phase, and actual phase shift information are substituted into the phase shift interference intensity relationship to reconstruct the fringe sequence frame by frame to obtain the reconstructed fringe sequence. The inter-frame reconstruction error between the calibration fringe sequence and the reconstructed fringe sequence is calculated frame by frame at the pixel position. The root mean square statistics are performed on all frames to obtain the phase fitting residual amplitude index. On the calibration fringe sequence, the local fringe contrast, fringe direction consistency index, and fringe frequency stability index are calculated pixel by pixel using a sliding window. These are then aligned with the phase fitting residual amplitude index at the pixel position and summarized to form the fringe quality feature.

[0049] The expression for calculating the inter-frame reconstruction error is as follows: ; in, Represents pixel coordinates In the Inter-frame reconstruction error; Represents pixel coordinates First The reconstructed image pixel values ​​of the frame.

[0050] The expression for calculating the magnitude index of the phase fitting residual is as follows: ; in, Represents pixel coordinates The phase fitting residual magnitude index.

[0051] It should also be noted that existing technologies directly calculate the phase by acquiring multiple frames of phase-shifted fringes and completing intensity correction and coordinate alignment. However, these calculations still involve phase shift deviations or deterioration in fringe quality, which can easily lead to phase errors and make it difficult to quantify the reliability. This solution uses pixel availability screening and phase shift self-calibration based on the reference fringe baseline, and combines the root mean square residual of the reconstruction error with fringe quality characteristics for consistency evaluation. This makes the phase inversion more closely match the actual phase shift state, reduces the error caused by phase shift deviations, suppresses the drag of deteriorated areas on the overall results, and outputs locatable residuals and quality characteristics, which facilitates the identification and traceability of the measurement quality distribution, and improves measurement stability, repeatability, and interpretability.

[0052] S3. Based on the wrapped phase, fringe quality features and reference fringe baseline, quality partitions are formed and routing rules are generated. Continuous phase recovery and edge repair are performed on different quality partitions according to the routing rules to obtain continuous phase and consistency markings.

[0053] S3.1. Based on the reference fringe baseline, compare the fringe quality features to form a quality score map. Based on the quality score map, perform quality partitioning on the pixel positions corresponding to the wrapped phase, and output a quality partition map and a quality partition set.

[0054] It should be noted that when the local stripe contrast is not lower than the pixel value of the reference contrast baseline at the same pixel position, the contrast score is recorded as 1. When the local stripe contrast is lower than the pixel value of the reference contrast baseline at the same pixel position, the ratio of the local stripe contrast to the pixel value of the reference contrast baseline at the same pixel position is recorded as the contrast score.

[0055] When the fringe frequency stability index is not lower than the pixel value of the reference frequency stability baseline at the same pixel position, the frequency score is recorded as 1. When the fringe frequency stability index is lower than the pixel value of the reference frequency stability baseline at the same pixel position, the ratio of the fringe frequency stability index to the pixel value of the reference frequency stability baseline at the same pixel position is recorded as the frequency score.

[0056] When the phase fitting residual magnitude index is not higher than the pixel value of the reference noise baseline at the same pixel position, the residual score is recorded as 1. When the phase fitting residual magnitude index is higher than the pixel value of the reference noise baseline at the same pixel position, the ratio of the pixel value of the reference noise baseline at the same pixel position to the phase fitting residual magnitude index is recorded as the residual score.

[0057] The fringe direction consistency index is marked as the direction score. The contrast score, frequency score, residual score, and direction score are arithmetically averaged at the current pixel position to obtain the quality score value for the current pixel position. The quality score values ​​of all pixel positions are arranged according to the pixel coordinates of the phase wrapping to obtain the quality score map. The positions marked as invalid in the observable pixel subset are marked with a quality score value of 0. All quality score values ​​of the positions where the measured effective pixels are valid are extracted from the quality score map and sorted from high to low according to the value of the quality score value. The observable pixel subset is divided into three equal parts into high-quality partitions, medium-quality partitions, and low-quality partitions. Each measured effective pixel is labeled with the corresponding partition number according to the interval to which its quality score value belongs in the sorted sequence, and arranged according to the pixel coordinates to obtain the quality partition map. The pixel coordinates are summarized according to the partition number to obtain the high-quality pixel coordinate set, the medium-quality pixel coordinate set, and the low-quality pixel coordinate set, which are then merged to form the quality partition set.

[0058] It should also be noted that the value of the trisection is based on the fact that the number of quality partitions corresponds to the three different processing methods used for high-quality, medium-quality, and low-quality pixels, respectively: region growing phase expansion, least squares phase expansion with boundary constraints, and edge-preserving interpolation repair.

[0059] S3.2 Perform partition classification and action mapping operations on the quality partition set to generate routing rules.

[0060] It should be noted that the process involves reading the high-quality pixel coordinate set, the medium-quality pixel coordinate set, and the low-quality pixel coordinate set from the quality partition set, and labeling these three types of pixel coordinate sets as high-quality partition category, medium-quality partition category, and low-quality partition category, respectively, to form a partition classification table. In the partition classification table, a corresponding processing action is assigned to each partition category, forming an action mapping table. The high-quality partition category is mapped to a region growing phase unfolding action, the medium-quality partition category is mapped to a least-squares phase unfolding action with boundary constraints, and the low-quality partition category is mapped to an edge-preserving interpolation repair action. The partition classification table and the action mapping table are associated according to the partition category field to obtain a routing rule composed of a triple of "partition category - pixel coordinate set - processing action".

[0061] S3.3 Extract the quality partition boundary based on the quality partition map, perform partition continuous phase recovery on the package phase according to the routing rules, perform edge repair at the quality partition boundary, and output the partition continuous phase.

[0062] It should be noted that in the quality partition map, the partition number of the current pixel is compared with the partition number of its four neighboring pixels. The coordinates of pixels whose partition numbers are different from any of the adjacent pixels are summarized to form a set of quality partition boundary pixels, and then organized according to connectivity to form a set of boundary segments.

[0063] For each high-quality connected component, select a starting pixel and set the continuous phase of the starting pixel to be equal to the wrapping phase. Traverse the pixels in the high-quality connected component along the four neighborhoods, calculate the wrapping phase difference between adjacent pixels and convert it to the range of (−π, π], then accumulate and update the continuous phase of adjacent pixels to obtain the continuous phase of the high-quality partition.

[0064] For each medium-quality connected component, at the boundary where it connects with the high-quality partition, select a ring of boundary pixels and let the continuous phase be the continuous phase of the high-quality partition. Inside the medium-quality connected component, the continuous phase of the pixels is the unknown quantity, and the phase difference of adjacent pixels is used as the difference constraint for iterative update. The sum of squares of the difference constraint is minimized under the condition that the boundary pixels have fixed values, thus obtaining the continuous phase of the medium-quality partition.

[0065] For each low-quality connected component, the continuous phase of non-low-quality pixels is collected within a sliding window as a reference sample. Weights are assigned according to the inverse of the spatial distance and phase fitting residual magnitude index, and a weighted average is calculated to obtain candidate continuous phases. The candidate continuous phases are aligned with the wrapping phase of the current pixel to the same 2π branch to obtain the low-quality partition continuous phases.

[0066] In the boundary segment set, the continuous phase difference between adjacent pixels on both sides of the boundary is compared segment by segment. The continuous phase difference is converted to (-π, π] and 2π is added or subtracted on one side of the boundary segment with a jump close to 2π to eliminate the whole cycle misalignment. At the boundary pixel set of the quality partition, the continuous phase on both sides of the boundary is used as a reference to perform edge-preserving interpolation repair on the boundary pixels to eliminate cracks. The continuous phases of the three partitions of high quality, medium quality and low quality are arranged and sorted according to pixel coordinates to obtain the partition continuous phase.

[0067] 2π is the integer period of the phase, and (-π, π] is the standard interval used to select the "minimum amplitude difference" in the integer period equivalence class.

[0068] S3.4 Perform consistency fusion on the continuous phase of the partition to generate continuous phase, and perform phase jump test and phase gradient continuity test at the boundary of the quality partition based on the continuous phase to generate consistency mark.

[0069] It should be noted that, on each boundary segment, adjacent pixel pairs with different partition numbers on both sides of the boundary are taken point by point. The continuous phase values ​​of the partitions on both sides are taken respectively, and a number of 2π are added or subtracted from the phase value on one side to make the phase difference between the two sides fall within the range of (-π, π]. The quality score values ​​corresponding to the pixels on both sides are used as weights to calculate the weighted average of the phases on both sides to obtain the fused phase at the boundary. For non-boundary pixels, the continuous phase value of the partition to which they belong is still taken as the non-boundary phase. The boundary fused phase and the non-boundary phase are arranged and sorted according to the pixel coordinates to obtain the continuous phase.

[0070] For the same boundary pixel pair, calculate the continuous phase difference on both sides and subtract the nearest integer multiple of 2π to make the continuous phase difference fall within the range of (-π, π]. Calculate the arithmetic mean of the absolute values ​​of the continuous phase differences of all pixel pairs within the boundary segment to obtain the phase jump test value of the current boundary segment. On both sides of the same boundary pixel pair, calculate the phase gradient components of the continuous phase in the horizontal and vertical directions using the central difference method, and calculate the magnitude of the gradient difference. Calculate the arithmetic mean of the magnitudes of the gradient differences of all pixel pairs within the boundary segment to obtain the phase gradient continuity test value of the current boundary segment.

[0071] Calculate the gradient magnitude of continuous phases in the neighborhood on both sides of the current boundary segment, and take the median as the local gradient benchmark of the current boundary segment. Compare the phase jump test quantity and the phase gradient continuity test quantity with the local gradient benchmark respectively, and give a consistency mark of consistency or inconsistency.

[0072] S4. The phase change is obtained based on the continuous phase and the reference phase, and the optical path difference distribution is generated based on the wavelength information of the light source. The corresponding quantitative conversion rule is executed according to the type of measurement object to obtain the quantitative measurement result, and the confidence characterization is generated by the phase fitting residual and the consistency mark.

[0073] S4.1 Under the premise that the continuous phase and the reference phase are at the same pixel coordinate position, calculate the difference between the continuous phase and the reference phase pixel by pixel to generate the phase change.

[0074] It should be noted that when the continuous phase and the reference phase have the same size and the same pixel coordinate definition, the continuous phase value and the reference phase value are taken as a pair for any pixel coordinate. When the pixel coordinate definitions of the continuous phase and the reference phase are inconsistent, one of the phases is resampled and aligned according to the coordinate mapping matrix so that the two phases form a point-to-point correspondence on the same pixel coordinate grid. After the two phases form a point-to-point correspondence, the continuous phase value and the reference phase value are taken for each pixel coordinate, and the difference between the continuous phase value and the reference phase value is calculated as the phase difference. The phase differences of all pixel coordinates are arranged and sorted according to the pixel coordinates to obtain the phase change.

[0075] S4.2 Combine the phase change with the light source wavelength information and perform conversion according to the phase-to-optical path difference conversion relationship to generate the optical path difference distribution. Select the corresponding quantitative conversion rule according to the type of measurement object and perform quantitative conversion on the optical path difference distribution to generate quantitative measurement results.

[0076] It should be noted that, based on the phase change and light source wavelength information, the optical path difference is calculated at each pixel coordinate according to the conversion relationship from phase to optical path difference. The optical path differences of all pixel coordinates are then arranged and organized according to the pixel coordinates to obtain the optical path difference distribution. The expression for calculating the optical path difference is as follows: ; in, Represents pixel coordinates Optical path difference at the location; Indicates the wavelength information of the light source; Represents pixel coordinates Phase change at that point.

[0077] The measurement object type identifier is read and the corresponding quantitative conversion rule is selected. When the measurement object type is a reflective morphology, the optical path difference distribution is converted into a surface height variation distribution by performing a pixel-by-pixel reflection double-path optical path difference halving conversion on the optical path difference distribution, and then arranged and organized according to pixel coordinates to obtain the quantitative measurement results. When the measurement object type is a transmission thickness, the refractive index of the measurement object material is read, and the optical path difference distribution is converted into a thickness distribution by performing a refractive index correction thickness conversion on the optical path difference distribution, and then arranged and organized according to pixel coordinates to obtain the quantitative measurement results. The expression for calculating the surface height variation distribution is as follows: ; in, Represents pixel coordinates Distribution of surface height variation at the location; The halving factor refers to the optical path difference that occurs twice the change in height during one round trip of the optical path in a reflection measurement.

[0078] The expression for calculating the thickness distribution is as follows: ; in, Represents pixel coordinates Thickness distribution at the location; It indicates the refractive index of the material being measured.

[0079] S4.3 Perform residual level classification on the phase fitting residuals and consistency level classification on the consistency labels, and combine the residual levels and consistency levels to generate a credibility characterization.

[0080] It should be noted that all phase fitting residual amplitude indices were extracted and sorted from smallest to largest. The number of effective pixels was divided into three equal groups: low residual group, medium residual group, and high residual group. Each measured effective pixel was labeled with the corresponding residual level according to the interval of the phase fitting residual amplitude index.

[0081] In the consistency tagging, locations marked as consistent are labeled as high consistency level, locations marked as inconsistent are labeled as low consistency level, and locations not covered by the consistency tagging are labeled as no consistency level. During the combined mapping, the no consistency level is treated as low confidence. The residual level and the consistency level are paired according to pixel coordinates to form a level combination. Low residual level and high consistency level correspond to high confidence, medium residual level and high consistency level correspond to medium confidence, and high residual level or low consistency level correspond to low confidence. The confidence labels are arranged and organized according to pixel coordinates to obtain the confidence characterization.

[0082] This embodiment also provides a computer device applicable to quantitative optical measurement methods based on optical interferometry, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the quantitative optical measurement method based on optical interferometry as proposed in the above embodiment.

[0083] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0084] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the quantitative optical measurement method based on optical interferometry as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0085] In summary, this invention improves the stability and consistency of phase results within the visual measurement field of view by performing phase shift self-calibration and phase inversion within a subset of observable pixels, isolating the interference of unobservable pixels and phase shift deviations from the computational domain. This reduces the amplification effect of deviations in subsequent phase changes, optical path differences, and quantitative conversions, thereby enhancing the accuracy and repeatability of quantitative measurements at the micro-nano scale. Furthermore, by partitioning the quality data and performing continuous phase recovery, boundary repair, and consistency checks on different partitions according to routing rules, this invention achieves targeted suppression of periodic misalignment and phase breakage at complex surface boundaries. This ensures that the continuous phase within the visual measurement field of view remains coherent even in high-gradient and steep edge regions, reducing boundary cracks and detail loss, and improving the usability and traceability of quantitative measurement results.

[0086] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A quantitative optical measurement method based on optical interferometry, characterized in that: include, Establish an interferometric visual measurement optical path, acquire light source wavelength information and collect calibration images, form correction alignment rules, collect reference fringe images and process them according to the correction alignment rules to generate reference phase and reference fringe baseline; Acquire and measure fringe images and process them according to the calibration alignment rules to generate a calibration fringe sequence. Select an observable pixel subset from the calibration fringe sequence based on the reference fringe baseline. Perform phase shift self-calibration and phase inversion on the observable pixel subset to obtain the wrapping phase, phase fitting residual and fringe quality characteristics. Based on the wrapping phase, fringe quality features and reference fringe baseline, quality partitions are formed and routing rules are generated. Continuous phase recovery and edge repair are performed on different quality partitions according to the routing rules to obtain continuous phase and consistency markers. The phase change is obtained based on the continuous phase and the reference phase, and the optical path difference distribution is generated based on the wavelength information of the light source. The corresponding quantitative conversion rules are executed according to the type of measurement object to obtain the quantitative measurement result, and the confidence characterization is generated by the phase fitting residual and the consistency mark.

2. The quantitative optical measurement method based on optical interferometry as described in claim 1, characterized in that: The specific steps for establishing the interferometric visual measurement optical path, acquiring light source wavelength information and collecting calibration images, and forming correction alignment rules are as follows. The interferometric vision measurement optical path components are connected and fixed in optical path sequence to form the interferometric vision measurement optical path, and the wavelength information of the light source is read from the light source instruction manual. Under the condition that the optical path of the interferometric vision measurement remains unchanged, the calibration image is acquired and the dark field calibration image, flat field calibration image and geometric calibration target calibration image are output. The calibration calculations and parameter processing are performed on the dark field calibration image, flat field calibration image, and geometric calibration target calibration image to generate correction alignment rules.

3. The quantitative optical measurement method based on optical interferometry as described in claim 2, characterized in that: The process of acquiring a reference fringe image and processing it according to correction and alignment rules to generate a reference phase and a reference fringe baseline is as follows: A reference fringe image is acquired under the interferometric visual measurement optical path and processed according to the correction alignment rules to obtain a reference corrected fringe image; Perform phase solving and fringe quality calculation on the reference corrected fringe image to generate the reference phase and reference fringe baseline.

4. The quantitative optical measurement method based on optical interferometry as described in claim 3, characterized in that: The specific steps for obtaining the wrapped phase, phase fitting residual, and fringe quality characteristics are as follows: Acquire measurement fringe images to form a measurement fringe image set, perform intensity correction and coordinate alignment on the measurement fringe image set according to the correction alignment rules, and output the correction fringe sequence; Based on the reference fringe baseline, the observability of the corrected fringe sequence is evaluated and a subset of observable pixels is selected. Phase shift self-calibration is then performed on the corrected fringe sequence within the subset of observable pixels to generate actual phase shift information. Phase inversion is performed on the corrected fringe sequence based on actual phase shift information to obtain the wrapping phase, and the phase fitting residual is calculated. Fringe quality features are then extracted based on the corrected fringe sequence and the phase fitting residual.

5. The quantitative optical measurement method based on optical interferometry as described in claim 4, characterized in that: The specific steps for forming quality partitions and generating routing rules based on the package phase, fringe quality features, and reference fringe baseline are as follows. A quality score map is generated by comparing the stripe quality characteristics against a reference stripe baseline. Based on the quality score map, perform quality partitioning on the pixel positions corresponding to the package phase, and output a quality partition map and a quality partition set; Perform partition classification and action mapping on the quality partition set to generate routing rules.

6. The quantitative optical measurement method based on optical interferometry as described in claim 5, characterized in that: The steps involve performing continuous phase recovery and edge repair on different quality partitions according to routing rules to obtain continuous phase and consistency tags. The specific steps are as follows: Based on the quality partition map, the quality partition boundaries are extracted. According to the routing rules, the continuous phase of the package is restored by partition and edge repair is performed at the quality partition boundaries. The continuous phase of the partition is then output. Consistency fusion is performed on the continuous phases of the partition to generate continuous phases, and phase jump test and phase gradient continuity test are performed at the boundary of the quality partition based on the continuous phases to generate consistency markers.

7. The quantitative optical measurement method based on optical interferometry as described in claim 6, characterized in that: The specific steps for obtaining the phase change based on the continuous phase and the reference phase, and generating the optical path difference distribution based on the light source wavelength information, are as follows: Given that the continuous phase and the reference phase are at the same pixel coordinate position, the phase change is generated by subtracting the continuous phase and the reference phase pixel by pixel. The phase change is combined with the wavelength information of the light source and converted according to the conversion relationship from phase to optical path difference to generate the optical path difference distribution.

8. The quantitative optical measurement method based on optical interferometry as described in claim 7, characterized in that: The quantitative measurement results are obtained by performing the corresponding quantitative conversion rules according to the type of measurement object, and a reliability characterization is generated by the phase fitting residual and consistency label. The specific steps are as follows: Select the corresponding quantitative conversion rule according to the type of measurement object and perform quantitative conversion on the optical path difference distribution to generate quantitative measurement results; The residuals of the phase fitting are classified into residual levels and the consistency labels are classified into consistency levels. The residual levels and consistency levels are combined and mapped to generate a credibility characterization.

9. 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 quantitative optical measurement method based on optical interference as described in any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the quantitative optical measurement method based on optical interference as described in any one of claims 1 to 8.

Images