Distributed cornea surface morphology measurement method and device

Abstract

The invention provides a distributed cornea surface morphology measurement method and device. The method comprises the following steps: acquiring special videos of left and right cornea reflection Placido plate patterns; performing image fusion on the video, and screening out a plurality of clear images meeting analysis requirements based on deep model learning; calculating positions of cornea vertexes and reflection points of the screened images, constructing a point cloud model covering the whole cornea, and calculating key parameters of the cornea; statistical analysis is carried out on cornea key parameters of a plurality of images, and a comprehensive result with the minimum dispersion is screened out. According to the invention, through application of a distributed architecture and a deep learning model, the accuracy and reliability of cornea surface morphology measurement can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of ophthalmic medical detection, and particularly relates to a distributed method and device for measuring corneal surface topography. Background Art

[0002] Corneal topography measurement is an examination method for obtaining corneal surface topography and curvature data through non-invasive techniques, which can provide key parameters such as corneal curvature (K value), astigmatism, refractive power, etc. It is widely used in preoperative evaluation of refractive surgery (such as excluding keratoconus), fitting of contact lenses, diagnosis of corneal diseases (ulcers / scars, etc.), monitoring of postoperative effects, and precise fitting of frame glasses. Among the currently used measurement principles, Placido disk reflection imaging has become the dominant technology in the mid- to low-end market due to its low cost and simple structure; although Scheimpflug imaging and OCT technology can obtain the anterior and posterior corneal surfaces and thickness data, they are limited by high cost, slow measurement, and complex structure, and their application range is relatively narrow. Existing devices based on the Placido disk principle have significant defects: the large disk design leads to a long measurement distance, and insufficient patient cooperation easily causes missing monitoring areas and data errors; it cannot handle image blurring caused by eye movement or blinking, and has poor dynamic adaptability; the measurement results rely on manual screening, with low automation (about 3 minutes for each measurement); the device is large in size (usually over 50 cm), requires professional operation, and lacks portability and ease of use. Existing technologies urgently need to be improved in terms of data accuracy, efficiency, and operation convenience to meet the clinical requirements for precise and efficient corneal topography measurement. Summary of the Invention

[0003] In view of the above problems, the present invention proposes a distributed method for measuring corneal surface topography, which solves the deficiencies of traditional corneal topographic devices in aspects such as data acquisition speed, integrity, dynamic adaptability, and automated analysis.

[0004] The specific solution of the present invention is as follows: A distributed method for measuring corneal surface topography includes the following steps: S1, obtaining videos of Placido disk patterns reflected by the left and right corneas; S2, performing image fusion on the videos, and screening out several images that meet the analysis requirements from the continuous videos based on a deep learning model. The specific operations are as follows: S21, converting the videos into an image sequence; S22, performing normalization processing on the images; S23, performing noise reduction on the images through a Gaussian filtering algorithm; S24, calculating the blur coefficient of the noise-reduced images using the Tenengrad algorithm, and initially screening the images according to the blur coefficient; S25. Input the pre-screened images into the CNN network for scoring, and further screen the images according to the scoring results. S3. Calculate the positions of the corneal apex and each reflection point for the screened images, construct a point cloud model covering the entire cornea using the apex and each reflection point, and further calculate the key parameters of the cornea based on the point cloud model. S4. Through statistical analysis of the key parameters of the cornea in multiple images, use the arithmetic mean and standard deviation methods to screen out the comprehensive result with the smallest dispersion.

[0005] Processing each frame of the image in the video, such as extraction, conversion to grayscale image, noise reduction, and calculation of the blur coefficient, can effectively remove noise and interference information, improve the quality and clarity of the image, and thus provide a more accurate image data basis for subsequent calculation of the corneal apex and reflection point positions. Through the synergistic effect of the Tenengrad algorithm and the CNN network, clear images that best meet the analysis requirements can be automatically screened out, avoiding the subjectivity and errors of manual screening, improving the automation degree and efficiency of measurement, ensuring the quality and representativeness of the selected images, and providing a strong guarantee for subsequent corneal parameter calculation.

[0006] Further, the S22 specifically includes: S221. Calculate the mean brightness, minimum brightness, and maximum brightness of each image. S222. Calculate the adaptive brightness adjustment parameter for each image according to the calculation results of S221 and the preset brightness parameters. S223. Perform normalization processing on the brightness of each image.

[0007] By calculating the mean brightness, minimum brightness, and maximum brightness of each image, and calculating the adaptive brightness adjustment parameter for each image according to the preset brightness parameters, so as to perform normalization processing on the brightness of each image, it can effectively solve the problem of large brightness differences between different images, make the image sequence have more uniform and stable brightness characteristics, facilitate subsequent image processing and analysis, further improve the accuracy and reliability of corneal measurement, and provide a better image data basis for subsequent calculation of the corneal apex and reflection point positions.

[0008] Further, the S24 specifically includes: S241. Use the Sobel operator to calculate the horizontal gradient and vertical gradient of the image respectively, and calculate the square of the gradient amplitude of each pixel in the image based on the horizontal gradient and vertical gradient of the image. S242. Take the 50% area in the center of the image as the central area, accumulate the square values of the gradient amplitudes of all pixels in the central area, and obtain the blur coefficient of the image.

[0009] The horizontal gradient and vertical gradient of the image are calculated using the Sobel operator, and based on this, the square of the gradient magnitude of each pixel is calculated. Then, the central 50% area of the image is taken as the central area for accumulation to obtain the blur coefficient of the image, which can more accurately evaluate the clarity of the image. This method focuses on the central area of the image, avoiding the influence of edge interference factors such as eyelashes and eyelids, making the calculation of the blur coefficient more accurate and reliable. Thus, clear images can be more effectively screened out, providing high-quality image data for subsequent corneal parameter calculation and improving the accuracy of the measurement results.

[0010] Further, the S3 specifically includes: S31, estimating the corneal vertex position using the Klyce method; S32, calculating the axial curvature radius of each angular meridian based on the corneal vertex position and the geometric relationship of adjacent reflection points, and then determining the three-dimensional coordinates of all reflection points; S33, establishing a one-to-one mapping relationship between the feature points on the Placido disk image and the corneal surface reflection points, and constructing a point cloud model covering the entire cornea using the three-dimensional coordinates of all reflection points; S34, calculating the key parameters of the cornea according to the point cloud model, including corneal curvature, corneal astigmatism, corneal refractive power, corneal irregularity index, corneal height map, corneal diameter, corneal eccentricity, corneal surface topography, corneal vertex position.

[0011] Estimating the corneal vertex position using the Klyce method and combining the geometric relationship of adjacent reflection points to obtain the positions of each reflection point can accurately determine the key geometric features of the corneal surface. By establishing a one-to-one mapping relationship between the feature points on the Placido disk image and the corneal surface reflection points, constructing a point cloud model covering the entire cornea, and calculating various key parameters including corneal curvature, corneal astigmatism, and corneal refractive power based on this.

[0012] A distributed corneal surface morphology measurement device, characterized by including: An acquisition device, which is used to acquire a dedicated video of the Placido disk pattern reflected by the left and right corneas and transmit the acquired video to the terminal; A terminal, which is used to upload the video received from the acquisition device to the server; display the data received from the server; The server is used to perform image fusion on the received video, and based on a deep learning model, select multiple clear images that meet the analysis requirements from the continuous video; calculate the positions of the corneal apex and each reflection point for the selected images, use these points to construct a point cloud model covering the entire cornea, and further calculate the key parameters of the cornea based on the point cloud model; screen out the comprehensive result with the smallest dispersion through statistical analysis of the key parameters of the cornea in multiple images; and send the result to the terminal.

[0013] By setting the acquisition device, terminal, and server, this distributed corneal surface morphology measurement device realizes the automation and high efficiency of the corneal measurement process. The acquisition device can collect special videos of the corneal reflection Placido disk patterns of the left and right eyes and transmit them to the terminal. The terminal is responsible for video uploading and data display, while the server completes core tasks such as image fusion, clear image screening, calculation of the positions of the corneal apex and reflection points, and calculation of key parameters. This modular design not only improves the stability and reliability of the device but also enables the modules to work together synergistically, giving full play to their respective advantages, effectively improving the efficiency and accuracy of corneal measurement, and providing a convenient and efficient corneal morphology measurement solution for clinical applications.

[0014] Further, the acquisition device specifically includes: The projection unit, including two Placido disks, is used to measure the topographic data of the left and right corneas; The imaging unit is used to collect videos of the reflected light points on the corneal surface and detect the pupil position in real time; The adjustment unit is used to automatically adjust the position of the Placido disk according to the pupil position to achieve centering alignment with the pupil; The communication unit is used to transmit the collected video to the terminal.

[0015] The projection unit, imaging unit, adjustment unit, and communication unit in the acquisition device cooperate with each other to achieve the efficient acquisition and transmission of the topographic data of the left and right corneas. The two Placido disks of the projection unit are used to measure the topographic data of the left and right corneas. The imaging unit can collect videos of the reflected light points on the corneal surface and detect the pupil position in real time. The adjustment unit automatically adjusts the position of the Placido disk according to the pupil position to achieve centering alignment with the pupil, ensuring the accuracy and reliability of the measurement. The communication unit is responsible for transmitting the collected video to the terminal, ensuring the timely transmission and subsequent processing of the data. This design makes the data acquisition process more automated and intelligent, improves the efficiency and accuracy of the measurement, and provides high-quality data support for the entire corneal surface morphology measurement device.

[0016] Furthermore, both of the two Placido disks use conical Placido disks with a diameter of 50 mm - 80 mm, which include at least 16 annuli. The annuli are composed of light-transmitting micropores. Each annulus contains 16 to 36 micropores, and the micropores are distributed in concentric circles. The spacing between adjacent annuli gradually increases.

[0017] Using conical Placido disks with a diameter of 50 mm - 80 mm, and each annulus contains 16 to 36 light-transmitting micropores. The micropores are distributed in concentric circles and the spacing between adjacent annuli gradually increases. This design can effectively reduce the size of the Placido disk while ensuring the measurement accuracy, making it more portable and easier to operate. In addition, this structure also helps to improve the quality and stability of the corneal reflection image, providing a better basis for subsequent image processing and parameter calculation. By optimizing the structure and parameters of the Placido disk, not only the cost and complexity of the device are reduced, but also the measurement efficiency and reliability are improved, making the corneal surface topography measurement more convenient and fast, providing strong support for clinical applications and daily monitoring.

[0018] Furthermore, the imaging unit includes an LED light board. After the pupil is centered and aligned, the adjustment unit activates the LED light board to emit white light. The white light passes through the micropores and enters the Placido disk, forming a reflection image composed of light points on the human cornea.

[0019] The imaging unit includes an LED light board. After the pupil is centered and aligned, the adjustment unit activates the LED light board to emit white light. The white light passes through the micropores and enters the Placido disk, forming a reflection image composed of light points on the human cornea. This design can provide a uniform and stable illumination source, ensuring the quality and clarity of the corneal reflection image. By controlling the start-stop and light intensity of the LED light board, it can better adapt to different eye conditions and measurement requirements, improving the flexibility and adaptability of the measurement. At the same time, this lighting method also helps to improve the contrast of the corneal reflection image and the extraction accuracy of feature points, providing a higher-quality data basis for subsequent image processing and corneal parameter calculation, further enhancing the accuracy and reliability of the corneal surface topography measurement.

[0020] Compared with the prior art, the beneficial effects of the present invention are as follows: The method and device can effectively improve the accuracy and reliability of corneal surface morphology measurement. By obtaining dedicated videos of the corneal reflection Placido disk patterns of the left and right eyes and through multiple steps of processing, clear images that meet the requirements can be accurately screened out, and a learning deep model is used to ensure the accuracy of screening. Subsequently, a point cloud model is constructed based on the corneal vertex and each reflection point, which can comprehensively and accurately calculate the key parameters of the cornea. The distributed architecture enables efficient data processing, and the acquisition device, terminal, and server work together to improve the overall measurement efficiency. The use of a specific Placido disk and related algorithms further enhances the scientificity and practicality of the measurement. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present drawings or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following-described drawings are only some embodiments of the present drawings. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on the structures shown in these drawings.

[0022] Figure 1 is a flowchart of the method of the present invention; Figure 2 is a projection diagram of a light spot on the human cornea; Figure 3 is a schematic diagram of a point cloud; Figure 4 is a schematic diagram of the structure of the acquisition device; Figure 5 is a schematic diagram of the structure of a Placido disk + infrared camera. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0023] In order to make the objectives, technical solutions, and advantages of the present invention clearer, the following will describe and explain the present invention in conjunction with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention. Based on the embodiments provided by the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

[0024] Embodiment 1: As Figure 1 shown, the present invention provides a distributed method for corneal surface morphology measurement, which specifically includes the following steps: S1, obtain dedicated videos of the corneal reflection Placido disk patterns of the left and right eyes.

[0025] A dedicated video of the corneal reflection Placido disk patterns of the left and right eyes is acquired simultaneously using a specially designed binocular aperture cone-shaped double Placido disk acquisition device. This device is small in size, light in weight, and easy to operate. It can be used by the examinee himself / herself, and the acquired video is transmitted in real time to a general terminal (such as a mobile phone or a PC) through a USB or WiFi interface.

[0026] S2. Perform image fusion on the video, and based on a deep learning model, select several images that meet the analysis requirements from the continuous video, that is, clear and correct images.

[0027] The terminal then uploads the video to the server for further processing. The server first performs image fusion on the left and right eye videos, and then based on a deep learning model optimized by the Tenengrad algorithm, automatically selects multiple clear images that meet the analysis requirements from the continuous video to ensure the effectiveness and integrity of the data.

[0028] Specifically, S2 includes the following steps: S21. Video image extraction: Extract each frame in the video and convert the video into a lossless image sequence.

[0029] S22. Image dimensionality reduction processing: Perform normalization processing on the images.

[0030] The original video image data is illuminated by an infrared light source and then collected by an infrared camera. The collected image is a monochromatic grayscale image with only the infrared light channel. To unify the brightness of all infrared images, it is necessary to perform normalization on each image, that is, calculate the average brightness, maximum, and minimum brightness of the image sequence, and then adaptively adjust the brightness of each image according to the calculated average brightness, maximum, and minimum brightness to ensure that the brightness of each picture is close to or equal to the average brightness.

[0031] The specific operations are as follows: 1) Calculate the average brightness, minimum brightness, and maximum brightness of each image: Average brightness : ； Minimum brightness ； Maximum brightness ； where H and W respectively represent the length and width of the image; (x, y) represents the coordinate position; represents the image at the pixel value at (x, y); Indicates the sum of all pixel values.

[0032] 2) Calculate the adaptive brightness adjustment parameter for each image according to the above calculation result and the preset brightness parameter of the acquisition device: Set the target brightness fixed value according to the preset brightness parameter ; Calculate the image brightness translation amount: ； Determine the brightness translation range: ， ； Among them, Indicates the brightness translation amount; Indicates the minimum translation amount; Indicates the maximum translation amount.

[0033] Perform normalization processing on the brightness of each image: ; Among them, Indicates the brightness after translation at (x, y) in the image; The clamp function is used to limit the pixel value within the range of 0 to 255.

[0034] S23, Image noise reduction processing: Denoise the gray image through the Gaussian filtering algorithm, remove the noise that may be generated during the compression and transmission of the video, and retain as many details of the Placido disk image as possible.

[0035] Generally, the noise generated by compression in an image belongs to the normal distribution (Gaussian distribution), so Gaussian filtering is adopted. The core algorithm is as follows: ; Among them, σ is the standard deviation, which is used to control the width of the Gaussian curve and is an important parameter for controlling the smoothness of the filter; μ is the mean value of the Gaussian function; p(z) represents the probability density at position z; e is the base of the natural logarithm; π is the circumference ratio.

[0036] S24, Use the Tenengrad algorithm to calculate the blur coefficient of the denoised image, and preliminarily screen the image according to the blur coefficient. Specifically, it includes the following steps: 1) Use the Sobel operator to calculate the horizontal gradient of the image respectively and the vertical gradient ; 2) Calculate the squared gradient magnitude of each pixel in the image ； 3) Take the 50% area at the center of the image as the central area to avoid interference from eyelashes, eyelids, etc.; 4) Accumulate the squared gradient magnitude values of all pixels in the central area to obtain the Tenengrad score Score, which is the blurring coefficient of the image. The higher the score, the clearer the image. The following formula is used for calculation: .

[0037] S25. Input the pre-screened images into the CNN network for scoring, and re-screen the images according to the scoring results.

[0038] The Tenengrad algorithm only judges the clarity of the image and does not judge the content and quality of the image. In order to avoid the subsequent algorithm using the Tenengrad alone for clarity judgment and mistakenly taking incorrect images (such as closed eyes, squints, eyelash occlusion, equipment failures, etc.) as input options to affect the calculation results, a small convolutional neural network (Convolutional Neural Networks, CNN) is introduced for training to cooperate with the Tenengrad algorithm to exclude clear but incorrect unqualified images: 1) Use a pre-trained lightweight CNN as the backbone network (ResNet-18); 2) Design a binary classification network to output a single probability value (the probability that the image is qualified); 3) Collect Placido disk projection videos at 120fps (frames per second) to ensure coverage of different ages, genders, and corneal states; 4) Collect 120,000 frames of original images containing samples with typical quality problems, such as blink occlusion, motion blur, ring breakage, abnormal reflection, etc.; 5) Manually label 70% of the images as the training set, and the remaining 30% of the images as the validation and test set. After training, work together with the Tenengrad algorithm.

[0039] To speed up the calculation, the CNN network is slower than the Tenengrad algorithm. Therefore, first score the images using the Tenengrad algorithm, select the top 10% of the images with the highest scores and input them into the CNN network, and finally select the 10 images with higher scores in the CNN network as the basis for subsequent calculations.

[0040] S3. Calculate the positions of the corneal apex and each reflection point for the selected images, use these points to construct a point cloud model covering the entire cornea, and further calculate the key parameters of the cornea based on the point cloud model.

[0041] For each selected image, the device automatically calculates the positions of the corneal apex and each reflection point, and uses these points to construct a point cloud model covering the entire cornea. Based on this model, key parameters including corneal curvature, corneal astigmatism, corneal refractive power, corneal irregularity index, corneal height map, corneal diameter, corneal eccentricity, and corneal surface topographic map are further calculated.

[0042] Currently, the Klyce method is used to estimate the position of the corneal apex. For the first ring (the innermost ring) of the Placido disk, the distance from the end face of the outermost ring of the Placido disk to the camera CCD is basically equal to the distance from the Placido disk surface to the corneal apex. Therefore, the radius of the first ring on the Placido image directly corresponds to the radius of the first reflection point on the cornea, and the radius of curvature of the corneal apex can be estimated using the position of the first reflection point on the cornea. : ; Among them, is the radius of the first reflection point on the i-th meridian; n is the number of meridians.

[0043] After determining the radius of curvature of the corneal apex according to Klyce, the position of the corneal apex is obtained according to the following formula : ; Among them, wd is the distance from the end face of the outermost ring of the Placido disk to the camera CCD; is the radius of the innermost ring of the Placido disk.

[0044] Based on the position of the corneal apex and the geometric relationship of adjacent reflection points, the axial radius of curvature of each angular meridian is calculated, and then the three-dimensional coordinates of all reflection points are determined.

[0045] 1) Axial radius of curvature : It reflects the curvature of the corneal surface along the axial direction and is defined as the distance between the corneal reflection point and the intersection of the normal line at this point and the axis: ; ; ; ; ; ; Among them, is the reflection point to be calculated; is the reflection point The radius of the corresponding ring of the Placido disk; They are the incident light length and the reflected light length respectively; is the included angle between the incident light and the reflected light; is the reflection point of the tangent slope angle.

[0046] Instantaneous curvature radius: It refers to the curvature radius of a local area on the corneal surface. By fitting three adjacent reflection points to the same circular arc, the curvature characteristics of this local area can be determined. Therefore, it can be obtained that: ; Among them, are three adjacent reflection points; is the instantaneous curvature radius; For Local curvature center.

[0047] Based on the above formula, solve the 。

[0048] Such as Figure 3 As shown, establish a one-to-one mapping relationship between the feature points on the Placido disk image and the corneal surface reflection points, and use the three-dimensional coordinates of all reflection points to construct a point cloud model covering the entire cornea.

[0049] The number of data points in the point cloud model depends on the number of feature points formed by the micropores on the specially designed Placido disk.

[0050] Calculate the key parameters of the cornea according to the point cloud model, including corneal curvature, corneal astigmatism, corneal refractive power, corneal irregularity index, corneal height map, corneal diameter, corneal eccentricity, corneal surface topographic map, corneal vertex position and other corneal surface morphology data.

[0051] S4. Through statistical analysis of the key parameters of the cornea in multiple images, use the arithmetic mean and standard deviation methods to select the comprehensive result with the smallest dispersion.

[0052] Embodiment 2: The present invention also provides a distributed corneal surface morphology measurement device, which specifically includes the following 3 modules: An acquisition device, which is used to acquire a dedicated video of the Placido disk pattern reflected by the left and right corneas and transmit the acquired video to the terminal.

[0053] Specifically, as Figure 4 shown, the acquisition device includes: The projection unit includes two Placido disks, the first Placido disk 1 and the second Placido disk 11, which are used to measure the topographic data of the left and right corneas.

[0054] Specifically, as Figure 5 shown, both Placido disks use conical Placido disks with a diameter of 50 mm - 80 mm, and at least include 16 annuli. The annuli are composed of light-transmitting micropores 10. Each annulus contains 16 to 36 micropores 10. The diameter of each micropore 10 is 0.5 ± 0.1 mm. The micropores 10 are distributed in concentric circles, and the spacing between adjacent annuli gradually increases.

[0055] The imaging unit is used to collect the video of the reflected light points on the corneal surface and detect the pupil position in real time.

[0056] Specifically, the imaging unit includes: At least two cameras, which collect the video of the reflected light points on the corneal surface at a frame rate of 120 frames per second; The infrared guiding device includes an infrared illumination light source and a pupil detection camera 2, which is used to detect the pupil position in real time; The LED light board 5. After the pupil is centered and aligned, the adjustment unit activates the LED light board 5 to emit white light. The white light passes through the micropores 10 and enters the Placido disk, forming a reflected image composed of light points on the human cornea.

[0057] The adjustment unit is used to automatically adjust the position of the Placido disk according to the pupil position to achieve centering and alignment with the pupil.

[0058] Specifically, the adjustment unit includes a guide rail 3, a motor 4 and a control board 6.

[0059] The communication unit is used to transmit the collected video to the terminal.

[0060] The acquisition device is integrated in a handheld portable device. By physically contacting and aligning the eyes of the subject, corneal topography measurement can be performed on both eyes simultaneously, and it communicates with the terminal through a USB interface 7 or a WiFi interface 8. It does not have a calculation and measurement function itself.

[0061] Specifically, the working process of the acquisition device is as follows: 1) First, ask the person to be measured to press the measurement hole of the acquisition device 12 tightly against both eyes. At this time, the eyes of the person to be measured will be covered by the first Placido disk 1 and the second Placido disk 11. The person to be measured looks at the infrared camera 9. The infrared camera 9 is provided with illuminating red light for gaze guidance, with a wavelength of about 850 nm, which can be observed by the human eye as a gaze guide and can also perform infrared illumination without affecting the pupil image.

[0062] 2) The control board 6 of the acquisition device is built-in with a pupil position recognition algorithm, which cyclically detects the images captured by the pupil detection camera 2.

[0063] 3) After identifying the pupil pattern in the captured image, the control board 6 drives the adjustment motor 4 to adjust the positions of the left and right Placido disks through the control rail 3, so that the left and right Placido disks are centered with the left and right eye pupils.

[0064] 4) When the pupils are centered, the control board 6 will activate multiple built-in LED light boards 5 of the device. The LED light boards 5 use white LED light sources, and the wavelengths do not conflict with the illumination red light built into the infrared camera 9. After the LED light boards 5 are activated, the light will penetrate through the micropores 10 on the specially designed conical Placido disk.

[0065] 5) The white light emitted by the LED light boards 5 passes through the micropores 10 and enters the Placido disk, and finally forms a reflected image composed of light points on the human cornea.

[0066] The Placido disk structure of this device is different from the traditional Placido disk design. The optical design mode of the traditional Placido disk generally adopts a reflective design (a high-reflection coating is applied to the white area of the disk surface) or a transmissive design (with the help of a backlight source, a transparent material is used to cover an opaque coating). The purpose of the traditional design is to ensure uniform brightness distribution of each ring during projection, and prevent the decrease of the image signal-to-noise ratio caused by uneven light intensity, thereby affecting the corneal curvature calculation. However, because the traditional design needs to ensure the machining accuracy of concentric rings (too thin lines cannot be machined), the line widths are relatively large. Since the line widths are large, in order to ensure the line density of the corneal projection, the Placido disk cannot be reduced and close to the pupil, which is one of the reasons why the volume of the traditional corneal topographer cannot be reduced.

[0067] Because the circular ring design of the traditional Placido disk is ultimately to extract the feature points on the corneal projection in the circular ring to calculate corneal parameters. This device directly simplifies the circular ring into light points to directly form feature points, and then uses the feature points formed by the light points to fit a single circular ring through ellipse calculation, and then forms corneal surface parameters from multiple fitted circular rings, and finally uses them for corneal morphology calculation. In this way, it can not only replace the circular ring design of the traditional Placido disk, but also significantly reduce the processing difficulty and accuracy. It is no longer necessary to use light-transmitting materials, and ordinary ABS injection molding or 3D printing can be used, which the traditional Placido disk design cannot achieve, and it can greatly reduce the volume of the Placido disk.

[0068] 6) After the projection images of the corneas of both eyes are formed in the Placido disk, the imaging unit records a video at a speed of 120 frames per second for 5 - 10 seconds, and then transmits the video through the USB interface 7 or WiF iThe interface 8 transmits to the terminal.

[0069] The terminal is used to upload the video received from the acquisition device to the server and display the data received from the server.

[0070] The terminal can be any device that supports USB communication or WIFI. After downloading the video generated by the acquisition device through the standard protocol, it is then uploaded to the server. After the server completes the video analysis and issues the calculation results of the corneal shape, the terminal can display them. Compared with traditional corneal topographic measurement devices, in this process, the terminal does not have any major computing work and has almost no requirements for hardware. Ordinary mobile phones or computers can be used. In this way, there is no need to equip the traditional corneal topographic measurement device with a computing device and a display device. This change can not only effectively reduce the volume of the corneal topographic measurement device, but also greatly reduce the cost and volume of the corneal measurement device.

[0071] The server is used to perform image fusion on the received video, screen out multiple clear images that meet the analysis requirements from the continuous video based on the deep learning model, calculate the positions of the corneal vertex and each reflection point for the selected images, use these points to construct a point cloud model covering the entire cornea, further calculate the key parameters of the cornea based on the point cloud model, screen out the comprehensive result with the smallest dispersion through statistical analysis of the corneal key parameters of multiple images, and send the result to the terminal.

[0072] In order to reduce the difficulty of corneal topographic measurement, the present invention places the calculation part in the traditional corneal topographic measurement in the server and abandons the method of manually screening qualified pictures in the traditional measurement. The server performs tasks such as measurement data quality analysis, vertex fitting, and corneal topographic parameter calculation, solving the problems of large data errors, poor dynamic adaptability, and low automation in traditional corneal topographic devices.

[0073] Because of the adoption of a double-disk structure and the use of an innovative automatic image screening mechanism, the single-eye measurement speed is increased by 20%, and the binocular measurement speed is increased by more than 50%. At the same time, due to the new Placido disk structure, the production cost is reduced by 90% compared with the traditional Placido disk. Because of the reduction of processing errors in the new structure Placido disk, the reliability during data measurement is increased by 5-10%.

[0074] It should be noted that the present invention is not limited to the above embodiments. The above embodiments are only examples, and embodiments having the same constitution and achieving the same effects as the technical idea within the technical scope of the present invention are all included in the technical scope of the present invention. In addition, within the scope of not departing from the gist of the present invention, various modifications that can be conceived by those skilled in the art to the embodiments, and other ways constructed by combining some constituent elements in the embodiments are also included in the scope of the present invention.

Claims

1. A distributed method for measuring corneal surface morphology, characterized in that It includes the following steps: S1. Obtain the video of the Placido disk pattern of the left and right corneal reflexes; S2. Perform image fusion on the video, and based on the deep learning model, screen out several images that meet the analysis requirements from the continuous video. The specific operations are as follows: S21. Convert the video into an image sequence; S22. Perform normalization processing on the images; S23. Denoise the images through the Gaussian filtering algorithm; S24. Use the Tenengrad algorithm to calculate the blur coefficient of the denoised images, and preliminarily screen the images according to the blur coefficient; S25. Input the preliminarily screened images into the CNN network for scoring, and screen the images again according to the scoring results; S3. Calculate the positions of the corneal apex and each reflection point for the screened images, use the apex and each reflection point to construct a point cloud model covering the entire cornea, and further calculate the key parameters of the cornea based on the point cloud model; S4. Through statistical analysis of the key parameters of the cornea in multiple images, use the arithmetic mean and standard deviation methods to screen out the comprehensive result with the smallest dispersion.

2. The distributed corneal surface morphology measurement method according to claim 1, characterized in that The specific content of S22 includes: S221. Calculate the mean brightness, minimum brightness, and maximum brightness of each image; S222. Calculate the adaptive brightness adjustment parameter for each image according to the calculation results of S221 and the preset brightness parameters; S223. Perform normalization processing on the brightness of each image.

3. A distributed corneal surface morphology measurement method according to claim 1, characterized in that, The specific content of S24 includes: S241. Use the Sobel operator to calculate the horizontal gradient and vertical gradient of the image respectively, and calculate the square of the gradient amplitude of each pixel in the image based on the horizontal gradient and vertical gradient of the image; S242. Take the 50% area in the center of the image as the central area, accumulate the square values of the gradient amplitudes of all pixels in the central area, and obtain the blur coefficient of the image.

4. A distributed corneal surface morphology measurement method according to claim 1, characterized in that The specific content of S3 includes: S31. Estimate the position of the corneal apex using the Klyce method; S32. Based on the geometric relationship between the corneal apex position and adjacent reflection points, calculate the axial curvature radius of each angular meridian, and then determine the three-dimensional coordinates of all reflection points; S33. Establish a one-to-one mapping relationship between the feature points on the Placido disk image and the corneal surface reflection points, and use the three-dimensional coordinates of all reflection points to construct a point cloud model covering the entire cornea; S34. Calculate the key parameters of the cornea according to the point cloud model, including corneal curvature, corneal astigmatism, corneal refractive power, corneal irregularity index, corneal height map, corneal diameter, corneal eccentricity, corneal surface topographic map, corneal apex position.

5. A distributed corneal surface morphology measurement device, characterized in that, It includes: An acquisition device, which is used to acquire the dedicated video of the Placido disk pattern of the left and right corneal reflexes and transmit the acquired video to the terminal; A terminal, which is used to upload the video received from the acquisition device to the server; Display the data received from the server; A server, which is used to perform image fusion on the received video, and based on the deep learning model, screen out multiple clear images that meet the analysis requirements from the continuous video; Calculate the positions of the corneal apex and each reflection point for the screened images, use these points to construct a point cloud model covering the entire cornea, and further calculate the key parameters of the cornea based on the point cloud model; By statistically analyzing the key corneal parameters of multiple images, the comprehensive result with the smallest dispersion is selected; and the result is sent to the terminal.

6. The distributed corneal surface topography measurement device according to claim 5, characterized in that The acquisition device specifically includes: A projection unit, including two Placido disks, for measuring the corneal topographic data of the left and right eyes; An imaging unit, for collecting the video of the reflected light points on the corneal surface and detecting the pupil position in real time; An adjustment unit, for automatically adjusting the position of the Placido disk according to the pupil position to achieve centering alignment with the pupil; A communication unit, for transmitting the collected video to the terminal.

7. A distributed corneal surface topography measurement device according to claim 6, characterized in that, Both of the two Placido disks use conical Placido disks with a diameter of 50 mm - 80 mm, and at least include 16 annuli. The annuli are composed of light-transmitting micropores. Each annulus contains 16 to 36 micropores, and the micropores are distributed in concentric circles. The spacing between adjacent annuli gradually increases.

8. A distributed corneal surface topography measurement device according to claim 7, wherein, The imaging unit includes an LED light board. After the pupil is centered and aligned, the adjustment unit activates the LED light board to emit white light. The white light passes through the micropores and enters the Placido disk, forming a reflected image composed of light points on the human cornea.

Images