Computer implemented methods and devices for determining dimensions and distances of head features

Patent metadata

Field	Value
Publication number	US-12586229-B2
Application number	US-202218066428-A
Country	US
Kind code	B2
Filing date	Dec 15, 2022
Priority date	Mar 16, 2020
Publication date	Mar 24, 2026
Grant date	Mar 24, 2026

Abstract

Computer implemented methods and devices for determining dimensions or distances of head features are provided. The method includes identifying a plurality of features in an image of a head of a person. A real dimension of at least one target feature of the plurality of features or a real distance between at least one target feature of the plurality features and a camera device used for capturing the image is estimated based on probability distributions for real dimensions of at least one feature of the plurality of features and a pixel dimension of the at least one feature of the plurality of features.

First claim

The invention claimed is:

1 . A computer implemented method for estimating or determining dimensions or distances of head features, the method comprising:
providing an image of a head of a person; identifying a plurality of features in the image including a landmark; obtaining a probability distribution for a number of pixels per millimeter for the image based on a machine learning approach including a Gaussian process, wherein the probability distribution for the number of pixels per millimeter for the image is further based on pixel dimensions of at least two features of the plurality of features and a respective probability distribution of a real dimension for each of the at least two features of the plurality of features; and estimating a real distance between at least one target feature of the plurality of features and a camera device used for capturing the image based on the probability distribution for the number of pixels per millimeter for the image and a pixel dimension of at least one feature of the plurality of features, wherein the machine learning approach includes extracting an image gradient orientation from training images and aligning respective pixels of the landmark for the training images.

2 . The method of claim 1 , wherein the machine learning approach is trained based on training data where a real dimension is known.

3 . The method of claim 1 , wherein the at least one target feature of the plurality of features is an interpupillary distance.

4 . The method of claim 3 , further comprising:
updating a multivariate Gaussian distribution.

5 . The method of claim 4 , further comprising:
estimating the interpupillary distance based on the multivariate Gaussian distribution.

6 . The method of claim 1 , wherein the plurality of features comprises one or more features taken from the group consisting of:
an interpupillary distance, an iris diameter, a pupil diameter, a vertical ear length, a Menton-Sellion distance, a bizygomatic breadth, an Euryon breadth, an eye width, and a head height.

7 . The method of claim 1 , further comprising:
providing additional information regarding the person; and selecting the respective probability distribution of the real dimension for each of the at least two features of the plurality of features based on the additional information.

8 . The method of claim 7 , wherein providing the additional information comprises:
receiving the additional information as a user input; and/or determining the additional information based on the image.

9 . The method of claim 7 , wherein the additional information comprises one or more of a sex of the person, an age of the person, an ethnicity of the person, or a size of the person.

10 . The method of claim 1 , wherein providing the image comprises providing a plurality of images, wherein the estimating is performed based on the plurality of images.

11 . The method of claim 1 , further comprising one or more of:
fitting a spectacle frame to the head of the person based on the estimating; manufacturing spectacle glasses based on the estimating; and performing an eye examination based on the estimating.

12 . A computer program being stored on a non-transitory storage medium and comprising instructions which, when carried out on one or more processors, cause execution of the method of claim 1 .

13 . A device, comprising:
means for providing an image of a head of a person, means for identifying a plurality of features in the image including a landmark, means for calculating a probability distribution for a number of pixels per millimeter for the image based on a machine learning approach including a Gaussian process, wherein the probability distribution for the number of pixels per millimeter for the image is further based on pixel dimensions of at least two features of the plurality of features and a respective probability distribution of a real dimension for each of the at least two features of the plurality of features, and means for estimating a real distance between at least one target feature of the plurality of features and a camera device used for capturing the image based on the probability distribution for the number of pixels per millimeter for the image and a pixel dimension of at least one feature of the plurality of features, wherein the machine learning approach includes extracting an image gradient orientation from training images and aligning respective pixels of the landmark for the training images.

14 . The device of claim 13 , wherein the device is configured to carry out a method for estimating or determining dimensions or distances of head features, the device being configured to:
provide the image of the head of the person; identify the plurality of features in the image; calculate the probability distribution for the number of pixels per millimeter for the image based on the machine learning approach including the Gaussian process, the pixel dimensions of the at least two features of the plurality of features and the respective probability distribution of the real dimension for each of the at least two features of the plurality of features; and estimate the real distance between the at least one target feature of the plurality of features and the camera device used for capturing the image based on the probability distribution for the number of pixels per millimeter for the image and the pixel dimension of the at least one feature of the plurality of features.

15 . A computer implemented method for estimating or determining dimensions or distances of head features, the method comprising:
providing an image of a head of a person; identifying at least two features and at least one target feature in the image; obtaining a probability distribution for a number of pixels per millimeter for the image based on a machine learning approach including a Gaussian process based on prior information, wherein the probability distribution for the number of pixels per millimeter for the image is further based on pixel dimensions of the at least two features and a respective probability distribution of a real dimension for each of the at least two features; and estimating at least one of a real dimension of the at least one target feature or a real distance between the at least one target feature and a camera device used for capturing the image based on the probability distribution for the number of pixels per millimeter for the image and a pixel dimension of at least one feature in the image, wherein the at least two features are different from the at least one target feature, wherein the prior information includes information about size distributions of the at least two features, and wherein the information about the size distributions is obtained from a plurality of heads.

16 . The method of claim 15 , wherein the prior information is based on statistics on interpupillary distances.

17 . The method of claim 15 , wherein the prior information further includes information about covariances between dimensions of different features.

18 . The method of claim 15 , wherein a single image of the head of the person is provided.

19 . A computer implemented method for estimating or determining dimensions or distances of head features, the method comprising:
providing an image of a head of a person; identifying a plurality of features in the image; calculating a probability distribution for a number of pixels per millimeter for the image based on a machine learning a

Assignees

• Zeiss Carl Vision Int Gmbh

Inventors

• Hoyle Ben

Classifications

• A61B5/1072

  measuring distances on the body, e.g. measuring length, height or thickness (A61B5/1076 takes precedence) · CPC title

• A61B3/111

  for measuring interpupillary distance · CPC title

• G06V40/171 Primary

  Local features and components; Facial parts (eye characteristics G06V40/18); Occluding parts, e.g. glasses; Geometrical relationships · CPC title

• G06V40/161

  Detection; Localisation; Normalisation · CPC title

• G01B11/026

  by measuring distance between sensor and object (G01B11/0608 takes precedence) · CPC title