Method capable of measuring peripheral aberration of retina

Abstract

The invention relates to a method capable of measuring peripheral aberration of a retina. The method comprises the following steps: acquiring wavefront images at a plurality of positions on a retina, and for each wavefront image, carrying out the following processing: searching each light spot in the wavefront image; the light spot located at the centermost position in all the light spots serves as a standard light spot, the wavefront reference pattern is integrally moved, the center position of the wavefront reference pattern is made to coincide with the center of the standard light spot, and a corrected reference pattern is obtained; calculating the deviation condition of each light spot and the corresponding reference position in the correction reference pattern; and determining the aberration of the retina at the position according to each offset condition. The method can realize rapid and accurate wavefront image acquisition in multiple scanning modes, and is high in automation level and convenient to operate.

Description

Technical Field

[0001] The present invention relates to the technical field of optical imaging, and in particular to a method for measuring retinal peripheral aberration. Background Art

[0002] Commonly used ophthalmic equipment can only obtain eye data in the central area of ​​the macula of the retina, but cannot obtain eye data in the peripheral retinal area. Later, the use of fundus cameras to collect fundus images appeared. After processing the collected fundus images, the refraction of each point on the retina can be calculated, but the calculation error of this technology is large and the accuracy is difficult to guarantee. In order to solve this technical problem, the wavefront aberration principle of Hartmann-Shack can be used to improve the accuracy of the calculation. Among the existing technologies that use this principle, one is to obtain peripheral field eye data by changing the gaze direction of the subject. This method requires the subject to change the gaze direction once for each measurement, resulting in too slow measurement speed and low detection efficiency. In addition, multiple measurements are easily disturbed by changes in the state of the human eye, resulting in low accuracy of eye data detection; another technology is to collect peripheral retinal data by galvanometer scanning, but this method has the following defects:

[0003] 1. Using a galvanometer mirror and the human eye pupil conjugate will introduce inherent aberrations in the system, affecting the dynamic range of the entire optical path. Especially at large angles, the afocal system itself will have large aberrations.

[0004] 2. When the galvanometer scans, the light path at each angle is different, so the aberration of the afocal system at each angle is different. Therefore, it is necessary to collect reference values ​​at each angle to eliminate the inherent aberration of the system at each angle, which requires a lot of calculation.

[0005] 3. The control accuracy of the galvanometer is required to be high. Due to the high-frequency vibration of the galvanometer, a high-performance wavefront camera needs to be selected to match the galvanometer. This results in high cost not only for the galvanometer but also for the wavefront camera.

[0006] In addition, the optical system using the Hartmann-Shack principle has a microlens. According to the position of the parallel light passing through the microlens array and hitting the wavefront camera, a reference frame (i.e., reference box) can be drawn. The reference frame represents the range of activity of the light spot that passes through the microlens array and hits the wavefront camera under the preset dynamic range. All reference frames constitute a wavefront reference pattern. The shape and arrangement of the reference frame are related to the arrangement of the microlens array. It is generally a square, but it can also be any shape such as a regular hexagon or a regular triangle. Of course, the reference frame is not necessary, and the shape is not fixed. It mainly plays the role of assisting the search for the light spot. When processing the collected light spot, it is necessary to calculate the offset of the light spot and the reference position corresponding to it in the wavefront reference pattern. The reference position is generally the center position of the corresponding reference frame. However, when processing, the prior art will automatically match the reference frame where the center of the light spot is located with it, calculate the offset of the two, and then calculate the aberration. However, due to the existence of human eye aberrations, the reference frame where the center of the light spot is located is not necessarily the reference frame it should match. The error in matching the reference frame with the light spot will cause errors in the calculation of the aberration. This error becomes more pronounced because the aberration in the periphery of the human eye increases the farther away from the center of the retina.

[0007] In addition, because of the reflection of the human cornea, stray light appears in the spot image. The stray light will affect the determination of the centroid of the normal spot and make the matching error between the reference frame and the spot more serious, resulting in lower accuracy of the aberration calculation. Summary of the invention

[0008] The embodiment of the present invention provides a method for measuring retinal peripheral aberration to solve the problems of large calculation error, low detection efficiency or high cost in the prior art.

[0009] In order to achieve the above purpose, the embodiment of the present invention provides the following technical solutions:

[0010] A method for measuring peripheral retinal aberrations comprising:

[0011] Wavefront images at multiple locations on the retina are collected, and each wavefront image is processed as follows:

[0012] Searching for each light spot in the wavefront image;

[0013] A light spot located at the most center of each light spot is used as a reference light spot, and the wavefront reference pattern is moved as a whole so that the center position of the wavefront reference pattern coincides with the center of the reference light spot, thereby obtaining a corrected reference pattern;

[0014] Calculating the offset of each light spot from the corresponding reference position in the corrected reference pattern;

[0015] The aberration at the retinal position is determined according to each shift situation.

[0016] Furthermore, collecting wavefront images at multiple positions on the retina includes:

[0017] Using an optical acquisition module to acquire a wavefront image at a certain position on the retina;

[0018] Adjust the position of the optical acquisition module on the arc-shaped moving device, and / or adjust the angle at which the arc-shaped moving device is driven by the rotary arm to rotate;

[0019] The above steps are repeated until the optical acquisition module moves along the spherical surface to acquire wavefront images at multiple positions on the retina.

[0020] Further, the plurality of positions include P points on the retina, and the P points are intersections of K concentric circles with the center of the macula as the center and L straight lines passing through the center of the macula;

[0021] The order in which the optical acquisition module acquires the P points is:

[0022] Adjust the optical acquisition module to reach the initial position of the first straight line, and acquire the position of each point on the first straight line along the first straight line;

[0023] The optical acquisition module is adjusted to reach the initial position of the second straight line, and the positions of each point on the second straight line are acquired along the second straight line; the above steps are repeated until L straight lines are traversed.

[0024] Further, the light spots in the search wavefront image include:

[0025] The first search searches for light spots located within each reference frame of the wavefront reference pattern in the wavefront image;

[0026] And / or, in a second search, each light spot obtained in the first search is taken as the center, a search box of a preset size is established around it, and the missing light spot is searched in the search box; if the missing light spot is found, a search box of a preset size is established around the missing light spot as the center, and the missing light spot is searched again in the search box; repeat this step until no missing light spot is found;

[0027] And / or, after obtaining the corrected reference pattern, the method further comprises:

[0028] The third search searches the wavefront image for the missing light spots within each reference frame of the corrected reference pattern.

[0029] Furthermore, before calculating the offset between each light spot and the corresponding reference position in the corrected reference pattern, the method further includes:

[0030] Determine the light spot located at the edge among the light spots;

[0031] Perform ellipse fitting on the light spot at the edge;

[0032] The circular area is determined according to the result of ellipse fitting, and the light spots outside the circular area are deleted.

[0033] Furthermore, before calculating the offset between each light spot and the corresponding reference position in the corrected reference pattern, the method further includes:

[0034] Determine whether each light spot is stray light, and if so, delete the light spot; wherein, the method for determining whether a light spot is stray light is:

[0035] A circle is determined with the center of the light spot as the center and a preset radius. If the difference between the highest pixel value and the lowest pixel value on the circumference is greater than a threshold, the light spot is judged to be stray light.

[0036] Furthermore, before calculating the offset between each light spot and the corresponding reference position in the corrected reference pattern, the method further includes:

[0037] Determine whether there is stray light in each reference frame, and if so, delete the stray light;

[0038] Among them, the method for judging whether there is stray light in a reference frame is:

[0039] The reference frame is divided into 4 regions, and the highest pixel value in each region is compared with the first threshold. There are N regions with values ​​greater than the first threshold.

[0040] If N = 0, there is no stray light;

[0041] If N=1, starting from the position of the highest pixel value in the area, a preset number of pixel points are obtained in the preset M directions respectively, and whether it is stray light is determined according to the difference in pixel values ​​of the pixel points in the M directions;

[0042] If N≥2, then construct lines from the position of the highest pixel value in the N areas to the position of the highest pixel value in the remaining N-1 areas, and determine whether there is stray light based on the relationship between the pixel value of the pixel point on the line and the second threshold.

[0043] Further, the calculating of the offset between each light spot and the corresponding reference position in the corrected reference pattern includes:

[0044] Taking the reference light spot as the origin, adding a light spot index to each surrounding light spot using an index rule, wherein the light spot index is determined based on the relative position relationship between each light spot and the reference light spot;

[0045] Taking the center position of the modified reference pattern as the origin, adding reference indexes to the surrounding reference positions using the index rule, wherein the reference indexes are determined based on the relative positional relationship between each reference position and the center position of the reference pattern;

[0046] The light spot with the same light spot index value and reference index value is determined as a corresponding relationship with the reference position;

[0047] Calculate the offset of each light spot from its corresponding reference position.

[0048] Furthermore, before collecting the wavefront images at multiple positions on the retina, pupil positioning is also performed, including:

[0049] Perform downsampling processing on the image at multiple different scales to obtain multiple images to be processed;

[0050] Detecting candidate points of the pupil from each of the images to be processed;

[0051] Mapping all candidate points obtained from each of the images to be processed into the image according to corresponding scales to obtain mapped candidate points;

[0052] Determine pupil edge points based on the mapped candidate points and the image, and use the center position of each pupil edge point as a pupil positioning point;

[0053] The position of the eye to be tested or the acquisition device is adjusted according to the position of the pupil positioning point so that the optical axis of the acquisition device is aligned with the pupil positioning point.

[0054] Furthermore, after determining the aberration at the retina position according to each offset situation, the method further includes:

[0055] The aberrations at multiple locations on the retina are summarized to obtain a distribution map including the aberrations in the retinal periphery.

[0056] The present invention has the following beneficial effects:

[0057] The present invention provides a method for measuring peripheral retinal aberrations, wherein a wavefront reference pattern is moved according to a reference light spot to obtain a modified reference pattern, and the modified reference pattern can better correspond the light spot to the reference frame, and when calculating the offset of each light spot and the corresponding reference position in the modified reference pattern, the calculation error caused by the matching error is reduced, thereby improving the accuracy, and being able to measure more realistic retinal aberrations, especially peripheral retinal aberrations. The method can realize fast and accurate wavefront image acquisition in multiple scanning modes, has a high level of automation, and is easy to operate.

[0058] Compared with the galvanometer scanning acquisition method, this method does not introduce the inherent aberration that exists in the galvanometer system and is easy to operate. In addition, this method does not need to collect reference values ​​at each angle, which reduces the amount of calculation. It can also scan the retina at different granularities according to different needs, and does not use high-frequency galvanometers, and does not require high-performance camera adaptation, which can greatly reduce equipment costs.

[0059] This method also solves the problems of pupil positioning and light spot omission and excludes stray light in the light spot image, greatly improving the accuracy of aberration calculation and making the result more objective and accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to more clearly illustrate the implementation methods of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the implementation methods or the description of the prior art. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other implementation drawings can be derived from the provided drawings without creative work.

[0061] The structures, proportions, sizes, etc. illustrated in this specification are only used to match the contents disclosed in the specification so as to facilitate understanding and reading by persons familiar with the technology. They are not used to limit the conditions under which the present invention can be implemented, and therefore have no substantial technical significance. Any structural modification, change in proportion or adjustment of size shall still fall within the scope of the technical contents disclosed in the present invention without affecting the effects and purposes that can be achieved by the present invention.

[0062] Figure 1 is a flow chart of the method of an embodiment;

[0063] Figure 2 It is a structural diagram of the acquisition device of the embodiment;

[0064] Figure 3 This is a schematic diagram of a custom collection trajectory, where the intersections are the collection locations of the dot graph;

[0065] Figure 4 This is a schematic diagram of the results of the first search;

[0066] Figure 5 This is a schematic diagram of the results of the second search;

[0067] Figure 6 This is a schematic diagram before correcting the reference pattern;

[0068] Figure 7 For Figure 6 Schematic diagram of the calculation offset situation;

[0069] Figure 8is a schematic diagram after the reference pattern is corrected;

[0070] Fig. 9 For Figure 8 Schematic diagram of the calculation offset situation;

[0071] Fig.10 It is a schematic diagram of a circular area;

[0072] Fig.11 is a two-dimensional spherical mirror distribution map;

[0073] Fig.12 It is a three-dimensional spherical mirror distribution map. DETAILED DESCRIPTION

[0074] The following is a description of the implementation of the present invention by specific embodiments. People familiar with the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0075] like Figure 1 A method for measuring peripheral retinal aberrations is shown, the method comprising:

[0076] Wavefront images at multiple locations on the retina are collected, and each wavefront image is processed as follows:

[0077] Searching for each light spot in the wavefront image;

[0078] A light spot located at the most center of each light spot is used as a reference light spot, and the wavefront reference pattern is moved as a whole so that the center position of the wavefront reference pattern coincides with the center of the reference light spot, thereby obtaining a corrected reference pattern;

[0079] Calculating the offset of each light spot from the corresponding reference position in the corrected reference pattern;

[0080] The aberration at the retinal position is determined according to each shift situation.

[0081] The technical solutions and beneficial effects of the above steps are further explained below with reference to specific examples:

[0082] Before collecting the wavefront images at multiple positions of the retina, pupil positioning is performed first. A monocular pupil positioning method or a binocular pupil positioning method can be used, such as relying on the clarity in the pupil edge for front and back positioning, and its up, down, left and right positions are determined by the reference position calibrated by the pupil camera. The monocular pupil positioning method for adjusting the eye position according to the calibrated coordinates, all existing pupil positioning methods fall within the protection scope of the present invention, but for this technology, it is preferred to use a binocular pupil positioning method, because when the motor is not moving, the front and back errors will only affect the image quality. After the motor moves, the eye is not at the center of rotation, which will inevitably cause the dot pattern to shift. The use of a binocular pupil positioning method can greatly alleviate the problem of eye position movement. Since this technology is a through-hole motor driving the optical system to move in an arc, the spatial position of the optical system during pupil positioning is particularly important. In this embodiment, it is preferred to use the center of the sphere where the motor moves as the spatial position for pupil positioning.

[0083] In addition, this embodiment also provides a pupil positioning method based on multi-scale images, which can be applied to monocular or binocular positioning, including:

[0084] Perform downsampling processing on the image at multiple different scales to obtain multiple images to be processed;

[0085] Detecting candidate points of the pupil from each of the images to be processed;

[0086] Mapping all candidate points obtained from each of the images to be processed into the image according to corresponding scales to obtain mapped candidate points;

[0087] Determine pupil edge points based on the mapped candidate points and the image, and use the center position of each pupil edge point as a pupil positioning point;

[0088] The position of the eye to be tested or the acquisition device is adjusted according to the position of the pupil positioning point so that the optical axis of the acquisition device is aligned with the pupil positioning point.

[0089] Wherein, determining pupil edge points based on the mapped candidate points and the image includes:

[0090] For any mapped candidate point, edge points are searched in multiple directions within the circumference of the candidate point: for any direction, if the difference between the average pixel value of each point on the arc determined by the first radius and the preset arc length in that direction and the average pixel value of each point on the arc determined by the second radius and the preset arc length is greater than a first threshold, then the point at the first radius or the second radius in that direction is taken as the edge point in that direction, and the first radius is longer than the second radius by a first preset number of pixels.

[0091] After detecting the candidate points of the pupil and / or obtaining the mapped candidate points, the candidate points are merged:

[0092] For each candidate point, select the candidate points whose distance to the candidate point is less than the threshold from the remaining candidate points, and form a set with the candidate point;

[0093] The following process is repeated: for each candidate point in the set, a circular area is drawn with the candidate point as the center and the threshold as the radius, and the candidate points in the circular area are included in the set; the process stops when the number of candidate points in the set remains unchanged;

[0094] The coordinate point corresponding to the average position of each candidate point in the obtained set is used as the merged candidate point.

[0095] Due to the influence of the complex and diverse environment in the actual scene and the movement of the human head or the eye movement reaction of the eyeball, the size of the pupil in the obtained image is not fixed. In this method, images of various scales are obtained, and candidate points corresponding to the pupil are detected from the images at various scales, which helps to locate the pupil more accurately. In addition, in the present invention, the pupil edge points are determined based on the mapped candidate points in the image, and the pupil position is determined according to the pupil edge points. It is applicable to the situation where the pupil is complete or partially blocked, and can adapt to the complex and diverse environment in the actual scene, reduce the influence of the movement of the human head or the eye movement reaction of the eyeball, and can quickly and accurately locate the pupil under various undesirable conditions. After that, the position of the eye to be tested or the acquisition device can be adjusted according to the position of the pupil positioning point, so that the optical axis of the acquisition device is aligned with the pupil positioning point. The accuracy of pupil positioning is improved, and the error caused by the deviation of the incident light can be reduced, thereby improving the accuracy of retinal aberration detection.

[0096] like Figure 3 As shown, the acquisition device used in this embodiment includes: a base 1, a rotary arm 2, an arc-shaped moving device 3, an optical system installation position 4, and an optical axis schematic position 5. The pupil of the human eye needs to be aligned with the optical axis schematic position 5. An optical acquisition module is installed at the optical system installation position 4. The optical acquisition module includes a scanning device and an image acquisition device. The scanning device is used to emit laser light, and the image acquisition device is used to collect wavefront images and transmit them to the image processing module.

[0097] The swivel arm is mounted on the base, and the swivel arm can drive the arc moving device to rotate, and the rotation center of the swivel arm overlaps with the projection of the optical axis schematic position on the base. The swivel arm can adjust its rotation angle manually, or a rotation drive device can be set to control the rotation angle of the swivel arm, so as to realize automatic adjustment of the rotation angle of the swivel arm and improve the intelligence of the system.

[0098] The optical acquisition module can move on the arc moving device, and the arc center of the arc moving device overlaps with the optical axis schematic position 5, so that the optical acquisition module can slide to any position of the arc moving device, and its optical axis passes through the optical axis schematic position 5. Specifically, the arc moving device includes an arc slide rail, a slider, and an arc driving device, the arc slide rail is arranged on a side wall of the swing arm, the slider is arranged on the arc slide rail and moves along the slide rail, and the optical acquisition module is installed on the slider.

[0099] The arc driving device includes an arc driving motor, a driving gear, and arc teeth. The output end of the arc driving motor is fixedly connected to the driving gear. The driving gear and the arc teeth are meshed and transmitted. The tooth surface of the arc teeth is arranged on the support frame. The curvature of the tooth surface of the arc teeth is the same as the curvature of the arc slide rail, that is, the positions of the arc centers of the two are the same. The body of the arc driving motor is directly or indirectly fixedly connected to the slider:

[0100] a. Direct fixation

[0101] The outer side of the slider is provided with a mounting surface parallel to the arc surface of the arc slide rail. The arc teeth are fixed on the support frame and are located on the outer arc surface side of the arc slide rail, so that the arc teeth and the arc slide rail are concentric arcs with each other. The body of the drive motor is fixedly connected to the slider to realize the synchronous movement of the arc drive motor and the slider.

[0102] b. Indirect fixation

[0103] The arc-shaped slide rail and the arc-shaped teeth are respectively arranged on the side walls of the support frame, the arc-shaped slide rail and the arc-shaped teeth are parallel to each other, and the arc-shaped drive motor and the slider are fixedly connected by a connecting frame to realize the synchronous movement of the arc-shaped drive motor and the slider. In this embodiment, the shape and number of the connecting frame are not limited, as long as the technical solution for fixing the connection between the drive motor and the slider falls within the scope of the embodiment.

[0104] One or more transmission gears can be added between the driving gear and the arc-shaped teeth. For example, a set of transmission gears can be added between the driving gear and the arc-shaped teeth. The transmission gears are single-axis double gears, one of which is meshed with the driving gear and the other is meshed with the arc-shaped teeth. This design can reduce the transmission speed and increase the transmission torque, so that a small-power motor can provide sufficient torque, reduce the resistance that needs to be overcome when the module moves, and improve the transmission accuracy to meet the adjustment accuracy of the optical module.

[0105] The arc moving device can adjust its movement angle manually, or set an arc driving device, which controls the movement angle of the arc moving device to improve the intelligence of the system. The acquisition equipment adopts a two-dimensional centripetal rotation spherical scanning mechanism to make the movement adjustment of the two cross-dimensional dimensions act on the optical acquisition module, realizing the scanning of the triangular pyramid area centered on the focus, meeting the needs of multi-dimensional data acquisition.

[0106] The wavefront images at multiple positions of the retina are collected using the above-mentioned collection device, including:

[0107] The optical acquisition module acquires a wavefront image at a certain position of the retina;

[0108] Adjust the position of the optical acquisition module on the arc-shaped moving device, and / or adjust the angle at which the arc-shaped moving device is driven by the rotary arm to rotate;

[0109] The above steps are repeated until the optical acquisition module moves along the spherical surface to acquire wavefront images at multiple positions on the retina.

[0110] Compared with the galvanometer scanning acquisition method, this method does not introduce the inherent aberration that exists in the galvanometer system and is easy to operate. In addition, this method does not need to collect reference values ​​at each angle, which reduces the amount of calculation. It can also scan the retina at different granularities according to different needs, and does not use high-frequency galvanometers, and does not require high-performance camera adaptation, which can greatly reduce equipment costs.

[0111] The scanning mode of the acquisition device can be set as needed, such as single-line scanning, cross scanning, M-shaped scanning and scanning mode with custom step length to determine the acquisition movement trajectory.

[0112] The target position of this embodiment includes P points on the retina, and the P points are the intersections of K concentric circles with the center of the macula as the center and L straight lines passing through the center of the macula. Figure 3 As shown in the figure, the scanning angle is (-25°, +25), with an interval of 1°, then the center point on the retina and 50 concentric circles can be collected, and there are 36 straight lines passing through the center of the macula, then the wavefront image at the intersection of the 25 concentric circles and 36 straight lines with the retinal fovea as the center will be collected. Figure 3 For ease of presentation, only five concentric circles are shown and only three intersection points are marked.

[0113] For each intersection point, a wavefront image is obtained, and for each wavefront image, the following processing is performed:

[0114] To search for each light spot in the wavefront image, the existing method can be used to directly extract all the light spots in the whole image according to the pixel features in the wavefront image, but this method takes a long time to calculate. In this embodiment, three searches are performed.

[0115] The first search, such as Figure 4 As shown, searching for light spots located within each reference frame of the wavefront reference pattern in the wavefront image;

[0116] and / or, a second search such as Figure 5 As shown, each light spot obtained in the first search is taken as the center, and a search box of a preset size is established around it, and the missing light spots are searched in the search box; the missing light spots are the light spots that have not been searched in the previous search steps, the same below;

[0117] If a missing light spot is found, a search box of a preset size is established around the missing light spot as the center, and the missing light spot is searched again in the search box; this step is repeated until no missing light spot is found;

[0118] And / or, after obtaining the corrected reference pattern, the method further comprises:

[0119] The third search searches the wavefront image for the missing light spots within each reference frame of the corrected reference pattern.

[0120] Among them, according to the change of pixel values ​​of each coordinate point in the reference frame, it is possible to identify whether there is a light spot in the reference frame. If there is no light spot in the reference frame, the reference frame is ignored. For example, the sum of all pixels in the reference frame is taken as an average value. If it is greater than a threshold, it means that there is a light spot. In this embodiment, the technical solution for how to identify the presence of a light spot in the reference frame is not limited. As long as the technical solution that can identify the presence of a light spot in the reference frame falls within the protection scope of the present invention. In addition, according to conventional operations in the field, the searched light spot position is represented by the center coordinates of the light spot. The calculation method of the light spot center can adopt existing technologies, such as Figure 5 As shown, the black dots shown in the figure are the collected center coordinates of the light spot.

[0121] The calculation method of the spot center coordinates (x, y) in this embodiment is:

[0122]

[0123]

[0124] Where (i, j) is the pixel coordinate in the rectangular area;

[0125] I(i, j) is the pixel value of the pixel point coordinates (i, j);

[0126] M is the length of the reference frame;

[0127] N is the width of the reference frame.

[0128] Put the detected spot center coordinates into the spot center set to facilitate subsequent operations.

[0129] It is possible that the center of the light spot is located on the edge line of the reference frame or in the gap between the reference frames. Therefore, the first search will miss this light spot center. In order to fully collect the light spot information, a second search is performed after the first search to put the missed light spot center coordinates into the light spot center set.

[0130] This method uses the most central spot among the light spots as the reference spot, and moves the wavefront reference pattern as a whole so that the center position of the wavefront reference pattern coincides with the center of the reference spot to obtain a corrected reference pattern. Specifically:

[0131] The first ellipse fitting is performed according to the light spot center set to obtain a first fitting ellipse region containing all the light spot center coordinates. The ellipse fitting method includes finding a light spot center coordinate located at the edge from the light spot center set. In this embodiment, the light spot center coordinate located at the far left is found and set as the initial starting point coordinate, that is, the minimum X coordinate is determined after the maximum Y coordinate is found, and then the distance and angle between the initial starting point coordinate and the remaining coordinates in the light spot center set are calculated respectively, and the angle is the angle between the positive semi-axis of the x-axis and the line connecting the two coordinates. If there are other coordinates whose distance is less than the distance threshold and whose angle is less than the angle threshold, the other coordinates are set as the second starting coordinate, and the above steps are repeated to find the next starting coordinate. If there are no other coordinates whose distance is less than the distance threshold and whose angle is less than the angle threshold, the angle threshold is expanded, such as by 5%, and the above steps are repeated to find the next starting coordinate, and so on, until a closed figure is formed to obtain the first fitting ellipse region.

[0132] After obtaining the first fitted elliptical area, the center E1 of the first fitted elliptical area is calculated, and the center coordinates of the spot closest to the center E1 in the spot center set are determined as the center coordinates of the reference spot, that is, the wavefront image center coordinates I1. The coordinate displacement (Δx, Δy) between the reference frame center coordinates R1 of the central reference frame and the wavefront image center coordinates I1 is calculated, and all reference frames are synchronized according to the coordinate displacement, that is, the reference frame coordinates (RefCenterX, RefCenterY) are updated to (RefCenterX+Δx, RefCenterY+Δy) to obtain the corrected reference frame.

[0133] The corrected reference pattern can better match the light spot with the reference frame. When calculating the offset between each light spot and the corresponding reference position in the corrected reference pattern, the calculation error caused by matching errors can be reduced, thereby improving accuracy and measuring more realistic retinal aberrations, especially peripheral retinal aberrations. Figure 6-9 As shown, Fig. 9 The calculated offset is better than Figure 8The offset in the image is closer to the center of the light spot, which is more in line with the actual situation and the result is more accurate.

[0134] After the corrected reference pattern is obtained, since the reference pattern has moved, a third search is performed to search the wavefront image for missed light spots within each reference frame of the corrected reference pattern, and to find in turn the centers of the light spots that may have been missed in the first two searches because they were located on the edge line of the reference frame or in the gap between the reference frames.

[0135] Since the light spots at the edge in the wavefront image have large errors, which will affect the accuracy during calculation, the light spots at the edge will be deleted in this embodiment to avoid calculation errors, including: determining the light spots at the edge among the light spots; performing ellipse fitting on the light spots at the edge; determining a circular area according to the result of the ellipse fitting, and deleting the light spots outside the circular area, specifically:

[0136] Perform a second ellipse fitting to obtain a second fitting ellipse region containing all the center coordinates of the light spots, and determine the center and the minor axis of the second fitting ellipse region, such as Fig.10 As shown, a circular area is calculated with the center as the center and the short axis as the radius, and all the spot center coordinates outside the circular area are deleted to obtain a corrected wavefront image.

[0137] The fitting method for the second ellipse fitting includes: selecting a spot center coordinate located at the edge in the corrected spot center set and setting it as the initial starting point coordinate, respectively calculating the distance and angle between the starting point coordinate and the remaining spot center coordinates in the corrected spot center set, if the distance is less than a distance threshold and the angle is less than an angle threshold, setting the spot center coordinate that meets the conditions as the next starting point coordinate, if the spot center coordinate that meets the conditions cannot be found, increasing the angle threshold, the angle threshold is less than 180°, repeating the detection until a closed figure is formed, and obtaining the coordinate set of the second fitting ellipse.

[0138] In addition, in order to avoid the influence of stray light, in this embodiment, before calculating the offset between each light spot and the corresponding reference position in the corrected reference pattern, it is also necessary to delete the stray light.

[0139] In one embodiment, it is determined whether each light spot is stray light, and if so, the light spot is deleted; wherein, the method for determining whether a light spot is stray light is: taking the center of the light spot as the center of the circle and determining a circle with a preset radius, if the difference between the highest pixel value and the lowest pixel value on the circumference is greater than a threshold value, the light spot is determined to be stray light.

[0140] In another embodiment, it is determined whether there is stray light in each reference frame, and if so, the stray light is deleted;

[0141] Among them, the method for judging whether there is stray light in a reference frame is:

[0142] The reference frame is divided into 4 regions, and the highest pixel value in each region is compared with the first threshold. There are N regions with values ​​greater than the first threshold.

[0143] If N = 0, there is no stray light;

[0144] If N=1, starting from the position of the highest pixel value in the area, a preset number of pixel points are obtained in the preset M directions respectively, and whether it is stray light is determined according to the difference in pixel values ​​of the pixel points in the M directions; whether it is stray light is determined according to the difference in pixel values ​​of the pixel points in the M directions. Specifically, if the number of the maximum and minimum value differences at the M positions at the same radius is greater than the threshold and is greater than 1 / 3 of the radius length, it is determined to be stray light, otherwise there is no stray light;

[0145] If N≥2, then construct lines from the position of the highest pixel value in the N areas to the position of the highest pixel value in the remaining N-1 areas, and determine whether there is stray light based on the relationship between the pixel value of the pixel point on the line and the second threshold.

[0146] When the proportion of pixels on the line whose pixel values ​​are lower than the second threshold (which represents the background) is greater than the third threshold (which represents the ratio), it is judged as stray light, otherwise there is no stray light. If there is stray light, refer to the method when N=1 to specifically determine the stray light, that is: starting from the position of the highest pixel value in the area where the stray light exists, obtain a preset number of pixel points in the preset M directions respectively, and judge whether it is stray light based on the difference in pixel values ​​of the pixel points in the M directions. Among them, the second threshold can be set to a pixel value between 0-20, and the third threshold can be set to a ratio between 0-30%.

[0147] In order to expand the dynamic range, all light spots and reference frames are assigned indexes to facilitate subsequent calculations. The calculation of the offset between each light spot and the corresponding reference position in the corrected reference pattern includes:

[0148] Taking the reference light spot as the origin, adding a light spot index to each surrounding light spot using an index rule, wherein the light spot index is determined based on the relative position relationship between each light spot and the reference light spot;

[0149] Specifically, the search box can be constructed by referring to the previous method for searching for light spots. The light spots are searched for in the periphery using the search box with the center coordinate I1 of the wavefront image as the center. Then, the search box is searched for in the periphery using the searched light spots as the center. This is repeated until all the light spots are searched. Unique light spot indexes are added to the center coordinates of the searched light spots in turn. In this embodiment, the peripheral search includes searching in four directions: up, down, left, and right. The spot index includes a horizontal sub-index and a vertical sub-index. For example, taking downward as positive and rightward as positive, the spot index of the wavefront image center coordinate I1 is (0, 0), the spot index of the spot above it is (0, -1), the spot index of the spot below it is (0, 1), the spot index of the spot to its left is (-1, 0), and the spot index of the spot to its right is (-1, 0). For another example, taking the spot with a spot index of (0, -1) as the center and searching around it, the spot index of the spot above it is (0, -2), the spot index of the spot below it is (0, 0), the spot index of the spot to its left is (-1, -1), and the spot index of the spot to its right is (-1, -1), and so on, until all the spots are marked and each spot has a unique spot index.

[0150] Taking the center position of the corrected reference pattern as the origin, the reference index is added to each surrounding reference position using the index rule, and the reference index is determined based on the relative position relationship between each reference position and the center position of the reference pattern; that is, the reference frame adopts the same index rule as above, and the center position of the centermost reference frame is set to (0, 0), then the reference index of the reference position above it (that is, the center position of the reference frame above it) is (0, -1), the reference index of the reference position below it (that is, the center position of the reference frame below it) is (0, 1), and the reference index of the reference position on its left (that is, the center position of the reference frame below it) is (0, 1). The reference index of the reference position is (-1, 0), and the reference index of the reference position to its right is (-1, 0). For example, taking the reference position with reference index (0, -1) as the center and searching around it, the reference index of the reference position above it is (0, -2), the reference index of the reference position below it is (0, 0), the reference index of the reference position to its left is (-1, -1), and the reference index of the reference position to its right is (-1, -1), and so on, until all reference positions are marked and each reference position has a unique reference index.

[0151] Afterwards, the light spot having the same light spot index value as the reference index value and the reference position are determined as a corresponding relationship;

[0152] The offset of each light spot and its corresponding reference position is calculated, and the aberration at the retinal position is determined according to each offset. The calculation can adopt the method of the prior art. In this embodiment, the offset S of the light spot center and the corresponding reference position in the xy direction is calculated. According to the existing Zernike polynomial, the Zernike coefficient A is fitted, S=DA, wherein D is a matrix composed of partial derivatives of the corresponding Zernike polynomial corresponding to the light spot position in two directions. The coefficient A can be solved by matrix operation, and the aberration is solved according to A and the pupil radius. The aberration may include spherical lens, cylindrical lens and axial position, and higher-order aberrations can also be solved as needed.

[0153] After determining the aberration at the retinal position according to each offset situation, the aberrations at multiple positions on the retina are summarized to obtain a distribution map of aberrations including the retinal center and periphery. A spherical distribution map, a cylindrical distribution map, etc. including the retinal center and / or the retinal periphery can also be obtained according to the aberrations. The distribution map can be a two-dimensional map or a three-dimensional map. Fig.11 The two-dimensional spherical mirror distribution diagram is shown in Fig.12 Shown in FIG. 1 is a three-dimensional spherical mirror distribution diagram.

[0154] Although the present invention has been described in detail above by general description and specific embodiments, it is obvious to those skilled in the art that some modifications or improvements can be made to the present invention. Therefore, these modifications or improvements made without departing from the spirit of the present invention all belong to the scope of protection claimed by the present invention.

Claims

1. A method for measuring peripheral retinal aberrations, characterized in that: include: Wavefront images at multiple locations on the retina are collected, and each wavefront image is processed as follows: Searching for each light spot in the wavefront image; A light spot located at the most center of each light spot is used as a reference light spot, and the wavefront reference pattern is moved as a whole so that the center position of the wavefront reference pattern coincides with the center of the reference light spot, thereby obtaining a corrected reference pattern; Calculating the offset of each light spot from the corresponding reference position in the corrected reference pattern; The aberration at the retinal position is determined according to each shift situation.

2. A method for measuring retinal peripheral aberrations according to claim 1, characterized in that: The collecting of wavefront images at multiple positions on the retina comprises: Using an optical acquisition module to acquire a wavefront image at a certain position on the retina; Adjust the position of the optical acquisition module on the arc-shaped moving device, and / or adjust the angle at which the arc-shaped moving device is driven by the rotary arm to rotate; The above steps are repeated until the optical acquisition module moves along the spherical surface to acquire wavefront images at multiple positions on the retina.

3. A method for measuring retinal peripheral aberration according to claim 2, characterized in that: The multiple positions include P points on the retina, where the P points are intersections of K concentric circles centered on the macula center and L straight lines passing through the macula center; The order in which the optical acquisition module acquires the P points is: Adjust the optical acquisition module to reach the initial position of the first straight line, and acquire the position of each point on the first straight line along the first straight line; The optical acquisition module is adjusted to reach the initial position of the second straight line, and the positions of each point on the second straight line are acquired along the second straight line; the above steps are repeated until L straight lines are traversed.

4. A method for measuring retinal peripheral aberrations according to claim 1, characterized in that: Each light spot in the search wavefront image includes: The first search searches for light spots located within each reference frame of the wavefront reference pattern in the wavefront image; And / or, in a second search, each light spot obtained in the first search is taken as the center, a search box of a preset size is established around it, and the missing light spot is searched in the search box; if the missing light spot is found, a search box of a preset size is established around the missing light spot as the center, and the missing light spot is searched again in the search box; repeat this step until no missing light spot is found; And / or, after obtaining the corrected reference pattern, the method further comprises: The third search searches the wavefront image for the missing light spots within each reference frame of the corrected reference pattern.

5. A method for measuring retinal peripheral aberrations according to claim 1, characterized in that: Before calculating the offset between each light spot and the corresponding reference position in the corrected reference pattern, the method includes: Determine the light spot located at the edge among the light spots; Perform ellipse fitting on the light spot at the edge; The circular area is determined according to the result of ellipse fitting, and the light spots outside the circular area are deleted.

6. A method for measuring retinal peripheral aberrations according to claim 1, characterized in that: Before calculating the offset between each light spot and the corresponding reference position in the corrected reference pattern, the method includes: Determine whether each light spot is stray light, and if so, delete the light spot; wherein, the method for determining whether a light spot is stray light is: A circle is determined with the center of the light spot as the center and a preset radius. If the difference between the highest pixel value and the lowest pixel value on the circumference is greater than a threshold, the light spot is judged to be stray light.

7. A method for measuring retinal peripheral aberrations according to claim 1, characterized in that: Before calculating the offset between each light spot and the corresponding reference position in the corrected reference pattern, the method includes: Determine whether there is stray light in each reference frame, and if so, delete the stray light; Among them, the method for judging whether there is stray light in a reference frame is: The reference frame is divided into 4 regions, and the highest pixel value in each region is compared with the first threshold. There are N regions with values ​​greater than the first threshold. If N = 0, there is no stray light; If N=1, starting from the position of the highest pixel value in the area, a preset number of pixel points are obtained in the preset M directions respectively, and whether it is stray light is determined according to the difference in pixel values ​​of the pixel points in the M directions; If N≥2, then construct lines from the position of the highest pixel value in the N areas to the position of the highest pixel value in the remaining N-1 areas, and determine whether there is stray light based on the relationship between the pixel value of the pixel point on the line and the second threshold.

8. A method for measuring retinal peripheral aberrations according to claim 1, characterized in that: The calculating of the offset between each light spot and the corresponding reference position in the corrected reference pattern includes: Taking the reference light spot as the origin, adding a light spot index to each surrounding light spot using an index rule, wherein the light spot index is determined based on the relative position relationship between each light spot and the reference light spot; Taking the center position of the modified reference pattern as the origin, adding reference indexes to the surrounding reference positions using the index rule, wherein the reference indexes are determined based on the relative positional relationship between each reference position and the center position of the reference pattern; The light spot with the same light spot index value and reference index value is determined as a corresponding relationship with the reference position; Calculate the offset of each light spot from its corresponding reference position.

9. A method for measuring retinal peripheral aberrations according to claim 1, characterized in that: Before collecting the wavefront images at multiple positions on the retina, pupil positioning is also performed, including: Perform downsampling processing on the image at multiple different scales to obtain multiple images to be processed; Detecting candidate points of the pupil from each of the images to be processed; Mapping all candidate points obtained from each of the images to be processed into the image according to corresponding scales to obtain mapped candidate points; Determine pupil edge points based on the mapped candidate points and the image, and use the center position of each pupil edge point as a pupil positioning point; The position of the eye to be tested or the acquisition device is adjusted according to the position of the pupil positioning point so that the optical axis of the acquisition device is aligned with the pupil positioning point.

10. The method for measuring retinal peripheral aberration according to claim 1, characterized in that: After determining the aberration at the retina position according to each offset situation, the method further includes: The aberrations at multiple locations on the retina are summarized to obtain a distribution map including the aberrations in the retinal periphery.

Images