Method for generating keratosclera feature map based on Scheimpflug image

Abstract

The invention relates to a method for generating a keratosclera feature map based on a Scheimpflug image. The method comprises the following steps: step 1, acquiring data of a Scheimpflug cross-section image and a front image; 2, extracting the edge of the keratosclera tissue; step 3, correcting image distortion; 4, reconstructing three-dimensional data of the keratosclera; 5, correcting coordinate axes of the three-dimensional data of the keratosclera; step 6, analyzing and expressing the three-dimensional data of the keratosclera; step 7, generating a keratosclera rise topographic map; 8, calculating the characteristics of the keratosclera data; and 9, generating a relative height topographic map of the keratosclera. The method solves the problems of single measurement angle, large error and few directions in the technology of measuring the three-dimensional geometrical morphology data of the single keratosclera before wearing the selected film, and overcomes the defect of high exposure of a Scheimpflug image in a sclera area through a high-low double exposure technology; establishing a keratosclera three-dimensional data structure through an exposure fusion technology and an image segmentation technology; and quantizing the optimal fitting spherical surface of the cornea and the optimal fitting spherical surface of the sclera to generate a rise and height topographic map.

Description

Technical Field

[0001] The present invention relates to the field of ophthalmic clinical medical technology, and in particular to a method for quantitatively processing corneal and scleral data in the field of scleral lens fitting. Background Art

[0002] Scleral lenses are large-diameter rigid gas-permeable contact lenses worn during the day. They are a type of contact lens suitable for people with high refractive errors, irregular corneas, and moderate to severe dry eyes. During wearing, the contact position, area, angle, and sub-lens gap between the scleral lens and the sclera play a key role in achieving the expected efficacy. Therefore, accurate quantification of the three-dimensional geometric morphology of the single cornea and sclera is an important data basis for early lens selection.

[0003] Under current technology, slit lamps are mostly used to observe and measure the scleral sagitta at different diameters at a certain angle. This measurement method can only obtain the scleral sagitta value of a single section, and the selection of manual measurement boundaries is also uncertain, introducing active errors; or anterior segment OCT technology (optical coherence tomography) is used to obtain anterior segment cross-sectional images at multiple angles, and then measure the scleral sagitta. This measurement method increases the number of cross-sectional angles, but is limited by the sweep speed and the cross-sectional angles are usually only 10-15, and the fixed angles cannot obtain the scleral geometry in all directions. Summary of the invention

[0004] Aiming at the problems of single angle, large error and few orientations in the measurement of three-dimensional geometric data of the cornea before lens selection, a method for generating cornea sclera feature map based on Scheimpflug image was proposed.

[0005] The technical solution of the present invention is:

[0006] A method for generating a corneal scleral feature map based on a Scheimpflug image comprises the following steps:

[0007] Step 1: Scheimpflug cross-sectional image and frontal image data acquisition

[0008] In order to collect the full-field corneal and scleral data of the anterior segment and effectively separate the corneal and scleral data, the full-field corneal and scleral data of the anterior segment includes corneal data and scleral data. A three-dimensional anterior segment analysis device based on the Scheimpflug image principle is used with two exposure settings in the same cross-sectional area;

[0009] Step 2: Extraction of corneal scleral tissue edge

[0010] Extract the anterior surface edge of the cornea from the corneal image data and the sclera edge from the sclera image data, and fuse the edge structures of the two images at the same cross-sectional position to obtain the corneal and sclera edge data;

[0011] Step 3: Image distortion correction

[0012] The edge data in the actual space coordinates is obtained by distortion correction of Scheimpflug image data;

[0013] Step 4: Reconstruction of corneal and scleral 3D data

[0014] According to the angular position relationship of each section in step one, the real corneoscleral edge data of each section position in step three are rotated around the vertex, and all the real corneoscleral edge data coordinates are transformed into three-dimensional space coordinates;

[0015] Step 5: Corneoscleral 3D data coordinate axis correction

[0016] Select the acquisition sections in the 0 degree direction and the 90 degree direction respectively, identify the angle between the corneoscleral boundary line and the horizontal direction of the image on the single image, and then generate a three-dimensional rotation matrix. Apply the matrix to all the three-dimensional data points obtained in step 4 to obtain the three-dimensional point data of the corneosclera after coordinate correction;

[0017] Step 6: Analysis and expression of corneal 3D data

[0018] The three-dimensional data points obtained in step 4 are numerically and analytically fitted using the Zernike polynomial fitting method;

[0019] Step 7: Generation of corneoscleral sagittal height topography

[0020] According to the numerical analytical fitting result obtained in step six, an equally spaced corneoscleral sagittal height data matrix is ​​generated, and then the matrix is ​​converted into a corneoscleral sagittal height topography map by a color bar mapping method;

[0021] Step 8. Calculation of corneal and scleral data features

[0022] The cornea-sclera three-dimensional data point set is divided into cornea three-dimensional data points and sclera three-dimensional data points, and the cornea data points and sclera data points are respectively used to fit a sphere or an ellipsoid using the least squares principle to obtain the best fitting spherical or ellipsoidal surface of the cornea and the best fitting spherical or ellipsoidal surface of the sclera, as well as the corresponding mathematical expressions of the spherical or ellipsoidal surfaces;

[0023] Step 9: Generation of corneal sclera relative height topography

[0024] According to the corneal best-fit sphere or ellipsoid obtained in step eight, generate a corneal best-fit sphere or ellipsoid data matrix of the same size and data interval as the corneal scleral vector height data matrix in step seven. Similarly, generate a scleral best-fit sphere or ellipsoid data matrix to establish a corneal-scleral relative height data matrix; similarly, generate the corneal-scleral relative height data matrix into a corneal-scleral relative height topography map through the color strip mapping method.

[0025] Further, the method specifically includes the following steps:

[0026] Step 1: Scheimpflug cross-sectional image and frontal image data acquisition

[0027] In order to collect the full-field corneal and scleral data of the anterior segment and effectively separate the corneal and scleral data, the full-field corneal and scleral data of the anterior segment includes corneal data and scleral data. A three-dimensional anterior segment analysis device based on the Scheimpflug image principle is used to adopt a two-exposure setting in the same cross-sectional area. The light intensity of the first exposure is recorded as s1, and the exposure time is t1. When the exposure is completed, the second exposure is immediately performed, and the light intensity is adjusted to s2, and the exposure time is t2; where s1<s2,t1> t2; and adjust the parameters appropriately by observing the imaging effects of the corneal area and the scleral area of ​​the two images; in one acquisition process, the front camera is used to monitor the forward field of view of the eye, and the front image data is triggered when the corneal vertex and the image center coincide and the image plane is clear; at the same time, the rotation interval angle of the Scheimpflug camera is set to Δd degrees, and within the range of 180 degrees, the acquisition positions are 0 degrees, 1*Δd degrees, 2*Δd, 3*Δd degrees, etc., and a total of N=180 / Δd corneal data images, i.e., the first exposure images, and the same number of scleral data images, i.e., the second exposure images, are collected; establish the angle position of each section and the relationship diagram;

[0028] Step 2: Extraction of corneal scleral tissue edge

[0029] Extract the anterior surface edge of the cornea from the corneal image data, extract the sclera edge from the sclera image data, and fuse the edge structures of the two images at the same cross-sectional position to obtain corneosclera edge data; during the processing, the overlapping part of the intersection of the cornea and the sclera can be processed by limiting the boundary direction derivatives and height differences of the intersection area; the other images collected at different cross-sectional positions are processed in the same way;

[0030] Step 3: Image distortion correction

[0031] Scheimpflug image distortion elimination is achieved by:

[0032] Step 3.1, place the standard checkerboard plate on the device's slit projection plane, and adjust the front and rear distance from the device until the observation window displays a complete checkerboard plate image;

[0033] Step 3.2: Locate the corner points on the grid plate image and record the position coordinates p of each corner point with the upper left corner of the image as (0,0) i (x,y);

[0034] Step 3.3, taking the center position of the grid plate as the actual coordinate (0,0), combined with the actual size of the grid plate and all the corner points obtained, calculate the single mapping matrix H1, the H1 matrix is ​​3*3 in size, reflecting the positional relationship between the image coordinates and the actual coordinates;

[0035] Step 3.4, the corneoscleral edge data segmented in step 2 at all cross-sectional angles are applied to the monographic matrix H1 to obtain the real corneoscleral edge data after object-image conversion; thus, the edge data in the actual space coordinates is obtained from the Scheimpflug image data through distortion correction;

[0036] Step 4: Reconstruction of corneal and scleral 3D data

[0037] According to the angular position relationship of each section in step 1, the real corneoscleral edge data of each section position in step 3 are rotated around the vertex, and all the real corneoscleral edge data coordinates are transformed into three-dimensional space coordinates to obtain the number of all three-dimensional data points Nu=N*M; where N is the number of sections, and M is the number of corneoscleral edges in a single section image;

[0038] Step 5: Corneoscleral 3D data coordinate axis correction

[0039] Select the acquisition section at 0 degree and 90 degree respectively, identify the angle between the corneoscleral boundary line and the horizontal direction of the image in a single image; record the above angles of the two section images at 0 degree and 90 degree as αx and αy respectively, and then generate a three-dimensional rotation matrix Applying matrices H2 and H3 to all the three-dimensional data points obtained in step 4 to obtain three-dimensional point data of the cornea after coordinate correction;

[0040] Step 6: Analysis and expression of corneal sclera 3D data

[0041] The three-dimensional data points obtained in step 4 are numerically and analytically fitted using the Zernike polynomial fitting method;

[0042] Zernike polynomials The form is as follows:

[0043]

[0044] Where ρ and θ represent the radius and angle, which are used to determine the position of any point in the polar coordinate system. n represents the order, and the value of m is -n, -n+2, -n+4, ..., n. is the normalization factor;

[0045] δ m0 =1 when m = 0,δ m0 =0 when m≠0

[0046] is a radially distributed polynomial:

[0047]

[0048] The fitted corneoscleral surface is expressed by the weighted Zernike polynomial W(ρ,θ):

[0049]

[0050] Convert the three-dimensional data points in step 4 from rectangular coordinates to polar coordinates, and use the least squares method to fit to obtain the coefficients of each basis term in the Zernike expression. In this way, the spatially resolved Zernike expression of the corneoscleral surface is analytically described;

[0051] Step 7: Generation of corneoscleral sagittal height topography

[0052] According to the expression W(ρ,θ) obtained in step 6, and the polar coordinate and rectangular coordinate conversion relationship x=p*cos(θ), y=ρ*sin(θ), the data range is defined as diameter D, the data interval Δs=D / nu, where nu is the number of data points in a single direction, and the corneoscleral sagittal height data matrix M0 with equal intervals in the x and y directions is generated. The size of M0 is nu*nu, and the data exceeding the diameter D is represented as invalid data; then the matrix M0 is generated into a corneoscleral sagittal height topography map by the color bar mapping method, that is, a certain range of values ​​is represented by a certain color;

[0053] Step 8. Calculation of corneal and scleral data features

[0054] During the rotation acquisition process of the Scheimpflug image, the forward camera provides alignment and plane dimension information, identifies the boundary range of the cornea and sclera on the image, and maps the cornea and sclera edge information of the front camera to the cornea and sclera 3D point data obtained in step 5 through the position matching of the cornea vertex. The cornea data is inside the cornea and the sclera data is outside the cornea and sclera edge. In this way, the cornea and sclera 3D data point set is divided into cornea and sclera 3D data points. The cornea and sclera data points are used to fit the sphere or ellipsoid using the least squares principle, respectively, to obtain the best fitting sphere or ellipsoid of the cornea BFS_C and the best fitting sphere or ellipsoid of the sclera BFS_S, as well as the corresponding mathematical expressions of the sphere or ellipsoid.

[0055] Step 9: Generation of corneal sclera relative height topography

[0056] According to the corneal best-fitting sphere or ellipsoid obtained in step eight, generate the corneal best-fitting sphere or ellipsoid data matrix M1 with the same size and data interval as the corneoscleral sagittal height data matrix M0 in step seven, and set the data outside the corneoscleral boundary range to invalid values; similarly, according to the scleral best-fitting sphere or ellipsoid obtained in step eight, generate the scleral best-fitting sphere or ellipsoid data matrix M2 with the same size and data interval as the corneoscleral sagittal height data matrix M0 in step seven, and set the data within the corneoscleral boundary range to invalid values; at the same time, the data outside the data range diameter D defined in step seven in the two data matrices generated are also set as invalid data; define the corneoscleral relative height data matrix M3 = M0-M1-M2; and similarly generate the matrix M3 as a corneoscleral relative height topography map through the color strip mapping method.

[0057] The beneficial effects of the present invention are:

[0058] The method of the present invention is based on the Scheimpflug image principle (an imaging technology for corneal topography detection), and overcomes the high exposure defect of Scheimpflug images in the sclera area through high and low double exposure technology; a single scan can be used to collect multiple corneal and scleral images in the same spatial coordinate system in conjunction with the motor position trigger; a three-dimensional data structure of the cornea and sclera is established through exposure fusion technology and image segmentation technology; the coordinate reference system is automatically corrected, and the visual axis is switched to the optical axis direction to establish a cornea analytical expression; the best fitting sphere of the cornea and the best fitting sphere of the sclera (or ellipsoid) are quantified to generate a sagittal height and height topography. The quantification of these data features provides sufficient data basis for the adaptation of personalized scleral lenses. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 It is the first exposure image at a certain cross-section position of the present invention;

[0060] Figure 2 It is the second exposure image of a certain cross-sectional position of the present invention;

[0061] Figure 3 This is a schematic diagram of the angle position relationship of a single acquisition section of the present invention;

[0062] Figure 4 This is the data diagram of the corneoscleral fusion edge of the present invention;

[0063] Figure 5 This is a schematic diagram of the corner points of the black and white checkered plate of the present invention;

[0064] Figure 6 It is a schematic diagram of the object-image conversion of the monoscopic moment distortion correction array of the present invention;

[0065] Figure 7 It is a data diagram of the corneoscleral edge in the physical coordinate system of the present invention;

[0066] Figure 8 It is a schematic diagram of the visual axis and the optical axis of the present invention;

[0067] Fig. 9 It is a three-dimensional corneoscleral edge data map in the correction coordinate system of the present invention;

[0068] Fig.10 It is the corneoscleral sagittal height topography of the present invention;

[0069] Fig.11 This is a diagram of the relative height of the cornea and sclera of the present invention. DETAILED DESCRIPTION

[0070] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0071] A method for generating a corneal scleral feature map based on a Scheimpflug image comprises the following steps:

[0072] Step 1: Scheimpflug cross-sectional image and frontal image data acquisition

[0073] In order to collect the full-field corneal and scleral data of the anterior segment and effectively separate the corneal and scleral data, the full-field corneal and scleral data of the anterior segment includes corneal data and scleral data. A three-dimensional anterior segment analysis device based on the Scheimpflug image principle is used to adopt two exposure settings in the same cross-sectional area. The light intensity of the first exposure is recorded as s1, and the exposure time is t1. When the exposure is completed, the second exposure is immediately performed, and the light intensity is adjusted to s2 and the exposure time is t2. Among them, s1<s2,t1> t2. And adjust the parameters appropriately by observing the imaging effects of the cornea and sclera areas of the two images, such as Figure 1 , Figure 2 As shown. In one acquisition process, the front camera is used to monitor the forward field of view of the eye. When the corneal vertex and the image center coincide and the image plane is clear, the front image data is triggered to be captured; at the same time, the rotation interval angle of the Scheimpflug camera is set to Δd degrees. Within the range of 180 degrees, the acquisition positions are 0 degrees, 1*Δd degrees, 2*Δd, 3*Δd degrees, etc., and a total of N=180 / Δd corneal data images (first exposure images) and the same number of sclera data images (second exposure images) are collected. Establish the angle position of each section and the relationship diagram, as shown in Figure 3 As shown in the figure, n0, n1, n n / 2 Represent the first, second and n / 2th cross-sectional exposure positions respectively.

[0074] Step 2: Extraction of corneal scleral tissue edge

[0075] The corneal anterior surface edge is extracted from the corneal image data, and the scleral edge is extracted from the scleral image data. The edge structures of the two images at the same cross-sectional position are fused to obtain the corneal and scleral edge data, such as Figure 4 As shown in the figure, for the overlapping part of the cornea and sclera, the repeated edge information in the two tissue structures can be processed by limiting the boundary direction derivatives and height differences of the intersection area. The same processing is performed on the other images collected at different cross-sectional positions.

[0076] Step 3: Image distortion correction

[0077] The Scheimpflug image has the characteristics of large depth of field and clear imaging, but due to the particularity of the optical path design, the distortion is serious, and effective positive means are needed to eliminate these distortions. This can be achieved by:

[0078] Step 3.1, place the standard checkerboard plate on the device's slit projection plane, and adjust the front and rear distance from the device until the observation window displays a complete checkerboard plate image;

[0079] Step 3.2: Locate the corner points on the grid plate image and record the position coordinates p of each corner point with the upper left corner of the image as (0,0) i (x,y), such as Figure 5 shown.

[0080] Step 3.3, take the center position of the grid plate as the actual coordinate (0,0), combine the actual size of the grid plate and all the corner points obtained, and calculate the single mapping matrix H1. The H1 matrix is ​​3*3 in size and reflects the positional relationship between the image coordinates and the actual coordinates. The process is as follows: Figure 6 .

[0081] Step 3.4: Apply the corneoscleral edge data segmented in step 2 at all cross-sectional angles to the monographic matrix H1 to obtain the real corneoscleral edge data after object-image conversion. In this way, the edge data in the actual space coordinates is obtained from the Scheimpflug image data through distortion correction.

[0082] The actual data of the cornea and sclera at a certain cross-section position is as follows Figure 7 shown.

[0083] Step 4: Reconstruction of corneal and scleral 3D data

[0084] According to the angular position relationship of each section in step one, the real corneoscleral edge data of each section position in step three is rotated around the vertex, and all the real corneoscleral edge data coordinates are transformed into three-dimensional space coordinates to obtain the number of all three-dimensional data points Nu = N*M; where N is the number of sections, and M is the number of corneoscleral edges in a single section image.

[0085] Step 5: Corneoscleral 3D data coordinate axis correction

[0086] During the image acquisition phase, the human eye is focused on the fixation point of the device. At this time, the visual axis of the human eye is coaxial with the rotation axis of the device. Therefore, the vertical direction of all the collected cross-sectional images is the visual axis of the human eye. When characterizing the corneoscleral data, we expect its vertical direction to be the optical axis of the corneoscleral geometry itself. Select the 0-degree and 90-degree directions to collect cross sections, and identify the angle between the corneoscleral boundary line and the horizontal direction of the image in a single image, such as Figure 8 The above angles of the two cross-sectional images at 0 degrees and 90 degrees are recorded as αx and αy respectively, and then a three-dimensional rotation matrix is ​​generated. All the three-dimensional data points obtained in step 4 are applied to matrices H2 and H3 to obtain the three-dimensional point data of the cornea after coordinate correction. The three-dimensional cornea and sclera edge data in the corrected coordinate system are as follows: Fig. 9 shown.

[0087] Step 6: Analysis and expression of corneal 3D data

[0088] To facilitate feature extraction, and because the geometric morphology of the cornea is similar to the concept of wavefront in optics, the three-dimensional data points obtained in step 4 are numerically fit using the Zernike polynomial fitting method. The Zernike polynomial fitting method is a pattern-based wavefront reconstruction method that is fitted using a set of basis functions. It was first proposed by F. Zernike and later derived by ABBhatia and E. Wolf.

[0089] Zernike polynomials The form is as follows:

[0090]

[0091] Where ρ and θ represent the radius and angle, which are used to determine the position of any point in the polar coordinate system. n represents the order, and the value of m is -n, -n+2, -n+4, ..., n. is the normalization factor.

[0092] δ m0 =1 when m = 0,δ m0 =0 when m≠0

[0093] is a radially distributed polynomial:

[0094]

[0095] The fitted corneoscleral surface is expressed by the weighted Zernike polynomial W(ρ,θ):

[0096]

[0097] Convert the three-dimensional data points in step 4 from rectangular coordinates to polar coordinates, and use the least squares method to fit to obtain the coefficients of each basis term in the Zernike expression. In this way, the spatially resolved Zernike expression of the corneoscleral surface is analytically described.

[0098] Step 7: Generation of corneoscleral sagittal height topography

[0099] According to the expression W(ρ,θ) obtained in step 6, and the polar coordinate and rectangular coordinate conversion relationship x=p*cos(θ), y=ρ*sin(θ), define the data range as diameter D, data interval Δs=D / nu, where nu is the number of data points in a single direction, and generate the corneoscleral sagittal height data matrix M0 with equal intervals in the x and y directions. The size of M0 is nu*nu, and the data exceeding the diameter D is represented as invalid data. Then, the matrix M0 is generated into a corneoscleral sagittal height topography through the color bar mapping method (i.e., a certain range of values ​​is represented by a certain color). The corneoscleral sagittal height topography result is as follows: Fig.10 shown.

[0100] Step 8. Calculation of corneal and scleral data features

[0101] During the rotational acquisition of the Scheimpflug image, the forward camera provides alignment and plane dimension information to identify the boundary range of the cornea and sclera on the image (such as Figure 3 The outer black ring position of the mark) and the corneal scleral edge information of the front camera is mapped to the corneal scleral 3D point data obtained in step 5 through the position matching of the corneal vertex. The mapping result is as follows Fig. 9 As shown in the figure, the data inside the corneoscleral edge is corneal data, and the data outside the corneoscleral edge is scleral data, so that the corneoscleral three-dimensional data point set is divided into corneal three-dimensional data points and scleral three-dimensional data points. The corneal data points and scleral data points are used to perform sphere (or ellipsoid) fitting using the least squares principle, respectively, to obtain the corneal best fitting sphere (or ellipsoid) BFS_C, and the sclera best fitting sphere (or ellipsoid) BFS_S, as well as the corresponding spherical (or ellipsoid) mathematical expressions.

[0102] Step 9: Generation of corneal sclera relative height topography

[0103] According to the corneal best-fitting sphere (or ellipsoid) obtained in step eight, generate the corneal best-fitting sphere (or ellipsoid) data matrix M1 with the same size and data interval as the corneoscleral sagittal height data matrix M0 in step seven, and set the data outside the corneoscleral boundary range to invalid values; similarly, according to the scleral best-fitting sphere (or ellipsoid) obtained in step eight, generate the scleral best-fitting sphere (or ellipsoid) data matrix M2 with the same size and data interval as the corneoscleral sagittal height data matrix M0 in step seven, and set the data within the corneoscleral boundary range (sclera data matrix M2, the edge data is actually invalid values) to invalid values; at the same time, the two data matrices generated have data outside the data range diameter D defined in step seven set to invalid data. Define the corneoscleral relative height data matrix M3 = M0-M1-M2; similarly, generate the matrix M3 into a corneoscleral relative height topography map through the color bar mapping method. Fig.11 shown.

[0104] The above-mentioned embodiment only expresses one implementation mode of the present invention, and its description is relatively specific and detailed, but it cannot be understood as limiting the scope of the invention patent. It should be pointed out that for ordinary technicians in this field, several modifications and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be based on the attached claims.

Claims

1. A method for generating a corneal scleral feature map based on a Scheimpflug image, characterized in that: The following steps are involved: Step 1: Scheimpflug cross-sectional image and frontal image data acquisition In order to collect the full-field corneal and scleral data of the anterior segment and effectively separate the corneal and scleral data, the full-field corneal and scleral data of the anterior segment includes corneal data and scleral data. A three-dimensional anterior segment analysis device based on the Scheimpflug image principle is used with two exposure settings in the same cross-sectional area; Step 2: Extraction of corneal scleral tissue edge Extract the anterior surface edge of the cornea from the corneal image data and the sclera edge from the sclera image data, and fuse the edge structures of the two images at the same cross-sectional position to obtain the corneal and sclera edge data; Step 3: Image distortion correction The edge data in the actual space coordinates is obtained by distortion correction of Scheimpflug image data; Step 4: Reconstruction of corneal and scleral 3D data According to the angular position relationship of each section in step one, the real corneoscleral edge data of each section position in step three are rotated around the vertex, and all the real corneoscleral edge data coordinates are transformed into three-dimensional space coordinates; Step 5: Corneoscleral 3D data coordinate axis correction Select the acquisition sections in the 0 degree direction and the 90 degree direction respectively, identify the angle between the corneoscleral boundary line and the horizontal direction of the image on the single image, and then generate a three-dimensional rotation matrix. Apply the matrix to all the three-dimensional data points obtained in step 4 to obtain the three-dimensional point data of the corneosclera after coordinate correction; Step 6: Analysis and expression of corneal sclera 3D data The three-dimensional data points obtained in step 4 are numerically and analytically fitted using the Zernike polynomial fitting method; Step 7: Generation of corneoscleral sagittal height topography According to the numerical analytical fitting result obtained in step six, an equally spaced corneoscleral sagittal height data matrix is ​​generated, and then the matrix is ​​converted into a corneoscleral sagittal height topography map by a color bar mapping method; Step 8. Calculation of corneal and scleral data features The cornea-sclera three-dimensional data point set is divided into cornea three-dimensional data points and sclera three-dimensional data points, and the cornea data points and sclera data points are respectively used to fit a sphere or an ellipsoid using the least squares principle to obtain the best fitting spherical or ellipsoidal surface of the cornea and the best fitting spherical or ellipsoidal surface of the sclera, as well as the corresponding mathematical expressions of the spherical or ellipsoidal surfaces; Step 9: Generation of corneal sclera relative height topography According to the corneal best-fit sphere or ellipsoid obtained in step eight, generate a corneal best-fit sphere or ellipsoid data matrix of the same size and data interval as the corneal scleral vector height data matrix in step seven. Similarly, generate a scleral best-fit sphere or ellipsoid data matrix to establish a corneal-scleral relative height data matrix; similarly, generate the corneal-scleral relative height data matrix into a corneal-scleral relative height topography map through the color strip mapping method.

2. The method for generating a corneal scleral feature map based on a Scheimpflug image according to claim 1, characterized in that: The specific steps include: Step 1: Scheimpflug cross-sectional image and frontal image data acquisition In order to collect the full-field corneal and scleral data of the anterior segment and effectively separate the corneal and scleral data, the full-field corneal and scleral data of the anterior segment includes corneal data and scleral data. A three-dimensional anterior segment analysis device based on the Scheimpflug image principle is used to adopt a two-exposure setting in the same cross-sectional area. The light intensity of the first exposure is recorded as s1, and the exposure time is t1. When the exposure is completed, the second exposure is immediately performed, and the light intensity is adjusted to s2, and the exposure time is t2; where s1<s2,t1> t2; and adjust the parameters appropriately by observing the imaging effects of the corneal area and the scleral area of ​​the two images; in one acquisition process, the front camera is used to monitor the forward field of view of the eye, and the front image data is triggered when the corneal vertex and the image center coincide and the image plane is clear; at the same time, the rotation interval angle of the Scheimpflug camera is set to Δd degrees, and within the range of 180 degrees, the acquisition positions are 0 degrees, 1*Δd degrees, 2*Δd, 3*Δd degrees, etc., and a total of N=180 / Δd corneal data images, i.e., the first exposure images, and the same number of scleral data images, i.e., the second exposure images, are collected; establish the angle position of each section and the relationship diagram; Step 2: Extraction of corneal scleral tissue edge Extract the anterior surface edge of the cornea from the corneal image data, extract the sclera edge from the sclera image data, and fuse the edge structures of the two images at the same cross-sectional position to obtain corneosclera edge data; during the processing, the overlapping part of the intersection of the cornea and the sclera can be processed by limiting the boundary direction derivatives and height differences of the intersection area; the other images collected at different cross-sectional positions are processed in the same way; Step 3: Image distortion correction Scheimpflug image distortion elimination is achieved by: Step 3.1, place the standard checkerboard plate on the device's slit projection plane, and adjust the front and rear distance from the device until the observation window displays a complete checkerboard plate image; Step 3.2: Locate the corner points on the grid plate image and record the position coordinates p of each corner point with the upper left corner of the image as (0,0) i (x,y); Step 3.3, taking the center position of the grid plate as the actual coordinate (0,0), combined with the actual size of the grid plate and all the corner points obtained, calculate the single mapping matrix H1, the H1 matrix is ​​3*3 in size, reflecting the positional relationship between the image coordinates and the actual coordinates; Step 3.4, the corneoscleral edge data segmented in step 2 at all cross-sectional angles are applied to the monographic matrix H1 to obtain the real corneoscleral edge data after object-image conversion; thus, the edge data in the actual space coordinates is obtained from the Scheimpflug image data through distortion correction; Step 4: Reconstruction of corneal and scleral 3D data According to the angular position relationship of each section in step 1, the real corneoscleral edge data of each section position in step 3 are rotated around the vertex, and all the real corneoscleral edge data coordinates are transformed into three-dimensional space coordinates to obtain the number of all three-dimensional data points Nu=N*M; where N is the number of sections, and M is the number of corneoscleral edges in a single section image; Step 5: Corneoscleral 3D data coordinate axis correction Select the acquisition section at 0 degree and 90 degree respectively, identify the angle between the corneoscleral boundary line and the horizontal direction of the image in a single image; record the above angles of the two section images at 0 degree and 90 degree as αx and αy respectively, and then generate a three-dimensional rotation matrix Applying matrices H2 and H3 to all the three-dimensional data points obtained in step 4 to obtain three-dimensional point data of the cornea after coordinate correction; Step 6: Analysis and expression of corneal sclera 3D data The three-dimensional data points obtained in step 4 are numerically and analytically fitted using the Zernike polynomial fitting method; Zernike polynomials The form is as follows: Where ρ and θ represent the radius and angle, which are used to determine the position of any point in the polar coordinate system. n represents the order, and the value of m is -n, -n+2, -n+4, ..., n. is the normalization factor; is a radially distributed polynomial: The fitted corneoscleral surface is expressed by the weighted Zernike polynomial W(ρ,θ): Convert the three-dimensional data points in step 4 from rectangular coordinates to polar coordinates, and use the least squares method to fit to obtain the coefficients of each basis term in the Zernike expression. In this way, the spatially resolved Zernike expression of the corneoscleral surface is analytically described; Step 7: Generation of corneoscleral sagittal height topography According to the expression W(ρ,θ) obtained in step 6, and the polar coordinate and rectangular coordinate conversion relationship x=p*cos(θ), y=ρ*sin(θ), the data range is defined as diameter D, the data interval Δs=D / nu, where nu is the number of data points in a single direction, and the corneoscleral sagittal height data matrix M0 with equal intervals in the x and y directions is generated. The size of M0 is nu*nu, and the data exceeding the diameter D is represented as invalid data; then the matrix M0 is generated into a corneoscleral sagittal height topography map by the color bar mapping method, that is, a certain range of values ​​is represented by a certain color; Step 8. Calculation of corneal and scleral data features During the rotation acquisition process of the Scheimpflug image, the forward camera provides alignment and plane dimension information, identifies the boundary range of the cornea and sclera on the image, and maps the cornea and sclera edge information of the front camera to the cornea and sclera 3D point data obtained in step 5 through the position matching of the cornea vertex. The cornea data is inside the cornea and the sclera data is outside the cornea and sclera edge. In this way, the cornea and sclera 3D data point set is divided into cornea and sclera 3D data points. The cornea and sclera data points are used to fit the sphere or ellipsoid using the least squares principle, respectively, to obtain the best fitting sphere or ellipsoid of the cornea BFS_C and the best fitting sphere or ellipsoid of the sclera BFS_S, as well as the corresponding mathematical expressions of the sphere or ellipsoid. Step 9: Generation of corneal sclera relative height topography According to the corneal best fitting sphere or ellipsoid obtained in step eight, a corneal best fitting sphere or ellipsoid data matrix M1 with the same size and data interval as the corneoscleral sagittal height data matrix M0 in step seven is generated, and the data outside the corneoscleral boundary range is set to invalid values; similarly, according to the scleral best fitting sphere or ellipsoid obtained in step eight, a scleral best fitting sphere or ellipsoid data matrix M2 with the same size and data interval as the corneoscleral sagittal height data matrix M0 in step seven is generated, and the data within the corneoscleral boundary range is set to invalid values; at the same time, the data outside the data range diameter D defined in step seven of the two data matrices generated are also set as invalid data; The corneoscleral relative height data matrix M3 is defined as M0-M1-M2; the matrix M3 is also generated into a corneoscleral relative height topography map by the color strip mapping method.

Images