Methods and systems for determining refractive characteristics of a subject's eye using an image capture device

Abstract

The invention relates to a method for determining a refractive feature of an eye (13) of a subject (12) using an image capturing device (2). The method comprises the steps of: - taking at least one picture of a retina of an eye (13) of a subject (12), and - determining said refractive feature based on a blur level of said taken picture of the retina.

Description

TECHNICAL FIELD

[0001] The present invention relates to a method and a system for determining a refractive feature of an eye of a subject. BACKGROUND

[0002] Many documents describe devices and methods for determining such a refractive feature.

[0003] In particular, automatic refraction methods for determining an objective value of the refraction of a subject are known. These methods are complex and sometimes time-consuming. These methods generally imply the use of large and expensive devices that need to be manipulated by qualified personnel.

[0004] In particular, these methods require the use of a specific instrument having a camera and a plurality of light sources placed on the same plane.

[0005] The access to these automatic refraction methods is therefore limited and the majority of the world population does not benefit from them. SUMMARY

[0006] It is therefore an object of the present invention to provide a new method for determining a refractive feature of an eye of a subject that will be simplified since it does not require the use of specific materials or the intervention of qualified personnel.

[0007] More precisely, the invention consists in a method for determining a refractive feature of an eye of a subject using an image capturing device. The method comprises the following steps:

[0008] - taking at least one photo of the retina of the eye of the subject; and

[0009] - determining the refractive feature based on a blur level of the taken photo of the retina, the blur level of the photo being determined using a value of a point spread function associated with the photo, the refractive feature being determined by calculating a modified photo having a reduced blur level compared to an initial blur level of the taken photo of the retina.

[0010] This method can be implemented by the subject himself and only using a smartphone or a tablet without the need to add optical components or to use an augmented reality display. This method can therefore be accessed by a wide range of people, including some people who are excluded from the access to existing methods.

[0011] Other advantageous features of the method are the following features:

[0012] - the refractive feature is determined by using a model relating the blur level of the taken photo of the retina to a refractive feature;

[0013] - the method further comprises a step of acquiring at least a photo of the pupil of the eye of the subject, the step of determining said refractive feature being dependent on a pupil diameter determined from said photo of the pupil;

[0014] - the blur level is determined using a convolutional neural network;

[0015] - the neural network is trained using a dataset of pairs of images, each pair of images being associated with a specific refractive feature;

[0016] - the method further comprises a step of determining a value of a point spread function associated with the acquired photo of the retina by using the training of the neural network with the dataset of pairs of images;

[0017] - the method further comprises a step of determining a refractive feature associated with the acquired photo of the retina by using the training of the neural network with the dataset of pairs of images;

[0018] - the modified photo is computed using a blind deconvolution method of the acquired photo of the retina;

[0019] - the blind deconvolution method is based on a dataset of values of a point spread function, each value of the point spread function being associated with a specific refractive feature;

[0020] - the modified photo is determined by selecting in the dataset of said values of the point spread function a best value of the point spread function, the best value of the point spread function corresponding to an improved contrast level in the modified photo compared to an initial contrast level of the acquired photo of the retina;

[0021] - the photo of the retina is acquired at an infinite focus;

[0022] - the distance between the eye of the subject and the image capture device is greater than 20 mm; and

[0023] - the method further comprises a step of acquiring another photo of the retina of the eye of the subject, said another photo being acquired at another focus comprised between the focus of said acquired photo of the retina and the focus of said photo of the pupil of the eye, said refractive feature being further dependent on said another photo of the retina, said another acquired photo being suitable for determining a sign of the refractive feature, the sign of the refractive feature being determined by comparing a blur level between said another acquired photo of the retina and said acquired photo of the retina.

[0024] The invention also comprises a system for determining a refractive feature of an eye of a subject, comprising:

[0025] - an image capture device suitable for acquiring at least one photo of the retina of the eye of the subject; and

[0026] - A data processor adapted to determine the refractive features based on the blur level of the acquired photograph of the retina, the blur level of the photograph being determined using the value of a point spread function associated with the photograph, and the refractive features being determined by calculating a modified photograph having a reduced blur level compared to the initial blur level of the acquired photograph of the retina.

[0027] The system further includes a housing suitable for accommodating the image capture device and the data processor.

[0028] The system further includes a housing adapted to house the image capturing device, the housing also including an electronic entity adapted to send the acquired photograph of the retina to a data processor and to receive the refractive features from the data processor.

[0029] The system further includes a light source suitable for illuminating the pupil of the subject's eye. Detailed Implementation

[0030] The following description, given in conjunction with the accompanying drawings by way of non-limiting example, will allow for an understanding of the composition of the invention and how it can be practiced.

[0031] In the attached diagram:

[0032] - Figure 1 An exemplary system according to the present invention suitable for determining the refractive characteristics of a subject's eye is shown;

[0033] - Figure 2 Another exemplary system according to the invention is shown, suitable for determining the refractive characteristics of a subject's eye;

[0034] - Figure 3 A first exemplary flowchart corresponding to a method for determining the refractive characteristics of a subject's eye according to the present invention is shown;

[0035] - Figure 4 This is a schematic diagram of the system configuration for determining the refractive characteristics of a subject's eye; and

[0036] - Figure 5 A second exemplary flowchart corresponding to the method for determining the refractive characteristics of a subject's eye according to the present invention is shown.

[0037] Figure 1 and Figure 2 This shows the appropriate method for determining subject 12 ( Figure 4An exemplary system 1, 10 for determining a refractive feature of an eye 13 of a subject 12. In the present description, the refractive features that can be determined include the spherical power Sph of the refractive error of the eye of the subject, the cylindrical power C of this refractive error, and the orientation Θ of the axis of the corresponding cylinder.

[0038] In the following, elements shared by both examples of system 1, 10 have the same reference numerals and are only described once.

[0039] System 1, 10 comprises an image capturing device 2 and a data processor 4. Optionally, system 1, 10 comprises a light source 8.

[0040] Image capturing device 2 is a small, general-purpose digital camera. Image capturing device 2 is suitable for taking at least one picture of the retina of eye 13 of subject 12.

[0041] Image capturing device 2 is for example located on a face of system 1, 10 oriented in the direction of subject 12. Image capturing device 2 generally comprises a lens, for example defined by an optical axis. Image capturing device 2 is for example a camera included in system 1, 10. Image capturing device 2 comprises for example a motorized focusing device (not shown) allowing to change the focal distance while taking a picture.

[0042] Optional light source 8 comprises for example one or several LEDs ("Light Emitting Diodes") or small arc lamps. Light source 8 is suitable for illuminating the pupil of eye 13 of subject 12.

[0043] Optional light source 8 is located in the vicinity of image capturing device 2, in particular in the vicinity of the optical axis of the lens of said image capturing device 2. In the case of a camera, optional light source 8 can be a built-in flash included in the camera.

[0044] Preferably, light source 8 uses near-infrared light to avoid pupil reflex and keep a good diameter of the pupil of eye 13 of subject 12. As an alternative, light source 8 can use white light or red wavelengths.

[0045] Data processor 4 is suitable for determining a refractive feature based on the picture taken by image capturing device 2. Data processor 4 is associated with a memory (not shown). The memory stores instructions which, when executed by data processor 4, allow system 1, 10 to implement the method for determining a refractive feature of eye 13 of subject 12 as described below.

[0046] According to Figure 1 The exemplary system 1 illustrated in the figure comprises a housing 5. This housing 5 is suitable for housing image capturing device 2 and data processor 4. In this case, the housing is for example a portable electronic device, such as a smartphone or a tablet computer.

[0047] According toFigure 2 Another exemplary system 10 is illustrated, the system 10 comprising a housing 7. The housing 7 is suitable for housing the image capturing device 2. In this case, the data processor 4 is located outside the housing 7 of the system 10.

[0048] Here, the housing 7 further comprises the electronic entity 6. This electronic entity 6 is suitable for sending the acquired photo of the eye 13 of the subject 12 to the data processor 4 and for receiving the refractive features (after their determination) from the data processor 4. In this case, the housing 7 is for example a portable electronic device, such as a basic camera.

[0049] As an alternative, the system can comprise an additional lens 9. The optical axis of this additional lens 9 is here aligned with the optical axis of the lens of the image capturing device 2. This additional lens 9 serves to increase the field of view of the photo acquired by the image capturing device 2.

[0050] Figure 1 and Figure 2 The systems 1, 10 illustrated and previously described are suitable for carrying out a method for determining the refractive features of the eye 13 of the subject 12. Figure 3 and Figure 5 An exemplary flowchart corresponding to the method for determining the refractive features of the eye 13 of the subject 12 according to the application is illustrated.

[0051] Figure 3 A schematic diagram of the main steps of a first embodiment of the method for determining the refractive features of the eye 13 of the subject 12 is described. Figure 5 A schematic diagram of the main steps of a second embodiment of the method for determining the refractive features of the eye 13 of the subject 12 is described.

[0052] In the following, the steps shared by the two embodiments have the same reference and are only described once.

[0053] As Figure 3 The first embodiment of the method is illustrated as starting at step S2, in which the subject 12 arranges the housing 5, 7 of the system 1 in front of its face. In particular, the direction of the image capturing device 2 is determined so that the image capturing device can directly acquire a photo of the eye 13 (retina and pupil) of the subject 12. Figure 4 This configuration is illustrated in Fig. 2. Indeed, the housing 5, 7 of the system 1 can be held by the subject 12 or by a support (not illustrated).

[0054] Indeed, the image capturing device 2 is located in the vicinity of the eye 13 of the subject 12. The distance D between the eye 13 of the subject 12 and the image capturing device 2 is typically greater than 20 millimeters (mm), for example between 20 mm and 100 mm.

[0055] AsFigure 3 As can be seen, the method then comprises a step S4 of illuminating the pupil of the eye 13 of the subject 12. In practice, the illumination is adapted to avoid a small diameter of the pupil. Preferably, the illumination is a weak light from the light source 8 or the indoor lighting.

[0056] The method continues with a step S6 of acquiring at least one photo of the eye 13 of the subject 12 illuminated in the previous step S4. In order to reduce the accommodation of the eye 13, the subject 12 looks very far ahead in this step.

[0057] During this step S6, in a sub-step S6a, at least one photo of the retina of the eye 13 of the subject 12 is acquired. Preferably, here, two different photos of the retina of the eye 13 of the subject 12 are acquired.

[0058] The first photo of the retina (hereinafter also referred to as "reference photo of the retina") is acquired at an infinite focus. In order to acquire this reference photo of the retina, the image capture device 2 is located at a large focus distance from the eye 13 of the subject 12, for example at a focus distance greater than 4 m.

[0059] The second photo of the retina of the eye 13 of the subject 12 is acquired at another focus distance between the focus distance of the acquired reference photo of the retina and the image capture device 2.

[0060] The step S6 also comprises a sub-step S6b of acquiring at least one photo of the pupil of the eye 13 of the subject 12. The photo of the pupil of the eye 13 of the subject 12 is acquired at a focus distance smaller than the focus distance at which the photo of the retina of the eye 13 is acquired.

[0061] As an example, the motorized focusing device comprised in the image capture device 2 can be used to change the focus distance in order to acquire the photo of the pupil of the eye 13, the reference photo and the second photo of the retina.

[0062] Here, the sequence of three photos (combining the sub-steps S6a and S6b) is taken within a short time interval (for example less than 0.5 seconds (s)). This short time interval ensures identical conditions for the three photos and thus improves the accuracy of the further analysis.

[0063] With reference to the different focus distances at which the photos are acquired as described previously, in practice, for this step S6, the motorized focusing device can be positioned at a first focus distance (for example the closest focus distance or the farthest focus distance) and the photo is acquired, then the motorized focusing device is moved to a second focus distance (for example the intermediate focus distance) and another photo is acquired, and finally the motorized focusing device is moved again to the last focus distance and the last photo is acquired (without exceeding the time limit).

[0064] As Figure 3The method then comprises, as illustrated, a step S8. During this step, the acquired photos are transmitted to the data processor 4 for analysis.

[0065] Here, the analysis (and thus the determination of the refractive features) is based on the blur level of the reference photo of the retina.

[0066] In the present description, the expression "blur level" of a photo is associated with a measure of the degree of focus of the image, for example as introduced in the article "Analysis of focus measure operators in shape-from-focus" by S. Petruz, D. Puig, M. A. Garcia (Pattern Recognition, 46, 2013, 1415-1432). A high degree of focus is associated with a low blur level (thus, the image is sharp).

[0067] As an alternative, the contrast level of the photo can be associated with the blur level. The contrast level is determined, for example, using a Fourier transform analysis. A high contrast level is associated with a low blur level (thus, the image is sharp). As an example, the contrast level of the photo can be determined pixel by pixel.

[0068] Advantageously, thanks to the invention, the determination of the blur level makes it possible to derive the refractive features, such as the spherical power Sph of the refractive error of the eye 13 of the subject 12.

[0069] In practice, the blur level of the photo is determined using the values of the point spread function (also known as PSF) associated with the photo. In fact, the surface of the point spread function (also known as "spread") is directly related to the blur level, since the point spread function depends on the refractive error of the eye. A broad point spread function is associated with a very blurred image, while a narrow point spread function is associated with a sharp image.

[0070] However, the refractive features cannot be derived easily and directly from the values of the point spread function in a classical way. The plurality of values of the point spread function is stored in the memory of the data processor 4. This plurality of values of the point spread function is determined, for example, theoretically from any refractive features as described below.

[0071] The classical sphero-cylindrical features (spherical power S, cylindrical power C, orientation Θ) are generally represented by a triplet of orthogonal values (Sph, Jo, J45) defined as the spherical power Sph = S + C, the Jackson cross cylinder with power Jo = C x cos(2Θ) at 0° axis and the Jackson cross cylinder with power J45 = C x sin(2Θ) at 45° axis, as a decomposition of the astigmatism (C, Θ) in polar coordinates. ​

[0072] This triplet of orthogonal values (Sph, Jo, J45) is then used to determine the Zernike coefficients: where R is the pupil ray. Finally, from these Zernike coefficients, an expression of the wavefront W(p, q) can be derived as follows: Next, by determining the Fourier transform and the square modulus, the value of the point spread function is theoretically directly calculated from the wavefront. Next, before performing the method according to the application, a plurality of theoretical values of the point spread function are determined.

[0073] Here, at step S10, using the determined plurality of theoretical values of the point spread function and the corresponding refractive features, a measured value of the point spread function of the reference photograph of the retina of the eye 13 of the subject 12 is determined.

[0074] Advantageously, according to Figure 3 According to a first embodiment of the method illustrated, the measured value of the point spread function associated with the reference photograph of the retina is determined using a convolutional neural network.

[0075] As for the classical method, the convolutional neural network is trained before use. Here, the convolutional neural network is trained before performing the method according to the application.

[0076] The convolutional neural network is trained using a dataset of pairs of images for a given pupil diameter. Each pair of images comprises a blurred image of the retina of the eye of a subject and an associated unblurred image. In practice, the unblurred image corresponds to a modified photograph having a reduced blur level compared to the initial blur level of the associated blurred image. In practice, the modified photograph presents a better degree of focus or a better contrast level.

[0077] Each pair of images is associated with a particular value of the point spread function and thus with a particular refractive feature. Different pairs of images of the retina are simulated with the model, thus allowing to produce many pairs of images. For the training part, the blurred image is generated using the unblurred image convolved with the associated point spread function.

[0078] Here, the convolutional neural network is trained for example using the blurred image of the pair of images as input data and using the associated value of the point spread function as output data.

[0079] At step S10, the reference photograph of the retina of the eye 13 of the subject 12 is introduced as input of the convolutional neural network. Thanks to the training previously performed, a measured value of the point spread function associated with this reference photograph of the retina is obtained at the output of the convolutional neural network.

[0080] In the case where the pupil diameter of the subject 12 is known by the photo of the pupil of the eye 13 of the subject 12 acquired in step S6b, the refractive feature can then be extracted, for example by using Zernike coefficients. Indeed, this step S10 comprises a step of minimizing the difference between the measured value of the point spread function and one of the plurality of theoretical values of the point spread function stored in the memory of the data processor 4. The minimum difference allows to determine the corresponding theoretical value of the point spread function and thus the associated refractive feature.

[0081] As an alternative, the convolutional neural network can be trained using the blurred image of the pair of images as input data and using the associated refractive feature as output data. In step S10, the refractive feature is obtained directly at the output of the convolutional neural network by introducing in the convolutional neural network the reference photo of the retina of the eye 13 of the subject 12.

[0082] As another alternative, the convolutional neural network can be trained using the blurred image of the pair of images as input and using the unblurred image of the pair of images as output. In this case, by introducing in the convolutional neural network the reference photo of the retina, an associated modified photo with a reduced blur level compared to the initial blur level of the reference photo of the retina is obtained at the output of the convolutional neural network. Because by definition, the reference photo of the retina of the eye 13 of the subject 12 is determined by deconvolution of the modified photo and of the associated value of the point spread function, this value can be derived from the two pictures using a conventional deconvolution method. The refractive feature can then be extracted using the pupil diameter, for example by using Zernike coefficients as previously described.

[0083] Finally, in step S12, the refractive feature is elucidated using the acquired second photo of the retina of the eye 13 of the subject 12. In particular, this second photo of the retina is suitable for determining the sign of the refractive feature and thus the sign of the refractive error (in order to identify myopic or hyperopic eyes). Because the value of the point spread function is the same for positive and negative refractive error parameters, the refractive anomaly is completely identified thanks to this second photo of the retina.

[0084] The sign of the refractive feature is determined by comparing the blur level, for example based on the contrast level, between the second photo of the retina and the reference photo of the retina. If the blur level of the second photo of the retina is higher than the blur level of the reference photo of the retina, a myopic eye is identified. If the blur level of the second photo of the retina is lower than the blur level of the reference photo, the refractive error is associated with a hyperopic eye.

[0085] As another example, if the housing 7 is a smartphone, the focal range is between 70 mm and infinity. This parameter is positive because the image of the retina is formed at a focal length which is the inverse of the reciprocal of the spherical power.

[0086] Figure 5 A schematic representation of the main steps of a second embodiment of the method for determining the refractive features of the eye 13 of the subject 12 is described.

[0087] As Figure 5 illustrated, this second embodiment of the method for determining the refractive features of the eye 13 of the subject 12 also comprises the previously described steps S2 to S8.

[0088] As previously mentioned, here, the analysis (and thus the determination of the refractive features) is based on an analysis of the blur level of the reference photograph of the retina.

[0089] Advantageously, with the present invention, the determination of the blur level makes it possible to derive the refractive features, such as the spherical power of the refractive error of the eye 13 of the subject 12.

[0090] In this second embodiment of the method, the blur level of the reference photograph is determined using the value of the point spread function associated with this photograph. Thus, at step S20, the value of the point spread function of the reference photograph of the retina of the eye 13 of the subject 12 is determined.

[0091] Advantageously, according to Figure 5 the second embodiment of the method illustrated, this value of the point spread function of the reference photograph of the retina is determined using a blind deconvolution method. This blind deconvolution method is applied to the reference photograph of the retina in order to simultaneously determine a modified photograph having a reduced blur level compared to the initial blur level of the reference photograph of the retina and an associated value of the point spread function. The blind deconvolution method used here is for example an iterative method, such as the Lucy-Richardson deconvolution method or the Bayesian method deconvolution method. As another example, the blind deconvolution method can be a non-iterative method, such as the SeDDaRa deconvolution method or the cepstrum-based method.

[0092] Here, the deconvolution method is based on a dataset of values of the point spread function. Different values of the point spread function are calculated for different refractive features for a given pupil diameter. Indeed, here, each value of the point spread function of the dataset is associated with a particular refractive feature.

[0093] By determining the best value of the point spread function in the dataset of values of the point spread function, the modified photograph obtained at the output of the method is derived. In other words, the modified photograph is obtained by optimizing the value of the point spread function in the dataset of values of the point spread function.

[0094] Here, the optimal value of the point spread function is determined taking into account the deconvolved photo having an improved contrast level compared to the initial contrast level of the reference photo of the retina. This photo, also called "modified photo", is thus sharper (its blur level is thus lower) than the reference photo of the retina.

[0095] Once the value of the point spread function has been determined, and the pupil diameter of the eye 13 of the subject 12 is known thanks to the photo of the pupil of the eye 13 of the subject 12 acquired at step S6b, the refractive feature can be extracted, for example by using the Zernike coefficients as previously described.

[0096] Finally, at step S22 (similar to step S12 previously described), the refractive feature is elucidated using the acquired second photo of the retina of the eye 13 of the subject 12 in order to determine the sign of the refractive feature, and thus the sign of the refractive error (in order to identify a myopic or hyperopic eye).

[0097] The application has been described in relation to what is believed to be the most practical and preferred embodiments. It is understood that the application is not limited to the disclosed embodiments, but is intended to cover various arrangements included within the spirit and scope of the broadest interpretation of the appended claims.

Claims

1. A method for determining the refractive characteristics of the eye (13) of a subject (12) using an image capture device (2), The method includes the following steps: - Obtain at least one photograph of the retina of the eye (13) of the subject (12); as well as - The refractive features are determined based on the blur level of the acquired photograph of the retina, the blur level of the photograph being determined using the value of a point spread function associated with the photograph, and the refractive features are determined by: - Calculate a modified image with a reduced blur level compared to the initial blur level of the acquired image on the retina, wherein the value of the point spread function associated with the acquired image can be derived from the acquired image and the modified image with the reduced blur level using a deconvolution method, and - Minimize the difference between the value of the point spread function associated with the photograph and one of a plurality of theoretical values ​​of the point spread function to determine the corresponding theoretical value of the point spread function and thus determine the associated refractive feature, or - Optimize the value of the point spread function associated with the photograph among the plurality of theoretical values ​​of the point spread function to determine the corresponding theoretical value of the point spread function and thus determine the associated refractive features. The multiple values ​​of the point spread function are theoretically determined by arbitrary refractive features, which are represented by triplets used to determine the orthogonal values ​​of the Zernike coefficients, from which the wavefront is derived, and from the wavefront the values ​​of the point spread function are theoretically calculated.

2. The method according to claim 1, wherein, The refractive features are determined using a model that correlates the blur level of the acquired photograph of the retina with the refractive features.

3. The method according to claim 1 or 2, further comprising the step of acquiring at least a photograph of the pupil of the eye (13) of the subject (12), wherein the step of determining the refractive characteristics depends on the pupil diameter determined from the photograph of the pupil.

4. The method according to claim 1, wherein, The blur level is determined using a convolutional neural network.

5. The method according to claim 4, wherein, The neural network is trained using a dataset of multiple pairs of images, each pair of images being associated with a specific refractive feature.

6. The method of claim 5, further comprising the step of determining the value of the point spread function associated with the acquired photograph of the retina by training the neural network using the dataset of the multiple pairs of images.

7. The method of claim 5, further comprising the step of determining the refractive features associated with the acquired photograph of the retina by training the neural network using the dataset of the multiple pairs of images.

8. The method according to claim 1, wherein, The modified photograph was calculated using a blind deconvolution method on the photograph acquired from the retina.

9. The method according to claim 8, wherein, The blind deconvolution method is based on a dataset of values ​​of the point spread function, each value of which is associated with a specific refractive feature.

10. The method according to claim 9, wherein, The modified photograph was determined by selecting the optimal value of the point spread function from the dataset of the values ​​of the point spread function, the optimal value of the point spread function corresponding to the increased contrast level in the modified photograph compared to the initial contrast level of the acquired photograph of the retina.

11. The method according to claim 1, wherein, The photograph of the retina was taken at an infinity focal length.

12. The method according to claim 1, wherein, The distance between the subject's (12) eye (13) and the image capture device (2) is greater than 20 mm.

13. The method of claim 1, further comprising the step of acquiring another photograph of the retina of the eye (13) of the subject (12), the other photograph being acquired at a focal length between the focal length of the acquired photograph of the retina and the focal length of the photograph of the pupil of the eye, the refractive feature also depending on the other photograph of the retina, the acquired other photograph being adapted to determine a sign of the refractive feature, the sign of the refractive feature being determined by comparing the blur level between the acquired other photograph of the retina and the acquired photograph of the retina.

14. A system (1, 10) for determining the refractive characteristics of the eye (13) of a subject (12), comprising: - Image capture device (2), the image capture device being adapted to acquire at least one photograph of the retina of the eye (13) of the subject (12); as well as - A data processor (4) adapted to determine the refractive features based on the blur level of the acquired photograph of the retina, the blur level of the photograph being determined using the value of a point spread function associated with the photograph, the refractive features being determined by calculating a modified photograph with a reduced blur level compared to the initial blur level of the acquired photograph of the retina, the value of the point spread function being deriveable from both photographs using conventional deconvolution methods, and - Minimize the difference between the value of the point spread function associated with the photograph and one of a plurality of theoretical values ​​of the point spread function stored in memory to determine the corresponding theoretical value of the point spread function and thus determine the associated refractive feature, or - Optimize the value of the point spread function associated with the photograph among the plurality of theoretical values ​​of the point spread function to determine the corresponding theoretical value of the point spread function and thus determine the associated refractive features. The multiple values ​​of the point spread function stored in memory are theoretically determined by arbitrary refractive features, which are represented by triplets used to determine the orthogonal values ​​of the Zernike coefficients, from which the wavefront is derived, and from which the value of the point spread function is theoretically calculated.

Images