Analog eye with adjustable diopter and real-time retinal imaging detection

Abstract

The application discloses a diopter-adjustable and real-time retina imaging detection simulation eye, comprising: an emulated cornea, an emulated lens and an adjustable vitreous structure arranged in sequence along an optical axis; the adjustable vitreous structure comprises: a vitreous cavity filled with an emulated vitreous, a curved retina arranged at the end of the vitreous cavity, and a micrometer adjustment system for driving the curved retina to move forward and backward along the optical axis; the foveal region of the curved retina is integrated with an attenuation module and an imaging module, which are used for collecting a light spot image on the foveal region in real time; the simulation eye further comprises an analysis module electrically connected with the imaging module, and the analysis module is configured to process the collected light spot image to extract optical parameters on the curved retina.

Description

Technical Field

[0001] This invention belongs to the field of biomedical optics and visual simulation technology, and specifically relates to a simulated eye with adjustable diopter and real-time retinal imaging detection. Background Technology

[0002] Simulated eyes have important applications in ophthalmic equipment calibration, visual optics research, biomedical effect testing, safety assessment of low-intensity red light therapy, and visual function simulation. Existing simulated eyes are mainly used to simulate the optical functions of the eye, especially in refractive error calibration, ophthalmic imaging testing, and phototherapy evaluation.

[0003] The innovations of existing simulated eye devices mainly lie in structural simulation and basic refractive power adjustment. However, they generally suffer from two key shortcomings: First, in terms of refractive power adjustment, they either lack continuous adjustment capabilities or have limited adjustment accuracy, typically no less than ±0.5D, while modern high-precision ophthalmic diagnostic equipment, such as computerized refractometers and comprehensive refractometers, typically require measurement accuracy of ±0.25D to ±0.12D, or even higher. As a calibration benchmark, the simulated eye's own adjustment accuracy and repeatability must be significantly superior to the calibrated device, a requirement that current technologies struggle to meet. Second, most devices lack the ability to perform real-time, quantitative imaging detection of the retinal surface light field distribution, such as light intensity, irradiance, and spot morphology, failing to provide crucial data for applications such as low-intensity phototherapy safety assessment. Therefore, this invention provides a high-precision simulated eye with continuously adjustable refractive power and real-time detection capabilities of retinal optical parameters. This overcomes the shortcomings of existing simulated eyes in refractive power adjustment and real-time optical parameter detection, providing a more accurate solution for safety assessment of low-intensity red light therapy and precise calibration of ophthalmic equipment. Summary of the Invention

[0004] This application provides a high-precision adjustable refractive-real-time retinal imaging simulation eye, aiming to solve the technical problems of insufficient refractive power adjustment precision, poor continuity, and inability to quantitatively detect retinal optical parameters in real time in existing simulation eyes. This application introduces a precisely adjustable axial distance between the retina and lens, and integrates a high-resolution imaging sensor to simulate the foveal region of the retina, achieving precise and continuous adjustment of refractive power and synchronous real-time capture of retinal optical field parameters. This provides a powerful tool for safety assessment of low-intensity red light therapy, high-precision calibration of ophthalmic equipment, and visual function simulation.

[0005] This application discloses a simulated eye with adjustable refractive power and real-time retinal imaging detection, comprising: a simulated cornea, a simulated lens, and an adjustable vitreous structure arranged sequentially along the optical axis; the adjustable vitreous structure includes: a vitreous cavity filled with the simulated vitreous, a curved retina disposed at the end of the vitreous cavity, and a micrometer adjustment system for driving the curved retina to move back and forth along the optical axis; the foveal region of the curved retina integrates an attenuation module and an imaging module for real-time acquisition of light spot images on the foveal region; the attenuation module is used to reduce light intensity, ensure that the imaging module is not saturated, and effectively acquire the distribution of the light field; the simulated eye also includes an analysis module electrically connected to the imaging module, the analysis module being configured to process the acquired light spot images to extract optical parameters on the curved retina, the optical parameters including at least one of light intensity distribution, irradiance, light spot morphology, and single pixel power density.

[0006] According to the simulated eye disclosed in this application, a micrometer-scale adjustment system drives the curved retina to move along the optical axis, achieving precise and continuous adjustment of the axial length of the vitreous cavity, thereby simulating various refractive states such as emmetropia, myopia, and hyperopia. At the same time, the imaging module integrated in the fovea region of the retina can acquire light spot images in real time, and extract optical parameters in conjunction with the analysis module. The simulated eye integrates high-precision refractive adjustment and quantitative detection of the retinal surface light field, providing a comprehensive testing platform for the safety assessment of low-intensity red light therapy and the precise calibration of ophthalmic equipment.

[0007] In the simulated eye disclosed in this application, the simulated cornea is a positive meniscus lens, and the simulated lens is a biconvex lens.

[0008] According to the simulated eye disclosed in this application, a positive meniscus lens is used as a simulated cornea and a biconvex lens is used as a simulated lens. The combination of the two can accurately simulate the optical power distribution of the main refractive elements of the human eye, making the optical system of the simulated eye closer to the real human eye in terms of equivalent focal length, spherical aberration and other characteristics. This improves the physiological authenticity of the simulated eye in the safety assessment of low-intensity red light therapy and the precise calibration of ophthalmic equipment.

[0009] The simulated eye disclosed in this application further includes an aqueous cavity located between the simulated cornea and the simulated lens, the aqueous cavity being filled with anterior chamber fluid, and both the anterior chamber fluid and the simulated vitreous body being refractive index matching fluids with a refractive index of 1.336.

[0010] According to the simulated eye disclosed in this application, by setting an aqueous cavity filled with a refractive index matching fluid between the simulated cornea and the simulated lens, and filling the vitreous cavity with a simulated vitreous body of the same refractive index, not only is the optical environment of the aqueous humor and vitreous body of the human eye simulated, but also the reflection loss between different media interfaces is eliminated, the light energy transmission efficiency is improved, and the light spot image acquired by the imaging module can truly reflect the energy distribution of the retinal surface.

[0011] In the simulated eye disclosed in this application, the curved retina is a spherical structure with a radius of curvature of 12.5 mm, and its surface is coated with an optical film to reduce reflectivity.

[0012] According to the simulated eye disclosed in this application, the curved retina is designed as a spherical structure with a curvature radius of 12.5 mm, which matches the average curvature of the human retina, thus avoiding off-axis aberrations introduced by planar sensors; the optical film coated on the surface can effectively reduce reflectivity and reduce stray light entering the imaging module, thereby improving the contrast and signal-to-noise ratio of the spot image, which is beneficial for the accurate extraction of subsequent optical parameters.

[0013] In the simulated eye disclosed in this application, the imaging module includes a CMOS image sensor with a pixel size of 1.3 to 1.5 μm.

[0014] According to the simulated eye disclosed in this application, the imaging module uses a CMOS image sensor (preferably 1.4μm) with a pixel size of 1.3-1.5μm, which is comparable to the diameter of cone cells in the fovea of ​​the retina, and can capture the microscopic energy distribution of the light spot with a spatial sampling rate close to the resolution of the human eye.

[0015] In the simulated eye disclosed in this application, the micrometer adjustment system moves by 1 mm, and the refractive power of the simulated eye changes by 3.46D, thereby achieving continuous adjustment of the refractive power within the range of -6.92D to +6.92D.

[0016] According to the simulated eye disclosed in this application, the micrometer adjustment system corresponds to a refractive power change of 3.46D for every 1mm movement. Based on this proportional relationship, by setting the axial movement range of the curved retina to a distance from the simulated lens from 11.00mm to 15.00mm, continuous and precise adjustment of the refractive power from -6.92D to +6.92D can be achieved, with an adjustment step of approximately 0.0346D. This is significantly better than the ±0.5D adjustment accuracy of the conventional simulated eye and meets the fine-tuning calibration requirements of high-end ophthalmic equipment (accuracy ±0.12D).

[0017] In the simulated eye disclosed in this application, the processing of the spot image by the analysis module includes: preprocessing the spot image, the preprocessing including at least one of dark frame correction, background subtraction and median filtering; extracting the spot region based on an adaptive threshold algorithm combined with morphological operations and obtaining a spot mask; based on the spot mask, coarsely locating the centroid of the spot region by the gray-scale centroid method, and then obtaining the sub-pixel level coordinates of the centroid of the spot region by the two-dimensional Gaussian surface fitting method.

[0018] According to the simulated eye disclosed in this application, the analysis module first preprocesses the original image through dark frame correction, background subtraction, and median filtering, effectively eliminating sensor dark current noise, ambient stray light, and isolated noise. Then, it accurately extracts the spot mask using adaptive threshold segmentation and morphological operations. Finally, it improves the centroid positioning accuracy to the sub-pixel level (0.01-0.1 pixels) by combining coarse positioning using the gray-scale centroid method with fine positioning using the two-dimensional Gaussian surface fitting method, significantly improving the repeatability and reliability of optical parameter measurements.

[0019] In the simulated view disclosed in this application, the analysis module is further configured to: calculate the area of ​​the spot region based on the spot mask, and statistically analyze the cumulative gray value, maximum gray value, and average gray value within the spot region.

[0020] According to the simulated eye disclosed in this application, based on a light spot mask, the analysis module can quantitatively calculate the physical area of ​​the light spot (converted from pixel count and pixel size), the cumulative gray value (proportional to the total light energy), the maximum gray value (corresponding to peak irradiance), and the average gray value (reflecting average energy density). These parameters provide direct and quantitative indicators for evaluating the retinal surface radiation dose, determining the focusing performance of the optical system, and calibrating the light source output of ophthalmic equipment.

[0021] In the simulated view disclosed in this application, the analysis module is further configured to: divide the gray values ​​within the area of ​​the light spot mask into multiple gray ranges according to their size, and count the number of pixels and the cumulative gray value in each gray range to quantify the concentration of energy distribution within the light spot area.

[0022] According to the simulated eye disclosed in this application, by dividing the gray values ​​within the light spot mask into multiple gray-level intervals (e.g., each gray-level segment is 5%), and statistically analyzing the number of pixels and the cumulative gray-level value within each gray-level interval, the analysis module can quantify the concentration of energy distribution within the light spot, such as identifying the energy proportion of the high-gray-level core region. This feature can be used to determine whether the light spot exhibits abnormal states such as multi-peak, edge diffusion, or central overexposure, providing a refined analytical tool for the safety assessment of low-intensity red light therapy and the precise calibration of ophthalmic equipment.

[0023] In the simulated view disclosed in this application, the analysis module is further configured to: generate a two-dimensional pseudo-color distribution map for displaying the spatial distribution of light intensity and / or a three-dimensional intensity distribution map for displaying the local light intensity variation trend based on the gray-scale matrix of the light spot region.

[0024] According to the simulated eye disclosed in this application, the analysis module generates a two-dimensional pseudo-color distribution map and a three-dimensional intensity distribution map based on the grayscale matrix of the light spot, providing a visual basis for decision-making. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of a simulated eye structure with adjustable refractive power and real-time retinal imaging detection as described in this application.

[0026] Figure 2 A schematic diagram of the equivalent optical structure of the human eye to illustrate the principle of diopter accommodation;

[0027] Figure 3 This is a schematic diagram showing the correspondence between the axial position change relative to the emmetropic state and the refractive power of the simulated eye.

[0028] Figure 4 This is a schematic diagram of the pixel distribution of a spot image in every 5% grayscale interval in one embodiment of this application;

[0029] Figure 5 This is a schematic diagram illustrating the relationship between the measured refractive error and the power density of the light spot region in one embodiment of this application.

[0030] Figure 6 This is a schematic diagram of the two-dimensional pseudo-color distribution of a light spot image in one embodiment of this application;

[0031] Figure 7 This is a schematic diagram of the three-dimensional distribution of the light spot image in one embodiment of this application. Detailed Implementation

[0032] The present application will be further described below with reference to specific embodiments and accompanying drawings. It is to be understood that the illustrative embodiments of this disclosure are merely for explaining the present application and not for limiting it. Furthermore, for ease of description, the accompanying drawings show only the parts relevant to the present application, and not all of the structures or processes.

[0033] The following specific embodiments illustrate the implementation of this application. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. Although the description of this application is presented in conjunction with preferred embodiments, this does not mean that the features of this invention are limited to this embodiment. On the contrary, the purpose of describing the invention in conjunction with embodiments is to cover other options or modifications that may be derived based on the claims of this application. To provide a thorough understanding of this application, many specific details will be included in the following description. This application may also be implemented without using these details. Furthermore, to avoid confusion or obscuring the focus of this application, some specific details will be omitted in the description. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other.

[0034] Unless the context otherwise specifies, the terms “contains,” “has,” and “includes” are synonyms. The phrase “A / B” means “A or B.” The phrase “A and / or B” means “(A and B) or (A or B).”

[0035] It should be understood that although terms such as "first," "second," etc., may be used herein to describe various components, units, or data, these components, units, or data should not be limited by these terms. These terms are used merely to distinguish one feature from another. For example, without departing from the scope of the exemplary embodiments, a first feature may be referred to as a second feature, and similarly, a second feature may be referred to as a first feature.

[0036] It should be understood that although directional terms such as "up," "down," "left," and "right" may be used here to describe the positional relationship between the various components, these directional terms are only for the convenience of understanding and are not intended to limit the scope of protection of this application.

[0037] It should be noted that in this specification, similar reference numerals and letters in the accompanying drawings indicate similar items. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0038] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0039] Figure 1 This is a schematic diagram of a simulated eye structure for adjustable refractive power and real-time retinal imaging detection as described in this application. Figure 1As shown, the simulated eye includes multiple components arranged sequentially along the optical axis: a simulated cornea 2, a simulated lens 5, and an adjustable vitreous structure 7. Each component of the simulated eye is fixed within a rigid metal housing 1. The simulated eye further includes an aqueous humor cavity 3 and a pupil structure 4 located between the simulated cornea 2 and the simulated lens 5. The adjustable vitreous structure 7 includes a vitreous cavity 16 filled with a simulated vitreous, a curved retina 13 located at the end of the vitreous cavity 16, and a micrometer adjustment system for driving the curved retina 13 to move back and forth along the optical axis. An imaging module 12 is integrated into the foveal region of the curved retina 13. The imaging module includes a CMOS image sensor for real-time acquisition of light spot images on the foveal region. All components are arranged coaxially along the optical axis to ensure the consistency of the optical path. A corneal fixation ring 6 is provided at the front end of the housing 1 to fix the simulated cornea 2.

[0040] The simulated cornea 2 is a positive meniscus lens, designed with reference to the refractive power of the human cornea, with a target refractive power of +43D and an equivalent focal length of 23.4mm, used to simulate the main refractive characteristics of the human cornea. The simulated cornea 2 is preferably made of precision-machined quartz material. The advantages of quartz include its high hardness, ability to withstand a certain degree of wear, and longer service life. Additionally, quartz has a wide transmission spectrum, making it suitable for testing at more wavelengths. However, this invention is not limited to this; it may also be made of optical glass, such as BK7 glass, or optical plastics / resins, such as polymethyl methacrylate (PMMA).

[0041] Located behind the optical path of the simulated cornea 2 is the aqueous humor cavity 3, which is filled with anterior chamber fluid. The aqueous humor cavity 3 has a dimension of 3.0 mm along the optical axis. The anterior chamber fluid is preferably a refractive index matching fluid with a refractive index of 1.336, which is used to reproduce the optical environment of the anterior chamber of the human eye and reduce interface reflection.

[0042] The pupil structure 4 is located between the aqueous humor cavity 3 and the pseudo-lens 5. The pupil structure 4 adopts a structure with a central light-transmitting aperture, preferably made of aluminum alloy material with a light-transmitting aperture of 4mm. Its surface facing the light path is treated with black matte finish to absorb stray light deviating from the optical axis to the maximum extent and ensure clear entry pupil.

[0043] The simulated lens 5 is a biconvex lens designed with reference to the refractive power of the human eye's lens, with a target refractive power of +19.7D and an equivalent focal length of 50.8mm, used to simulate the refractive characteristics of the human eye's lens. Similar to the simulated cornea 2, the simulated lens 5 is preferably made of precision-machined quartz material, but the invention is not limited to this, and it may also be made of optical glass or optical plastic / optical resin.

[0044] In one embodiment of the present invention, a positive meniscus lens and a biconvex lens are used as refractive characteristic components in simulating the structure of the human eye. The refractive characteristics simulate the refractive tissue of the actual human eye, with an equivalent air focal length of 17mm.

[0045] The adjustable vitreous structure 7 includes a vitreous cavity 16 filled with a vitreous-like filling fluid. The vitreous cavity 16 is a chamber with a variable length along the optical axis, and the inner walls of the chamber are treated with a high-absorption matte black finish. The vitreous-like filling fluid filling the cavity 16 is preferably a refractive index-matching fluid with the same refractive index as the anterior chamber fluid, with a refractive index of 1.336, used to simulate the refractive characteristics of the human eye's vitreous body and reduce interfacial reflection loss. An injection tube 15 can also be provided on the side wall of the vitreous cavity 16 for injecting or withdrawing the vitreous-like filling fluid into or from the cavity, which can be used to change the axial length of the vitreous cavity 16 and also facilitates maintenance or media replacement.

[0046] The sealing ring 14 is used to ensure the airtightness of the aqueous humor cavity 3 and the vitreous cavity 16, and to prevent leakage of the anterior chamber fluid and the pseudo-vitreous filling fluid.

[0047] The curved retina 13 is located at the end of the vitreous cavity 16, where "end" refers to the end furthest from the light incident end in the optical axis direction, and is used to simulate the imaging surface of the human eye's retina. The curved retina 13 is preferably a spherical structure with a radius of 12.5 mm, but aspherical or freeform surfaces can also be used. To reduce the reflectivity of the curved retina 13, an optical thin film is sprayed onto its surface, which can effectively improve transmittance and reduce stray light.

[0048] The micrometer adjustment system includes a micrometer base 10 and a flat-head micrometer 11. The flat-head micrometer 11 is connected to the retina housing 9. By pushing or pulling the flat-head micrometer 11, the retina housing 9 and the curved retina 13 connected to it can be moved back and forth along the optical axis. The flat-head micrometer 11 can be adjusted manually or electrically by a stepper motor.

[0049] The flat-head micrometer 11 is a key component for simulating the refractive adjustment of the eye. By pushing and pulling the flat-head micrometer 11, the distance between the curved retina 13 and the posterior surface of the simulated lens 5 can be precisely adjusted, thereby simulating the axial length of the vitreous cavity. This adjustment directly affects the focal position of the retina, thus changing the refractive power. Through precise dynamic adjustment of the refractive power, different refractive states of the human eye, such as myopia, hyperopia, and emmetropia, can be simulated.

[0050] The diopter adjustment in this application is based on the principle of defocus control. Figure 2 A schematic diagram of the equivalent optical structure of the human eye to illustrate the principle of refractive adjustment. (e.g.) Figure 2As shown, the pseudo-cornea 2 and pseudo-lens 5 together constitute an optical system with a fixed equivalent focal length. The focal position is fixed relative to the posterior surface of the pseudo-lens 5. Here, AL (Axial Length) represents the axial length, the distance from the anterior surface of the cornea to the retina, and ACD (Anterior Chamber Depth) represents the anterior chamber depth. By changing the distance between the curved retina 13 and the pseudo-lens 5, the refractive state of the eye is simulated. When the curved retina 13 moves away from the pseudo-lens 5 along the optical axis, the focal point falls in front of the retina, resulting in myopia. When the curved retina 13 moves closer to the pseudo-lens 5 along the optical axis, the focal point falls behind the retina, resulting in hyperopia. When the distance between the curved retina 13 and the pseudo-lens 5 is appropriate, the focal point falls precisely on the retina, thus simulating emmetropia (normal vision).

[0051] Optical calculations show that when the equivalent focal length of the simulated eye is 17.00 mm, the simulated eye is in an emmetropic state. For every 1 mm movement of the flat-head micrometer 11, the central refractive power of the simulated eye changes by approximately 3.46 D. Figure 3 This is a schematic diagram illustrating the correspondence between the axial position change relative to the emmetropic state and the simulated eye's refractive power, as shown below. Figure 3 As shown, by continuously adjusting the axial length of the vitreous cavity, continuous and precise adjustment of the refractive power can be achieved. The scale value of the flat-head micrometer 11 can intuitively indicate the refractive power of the simulated eye corresponding to different advancement depths.

[0052] The imaging module 12 is tightly integrated into the foveal region of the curved retina 13 for real-time acquisition of light spot images on the curved retina 13. The imaging module 12 is preferably a high-resolution CMOS image sensor. The pixel size of the CMOS image sensor is 1.3 to 1.5 μm, preferably 1.4 μm, to simulate visual cells. The CMOS image sensor has a resolution of 1280(H) × 1024(V) and a frame rate of 60 frames per second.

[0053] In one embodiment of this application, an attenuation module is further provided between the imaging module 12 and the curved retina 13. The attenuation module is preferably a neutral density filter, which is used to adjust the light intensity of the light source, avoid the influence of excessive light source on the image sensor on the curved retina, ensure that the imaging module is not saturated, thereby effectively acquiring the distribution of the light field and ensuring the stability of the light spot image quality.

[0054] The simulated eye also includes an analysis module 17 electrically connected to the imaging module 12. The analysis module 17 is configured to process the acquired light spot image to extract optical parameters on the curved retina 13. The optical parameters that the analysis module 17 can extract include at least one of light intensity distribution, irradiance, light spot morphology, and single-pixel power density. Light intensity distribution refers to the spatial distribution of grayscale values ​​of each pixel within the light spot area, usually represented as a grayscale matrix, reflecting the relative intensity variation of light energy in space, and is the basic data for subsequent calculations of irradiance and single-pixel power density. Irradiance refers to the radiant power received per unit area, which can be calculated based on the light intensity distribution. For example, by combining known incident light power, sensor response curves, and pixel sizes for radiometric calibration, grayscale values ​​can be converted into actual power density values. Irradiance reflects the energy deposition rate at each point on the retinal surface and is a core indicator for assessing photobiological safety. Light spot morphology may include information such as the boundary contour of the light spot area, the area of ​​the light spot area, and the centroid position of the light spot area. Single-pixel power density refers to the light power received on the actual area corresponding to a single pixel. It is equal to the irradiance at that pixel multiplied by the area of ​​a single pixel. In this application, the size of each pixel of the CMOS sensor is approximately equal to the size of a visual cell to simulate a visual cell. Therefore, single-pixel power density can directly reflect the absolute light energy received by a single visual cell and can be used to determine whether there are risk points of excessively high local energy. The hardware of the analysis module 17 can be implemented using an external general-purpose computer or an embedded system. An external general-purpose computer can connect to the simulated eye through a data interface, such as a USB interface, and run image processing algorithms using CPU and / or GPU resources, suitable for high-computing-power scenarios in laboratories. With an embedded system, an ARM processor and an FPGA can be integrated into the simulated eye housing to achieve integrated simulation of the eye. Alternatively, a DSP or a high-performance MCU can also be used to implement the functions of the analysis module 17.

[0055] The analysis module 17 first preprocesses the spot image acquired by the CMOS image sensor, wherein the preprocessing of the spot image includes at least one of dark frame correction, background subtraction and median filtering.

[0056] Dark frame correction aims to eliminate the inherent dark current noise in CMOS image sensors. Dark current noise originates from electron-hole pairs generated by thermal excitation under no-light conditions, resulting in non-zero grayscale values ​​in pixel output, which intensifies with increasing temperature. To achieve accurate correction, one or more frames are acquired with the same exposure time and gain parameters under conditions simulating complete eye shading (no incident light). The average of these frames is then used as a standard dark frame image and pre-stored. During actual measurement, the currently acquired spot image is subtracted pixel-by-pixel from the pre-stored standard dark frame image; that is, the effective signal at each pixel location equals the original grayscale value minus the dark frame grayscale value. Due to the spatially non-uniform distribution of dark current noise, pixel-by-pixel subtraction effectively suppresses fixed-pattern noise, significantly improves the image signal-to-noise ratio, and provides a high-fidelity grayscale data foundation for subsequent quantitative optical parameter calculations.

[0057] Background subtraction aims to suppress the impact of stray ambient light on the image quality of light spots. Even after dark frame correction, the image may still contain background light intensity introduced by internal reflections of the optical system, light leakage from the external environment, or non-target light sources. This stray light, superimposed on the real light spot signal, interferes with the accuracy of light spot boundary recognition and grayscale statistics. The background subtraction in this application employs an adaptive background estimation method based on image edge regions. First, in the dark frame-corrected light spot image, pixels located in the outer edge regions of the image (theoretically, these regions should have no light spot signal, only background light intensity) are extracted, and the average grayscale value of these pixels is calculated as an estimate of the current ambient background light intensity. Subsequently, the estimated background light intensity is subtracted from the grayscale values ​​of all pixels in the entire image, and any negative values ​​are clamped to zero.

[0058] Median filtering is a non-linear smoothing technique used to remove isolated noise points in an image while preserving edge details. In a blemish image after background subtraction, there may be randomly occurring isolated bright or dark noise. Median filtering works by traversing each pixel of the blemish image through a sliding window (e.g., 3×3 or 5×5 pixels), sorting the gray values ​​of all pixels within the window by size, and taking the median value as the new gray value for that pixel. Since noise points typically exhibit extreme gray values ​​(extremely large or small), they are located at the ends of the sorted sequence. Therefore, median filtering effectively removes these outliers, while the pixel gray values ​​of normal blemish areas within the window are similar, making the median value close to the true signal.

[0059] In a preferred embodiment of this application, the preprocessing of the spot image includes three steps performed sequentially: dark frame correction, background subtraction, and median filtering. However, this application is not limited to this. The preprocessing of the spot image may include only one of the three steps: dark frame correction, background subtraction, and median filtering, or it may include any two of them. Moreover, the order of the two steps, dark frame correction and background subtraction, can be interchanged.

[0060] After obtaining the preprocessed spot image, the analysis module 17 performs spot region extraction and localization. First, an adaptive thresholding algorithm, such as the Sauvola method based on local neighborhood mean or the Otsu global thresholding method, is used to dynamically calculate the binarization threshold, avoiding the failure of the fixed threshold due to changes in the absolute brightness of the spot. Pixels with grayscale values ​​higher than the threshold are marked as foreground, and those lower are marked as background, generating an initial binary image. Then, morphological operations are applied, including opening (erosion followed by dilation) to eliminate fine noise and gaps, and closing (dilation followed by erosion) to close tiny holes inside the spot, thus obtaining a continuous and complete spot region. Based on this, a contour tracking algorithm, such as the Suzuki algorithm, is used to extract the boundary contours of all foreground regions. To remove artifacts caused by residual noise or stray light, the geometric feature parameters of each boundary contour are calculated, including area (number of pixels), perimeter, and roundness (4π·area / perimeter). 2 The aspect ratio is also considered. Only boundary contours that meet preset geometric feature criteria, such as area greater than or equal to the minimum threshold, circularity greater than or equal to 0.6, and aspect ratio close to 1, are retained as valid spot regions. The corresponding binary region is the spot mask, which is used for subsequent quantitative calculation of optical parameters.

[0061] Based on the obtained spot mask, coarse localization is first performed using the gray-level centroid method to obtain the preliminary coordinates of the energy center of the spot region. That is, the geometric center is calculated using gray-level values ​​as weights, and the calculation formula is as follows:

[0062] ,

[0063] Where G(x,y) represents the gray value of the preprocessed spot image at pixel (x,y), and M(x,y) represents the spot mask, (x... c ,y c The coordinates () represent the coarse centroid coordinates of the calculated spot region. This calculation method uses grayscale values ​​as weights to perform a weighted average of the coordinates of all spot pixels. Since the energy distribution of the spot is usually approximately symmetrical, the weighting center basically coincides with the energy center. This calculation method is simple and fast, but its accuracy is limited by pixel dispersion, typically only achieving pixel-level accuracy.

[0064] Based on the coarse localization, a local window is further extracted in the center of the spot region and fitted with a two-dimensional Gaussian surface to obtain the sub-pixel-level localization result of the spot centroid (x0, y0), which can significantly improve the localization repeatability and accuracy. The two-dimensional Gaussian function model is as follows, which is used to fit the gray-level distribution within the local window of the spot region:

[0065] ,

[0066] Where B represents the Gaussian peak amplitude, corresponding to the maximum gray value at the center of the light spot. and represents the standard deviation of the light spot in the x and y directions, respectively, reflecting the width or dispersion of the light spot. C represents the background baseline gray value (usually subtracted through preprocessing, but retained in the model to compensate for residual background). Fitting a two-dimensional Gaussian surface can improve the positioning accuracy from pixel level to sub-pixel level (up to 0.01-0.1 pixels).

[0067] Based on the centroid position of the spot mask and the spot region, the analysis module 17 performs quantitative calculations of optical parameters. The physical area A of the spot region can be obtained through the relationship between the number of pixels in the spot mask and the known pixel size p:

[0068] ,

[0069] Where N p This indicates the total number of pixels in the light spot mask.

[0070] Analysis module 17 can also calculate the cumulative grayscale value, maximum grayscale value, and average grayscale value of all pixels within the light spot area to obtain the overall optical energy index. (Cumulative grayscale value) It can be calculated using the following formula:

[0071] .

[0072] To quantify the light intensity distribution within the spot area, the analysis module 17 can also divide the gray values ​​of the spot mask area into a preset number of gray-level intervals, such as 5% intervals, and count the number of pixels in each corresponding gray-level interval and the cumulative gray value in that interval. Figure 4 As shown, the system can identify the energy proportion and total energy value of the high grayscale core region based on grayscale interval statistical results, and can further calculate the power density of a single pixel. The above statistical analysis results can quantitatively reflect the non-uniformity of light intensity within the spot area and the characteristics of concentrated light intensity at the center, providing a basis for spot distribution pattern determination, optical system calibration, and safety evaluation of low-intensity phototherapy.

[0073] Figure 5 This is a schematic diagram illustrating the relationship between the measured refractive error and the power density of the light spot region in one embodiment of this application. Using the simulated eye of this application, the refractive power of the simulated eye can be precisely adjusted using a flat-head micrometer 11. Simultaneously, the imaging module 12 and analysis module 17 can analyze and extract the optical parameters on the curved retina 13 under different refractive power conditions in real time, thereby intuitively demonstrating the imaging differences on the curved retina between minute refractive power differences. For example, as... Figure 5As shown, the average power density of the light spot region on the curved retina 13 is displayed under different refractive deviations (where a refractive deviation of 0 indicates emmetropia). The refractive deviation and power density exhibit an almost linear relationship. The goodness of fit R between the actual measured values ​​(the solid line connecting the measurement points) and the fitted line (represented by the dashed line) is shown. 2 It reached 0.9538.

[0074] After quantitatively extracting the optical parameters, the analysis module 17 can also perform light field visualization reconstruction and morphological analysis based on the grayscale matrix of the spot area, specifically including the generation of a two-dimensional pseudo-color distribution map and a three-dimensional intensity distribution map. Here, the grayscale matrix refers to a two-dimensional array obtained from a CMOS image sensor after preprocessing such as dark frame correction, background subtraction, and median filtering. Each element of this grayscale matrix corresponds to the grayscale value of a pixel in the spot image, and its value is proportional to the light intensity received by that pixel. The row and column dimensions of the grayscale matrix are consistent with the sensor resolution (e.g., 1280×1024), where the elements located within the spot mask area reflect the light energy distribution on the simulated fovea of ​​the retina. Figure 6 As shown, the two-dimensional pseudo-color distribution map applies gradient color mapping to the grayscale matrix, mapping different grayscale levels to color levels, thus visually displaying the spatial distribution of light intensity on the retinal surface. Figure 7 As shown, the three-dimensional intensity distribution map uses an interpolation algorithm to spatially reconstruct the grayscale matrix, generating a three-dimensional surface map that reflects the local light intensity variation trend.

[0075] The high-precision adjustable refractive power and real-time retinal imaging simulation eye provided in this application achieves two core technological innovations: First, it drives the curved retina to move precisely along the optical axis through a micrometer adjustment system. Each 1mm movement corresponds to a 3.46D change in refractive power, and it can be continuously adjusted within the range of -6.92D to +6.92D, with an adjustment step of 0.0346D, which is significantly better than the ±0.5D accuracy of conventional simulation eyes and meets the ±0.12D calibration requirements of high-end ophthalmic equipment; Second, it integrates a CMOS sensor in the fovea region of the retina. An image sensor (preferably with a pixel size of 1.4 μm) acquires spot images in real time. The analysis module performs preprocessing such as dark frame correction, background subtraction, and median filtering. It extracts the spot mask through adaptive threshold segmentation and achieves sub-pixel-level centroid localization (accuracy 0.01-0.1 pixels) by combining the gray-level centroid method with two-dimensional Gaussian surface fitting. Then, it quantitatively calculates the spot area, cumulative gray value, peak value, and average gray value. Furthermore, it statistically quantifies the energy concentration through gray-level intervals, ultimately generating a two-dimensional pseudo-color image and a three-dimensional intensity distribution map for light field visualization. This simulated eye integrates high-precision refractive accommodation and quantitative detection of the retinal surface light field, providing a high-precision and high-reliability comprehensive testing platform for ophthalmic equipment calibration, optical system aberration diagnosis, and safety assessment of low-intensity red light therapy.

[0076] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions 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. Where there is no conflict, the embodiments and features described in the embodiments of this application can be combined with each other. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A simulated eye with adjustable diopter and real-time retinal imaging detection, characterized in that, include: The structure consists of a corneal-like structure, a lens-like structure, and an adjustable vitreous body, arranged sequentially along the optical axis. The adjustable vitreous structure includes: a vitreous cavity filled with a vitreous body, a curved retina disposed at the end of the vitreous cavity, and a micrometer adjustment system for driving the curved retina to move back and forth along the optical axis. The concave region of the curved retina integrates an attenuation module and an imaging module for real-time acquisition of light spot images on the concave region. The simulated eye also includes an analysis module electrically connected to the imaging module. The analysis module is configured to process the acquired light spot image to extract optical parameters on the curved retina. The optical parameters include at least one of light intensity distribution, irradiance, light spot morphology, and single pixel power density.

2. The simulated eye according to claim 1, characterized in that, The simulated cornea is a positive meniscus lens, and the simulated lens is a biconvex lens.

3. The simulated eye according to claim 1, characterized in that, The simulated eye also includes an aqueous humor cavity located between the simulated cornea and the simulated lens, the aqueous humor cavity being filled with anterior chamber fluid, and both the anterior chamber fluid and the simulated vitreous body being refractive index-matching fluids with a refractive index of 1.

336.

4. The simulated eye according to claim 1, characterized in that, The curved retina is a spherical structure with a curvature radius of 12.5 mm, and its surface is coated with an optical film to reduce reflectivity.

5. The simulated eye according to claim 1, characterized in that, The imaging module includes a CMOS image sensor with a pixel size of 1.3 to 1.5 μm.

6. The simulated eye according to claim 1, characterized in that, For every 1 mm movement of the micrometer adjustment system, the refractive power of the simulated eye changes by 3.46 D, thereby achieving continuous adjustment of the refractive power within the range of -6.92 D to +6.92 D.

7. The simulated eye according to claim 1, characterized in that, The analysis module performs the following processing on the spot image: The light spot image is preprocessed, and the preprocessing includes at least one of dark frame correction, background subtraction, and median filtering; The spot region is extracted and the spot mask is obtained based on an adaptive threshold algorithm combined with morphological operations; Based on the light spot mask, the centroid of the light spot region is coarsely located using the gray-scale centroid method, and then the sub-pixel level coordinates of the centroid of the light spot region are obtained using the two-dimensional Gaussian surface fitting method.

8. The simulated eye according to claim 7, characterized in that, The analysis module is also configured to: calculate the area of ​​the spot region based on the spot mask, and statistically analyze the cumulative gray value, maximum gray value, and average gray value within the spot region.

9. The simulated eye according to claim 7, characterized in that, The analysis module is further configured to: divide the gray values ​​within the area of ​​the light spot mask into multiple gray ranges according to their size, and count the number of pixels and the cumulative gray value in each gray range to quantify the concentration of energy distribution within the light spot area.

10. The simulated eye according to claim 7, characterized in that, The analysis module is also configured to generate a two-dimensional pseudo-color distribution map for displaying the spatial distribution of light intensity and / or a three-dimensional intensity distribution map for displaying the local light intensity variation trend, based on the grayscale matrix of the light spot region.

Images