IP Library › Granted Patent US 12,622,582
Granted Patent B2
US 12,622,582 · App. 17/618,262 · Granted May 12, 2026

Method and device for determining a refraction feature of an eye of a subject using an image-capture device

Inventor: Stephane Boutinon (Charenton-le-Pont, FR)
Assignee: Essilor International
A61B3/103A61B3/14G06T5/73G06V10/454G06V10/82G06V40/19G06V40/193G06T2207/20084G06V40/18
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,622,582
App. No.
17/618,262
Granted
May 12, 2026
Kind
B2
Abstract

A system and method for determining a refraction feature of an eye of a subject using an image-capture device, the method including acquiring at least one picture of the retina of the eye of the subject, and determining the refraction feature based on a blur level of the acquired picture of the retina.

Claims (26)

1 . A method for determining a refraction feature of an eye of a subject using an image-capture device, said method comprising:

acquiring at least one picture of a retina of the eye of the subject; and

determining said refraction feature based on a blur level of said acquired picture of the retina, the blur level of a picture being determined using a value of a point spread function associated with said acquired picture, said refraction feature being determined by:

calculating a modified picture having a reduced blur level compared to an initial blur level of said acquired picture of the retina, said value of a point spread function associated with said acquired picture is derived from the acquired picture and the modified picture using a deconvolution method, and

identifying one theoretical value of a point spread function of a plurality of theoretical values of the point spread function which has a minimal difference between said value of the point spread function associated with said acquired picture and the one theoretical value of the point spread function of the plurality of theoretical values of the point spread function, to determine the corresponding theoretical value of the point spread function and the refraction feature associated with the corresponding theoretical value of the point spread function and thus the refraction feature of the eye of the subject, or

optimizing the value of the point spread function associated with said acquired picture among the plurality of theoretical values of the point spread function to determine the corresponding theoretical value of the point spread function and the refraction feature associated with the corresponding theoretical value of the point spread function and thus the refraction feature of the eye of the subject, an optimal value of the point spread function being determined considering the modified picture with a different contrast level compared to an initial contrast level of the acquired picture,

each theoretical value of the point spread function of the plurality of theoretical values of the point spread function being determined theoretically from an arbitrary refraction feature which is thus associated with said theoretical value of the point spread function, said arbitrary refraction feature being used to determine Zernike coefficients and derive an expression of a wavefront to calculate said theoretical value of the point spread function of the plurality of theoretical values of the point spread function, the plurality of theoretical values of the point spread function being determined before an execution of the method for determining a refraction feature.

2 . The method according to claim 1 , wherein said refraction feature is determined by using a model relating said blur level of the acquired picture of the retina to the refraction feature.

3 . The method according to claim 1 , further comprising acquiring at least a picture of a pupil of the eye of the subject, the determining said refraction feature depending on a pupil diameter determined from said picture of the pupil.

4 . The method according to claim 3 , further acquiring another picture of the retina of the eye of the subject, said another picture is acquired at another focus distance comprised between the one of said acquired picture of the retina and the one of said picture of the pupil of the eye, said refraction feature also depending on said another picture of the retina, said another acquired picture being suitable for determining a sign of the refraction feature, the sign of the refraction feature being determined comparing the blur level between said another acquired picture of the retina and said acquired picture of the retina.

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

6 . The method according to claim 5 , wherein the convolutional neural network is trained using a dataset of couples of images, each couple of images being associated with a specific refraction feature.

7 . The method according to claim 6 , further comprising determining a value of a point spread function associated with the acquired picture of the retina by using the training of the convolutional neural network with the dataset of couples of images.

8 . The method according to claim 6 , further comprising determining the refraction feature associated with the acquired picture of the retina by using the training of the convolutional neural network with the dataset of couples of images.

9 . The method according to claim 1 , wherein the modified picture is calculated using a blind deconvolution method of the acquired picture of the retina.

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

11 . The method according to claim 10 , wherein the modified picture is determined selecting an optimal value of the point spread function among said dataset of values of the point spread function, the optimal value of the point spread function corresponding to an improved contrast level in the modified picture compared to an initial contrast level of the acquired picture of the retina.

12 . The method according to claim 1 , wherein the picture of the retina is acquired with an image capture device with a focus distance at infinity.

13 . The method according to claim 1 , wherein a distance between the eye of the subject and the image-capture device is higher than 20 mm.

14 . A system for determining a refraction feature of an eye of a subject comprising:

an image-capture device suitable for acquiring at least one picture of a retina of the eye of the subject; and

a data processor suitable for determining said refraction feature based on a blur level of said acquired picture of the retina, a blur level of a picture being determined using a value of a point spread function associated with said acquired picture, said data processor being suitable for determining the refraction feature by:

calculating a modified picture having a reduced blur level compared to an initial blur level of said acquired picture of the retina, said value of a point spread function associated with said acquired picture being derived from the acquired picture and the modified picture using a deconvolution method, and

identifying one theoretical value of a point spread function of a plurality of theoretical values of the point spread function which has a minimal difference between said value of the point spread function associated with said acquired picture and the one theoretical value of the point spread function of the plurality of theoretical values of the point spread functions, to determine the corresponding theoretical value of the point spread function and the refraction feature associated with the corresponding theoretical value of the point spread function and thus the refraction feature of the eye of the subject, or

optimizing the value of the point spread function associated with said acquired picture among the plurality of theoretical values of the point spread function to determine the corresponding theoretical value of the point spread function and the refraction feature associated with the corresponding theoretical value of the point spread function and thus the refraction feature of the eye of the subject, an optimal value of the point spread function being determined considering the modified picture with a different contrast level compared to an initial contrast level of the acquired picture,

said data processor being suitable for determining theoretically each theoretical value of the point spread function of the plurality of theoretical values of the point spread function from an arbitrary refraction feature which is thus associated with said theoretical value of the point spread function, said arbitrary refraction feature being used to determine Zernike coefficients and derive an expression of a wavefront to calculate said theoretical value of the point spread function of the plurality of theoretical values of the point spread function, the plurality of theoretical values of the point spread function being determined before an execution of determining a refraction feature.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Dec 29, 2021
From: BOUTINON, STEPHANE
To: ESSILOR INTERNATIONAL
Reel/Frame 058497/0728 →
Priority Claims (1)
EP 19305754 · Jun 13, 2019 · regional
Continuity (1)
Related Publication 20220301294A1 · Sep 22, 2022
References Cited (25)
US 7261412B2 · Somani · 2007 [cited by examiner]
US 7654674B2 · Hegels · 2010 [cited by examiner]
US 7997731B2 · Dai · 2011 [cited by examiner]
US 10548473B2 · Escalier · 2020 [cited by examiner]
US 11659989B2 · Marin · 2023 [cited by examiner]
US 11944383B2 · Suchkov · 2024 [cited by examiner]
US 20020140902A1 · Guirao et al. · 2002 [cited by applicant]
US 20070002274A1 · Somani · 2007 [cited by examiner]
US 20080246916A1 · Hegels · 2008 [cited by examiner]
US 20100103376A1 · Dai · 2010 [cited by examiner]
US 20180116500A1 · Escalier · 2018 [cited by examiner]
US 20200069174A1 · Marin · 2020 [cited by examiner]
US 20230380681A1 · Suchkov · 2023 [cited by examiner]
JP 2001120504A · 2001 [cited by applicant]
WO WO0230273A1 · 2002 [cited by applicant]
WO WO2007005261A2 · 2007 [cited by applicant]
Vasyl V. Molebny, Igor H. Chyzh, Vyacheslav M. Sokurenko, Ioannis G. Pallikaris, Leonidas P. Naoumidis, “Eye aberration analysis with Zernike polynomials,” Proc. SPIE 3246, Ophthalmic Technologies VIII, (Jun. 1, 1998); … [cited by examiner]
Lakshminarayanan, Vasudevan, and L. Srinivasa Varadharajan. “Chapter 14: Zernike Polynomials.” Special Functions for Optical Science and Engineering, SPIE, Bellingham, WA, 2015, pp. 321-338. (Year: 2015). [cited by examiner]
International Search Report issued on Sep. 4, 2020 in PCT/EP2020/066211 filed Jun. 11, 2020, 4 pages. [cited by applicant]
Lian, J., et al., “Deblurring retinal optical coherence tomography via a convolutional neural network with anisotropic and double convolution layer”, IET Computer Vision, The Institution of Engineering and Technology, v… [cited by applicant]
Chan, T.F., et al., “Total Variation Blind Deconvolution”, Transactions on Image Processing, vol. 7, No. 3, 1998, XP-000738213, pp. 370-375. [cited by applicant]
Uvais, Q., et al., “Blind Deconvolution for Retinal Image Enhancement”, Conference on Biomedical Engineering & Sciences, 2010, XP031936712, pp. 20-25. [cited by applicant]
Pamplona, V., et al., “NETRA: Interactive Display for Estimating Refractive Errors and Focal Range”, ACM Transactions on Graphics, vol. 29, No. 4, 2010, XP058157964, pp. 1-8. [cited by applicant]
Pertuz, S., et al., “Analysis of focus measure operators for shape-from-focus”, Pattern Recognition, vol. 46, 2013, pp. 1415-1432. [cited by applicant]
Japanese Office Action mailed Mar. 18, 2024 in corresponding Japanese Patent Application No. 2021-573910 (with English translation), 13 pages. [cited by applicant]