A method and apparatus for measuring the diopter of a lens based on a diffractive optical element

Abstract

The present application belongs to the technical field of computational optical imaging and spectacle lens detection, and particularly relates to a lens diopter measurement method and device based on a diffractive optical element. Before measurement, calibration is performed, system calibration parameters are determined by simulation through an angular spectrum method and taking a normalized cross correlation as a target function, a single frame of lens speckle image is collected during measurement, the lens is modeled as a pure phase element, phase delay is constructed by spherical power, cylindrical power and axis angle, simulation speckle is obtained through two-stage angular spectrum propagation and phase modulation of the diffractive optical element, and optimal parameters are searched by maximizing NCC. The diopter is output through multi-stage hierarchical search and astigmatism determination. The device sequentially includes a narrow-band green LED light source, a beam expansion collimating lens, a diffractive optical element and an image acquisition camera along an optical path. The present application does not need a microlens array and an interference reference light path, only a single frame needs to be collected at a single field position, the full-aperture diopter distribution can be obtained through multi-position measurement, and the present application is suitable for free-form surface and progressive multifocal lens detection.

Description

Technical Field

[0001] This invention relates to the field of computational optical imaging and spectacle lens testing, and particularly to a method and apparatus for measuring lens refractive power based on diffractive optical elements. Background Technology

[0002] The accurate measurement of the refractive power (including spherical power, cylindrical power, and axis angle) of spectacle lenses is a fundamental aspect of optometry, lens quality control, and ophthalmic optical research. With the widespread use of freeform and progressive multifocal lenses, higher demands are placed on lens refractive power measurement methods in terms of measurement accuracy, dynamic range, full-aperture testing capability, and system complexity.

[0003] Currently, existing technologies for measuring the refractive power of spectacle lenses mainly fall into three categories. The first category is the automatic focimeter method, which uses a pinhole or multi-point aperture to confine the measurement beam to a limited number of narrow beams. This method can only collect deflection signals at limited locations on the lens, failing to capture the full-aperture refractive power distribution, and suffers from defocusing errors when measuring high-prescription lenses. The second category is the Hartmann-Shack wavefront sensor method, which utilizes a microlens array to segment the wavefront and infers the local wavefront slope by measuring the focal spot displacement of each sub-aperture. Its dynamic range is limited by the sub-aperture size, and it is prone to crosstalk between sub-apertures when measuring large curvature wavefronts, leading to failure. Furthermore, the spatial sampling density is limited by the physical number of microlenses. The third category is the interferometry method, which uses the optical path difference between the reference light and the test light to form interference fringes to invert the wavefront. This method has a complex system structure, is sensitive to vibration, and is difficult to miniaturize and integrate into production lines.

[0004] However, the aforementioned existing technologies have shortcomings such as limited dynamic range, spatial resolution constrained by hardware, complex systems that are difficult to integrate, and inability to perform full-caliber detection. Furthermore, some methods rely on mechanical scanning or multi-frame acquisition, which limits the measurement speed. Summary of the Invention

[0005] The purpose of this invention is to provide a method and device for measuring the refractive power of lenses based on diffractive optical elements, aiming to solve the technical problems existing in the current lens refractive power detection technology, such as the inability to measure the entire aperture, limited dynamic range, complex system that is difficult to miniaturize, and slow measurement speed.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides a method for measuring the refractive power of a lens based on a diffractive optical element. The method employs a narrow-band green light source, a beam expander collimating lens, a diffractive optical element, and an image acquisition camera arranged sequentially along the optical path. The method includes: S1, acquiring a single-frame image of the spatial speckle intensity without placing the lens under test; using the designed phase distribution of the diffractive optical element as input, propagating a first axial distance via angular spectrum method to the detector plane of the image acquisition camera to simulate the spatial speckle intensity image; determining system calibration parameters using the normalized cross-correlation value between the simulated spatial speckle intensity image and the measured spatial speckle intensity image as the objective function; S2, placing the lens under test between the beam expander collimating lens and the diffractive optical element. In the optical path between them, a single frame of speckle intensity image of the lens is acquired; S3, the lens under test is modeled as a pure phase element, and the phase delay distribution of the lens under test is constructed according to the spherical power, cylindrical power and axis angle of the lens under test; starting from the phase delay distribution, the second axial distance is propagated to the front surface of the diffractive optical element through the angular spectrum method, and after being modulated by the phase distribution of the diffractive optical element, the first axial distance is propagated to the detector plane through the angular spectrum method to obtain the simulated speckle intensity distribution; S4, under the system calibration parameters, the normalized cross-correlation value between the simulated speckle intensity distribution and the lens speckle intensity image is calculated, and the spherical power, cylindrical power and axis angle that maximize the normalized cross-correlation value are searched as the measurement result of the lens under test.

[0007] In step S3, the phase delay distribution is formed by the superposition of the quadric wavefront term and the tilted wavefront term. The quadric wavefront term is determined by the spherical power, cylindrical power, and axis angle through the following diopter matrix transformation: Let Where S is the spherical power and C is the cylindrical power, then the elements of the diopter matrix are... , , Determined by the following formula:

[0008]

[0009]

[0010] in, This is the axis position angle.

[0011] Step S4 employs a multi-stage hierarchical search strategy, including: First stage: within a preset spherical power range, scanning the spherical power with a first search step size, fixing the cylindrical power and axis angle to zero, to obtain a coarse-searched spherical power; Second stage: fixing the coarse-searched spherical power, scanning the second axial distance with a second search step size to obtain the optimal distance; Third stage: fixing the coarse-searched spherical power and the optimal distance, scanning the tilt angle parameter with a third search step size to obtain the optimal tilt angle; Fourth stage: fixing the optimal distance and the optimal tilt angle, finely scanning the spherical power with a fourth search step size smaller than the first search step size to obtain the optimal normalized cross-correlation value of the pure spherical model; Fifth stage: near the optimal spherical power obtained in the fourth stage, performing a three-dimensional joint search on the spherical power, cylindrical power, and axis angle to obtain the optimal normalized cross-correlation value of the spherical-cylindrical model.

[0012] After completing the fifth stage, the process also includes an astigmatism determination step: calculate the difference between the optimal normalized cross-correlation value of the spherical-cylindrical lens model and the optimal normalized cross-correlation value of the pure spherical lens model. If the difference is greater than or equal to 0.020, the lens under test is determined to have measurable astigmatism, and the spherical power, cylindrical power, and axis angle corresponding to the spherical-cylindrical lens model are adopted as the final measurement results; otherwise, it is determined to be a pure spherical lens, and the spherical power corresponding to the pure spherical lens model is adopted as the final measurement result, with the cylindrical power being zero.

[0013] The system calibration parameters in step S1 include the in-plane rotation angle of the diffractive optical element relative to the image acquisition camera, the first axial distance from the diffractive optical element to the image acquisition camera, and the equivalent phase modulation depth of the diffractive optical element.

[0014] After determining the system calibration parameters in step S1, a calibration verification step is also included: using the central region of interest in the open field speckle intensity image as a template, a sliding window matching is performed on the simulated open field speckle intensity image to determine the coordinate mapping offset between the simulated image and the measured image; using the coordinate mapping offset as a reference, multiple regions of interest are selected in the open field speckle intensity image, and each region of interest is matched and verified. If the matching result of each region of interest meets the preset spatial consistency criterion, the system calibration parameters are confirmed to be valid.

[0015] After completing step S1, the process also includes a full-aperture measurement step: at different field-of-view positions of the lens under test, steps S2, S3, S4 and astigmatism determination steps are executed respectively to obtain the final measurement results at each field-of-view position; the final measurement results at each field-of-view position are arranged according to spatial coordinates to generate a full-aperture refractive power distribution map of the lens under test.

[0016] After completing the astigmatism determination step, the procedure also includes a clinical output conversion step: The spherical and cylindrical powers in the final measurement results are multiplied by a preset sign conversion factor to obtain the spherical and cylindrical powers under the clinical sign; if the converted cylindrical power is positive, the spherical power is updated to the sum of the spherical and cylindrical powers, the cylindrical power is updated to its negative value, and the axis angle is increased by 90 degrees. The equivalent spherical power of the lens under test is then calculated according to... Calculation, where This is the converted spherical power. This is the converted cylinder power.

[0017] Secondly, the present invention also provides a lens refractive power measurement device based on a diffractive optical element, comprising, in sequence along the optical path: a light source, wherein the light source is a green light-emitting diode with a center wavelength of 500nm to 560nm and a spectral width of less than 60nm; a beam-expanding and collimating lens, disposed in the output optical path of the light source, for expanding and collimating the light beam emitted by the light source into a plane wavefront; a lens mounting position, disposed in the output optical path of the beam-expanding and collimating lens, for placing the lens to be measured, the lens mounting position having a bearing surface perpendicular to the optical axis to position the lens to be measured; and a diffractive optical element, disposed in the lens mounting position. The diffractive optical element, positioned on the outgoing optical path and spaced a second axial distance from the lens mounting position along the optical axis, has a preset phase distribution pattern for phase encoding of the wavefront modulated by the lens under test. An image acquisition camera, positioned on the outgoing optical path of the diffractive optical element and spaced a first axial distance from the diffractive optical element along the optical axis, is used to acquire a single-frame speckle intensity image formed after encoding by the diffractive optical element. A calculation and processing unit, electrically connected to the image acquisition camera, is used to receive the single-frame speckle intensity image and execute the aforementioned lens diopter measurement method based on the diffractive optical element.

[0018] The diffractive optical element is a microstructure phase plate with quartz glass as the substrate, which is etched by photolithography to form steps of different depths. The preset phase distribution pattern is stored in the computing processing unit.

[0019] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention achieves significant simplification and cost reduction at the hardware structure level. The optical path only includes a green LED light source, a beam expander and collimating lens, diffractive optical elements, and an image acquisition camera. It eliminates the need for the reference optical path and vibration isolation platform required for interferometry, and also eliminates the need for the high-precision microlens array relied upon by Hartmann sensors. This significantly reduces material costs and system complexity while ensuring measurement accuracy. The compact size makes it suitable for production line integration and miniaturized equipment applications. Furthermore, the light source uses a narrowband green LED with a center wavelength of 500nm to 560nm and a spectral width of less than 60nm, effectively suppressing the influence of dispersion effects on speckle contrast and improving the signal-to-noise ratio and matching stability of the speckle image.

[0020] 2. This invention achieves a breakthrough improvement in measurement dynamic range and spatial resolution. Compared to the spatial sampling method of Hartmann-Shack sensors that uses microlens arrays to segment the wavefront, this invention uses diffractive optical elements to perform deterministic phase encoding on the full-aperture wavefront. The equivalent wavefront sampling density is determined by the pixel resolution of the image detector and the sub-aperture partitioning strategy during computational reconstruction, fundamentally eliminating the limitation of sub-aperture boundaries on the measurable wavefront curvature, enabling refractive power measurement with a large dynamic range. Based on the normalized cross-correlation matching of simulated and measured speckle images, a multi-stage hierarchical search strategy is used to simultaneously optimize parameters such as spherical power, cylindrical power, axis angle, effective lens-DOE distance, and lens tilt in the seven-dimensional parameter space, achieving large dynamic range and high-precision refractive power reconstruction in a single frame measurement.

[0021] 3. This invention requires only a single frame acquisition at a single field of view location. By measuring different field of view locations separately, the full-aperture refractive power distribution can be obtained without multi-frame parameter switching. Unlike phase retrieval methods that require multiple changes in imaging parameters for multi-frame acquisition, and translation diffuser methods that require acquiring speckle images at multiple locations on the diffuser, this invention only requires a single frame of speckle image to complete wavefront reconstruction. After calibration, single-frame blind testing can be performed at any field of view location. By performing acquisition and reconstruction separately at different field of view locations through full-aperture measurement steps, a full-aperture refractive power distribution map of the lens under test is generated. This can effectively characterize the continuous refractive power variation characteristics of freeform lenses and progressive multifocal lenses, overcoming the inherent distortion of automatic focimeter sampling point measurements for complex surface lenses.

[0022] 4. This invention possesses a complete closed loop in terms of the clinical applicability of the measurement results. By setting a preset astigmatism threshold (0.020 in one embodiment) through the astigmatism determination step, false positives caused by parameter overfitting and noise fluctuations are effectively suppressed, ensuring the reliability of astigmatism determination. The clinical output conversion step maps the model's internal symbols to clinical symbols and automatically performs the expression transformation from positive to negative cylindrical lenses and the calculation of equivalent spherical power. The output results are directly compatible with the eyeglass prescription standards, eliminating the need for manual secondary conversion. Simultaneously, the multi-ROI calibration and verification step ensures the spatial consistency of the system calibration parameters across the entire field of view, further guaranteeing the accuracy and repeatability of the full-aperture measurement results. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of a lens refractive power measurement device based on diffractive optical elements provided in an embodiment of this application. Detailed Implementation

[0024] 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 of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0025] See Figure 1 The present invention provides a lens refractive power measurement device based on diffractive optical elements, which includes, in sequence along the optical path: a light source 1, a beam expander collimating lens 2, a lens mounting position (for placing the lens to be measured 3), a diffractive optical element 4, and an image acquisition camera 5.

[0026] In some embodiments, the light source 1 is a narrowband green light-emitting diode with a center wavelength range of 500 nm to 560 nm and a spectral width of less than 60 nm. Since the green light band is within the range of high visual sensitivity of the human eye, and the narrowband spectrum can effectively suppress the influence of dispersion on speckle contrast, using a narrowband green light source can improve the signal-to-noise ratio and matching stability of speckle images.

[0027] The beam expander and collimator lens 2 is disposed in the output optical path of the light source 1 to expand and collimate the beam emitted by the light source 1 into a plane wavefront. The diameter of the expanded and collimated beam must be larger than the effective aperture of the lens under test 3 to ensure that the lens under test can be illuminated in full aperture mode.

[0028] The lens mounting position is located on the output optical path of the beam expander collimating lens 2 and is used to place the lens to be tested 3. The lens mounting position has a bearing surface perpendicular to the optical axis to position the lens to be tested 3 and ensure that the optical axis of the lens to be tested 3 is substantially coincident with the optical axis of the measurement optical path.

[0029] The diffractive optical element 4 is disposed in the outgoing light path of the lens mounting position, and is spaced apart from the lens mounting position along the optical axis by a second axial distance. The diffractive optical element 4 is a microstructured phase plate with a quartz glass substrate, etched with steps of different depths using photolithography. It has a preset phase distribution pattern and is used for deterministic phase encoding of the wavefront modulated by the lens 3 under test. The feature size of the diffractive optical element 4 is preferably 14.4 μm, and its phase distribution... The information is known and stored in the computing unit, phase modulation depth .

[0030] The image acquisition camera 5 is positioned in the output light path of the diffractive optical element 4, and is spaced from the diffractive optical element 4 by a first axial distance along the optical axis. It is used to acquire a single-frame speckle intensity image formed after being encoded by diffractive optical element 4. The image acquisition camera 5 adopts an area array image detector, and its pixel resolution determines the upper limit density of the equivalent wavefront sampling.

[0031] The computing and processing unit is electrically connected to the image acquisition camera 5, and is used to receive a single-frame speckle intensity image acquired by the image acquisition camera 5, and execute the lens diopter measurement method provided by this invention. In the optical path of the embodiment provided in this application, no microlens array or interference reference optical path is provided.

[0032] System calibration is the first step in implementing the lens refractive power measurement method based on diffractive optical elements provided in this application. It is performed without the lens under test and is used to determine the inherent parameters of the system. First, an open-field speckle image is acquired: with the light source 1 turned off and the lens under test 3 not placed, the image acquisition camera 5 acquires a dark-field background image. This image is used to subsequently subtract the effects of ambient stray light and camera dark current. With light source 1 illuminated, and without the lens under test 3 placed, the beam emitted from light source 1 is expanded and collimated by beam-expanding collimating lens 2, and then directly passes through diffractive optical element 4. The resulting image of the ambient speckle intensity is captured by image acquisition camera 5. .

[0033] Next, a coarse calibration of the central region of interest is performed: using the designed phase distribution of the diffractive optical element 4 as input, the open field speckle intensity image is simulated in the forward direction using the angular spectrum method, and the initial values ​​of the system calibration parameters are determined by using the normalized cross-correlation value between the simulated image and the measured open field speckle intensity image as the objective function.

[0034] The complex amplitude transmittance function of diffractive optical element 4 is known to be:

[0035] in, The imaginary unit, The equivalent phase modulation depth of the diffractive optical element 4. The phase distribution (known quantity) is designed for diffractive optical element 4.

[0036] For complex amplitude transmittance function The angular spectrum is obtained by performing a Fourier transform. :

[0037] According to the angular spectrum propagation theory, the complex amplitude distribution of the photosensitive surface of the image acquisition camera 5 is as follows:

[0038] Among them, ℱ -1 Indicates the inverse Fourier transform. and Spatial frequency coordinates (unit: mm) -1 ), is the angular spectrum transfer function.

[0039] The expression for the angular spectrum transfer function is:

[0040] in, The sign parameter for the propagation direction (+1 indicates forward propagation, and -1 indicates backward propagation). λ is the wavelength of the light source (in nm), and z is the propagation distance (in mm). and For spatial frequency coordinates. When When ≤1, the transfer function represents a propagating wave; when When the value is greater than 1, it indicates an evanescent wave, which is usually ignored in actual calculations.

[0041] The speckle intensity distribution acquired by image acquisition camera 5 is a simulated speckle intensity distribution:

[0042] To compensate for the in-plane rotation angle of diffractive optical element 4 Rotate the simulated speckle pattern in reverse:

[0043] Here, rotate represents the image rotation operation. The in-plane rotation angle of the diffractive optical element 4 relative to the image acquisition camera 5.

[0044] A region of interest (ROI) of size M×M pixels is extracted from the measured spatial speckle intensity image as a matching template T. Sliding window matching is then performed on the rotated simulated image. After Gaussian high-pass filtering and zero-mean normalization preprocessing of the template T and the simulated image, the normalized cross-correlation value is calculated.

[0045] in, Let T be the average pixel grayscale value of template T. Let be the average pixel grayscale value of the matching window starting at position (u,v) in the simulated image. It is a very small positive number (used to prevent the denominator from being zero, usually taken as 10). -6 ), To match the starting pixel coordinates of the window in the simulation image.

[0046] The position (u, v) corresponding to the maximum normalized cross-correlation value is selected as the optimal matching position, and the coordinate mapping offset is:

[0047] in,( , () represents the starting pixel coordinates of the measured region of interest in the original image.

[0048] Determine the in-plane rotation angle of diffractive optical element 4 The first axial distance from the diffractive optical element 4 to the image acquisition camera 5 and the equivalent phase modulation depth of diffractive optical element 4 The process is divided into four stages: Phase 1 (All-round perspective) (Coarse sweep): Fixed and The set value before calibration, for Equal-interval scans were performed within the range of 0° to 358° with a step size of 2°. For each scan, a forward propagation simulation and normalized cross-correlation matching were performed, and the normalized cross-correlation value was recorded as... The changing curve. Select the value corresponding to the maximum normalized cross-correlation. This serves as the result of the coarse calibration.

[0049] Phase Two ( (Detailed scan): The first stage is optimal. Centered on the target, a fine scan is performed within a range of ±4° with a step size of 0.25°, and the value corresponding to the maximum normalized cross-correlation value is selected. This serves as the result of fine calibration.

[0050] Phase Three ( and Joint Refinement): Fixed For the optimal value in the second stage, and Perform a two-dimensional joint scan. The search range is approximately ±1.25 mm from the previous calibration value, with a step size of 0.1 mm. The search range is approximately 0.74π to 0.88π, with a step size of 0.01π. The value corresponding to the maximum normalized cross-correlation is selected. and The combination serves as the refined result. This stage is only performed when the normalized cross-correlation value from the second stage reaches a preset threshold (e.g., 0.12) to ensure... A reliable main peak has been found.

[0051] Phase Four (Final) (Re-examination): Fixed and The optimal value for the third stage, in the optimal... Rescan within a range of ±1.2° with a step size of 0.1°. The final calibration result is obtained.

[0052] Through the above four-stage calibration, the system calibration parameters include the in-plane rotation angle of the diffractive optical element 4 relative to the image acquisition camera 5. The first axial distance from the diffractive optical element 4 to the image acquisition camera 5 and the equivalent phase modulation depth of diffractive optical element 4 -- All of them were determined to be the optimal values.

[0053] Based on the initial parameters obtained from coarse calibration in a single region, this application performs global optimization of the calibration parameters through spatial consistency verification of multiple regions of interest.

[0054] Specifically, on the measured spatial speckle intensity image, using the central region of interest as a reference, it expands in four directions (left, right, up, and down) at preset intervals (e.g., 900 pixels) to form a cross-shaped array of regions of interest: center + left + right + up + down. The coordinate mapping offset is based on the coordinates obtained from the coarse calibration. Calculate the predicted matching position of each region of interest in the simulated image:

[0055] Where i represents the i-th region of interest, ( , ) represents the starting pixel coordinates of the i-th measured region of interest in the original image.

[0056] For each region of interest, perform normalized cross-correlation template matching within the local search range near the predicted location, and record the matching degree. Actual offset ( And whether the search boundary has been touched. Spatial consistency criteria include: Center match degree: Average matching degree of edge regions: Average matching degree across the entire region: Maximum position offset: Boundary touch mark: in, The boundary touch flag for the i-th region of interest is set (1 indicates boundary touch, 0 indicates boundary non-touch). If spatial consistency does not meet the requirements, parameter refinement is performed using the average normalized cross-correlation of multiple regions across the entire field of view as the objective function.

[0057] For the in-plane rotation angle of diffractive optical element 4 In the coarse calibration value Select a candidate scan list nearby (e.g.) ±1.0°, step size 0.2°). For each candidate The value is used to perform a match on all regions of interest and calculate the value. The corresponding comprehensive score:

[0058] Select the option that maximizes the average normalized cross-correlation. As the final calibration value:

[0059] The physical basis of this method is that the pixel displacement of the edge region of interest is more sensitive to the θ deviation than that of the central region of interest. Including the edge region of interest in the joint scoring makes the peak of the objective function sharper and the optimal value more accurately located.

[0060] For coordinate mapping offset After rematching each region of interest with the optimal parameters, the median of the local mapping offsets of all valid regions of interest is taken as the global optimum.

[0061] The median is naturally robust to individual anomalous regions of interest and can more accurately reflect the overall mapping relationship of the entire field of view.

[0062] For example, global optimization of calibration parameters is achieved using hierarchical criteria: The normalized cross-correlation value of the central region of interest is ≥0.35; the average normalized cross-correlation value of the peripheral regions of interest is ≥0.30; the average normalized cross-correlation value of all regions of interest is ≥0.35; no region of interest touches the search boundary.

[0063] If all the above criteria are met, the calibration parameters are valid and can be written into the blind lens diopter measurement program; if not, it is necessary to return to the single-region calibration stage and search again.

[0064] Through consistency verification and joint optimization across multiple regions of the full field of view, the effectiveness of the calibration parameters was ensured throughout the entire imaging field of view, avoiding the local optima problem that may exist in single-region calibration, and providing a guarantee for the accuracy of subsequent full-aperture refractive power measurements.

[0065] Once the system calibration is complete, a blind diopter test can be performed on the lens to be tested.

[0066] First, the speckle intensity image of the measured lens is acquired and preprocessed. Light source 1 is turned on, and the lens under test 3 is placed in the lens mounting position between the beam expander / collimator lens 2 and the diffractive optical element 4. The light beam emitted from light source 1 is expanded and collimated by the beam expander / collimator lens 2, passes through the lens under test 3, and continues to propagate to the diffractive optical element 4 after being modulated by the wavefront of the lens under test 3. After being encoded by the diffractive optical element 4, the image acquisition camera 5 captures the resulting speckle intensity image of the lens. .

[0067] Subtract the dark field image (the background image acquired with the light source turned off) from the original image to eliminate the effects of camera dark current and stray light:

[0068] If the difference results in a negative value, then set it to zero.

[0069] In the corrected large image, using the calibrated center coordinates Starting from a point, extract a region of interest with a size of M×M pixels (e.g., 500×500):

[0070] The region of interest image was subjected to Gaussian high-pass filtering and zero-mean normalization, which served as the measured template for subsequent matching.

[0071]

[0072] in, For Gaussian kernel, The standard deviation of the Gaussian kernel is given by the asterisk (*), which indicates a convolution operation. for The average grayscale value of the pixels. for The standard deviation of pixel grayscale It is a very small positive number.

[0073] Then, the phase parameterization model of the lens is established. The lens under test, 3, is modeled as a pure phase element, and its phase delay is composed of the superposition of the quadratic wavefront term and the tilted wavefront term:

[0074] Where k is the wave number (k=2) / , (Wavelength of the light source, unit: mm) This is the lens symbol parameter (take +1 or -1 to distinguish between positive and negative refractive power). and The lens tilt angle parameter (unit: radians). , , These are elements of the diopter matrix.

[0075] Refractive power matrix elements , , The spherical power S (unit: diopter D), cylindrical power C (unit: diopter D), and axis angle are used to determine the spherical power. (Unit: degrees) is determined, and the relationship is as follows:

[0076]

[0077]

[0078]

[0079] in, The diopter is the refractive power in the axial direction. This refers to the refractive power in the direction perpendicular to the axis angle. The wavefront term accurately describes the wavefront curvature caused by spherical and cylindrical powers, while the tilt wavefront term describes the minute tilt during lens mounting. This comprehensively covers all lens-related parameters affecting the speckle pattern, providing an accurate physical model basis for subsequent parameter searches.

[0080] Next, the phase retardation distribution of the lens under test 3 is used. Starting from (x,y), the complex amplitude of the rear surface of the lens 3 under test is:

[0081] The complex amplitude propagates along the second axial distance via the angular spectrum method. Reaching the front surface of diffractive optical element 4:

[0082] in, For the distance of transmission The angular spectral transfer function. The complex amplitude, after modulation by diffractive optical element 4:

[0083] in, The phase distribution is designed for diffractive optical element 4.

[0084] The modulated complex amplitude is then propagated via the angular spectrum method to the first axial distance. Reaching the detector plane of image acquisition camera 5:

[0085] The simulated speckle intensity distribution is as follows:

[0086] The two-stage angular spectrum propagation fully describes the entire physical optical propagation process from the lens under test through the diffractive optical element to the image acquisition camera, ensuring that the simulated speckle intensity distribution and the actual measured speckle intensity image correspond strictly in terms of physical mechanism, thereby guaranteeing the physical rationality of parameter search based on normalized cross-correlation matching.

[0087] Next, normalized cross-correlation was used to evaluate the similarity between the simulated speckle intensity distribution and the measured speckle intensity image of the lens:

[0088] in, The template is the measured speckle intensity image of the lens after high-pass filtering and zero-mean normalization. Its mean, To simulate speckle intensity distribution, Its mean, It is a very small positive number. The normalized cross-correlation value ranges from [-1, 1]. The closer the value is to 1, the more similar the simulated speckle pattern is to the measured speckle pattern, and the closer the corresponding parameter combination is to the true value.

[0089] Then, a multi-stage, hierarchical blind test search is performed. For example, the parameter space to be searched is:

[0090] There are a total of 7 parameters. To decouple the parameters and avoid local optima, a multi-stage hierarchical search strategy from coarse to fine is adopted: Phase 1 (Full-range blind scan of spherical power S): Under conditions of no prior information, perform a wide-range scan of S (the search range is set according to the possible spherical power range of the lens under test, with a step size of 0.50D), with C=0 fixed. =0, other parameters take their initial values. Select the value corresponding to the peak of normalized cross-correlation. As a rough search of spherical mirror.

[0091] Second stage (Effective lens-DOE distance) Optimization): Fixed S = Hierarchical optimization The coarse adjustment range is 0.5mm to 35.0mm, with a step size of 0.5mm; the fine adjustment range is ±2.0mm, with a step size of 0.25mm. This step is used to resolve the relationship between the spherical power S and the distance. The strong coupling between them is because changes in spherical power and distance will alter the wavefront curvature reaching the diffractive optical element 4, producing a similar effect on the speckle pattern.

[0092] Phase 3 (Lens Tilt Optimization): Fixing S and Hierarchical optimization and The coarse adjustment range is ±0.030 rad with a step size of 0.006 rad; the fine adjustment range is ±0.008 rad with a step size of 0.002 rad.

[0093] Fourth stage (fine spherical power scan): fixation By adjusting the tilt angle and performing a fine scan S (±0.80D, step size 0.05D), the optimal normalized cross-correlation value of the pure spherical mirror model was obtained. .

[0094] Fifth stage (astigmatism model determination): Near the optimal spherical power obtained in the fourth stage, adjust the spherical power S, cylindrical power C, and axis angle. Perform a three-dimensional joint search. The search range covers S ± 0.75D, C ∈ [-2.50D, 0D], ∈[0°, 165°], with step sizes of 0.05D, 0.05D, and 5° respectively. The optimal normalized cross-correlation value for the spherical-cylindrical mirror model is obtained. .

[0095] By decomposing the 7-dimensional parameter space into multiple low-dimensional subspaces and optimizing them sequentially, the computational complexity is significantly reduced; by adopting a coarse-to-fine stepwise refinement strategy, the risk of getting trapped in local optima is effectively avoided; and by conducting a staged search first for pure spherical lenses and then for spherocylindrical lenses, the astigmatic parameters are accurately extracted.

[0096] After completing the fifth stage (three-dimensional joint search), the astigmatism determination step is performed: Calculate the optimal normalized cross-correlation value of the spherical-cylindrical mirror model. Optimal normalized cross-correlation value with pure spherical mirror model The difference .

[0097] like If the value is ≥0.020, it is determined that the lens under test at the current field of view has measurable astigmatism, and the spherical power corresponding to the spherical cylindrical lens model is adopted. Cylinder power and axis position angle This serves as the final measurement result for the location of that field of view.

[0098] like If the value is less than 0.020, the lens under test at the current field of view is determined to be a pure spherical lens, and the spherical power corresponding to the pure spherical lens model is adopted. As the final measurement result for this field of view location, the cylinder power is zero, and the axis angle is meaningless.

[0099] For example, an astigmatism determination threshold of 0.020 is set. This threshold is used to suppress false positive astigmatism determinations caused by parameter overfitting and noise fluctuations, thereby improving the stability of astigmatism determination.

[0100] For example, after completing system calibration, perform the full-caliber measurement step: At different field-of-view positions of the lens 3 under test, i.e., at different spatial positions within the aperture range of the lens 3 under test, the aforementioned lens speckle acquisition and preprocessing, lens phase parameterization modeling, physical optics forward propagation, normalized cross-correlation similarity evaluation, multi-stage graded blind test search, and astigmatism determination steps are performed respectively to obtain the final measurement results (including spherical power, cylindrical power, and axis angle) at each field-of-view position.

[0101] The final measurement results at each field of view are arranged according to their spatial coordinates to generate a full-aperture refractive power distribution map of the lens under test (3). By performing single-frame measurements and reconstructions at multiple field of view locations across the full aperture of the lens under test, the continuous refractive power variation characteristics of freeform lenses and progressive multifocal lenses can be fully characterized, overcoming the inherent distortion of traditional automatic focimeter sampling point measurements for complex surface lenses. Furthermore, each field of view location is acquired using a single frame, eliminating the need for multi-frame parameter switching and resulting in high measurement speed.

[0102] For example, after completing the astigmatism determination step, a clinical output conversion step is performed: the spherical power S and cylindrical power C in the final measurement result are multiplied by a preset sign conversion factor. (Take +1 or -1 to map the model's internal symbols to clinical symbols), to obtain the spherical diopter under the clinical symbols. and cylinder power :

[0103] If the converted cylinder power If the value is positive, perform an expression transformation from positive cylindrical lens to negative cylindrical lens:

[0104] Finally, the spherical equivalent (SE) of the tested lens 3 is:

[0105] The model's internal parameters are automatically converted into clinically applicable symbolic expressions, and the automatic conversion from positive to negative cylindrical lenses and the calculation of equivalent spherical power are completed. The output results are directly compatible with the eyeglass prescription standards, eliminating the need for manual secondary conversion and improving the clinical applicability of the measurement results.

[0106] This application also provides a specific embodiment: The lens diopter measurement device based on diffractive optical elements provided by this invention is used to measure the diopter of a lens with nominal values ​​of S=-3.00D, C=-1.00D, and axis angle. Measurements were taken using a cylindrical lens with a radius of 30°.

[0107] Light source 1 uses a green LED with a center wavelength of 532 nm and a spectral width of 50 nm. Diffractive optical element 4 uses a quartz glass substrate, with an 8-step microstructure phase plate etched using photolithography. The feature size is 14.4 μm, and the phase modulation depth is [not specified]. =0.8π. The image acquisition camera 5 uses a 5-megapixel area array CMOS camera with a pixel size of 3.45μm.

[0108] Calibration Phase: Following the aforementioned calibration method, acquire open-field speckle images without placing the lens under test, and perform four-stage calibration (full-angle). coarse sweeping fine sweeping and Joint refinement, final (Re-examination) and multi-region of interest verification were performed to obtain system calibration parameters: =2.5° =25.0mm =0.79π, coordinate mapping offset (Δx,Δy)=(12,-8) pixels.

[0109] Measurement Phase: The lens to be tested, 3, is placed in the lens mounting position, and a single-frame image of the lens speckle intensity is acquired. After image preprocessing (dark field subtraction, region of interest cropping, high-pass filtering, and zero-mean normalization), a multi-stage hierarchical search is performed: the first stage obtains the full-range blind scan spherical power S. =-3.00D; Second stage optimization Obtain the optimal distance =15.0mm; the tilt angle was obtained in the third stage of optimization. =0.002rad =-0.001rad; Fourth-stage fine scanning S yields S=-3.00D, =0.52; the fifth stage of the three-dimensional joint search yielded S=-3.00D, C=-1.00D, =30° =0.58.

[0110] Astigmatism determination: ΔNCC=0.58-0.52=0.06≥0.020, indicating the presence of measurable astigmatism, and the result of the spherical cylindrical lens model is adopted.

[0111] Clinical output conversion: After sign conversion and positive cylinder to negative cylinder conversion, the outputs are S=-3.00D, C=-1.00D. =30°, consistent with the nominal value marked on the lens.

[0112] In the description of this specification, specific features, structures, materials, or characteristics may be combined in any suitable manner in one or more embodiments or examples.

[0113] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for measuring the refractive power of a lens based on diffractive optical elements, comprising a narrow-band green light source, a beam-expanding collimating lens, a diffractive optical element, and an image acquisition camera arranged sequentially along the optical path, characterized in that... include: S1. Acquire a single frame of spatial speckle intensity image without placing the lens under test; Using the designed phase distribution of the diffractive optical element as input, the image is propagated along the first axial distance to the detector plane of the image acquisition camera via the angular spectrum method to simulate the ambient speckle intensity image; the normalized cross-correlation value between the simulated ambient speckle intensity image and the measured ambient speckle intensity image is used as the objective function to determine the system calibration parameters; S2, the lens under test is placed in the optical path between the beam expanding collimating lens and the diffractive optical element, and a single frame of lens speckle intensity image is acquired; S3, the lens under test is modeled as a pure phase element, and a system is constructed based on the spherical power, cylindrical power, and axial angle of the lens under test. The phase retardation distribution of the lens under test; starting from the phase retardation distribution, propagating a second axial distance via angular spectrum method to the front surface of the diffractive optical element, modulating the phase distribution of the diffractive optical element, and then propagating a first axial distance via angular spectrum method to the detector plane to obtain the simulated speckle intensity distribution; S4, under the system calibration parameters, calculating the normalized cross-correlation value between the simulated speckle intensity distribution and the speckle intensity image of the lens, and searching for the spherical power, cylindrical power and axis angle that maximize the normalized cross-correlation value as the measurement result of the lens under test.

2. The method for measuring lens refractive power based on diffractive optical elements according to claim 1, characterized in that, The phase delay distribution in step S3 is composed of the superposition of the quadric wavefront term and the tilted wavefront term. The quadric wavefront term is determined by the spherical power, the cylindrical power, and the axis angle through the following diopter matrix transformation: Let Where S is the spherical power and C is the cylindrical power, then the elements of the diopter matrix are... , , Determined by the following formula: in, This is the axis position angle.

3. The method for measuring lens refractive power based on diffractive optical elements according to claim 1, characterized in that, Step S4 employs a multi-stage hierarchical search strategy, including: First stage: within a preset spherical power range, scanning the spherical power with a first search step size, fixing the cylindrical power and axis angle at zero, to obtain a coarse-searched spherical power; Second stage: fixing the coarse-searched spherical power, scanning the second axial distance with a second search step size to obtain the optimal distance; Third stage: fixing the coarse-searched spherical power and the optimal distance, scanning the tilt angle parameter with a third search step size to obtain the optimal tilt angle; Fourth stage: fixing the optimal distance and the optimal tilt angle, finely scanning the spherical power with a fourth search step size smaller than the first search step size to obtain the optimal normalized cross-correlation value of the pure spherical power model; Fifth stage: near the optimal spherical power obtained in the fourth stage, performing a three-dimensional joint search on the spherical power, the cylindrical power, and the axis angle to obtain the optimal normalized cross-correlation value of the spherical-cylindrical power model.

4. The method for measuring lens refractive power based on diffractive optical elements according to claim 3, characterized in that, After completing the fifth stage, an astigmatism determination step is also included: calculating the difference between the optimal normalized cross-correlation value of the spherical-cylindrical lens model and the optimal normalized cross-correlation value of the pure spherical lens model. If the difference is greater than or equal to 0.020, it is determined that the lens under test has measurable astigmatism, and the spherical power, cylindrical power, and axis angle corresponding to the spherical-cylindrical lens model are adopted as the final measurement results; otherwise, it is determined to be a pure spherical lens, and the spherical power corresponding to the pure spherical lens model is adopted as the final measurement result, with the cylindrical power being zero.

5. The method for measuring lens refractive power based on diffractive optical elements according to claim 1, characterized in that, The system calibration parameters in step S1 include the in-plane rotation angle of the diffractive optical element relative to the image acquisition camera, the first axial distance from the diffractive optical element to the image acquisition camera, and the equivalent phase modulation depth of the diffractive optical element.

6. The method for measuring lens refractive power based on diffractive optical elements according to claim 5, characterized in that, After determining the system calibration parameters in step S1, a calibration verification step is also included: using the central region of interest in the air field speckle intensity image as a template, a sliding window matching is performed on the simulated air field speckle intensity image to determine the coordinate mapping offset between the simulated image and the measured image; Based on the coordinate mapping offset, multiple regions of interest are selected in the open field speckle intensity image, and each region of interest is matched and verified. If the matching result of each region of interest meets the preset spatial consistency criterion, the system calibration parameters are confirmed to be valid.

7. The method for measuring lens refractive power based on diffractive optical elements according to claim 4, characterized in that, After completing step S1, the method further includes a full-aperture measurement step: at different field-of-view positions of the lens under test, steps S2, S3, S4 and the astigmatism determination step are executed respectively to obtain the final measurement results at each field-of-view position; the final measurement results at each field-of-view position are arranged according to spatial coordinates to generate a full-aperture refractive power distribution map of the lens under test.

8. The method for measuring lens refractive power based on diffractive optical elements according to claim 4, characterized in that, After completing the astigmatism determination step, a clinical output conversion step is also included: multiplying the spherical and cylindrical power in the final measurement result by a preset sign conversion factor to obtain the spherical and cylindrical power under clinical signs; if the converted cylindrical power is positive, then the spherical power is updated to the sum of the spherical power and the cylindrical power, the cylindrical power is updated to its negative value, and the axis angle is increased by 90 degrees; the equivalent spherical power of the lens under test is calculated according to... Calculation, where This is the converted spherical power. This is the converted cylinder power.

9. A lens diopter measuring device based on diffractive optical elements, characterized in that, The device comprises, sequentially along the optical path: a light source, wherein the light source is a green light-emitting diode with a center wavelength of 500nm to 560nm and a spectral width of less than 60nm; a beam-expanding and collimating lens, disposed in the output optical path of the light source, for expanding and collimating the light beam emitted by the light source into a plane wavefront; a lens mounting position, disposed in the output optical path of the beam-expanding and collimating lens, for placing the lens under test, wherein the lens mounting position has a bearing surface perpendicular to the optical axis to position the lens under test; and a diffractive optical element, disposed in the output optical path of the lens mounting position and positioned relative to the lens mounting position along the optical axis. The diffractive optical element has a preset phase distribution pattern at a second axial distance, used for phase encoding of the wavefront modulated by the lens under test; an image acquisition camera is disposed on the outgoing light path of the diffractive optical element and spaced a first axial distance from the diffractive optical element along the optical axis, used for acquiring a single-frame speckle intensity image formed after being encoded by the diffractive optical element; a calculation and processing unit is electrically connected to the image acquisition camera, used for receiving the single-frame speckle intensity image and executing the lens diopter measurement method based on the diffractive optical element as described in any one of claims 1-8.

10. A lens diopter measuring device based on a diffractive optical element according to claim 9, characterized in that, The diffractive optical element is a microstructure phase plate with quartz glass as the substrate, which is etched by photolithography to form steps of different depths. The preset phase distribution pattern is stored in the computing and processing unit.

Images