A lithium niobate optical crystal polarization structure intelligent detection method and system

Abstract

This invention discloses an intelligent detection method and system for the polarization structure of lithium niobate optical crystals, relating to the field of optical crystal structure detection. The method includes: The invention excites the nonlinear coefficient component along the optical axis of the lithium niobate crystal by focusing radially polarized light to generate a longitudinally matched frequency-doubled optical signal; the obtained longitudinally matched frequency-doubled optical signal is subjected to interferometric detection to obtain a polarization phase comparison map, and the reverse polarization region is segmented by grayscale jump detection; after obtaining the axial polarization period and duty cycle of each polarization interval, they are arranged into a sequence according to their spatial position and subjected to frequency domain transformation; based on the proportion of low-frequency energy to high-frequency energy and the characteristics of isolated spectral peaks in the periodic spectrum and duty cycle spectrum, the causes of chemical inhomogeneity are distinguished; polarization structure parameters are extracted and the sources of process deviations are attributed.

Description

Technical Field

[0001] This invention belongs to the field of optical crystal structure detection technology, and relates to an intelligent detection method and system for the polarization structure of lithium niobate optical crystals. Background Technology

[0002] Lithium niobate optical crystals are widely used in frequency conversion, electro-optic modulation, and quantum optics. Periodically polarized lithium niobate, as a core component for achieving quasi-phase matching, directly determines the frequency conversion efficiency and output beam quality through the period and duty cycle of its polarization structure.

[0003] To this end, patent application CN118090605A discloses a method and system for detecting the polarization structure of lithium niobate. By focusing radially polarized light in a three-dimensional imaging device onto the lithium niobate sample and applying voltage to electrodes distributed on the sample surface, the frequency-doubled light generated by the sample is obtained, resulting in an image with polarization structure. The location of the reverse polarization region, the axial polarization period, and the duty cycle are obtained, thus avoiding damage to the sample from chemical corrosion.

[0004] However, the above-mentioned existing technologies still have the following shortcomings in practical applications: First, the method relies on applying an external voltage to the sample to establish the refractive index difference between the positive and negative polarization regions. However, the voltage is applied by arranging electrodes on the crystal surface and energizing them. The attenuation of the electric field along the thickness direction in the crystal and the distortion of the electric field at the edge of the electrode will cause the refractive index difference between the positive and negative polarization regions at different depths and different lateral positions to be inconsistent. This results in uneven distribution of the contrast between bright and dark fringes at the domain walls in the field of view during frequency doubling imaging. The signal contrast in some areas is insufficient, making it difficult to identify the polarization boundary, thus affecting the extraction accuracy of the location of the negative polarization region.

[0005] Secondly, the distribution of intrinsic defects such as lithium vacancies and anti-niobium during crystal growth is not uniform, resulting in differences in the magnitude of refractive index change in different regions under the same voltage. This makes it difficult to determine the source of process deviation when the voltage-based detection method is used to deal with the chemical inhomogeneity of the material itself, as the repeatability of the imaging results deviates from the theoretical expectation. Summary of the Invention

[0006] In view of this, in order to solve the problems mentioned in the background art, the present invention provides a method and system for intelligent detection of polarization structure of lithium niobate optical crystal.

[0007] The objective of this invention can be achieved through the following technical solutions: In the first aspect, this invention provides an intelligent detection method for the polarization structure of a lithium niobate optical crystal, comprising: S1, determining whether the longitudinal electric field direction at the focal point of radially polarized light is parallel to the optical axis of the lithium niobate crystal; if they are parallel, focusing the beam onto the shallow layer of the sample to excite the maximum nonlinear coefficient component, thereby obtaining a longitudinally matched frequency-doubled light signal. S2. If mismatched, adjust the crystal orientation and optical path configuration until the matching conditions are met, and focus to excite and obtain a longitudinally matched frequency-doubled optical signal. S3. Perform interference detection on the longitudinally matched frequency-doubled optical signal to obtain a polarization phase comparison map. Perform grayscale jump detection and segmentation on the polarization phase comparison map to obtain the reverse polarization region. S4. Complete the positive polarization region based on the negative polarization region, calculate the center-to-center distance between adjacent positive polarization regions as the axial polarization period, and the ratio of the width of the negative region to the axial polarization period is the duty cycle. S5. Arrange the axial polarization period and duty cycle of each polarization region according to their spatial position, and determine the cause of chemical inhomogeneity based on the frequency domain characteristics of the sequence.

[0008] In a second aspect, the present invention provides an intelligent detection system for the polarization structure of lithium niobate optical crystals, comprising the following modules: a matching judgment module, used to determine whether the longitudinal electric field direction at the focal point of radially polarized light is parallel to the optical axis of the lithium niobate crystal; if matched, the beam is focused to the shallow layer of the sample to excite the maximum nonlinear coefficient component to obtain a longitudinally matched frequency-doubled light signal; if mismatched, the crystal orientation and optical path configuration are adjusted to meet the conditions, and the longitudinally matched frequency-doubled light signal is obtained by focusing and excitation. The interference detection module is used to perform interference detection on longitudinally matched frequency-doubled optical signals to obtain a polarization phase comparison map; The image segmentation module is used to segment the grayscale jump detection of the polarization phase contrast image to obtain the reverse polarization region; The parameter calculation module is used to complete the positive polarization region based on the negative polarization region, calculate the center distance between adjacent positive polarization regions as the axial polarization period, and the ratio of the width of the negative region to the axial polarization period as the duty cycle. The attribution judgment module is used to arrange the axial polarization period and duty cycle of each polarization interval according to spatial position, and judge the cause of chemical inhomogeneity based on the frequency domain characteristics of the sequence.

[0009] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) The present invention generates a longitudinal electric field after focusing radially polarized light and matches it parallel to the optical axis of the lithium niobate crystal, thereby exciting the nonlinear coefficient component of the crystal along the optical axis and obtaining a longitudinally matched frequency-doubled light signal. In principle, this invention eliminates the problem of uneven distribution of refractive index difference caused by the attenuation of electric field along the thickness direction in the crystal and the distortion of electric field at the electrode edge due to the application of voltage. This ensures that the signal contrast is uniform and the domain wall boundary is clearly distinguishable in the entire detection field, thereby improving the extraction accuracy of the reverse polarization region position.

[0010] (2) The present invention performs interference detection on the obtained longitudinal matching frequency-doubled optical signal to obtain a polarization phase comparison map, and obtains the reverse polarization region by gray-scale jump detection. It does not rely on the refractive index change caused by electro-optic effect, and the imaging result is not affected by the local stoichiometry of the crystal and the distribution of intrinsic defects. It overcomes the problem of repeated deviation of imaging results caused by the chemical inhomogeneity of the material due to the applied voltage method, and ensures high repeatability of polarization structure detection.

[0011] (3) After obtaining the axial polarization period and duty cycle of each polarization interval, the present invention arranges them into a sequence according to their spatial position and performs frequency domain transformation. Based on the proportion of low-frequency energy and high-frequency energy in the period spectrum and duty cycle spectrum and the characteristics of isolated peaks, the causes of chemical inhomogeneity are distinguished. This overcomes the shortcomings of the voltage detection method in quantitatively tracing the source of process deviation, and provides a basis for the polarization process deviation of lithium niobate crystal. Attached Figure Description

[0012] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. 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.

[0013] Figure 1 This is a diagram illustrating the implementation steps of the method of the present invention; Figure 2 This is a flowchart of the process for obtaining the reverse polarization region in this invention; Figure 3 This is a schematic diagram showing the connections of the various modules in the system of the present invention. Detailed Implementation

[0014] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0015] Please see Figure 1 As shown, the first aspect of the present invention provides a method for intelligent detection of the polarization structure of a lithium niobate optical crystal, comprising: S1, determining whether the longitudinal electric field direction at the focal point of the radially polarized light is parallel to the optical axis of the lithium niobate crystal; if they are parallel, focusing the light beam onto the shallow layer of the sample to excite the maximum nonlinear coefficient component, thereby obtaining a longitudinally matched frequency-doubled light signal.

[0016] Because radially polarized light is tightly focused by a high numerical aperture objective lens, a longitudinal electric field component that mainly oscillates along the optical axis of the objective lens is generated at the focal point. Lithium niobate crystal has anisotropic second-order nonlinear optical coefficients, among which the nonlinear coefficient component along the optical axis of the crystal is the largest. When the longitudinal electric field direction at the focal point is parallel to the optical axis of the crystal, the intensity of the excited frequency-doubled light signal reaches its maximum.

[0017] Based on this, the longitudinal electric field direction and the crystal optical axis direction are obtained as follows: The linearly polarized beam emitted from the light source is introduced into a radial polarization converter to obtain radially polarized light, which is then introduced into a high numerical aperture objective lens after passing through a beam expander and collimating lens group. The direction of the objective lens optical axis is taken as the longitudinal electric field direction at the focal point, and the objective lens mount is adjusted to make this direction perpendicular to the stage surface. A lithium niobate crystal sample with a known cut is selected, and the sample is fixed to the stage through a crystal reference edge or orientation clamp, so that the projection of the crystal optical axis in the stage plane coincides with the preset mechanical marking line on the stage, and the direction of this marking line represents the direction of the crystal optical axis.

[0018] The above method for determining whether the longitudinal electric field direction at the focal point of radially polarized light is parallel to the optical axis of the lithium niobate crystal includes: a radial marking line passing through the center of rotation is engraved on the surface of the stage, and the sample is placed on the stage so that the optical axis direction is aligned with the marking line of the stage.

[0019] Specifically, a radial marking line passing through the center of rotation is etched on the surface of the stage. A lithium niobate crystal sample of known cut is taken. If a reference edge is machined on one side of the sample and the reference edge is parallel to the projection direction of the crystal optical axis in the wafer plane, the reference edge is aligned with the marking line so that the optical axis direction coincides with the marking line direction. When the sample does not have a reference edge, the sample is placed under crossed polarized light and rotated to the extinction position. At this time, the crystal optical axis direction is perpendicular to the polarization direction of the crossed polarized light. The sample is then rotated 90 degrees so that the optical axis direction is consistent with the marking line direction. The sample is then fixed to complete the alignment of the optical axis with the marking line.

[0020] Move the objective lens along the optical axis to shrink the reflected light spot to its minimum, then position it on the surface and adjust it further into the crystal to bring the focal point into the shallow layer.

[0021] During operation, a photodetector is set at the back focal plane of the objective lens to monitor the reflected light spot on the sample surface in real time. The objective lens is driven to approach the sample along the optical axis by a stepper motor. The step size is initially set to 2 to 5 times the depth of focus of the objective lens. When the diameter of the reflected light spot decreases, the step size is switched to 0.3 to 0.5 times the depth of focus of the objective lens. After each step, the image of the current reflected light spot is acquired and its diameter is calculated. When the light spot diameter changes from decreasing to starting to increase, the light spot diameter of the previous feed position is determined to be the minimum value. This position corresponds to the objective lens focal point coinciding with the sample surface. The axial position of the objective lens at this time is recorded as the surface reference zero point. Then, based on the surface reference zero point, the objective lens is driven further into the crystal. The feed amount is calculated as the product of the depth of focus (1 to 5 μm below the sample surface) and the refractive index of the crystal, which is then converted into the objective lens displacement, so that the focal point enters the shallow layer of the sample.

[0022] The intensity of the frequency-doubled light signal at the focal point is acquired. With the focal point fixed, the stage is rotated to change the angle between the crystal optical axis and the polarization direction of the radially polarized light. When the intensity of the frequency-doubled light signal reaches a maximum, it is determined to be a parallel match, and the stage scale position at this point is recorded as the match position. During continuous stage rotation, the stage is rotated in increments no greater than 1 degree, and the corresponding frequency-doubled light signal intensity is recorded. The maximum value is then compared and taken as the maximum value position. The aforementioned fixed focal point refers to keeping the objective lens axial position unchanged.

[0023] The principle of determining parallel matching by using signal maxima is as follows: After radially polarized light is focused, the longitudinal electric field direction at the focal point is unique; the direction of the maximum nonlinear coefficient of the lithium niobate crystal is unique along the optical axis. When the two are parallel, the electric field vector and the direction of the maximum nonlinear coefficient completely coincide, resulting in the strongest frequency-doubled signal; rotating the stage changes the angle between them, and the signal weakens accordingly. Therefore, the location of the signal maxima is the parallel matching position.

[0024] The process of obtaining the longitudinally matched frequency-doubled optical signal includes: keeping the stage fixed at the matching position, performing a depth scan along the optical axis into the crystal, and acquiring the intensity of the frequency-doubled optical signal.

[0025] The depth scan starts from the surface reference zero point and gradually drives the objective lens into the crystal along the optical axis. After each step, the intensity of the frequency-doubled light signal is acquired, and the axial displacement of the objective lens and the corresponding signal intensity value are recorded. The scanning depth range covers the shallow layer of the sample to the depth region where the polarization structure is located.

[0026] When the intensity of the frequency-doubled optical signal first increases and then decreases as the objective lens moves, and the intensity reaches its peak range during the transition from rising to falling, it is determined that the beam focus has entered the depth position for exciting the maximum nonlinear coefficient component.

[0027] When the objective lens focus is not yet in the polarization region with the maximum nonlinear coefficient component, the frequency-harmonic signal is weak. As the focus gradually approaches and enters the region, the overlap between the longitudinal electric field and the polarization direction within the crystal gradually increases, the effective nonlinear coefficient increases, and the frequency-harmonic signal intensity rises accordingly. When the focus is completely near the center of the polarization region, the signal reaches its maximum. As the focus continues to move and begins to leave the polarization region, the effective nonlinear coefficient decreases, and the signal begins to decline. Therefore, the transition region where the signal increases and decreases marks the point where the beam focus has entered the depth position for exciting the maximum nonlinear coefficient component.

[0028] The above-mentioned upward trend determination is based on continuously acquiring signal strength sequences along the depth direction. If the intensity values ​​of the current sampling point and at least two subsequent consecutive sampling points are successively greater than or equal to the previous sampling point, and the intensity value of the last sampling point in the sequence is strictly greater than the intensity value of the first sampling point in the sequence, then the interval is determined to be an upward trend. The downward trend determination is based on the following: If the intensity values ​​of the current sampling point and at least two subsequent consecutive sampling points are successively less than or equal to the previous sampling point, and the intensity value of the last sampling point in the sequence is strictly less than the intensity value of the first sampling point in the sequence, then the interval is determined to be a downward trend.

[0029] When the signal strength no longer meets the conditions for an upward trend but has not entered a downward trend, it is determined to be a plateau zone. When the signal transitions directly from an upward trend to a plateau zone or a downward trend, the turning point from rising to falling is where the peak range is located.

[0030] The peak interval refers to the depth range where the frequency-doubled optical signal value transitions from rising to falling, and the signal strength is at its maximum value within the transition section. It is determined by taking the depth range within the transition section where the signal strength is not less than 90% of the maximum signal strength within the transition section as the peak interval.

[0031] The frequency-doubled optical signal at the depth position is output as the longitudinally matched frequency-doubled optical signal.

[0032] If no intensity peak is detected within the depth scan range, the detection is terminated.

[0033] If no signal intensity peak is detected within the depth scan range, it indicates that the beam focus has not entered the region that can excite the largest nonlinear coefficient component throughout the entire scan, meaning that there is no longitudinally matched frequency doubling signal source under the current detection conditions, so the detection is terminated.

[0034] S2. If mismatched, adjust the crystal orientation and optical path configuration until the matching conditions are met, and focus to excite and obtain a longitudinally matched frequency-doubled optical signal.

[0035] Since the longitudinal electric field vector and the optical axis of the crystal are at an angle in the mismatched state, only a portion of the nonlinear coefficient components are involved in the excitation. Therefore, it is necessary to adjust the electric field direction to coincide with the optical axis direction in order to excite the maximum nonlinear coefficient component.

[0036] Based on this, if there is a mismatch, the crystal orientation and optical path configuration are adjusted until the matching conditions are met, including rotating the stage one revolution. If the intensity of the frequency-doubled optical signal does not reach a maximum value, it is determined that the mismatch is not met.

[0037] After focusing, the radially polarized light has a unique longitudinal electric field direction, and the maximum nonlinear coefficient of the lithium niobate crystal has a unique direction along the optical axis. When the two are parallel, the frequency doubling signal is the strongest and a maximum value should appear. There is no maximum value after rotating one revolution, which indicates that there is always an angular deviation between the two directions. The effective component of the longitudinal electric field in the optical axis direction has not reached the maximum, so it is determined that the matching is not satisfied.

[0038] If the curve of the frequency-doubled optical signal intensity changing with the rotation angle has a bulge, adjust the pitch of the stage along the direction corresponding to the bulge so that the signal at the bulge is amplified to the maximum value.

[0039] The protrusion is determined as follows: in the curve of the frequency-doubled optical signal intensity changing with the rotation angle, the first-order difference of the signal intensity between two adjacent points is calculated. If there are at least two consecutive sampling points whose difference value changes from positive to negative, and the average signal intensity in this interval is higher than 1.2 times the average signal intensity in the entire rotation period, then it is determined that there is a protrusion at that point.

[0040] The procedure for adjusting the stage pitch along the direction corresponding to the protrusion is as follows: record the rotation angle position of the stage corresponding to the protrusion and keep the angle unchanged. If the signal strength increases, continue to adjust along the pitch axis of the stage. If the signal strength decreases, adjust in the opposite direction until the signal strength no longer increases. This is the maximum value at the protrusion.

[0041] If there is no spike in signal intensity across the entire circumference of the curve, then alternately change the fast axis angle of the waveplate and rotate it again for comparison until a spike appears. If no spike appears after multiple consecutive changes, then terminate the detection.

[0042] The alternating change of the fast axis angle of the waveplate specifically includes: placing a rotatable half-wave plate between the light source and the radial polarization converter; alternating the fast axis angle of the waveplate means switching the fast axis of the half-wave plate in 45-degree steps between at least two different angular positions, and repeating the operation of rotating the stage one full revolution at each angular position while recording the intensity of the frequency-doubled light signal. The two different angular positions include 0 degrees and 45 degrees. If no signal spike is detected at all angular positions, it is determined that the optical path cannot be matched under the current detection conditions, and the detection is terminated.

[0043] The "multiple consecutive attempts" refers to the maximum number of attempts to change the fast axis angle of the waveplate, such as 3 to 5 times.

[0044] After a signal spike appears, adjust the pitch of the stage along the spike direction until the signal reaches its maximum value, indicating that a match has been achieved.

[0045] The specific operation involves recording the rotation angle position of the stage corresponding to the protrusion, keeping this rotation angle fixed, and gradually adjusting the pitch angle in a single direction along the pitch axis of the stage with a preset angular step size. The intensity of the frequency-doubled optical signal is collected after each adjustment step. If the signal intensity increases compared to before the adjustment, the adjustment continues in that direction; if the signal intensity decreases, the adjustment reverses. When the signal intensity no longer increases and the current value is the maximum value observed during the adjustment in that direction, this value is determined to be the signal maximum. The preset angular step size is the pitch angle corresponding to a single step of the stepper motor.

[0046] S3. Perform interference detection on the longitudinally matched frequency-doubled optical signal to obtain a polarization phase comparison map. Perform grayscale jump detection and segmentation on the polarization phase comparison map to obtain the reverse polarization region.

[0047] Considering that the frequency-doubled light signals in the forward and reverse polarization regions have the same intensity but opposite phase, they cannot be distinguished by intensity alone. Therefore, obtaining the polarization phase comparison diagram includes: dividing the longitudinally matched frequency-doubled light signal into a probe light and a reference light, focusing the probe light onto the sample surface, and introducing the reference light into an adjustable optical path delay line; specifically, the longitudinally matched frequency-doubled light signal is introduced into a beam splitter and divided into a transmitted probe light and a reflected reference light.

[0048] After the probe light is reflected by the sample, it interferes with the reference light at the beam combiner, forming an interference light field. The interference light field is received, and the focus of the probe light is scanned point by point along the crystal surface.

[0049] By changing the optical path length of the adjustable optical path delay line at each scanning point, multiple interference images with different optical path differences are acquired.

[0050] The phase offset of the same scanning point in multiple interferometric images is extracted, arranged according to spatial position and mapped to grayscale values ​​to form a polarization phase contrast map.

[0051] The phase offsets of each scanning point are arranged according to their actual spatial coordinates to form a two-dimensional phase distribution matrix. The two-dimensional phase distribution matrix is ​​normalized and linearly mapped to a grayscale range of 0 to 255. The phase offset of each point corresponds to a unique grayscale value, thus obtaining a polarization phase contrast map presented in the form of a grayscale image.

[0052] The steps for extracting the phase shift are as follows: For each scanning point, the adjustable optical path delay line is controlled to perform optical path difference modulation in steps of π / 2, and an interference image is acquired after each modulation, for a total of 4 interference images, to obtain the light intensity sequence I1, I2, I3, I4 for that pixel; a four-step phase shift algorithm is used, through the formula... The wrapped phase value is calculated, and the wrapped phase value is unwrapped. The result is the phase offset of the scanning point. Wherein, I1 is the light intensity value of the detector when 0 phase difference is introduced between the two beams, I2 is the light intensity value when π / 2 phase difference is introduced, I3 is the light intensity value when π phase difference is introduced, and I4 is the light intensity value when 3π / 2 phase difference is introduced; It is proportional to the sine value of the measured phase offset φ, and the denominator term is... It is proportional to the cosine value of φ.

[0053] refer to Figure 2 As shown, obtaining the reverse polarization region includes: performing a two-dimensional Fourier transform on the polarization phase contrast map, identifying the direction of the spectral peak, and drawing grayscale sampling lines along the direction perpendicular to the spectral peak. The steps for identifying the peak direction of the spectrum are as follows: Perform a two-dimensional discrete Fourier transform on the polarization phase comparison map to obtain a complex spectrum matrix of the same size as the image; calculate the amplitude of each element in the complex spectrum matrix to generate an amplitude spectrum map; establish a polar coordinate system (ρ, θ) with the center of the amplitude spectrum map as the origin, where ρ is the polar radius and θ is the polar angle; within the range of polar angle θ [0°, 180°), integrate the amplitude along the ray in the direction of each angle θ with a step size of Δθ=1° to obtain the directional energy; the angle corresponding to the maximum value of the directional energy is the peak direction of the spectrum.

[0054] The grayscale value is recorded pixel by pixel along the grayscale sampling line. Points where the grayscale value changes from bright to dark or from dark to bright are identified and marked as boundary candidate points.

[0055] The method for determining whether a grayscale value changes from bright to dark or from dark to bright is as follows: The grayscale value difference between adjacent pixels is calculated pixel by pixel along the grayscale sampling line. When the absolute value of the grayscale value difference between adjacent pixels exceeds the transition threshold, a grayscale transition is determined to have occurred. If the grayscale value changes from high to low, it is a change from bright to dark; if the grayscale value changes from low to high, it is a change from dark to bright. The transition threshold is 2 to 3 times the standard deviation of the grayscale values ​​on the sampling line.

[0056] A continuous pixel region between adjacent boundary candidate points with a gray value lower than the gray value of the surrounding background is marked as a dark line region, and a gray value higher than the gray value of the surrounding background is marked as a bright line region. For a continuous pixel region between adjacent boundary candidate points, the gray value of the surrounding background is referenced by its median gray value. If the pixel gray value is lower than the median, it is a dark line region, and if it is higher than the median, it is a bright line region.

[0057] Using the central skeleton line of the dark line region as the boundary contour line, the area between two adjacent boundary contour lines is determined as the reverse polarization region, and other areas are recorded as undetermined areas.

[0058] In the polarization phase contrast map, the dark line region appears as a continuous pixel strip with a gray value lower than the surrounding background gray value between adjacent boundary candidate points, and its central skeleton line is the spatial extension path of the low gray value strip. Since the frequency harmonic signals of the reverse polarization region and the forward polarization region are out of phase, they correspond to low gray value region and high gray value region respectively in the polarization phase contrast map. Therefore, the area surrounded or defined by the dark line region physically corresponds to the reverse polarization region.

[0059] The central skeleton line of the dark line region is extracted through morphological thinning. Specifically, the structuring element is repeatedly applied to the binary image of the dark line region for erosion operation until a center line with a single pixel width is obtained, which is the boundary contour line.

[0060] S4. Complete the positive polarization region based on the negative polarization region, calculate the center distance between adjacent positive polarization regions as the axial polarization period, and the ratio of the width of the negative region to the axial polarization period is the duty cycle.

[0061] The steps for completing the positive polarization region based on the reverse polarization region are as follows: pixels in the region to be determined whose gray values ​​are between those of the dark and bright lines are denoted as the transition pixel zone.

[0062] The pixel grayscale values ​​between the dark line area and the bright line area are between the two, and their polarization is not yet clear. Directly classifying them into the dark line or bright line area will introduce region division errors. Marking them separately as transition pixel bands can ensure the accuracy of positive polarization region completion.

[0063] Within the undefined region, the bright line area and the transition pixel band are merged to form a positive polarization region.

[0064] In the polarization phase contrast map, the bright line area appears as a continuous pixel area with a gray value higher than the background, forming a gray value contrast with the already determined dark line area, indicating that it physically corresponds to a region of another polarization state; the transition pixel band has a gray value between the bright line and the dark line, and is spatially adjacent to the bright line area, belonging to the transition area from the positive polarization region to the negative polarization region. Therefore, the two together constitute a complete positive polarization region.

[0065] The alternating arrangement of each positive and negative polarization region is checked along the direction perpendicular to the peak of the spectrum. If an abnormal arrangement of adjacent regions of the same type is found, the region whose gray value is closer to the gray value of the confirmed positive polarization region is retained as a positive polarization region, and the other is corrected as a negative polarization region, until all regions conform to the alternating positive and negative polarization arrangement, and the number of corrections does not exceed the maximum number of iterations.

[0066] The periodic polarization structure of lithium niobate crystals exhibits a strict alternation of positive and negative polarization regions in space, making it impossible for adjacent polarization regions to be of the same type. If adjacent regions of the same type appear in the preliminary determination, it indicates that at least one polarization region in that adjacent region has been incorrectly identified and needs to be corrected. Specifically, the average grayscale value of all pixels within the confirmed positive polarization region is used as the positive reference grayscale value. The closer the grayscale value is to the positive reference grayscale value, the more consistent the phase of the frequency doubling signal in that region is with the positive polarization region, and therefore it should be classified as a positive polarization region. Another region with a significantly deviated grayscale value indicates that its phase is opposite to that of the positive polarization region, and therefore it is corrected to be a negative polarization region.

[0067] The maximum number of iterations is determined as follows: based on the number of marked boundary contour lines in the polarization phase comparison diagram, the maximum number of iterations is set to the number of boundary contour lines. This ensures that each correction eliminates at least one pair of adjacent aberrations of the same type, and the number of aberrations does not exceed the total number of boundary contour lines. Therefore, using the number of boundary contour lines as the upper limit of the number of iterations ensures that all possible aberrations have a chance to be corrected, while avoiding infinite loops.

[0068] The calculation steps for the center spacing between adjacent positively polarized regions in step S4 above are as follows: For the completed and corrected polarization region arrangement sequence, locate the center position of each positively polarized region one by one along the direction perpendicular to the peak of the spectrum. The center position is taken as the midpoint between the two boundary lines of the positively polarized region along the direction perpendicular to the peak of the spectrum. Calculate the spatial distance between the center positions of two adjacent positively polarized regions in turn, which is the axial polarization period of the pair of adjacent positively polarized regions.

[0069] In step S4 above, the width of the reverse region is the spatial distance between the two boundary lines of the reverse polarization region along the same direction.

[0070] S5. Arrange the axial polarization period and duty cycle of each polarization region according to their spatial position, and determine the cause of chemical inhomogeneity based on the frequency domain characteristics of the sequence.

[0071] Considering that polarization distortion caused by chemical inhomogeneity usually exhibits a regular spatial distribution, such as gradual change or periodic fluctuation along a certain direction, the axial polarization period and duty cycle of each polarization region are arranged according to spatial position as follows: a main measurement axis is set on the sample surface along the light propagation direction, and the sample is divided into sequentially connected polarization regions along the main measurement axis.

[0072] Specifically, after setting the main measurement axis along the light propagation direction on the sample surface, the positive and negative polarization regions defined by the boundary contour line along the main measurement axis are divided sequentially with reference to the main measurement axis. Each group of adjacent positive and negative polarization regions is taken as a polarization interval. The polarization intervals are arranged sequentially from end to end along the main measurement axis until the entire area to be measured is covered.

[0073] The average value of all measured axial polarization periods within each polarization interval is taken as the representative axial polarization period, and the average value of the ratio of all reverse region widths to the corresponding axial polarization periods within each polarization interval is taken as the representative duty cycle.

[0074] Within a single polarization range, due to local process fluctuations or detection noise, there are slight deviations in the measured values ​​of axial polarization period and duty cycle. Taking the average value of all measurements as the representative value can suppress random errors.

[0075] The representative axial polarization period and representative duty cycle corresponding to the center position coordinates of each polarization interval on the main measurement axis are arranged in spatial order to form a period sequence and a duty cycle sequence.

[0076] In step S5 above, determining the cause of chemical inhomogeneity includes performing Fourier transforms on the periodic sequence and the duty cycle sequence respectively to obtain the periodic spectrum and the duty cycle spectrum.

[0077] The steps for obtaining the periodic spectrum and duty cycle spectrum are as follows: First, the axial polarization periodic values ​​arranged in spatial order are taken as the periodic sequence, and the corresponding duty cycle values ​​are taken as the duty cycle sequence, and one-dimensional discrete data sequences are constructed respectively. Then, zero-padding is performed on the periodic sequence and the duty cycle sequence to make the sequence length meet the sampling point requirement of the discrete Fourier transform. Next, one-dimensional discrete Fourier transform is performed on the processed periodic sequence and the duty cycle sequence respectively to obtain their respective complex spectrum sequences. Finally, the amplitude of each element in each complex spectrum sequence is taken to generate the periodic spectrum and the duty cycle spectrum.

[0078] The low-frequency energy ratio and high-frequency energy ratio of the periodic spectrum and duty cycle spectrum are extracted respectively. When the low-frequency energy ratio of the periodic spectrum exceeds the first frequency domain threshold and the high-frequency energy ratio is lower than the second frequency domain threshold, and there are no isolated spectral peaks in either spectrum, it is determined that the energy attenuation is caused by energy decay in the direction of the polarization electric field.

[0079] The energy decay along the polarization electric field direction manifests as a gradual weakening of the electric field intensity along the light propagation direction, resulting in a monotonic gradual change in the polarization period and duty cycle in space, without any local abrupt changes. This gradual change characteristic mainly corresponds to the low-frequency components in the frequency domain, manifested as concentrated energy in the low-frequency bands of the periodic spectrum and duty cycle spectrum, and lower energy in the two high-frequency bands; at the same time, since the decay process is continuous and smooth, without generating periodic disturbances at specific frequencies, no isolated spectral peaks appear in the spectrum.

[0080] The low-frequency band refers to the frequency range from zero to the first 20% of the total spectrum bandwidth, while the high-frequency band refers to the frequency range excluding the low-frequency band and zero-frequency components from the total spectrum bandwidth. The low-frequency band energy proportion refers to the proportion of the sum of the squares of the amplitudes of each frequency component in the low-frequency band to the sum of the squares of the amplitudes of the entire frequency band, while the high-frequency band energy proportion refers to the proportion of the sum of the squares of the amplitudes of each frequency component in the high-frequency band to the sum of the squares of the amplitudes of the entire frequency band.

[0081] When the energy proportion of the low-frequency band in the periodic spectrum exceeds the first frequency domain threshold, the energy proportion of the high-frequency band exceeds the second frequency domain threshold, and the energy proportion of the low-frequency band in the duty cycle spectrum exceeds the third threshold, it is determined that the polarization electrode edge effect is the cause.

[0082] The edge effect of polarization electrodes causes local distortion of the electric field distribution at the electrode edge, resulting in the following spatial variations in the polarization period and duty cycle: First, the edge effect has a large influence range, generating low-frequency gradual components along the light propagation direction; second, the concentration or non-uniformity of the electric field at the edge introduces more intense periodic or duty cycle fluctuations locally, generating high-frequency components. Therefore, the energy in both the low-frequency and high-frequency bands of the periodic spectrum is high. Simultaneously, the duty cycle is more sensitive to changes in the edge electric field, and the duty cycle fluctuations caused by the edge effect spatially exhibit a low-frequency-dominated deviation, thus correspondingly increasing the proportion of low-frequency energy in the duty cycle spectrum.

[0083] For multiple similar lithium niobate crystal samples, the detection process from step S1 to step S5 is performed respectively. The distribution range of the low-frequency energy ratio and high-frequency energy ratio of the periodic spectrum of each sample under normal process conditions is statistically analyzed. The lower limit of the distribution range of the low-frequency energy ratio is taken as the first frequency domain threshold, and the upper limit of the distribution range of the high-frequency energy ratio is taken as the second frequency domain threshold.

[0084] The distribution range of the energy proportion in the low-frequency band of the duty cycle spectrum of each sample under normal process conditions was statistically analyzed, and the upper limit of this distribution range was taken as the third threshold.

[0085] When there are discrete, isolated spectral peaks in any spectrum, it is determined that the peaks are caused by native defects in the crystal.

[0086] Because the native defects in a crystal are distributed in a dispersed and isolated manner, they can introduce aperiodic abrupt changes in the polarization period or duty cycle sequence. In the spectrum, this manifests as discrete, isolated spectral peaks with amplitudes higher than the adjacent baseline level at a specific frequency point or within a narrow frequency range. If the amplitude at a certain frequency point is higher than the amplitudes of its left and right adjacent frequency points, and the amplitude exceeds three times the average amplitude of the adjacent frequency band, then a discrete, isolated spectral peak is determined to exist at that location.

[0087] refer to Figure 3 As shown, a second aspect of the present invention provides an intelligent detection system for the polarization structure of a lithium niobate optical crystal, comprising a matching judgment module, an interference detection module, an image segmentation module, a parameter calculation module, and an attribution judgment module. All modules are connected in the order described above.

[0088] Specifically, the matching judgment module is used to determine whether the longitudinal electric field direction at the focal point of the radially polarized light is parallel to the optical axis of the lithium niobate crystal. If they match, the beam is focused to the shallow layer of the sample to excite the maximum nonlinear coefficient component and obtain a longitudinally matched frequency-doubled light signal. If they do not match, the crystal orientation and optical path configuration are adjusted to meet the conditions and the beam is focused to excite and obtain a longitudinally matched frequency-doubled light signal. The interference detection module is used to perform interference detection on longitudinally matched frequency-doubled optical signals to obtain a polarization phase comparison map; The image segmentation module is used to segment the grayscale jump detection of the polarization phase contrast image to obtain the reverse polarization region; The parameter calculation module is used to complete the positive polarization region based on the negative polarization region, calculate the center distance between adjacent positive polarization regions as the axial polarization period, and the ratio of the width of the negative region to the axial polarization period as the duty cycle. The attribution judgment module is used to arrange the axial polarization period and duty cycle of each polarization interval according to spatial position, and judge the cause of chemical inhomogeneity based on the frequency domain characteristics of the sequence.

[0089] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0090] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0091] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0092] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0093] Finally, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for intelligent detection of the polarization structure of a lithium niobate optical crystal, characterized in that: include: Determine whether the longitudinal electric field direction at the focal point of the radially polarized light is parallel to the optical axis of the lithium niobate crystal. If they are parallel, focus the beam to excite the maximum nonlinear coefficient component in the shallow layer of the sample to obtain the longitudinally matched frequency-doubled light signal. If they do not match, adjust the crystal orientation and optical path configuration until the matching conditions are met, and focus the excitation to obtain a longitudinally matched frequency-doubled optical signal; Interference detection is performed on the longitudinally matched frequency-doubled optical signal to obtain a polarization phase comparison map. Gray-scale jump detection is performed on the polarization phase comparison map to segment it and obtain the reverse polarization region. The positive polarization region is completed based on the negative polarization region, and the center-to-center distance between adjacent positive polarization regions is calculated as the axial polarization period. The ratio of the width of the negative region to the axial polarization period is the duty cycle. Arrange the axial polarization period and duty cycle of each polarization region according to their spatial position, and determine the cause of chemical inhomogeneity based on the frequency domain characteristics of the sequence. The obtained polarization phase contrast map includes: The longitudinally matched frequency-doubled optical signal is divided into a probe light and a reference light. The probe light is focused onto the sample surface, and the reference light is introduced into an adjustable optical path delay line. After the probe light is reflected by the sample, it interferes with the reference light at the beam combiner, forming an interference light field. The interference light field is received so that the focus of the probe light scans point by point along the crystal surface. By changing the optical path length of the adjustable optical path delay line at each scanning point, multiple interference images with different optical path differences are acquired. The phase shift at the same scanning point in multiple interferometric images is extracted, arranged according to spatial position, and mapped to gray values ​​to form a polarization phase contrast map. The obtained reverse polarization region includes: Perform a two-dimensional Fourier transform on the polarization phase contrast map to identify the direction of the spectral peaks, and draw grayscale sampling lines along the direction perpendicular to the spectral peaks. Record gray values ​​pixel by pixel along the gray value sampling line, identify points where gray values ​​change from bright to dark or from dark to bright, and mark them as boundary candidate points; The continuous pixel region between adjacent boundary candidate points with a gray value lower than the gray value of the surrounding background is marked as a dark line region, and the gray value with a gray value higher than the gray value of the surrounding background is marked as a bright line region. Using the central skeleton line of the dark line region as the boundary contour line, the area between two adjacent boundary contour lines is determined as the reverse polarization region, and other areas are recorded as undetermined areas. The step of completing the positively polarized region based on the reverse polarized region includes: Pixels in the region to be determined whose gray values ​​are between those of the dark and bright lines are denoted as the transition pixel zone. Within the undefined region, the bright line area and the transition pixel band are merged to form a positive polarization region; The alternating arrangement of each positive and negative polarization region is checked along the direction perpendicular to the peak of the spectrum. If an abnormal arrangement of adjacent regions of the same type is found, the region whose gray value is closer to the gray value of the confirmed positive polarization region is retained as a positive polarization region, and the other is corrected as a negative polarization region, until all regions conform to the alternating positive and negative polarization arrangement, and the number of corrections does not exceed the maximum number of iterations. The arrangement of the axial polarization period and duty cycle of each polarization region according to spatial position includes: A main measurement axis is set on the sample surface along the direction of light propagation, and the sample is divided into sequentially connected polarization regions along the main measurement axis; The average value of all measured axial polarization periods in each polarization interval is taken as the representative axial polarization period, and the average value of the ratio of all reverse region widths to the corresponding axial polarization periods in each polarization interval is taken as the representative duty cycle. The representative axial polarization period and representative duty cycle corresponding to the center position coordinates of each polarization interval on the main measurement axis are arranged in spatial order to form a period sequence and a duty cycle sequence. The causes of chemical inhomogeneity include: Perform Fourier transforms on the periodic sequence and the duty cycle sequence respectively to obtain the periodic spectrum and the duty cycle spectrum; The low-frequency energy ratio and high-frequency energy ratio of the periodic spectrum and duty cycle spectrum are extracted respectively. When the low-frequency energy ratio of the periodic spectrum exceeds the first frequency domain threshold and the high-frequency energy ratio is lower than the second frequency domain threshold, and there are no isolated spectral peaks in either spectrum, it is determined that the energy attenuation is caused by the polarization electric field direction. When the energy proportion of the low-frequency band in the periodic spectrum exceeds the first frequency domain threshold, the energy proportion of the high-frequency band exceeds the second frequency domain threshold, and the energy proportion of the low-frequency band in the duty cycle spectrum exceeds the third threshold, it is determined that it is caused by the polarization electrode edge effect. When there are discrete, isolated spectral peaks in any spectrum, it is determined that the peaks are caused by native defects in the crystal.

2. The intelligent detection method for the polarization structure of a lithium niobate optical crystal according to claim 1, characterized in that: The determination of whether the longitudinal electric field direction at the focal point of radially polarized light is parallel to the optical axis of the lithium niobate crystal includes: A radial marking line passing through the center of rotation is engraved on the surface of the stage. The sample is placed on the stage so that the optical axis is aligned with the marking line of the stage. Move the objective lens along the optical axis to shrink the reflected spot to its minimum, then position it on the surface and adjust it further into the crystal to bring the focal point into the shallow layer. The intensity of the frequency-doubled light signal at the focal point is collected. The focal point is fixed, and the stage is rotated to change the angle between the crystal optical axis and the polarization direction of the radially polarized light. When the intensity of the frequency-doubled light signal reaches a maximum value, it is determined to be parallel matching. The scale position of the stage at this time is recorded as the matching position.

3. The intelligent detection method for the polarization structure of a lithium niobate optical crystal according to claim 1, characterized in that: The obtained longitudinally matched frequency-doubled optical signal includes: Keep the stage fixed at the matching position, perform a depth scan into the crystal along the optical axis, and collect the intensity of the frequency-doubled light signal; When the intensity of the frequency-doubled optical signal first increases and then decreases as the objective lens moves, and the intensity reaches its peak range between the increase and decrease, it is determined that the beam focus has entered the depth position for exciting the maximum nonlinear coefficient component. The frequency-doubled optical signal at the depth position is output as the longitudinally matched frequency-doubled optical signal; If no intensity peak is detected within the depth scan range, the detection is terminated.

4. The intelligent detection method for the polarization structure of a lithium niobate optical crystal according to claim 1, characterized in that: If a mismatch is found, the crystal orientation and optical path configuration are adjusted until the matching conditions are met, including: If the intensity of the frequency-doubled optical signal does not reach a maximum value after rotating the stage one revolution, it is determined that the matching is not satisfied. If the curve of frequency-doubled optical signal intensity changing with rotation angle has a bulge, adjust the pitch of the stage along the direction corresponding to the bulge so that the signal at the bulge is amplified to the maximum value; If there is no spike in signal intensity across the entire circumference of the change curve, then alternately change the fast axis angle of the waveplate and rotate it again for comparison until a spike appears. If there is still no spike after multiple changes, then terminate the detection. After a signal spike appears, adjust the pitch of the stage along the spike direction until the signal reaches its maximum value, indicating that a match has been achieved.

5. A method for intelligent detection of polarization structure of lithium niobate optical crystal according to any one of claims 1-4, wherein the method employs an intelligent detection system for polarization structure of lithium niobate optical crystal to complete the relevant steps, characterized in that: The intelligent detection system for the polarization structure of lithium niobate optical crystals includes: The matching judgment module is used to determine whether the longitudinal electric field direction at the focal point of the radially polarized light is parallel to the optical axis of the lithium niobate crystal. If they match, the beam is focused to the shallow layer of the sample to excite the maximum nonlinear coefficient component and obtain the longitudinal matching frequency-doubled light signal. If they do not match, the crystal orientation and optical path configuration are adjusted to meet the conditions and the longitudinal matching frequency-doubled light signal is obtained by focusing and excitation. The interference detection module is used to perform interference detection on longitudinally matched frequency-doubled optical signals to obtain a polarization phase comparison map; The image segmentation module is used to segment the grayscale jump detection of the polarization phase contrast image to obtain the reverse polarization region; The parameter calculation module is used to complete the positive polarization region based on the negative polarization region, calculate the center distance between adjacent positive polarization regions as the axial polarization period, and the ratio of the width of the negative region to the axial polarization period as the duty cycle; the attribution judgment module is used to arrange the axial polarization period and duty cycle of each polarization interval according to spatial position, and judge the cause of chemical inhomogeneity based on the frequency domain characteristics of the sequence.

Images