Eye tracking method and apparatus

• US 5,231,674 A
• Filed: 05/13/1991
• Issued: 07/27/1993
• Est. Priority Date: 06/09/1989
• Status: Expired due to Term

• Current Assignee: LC Technologies Incorporated
• Original Assignee: LC Technologies Incorporated
• Inventors: Cleveland, Dixon, Cleveland, James H., Norloff, Peter L.
• Primary Examiner(s): Moore, David K.
• Assistant Examiner(s): Cammarata, Michael
• Application Number:
  US07/702,139
• Time in Patent Office: 806 Days
• Field of Search: 382/49, 382/29, 382/27, 382/22, 382/6, 382/54, 382/9, 382/48, 351/210, 351/209
• US Class Current: 382/117
• CPC Class Codes: A61B 3/113 for determining or recording eye movement
• CPC Class Codes: G06T 7/73 using feature-based methods
• CPC Class Codes: G06V 10/20 Image preprocessing
• CPC Class Codes: G06V 40/193 Preprocessing; Feature extraction

Abstract

In applications where it is desired to determine the locations of image features, such as eye monitoring to determine the direction that a person is gazing, determining the point at which he is gazing, or measuring the motions of his eye using a camera to capture an optical image of the eye and image processing to extract information about the eye'"'"'s gaze point and/or orientation, there is provided a method and apparatus for precise location of image features such as edge coordinates between the pupil anc iris of the eye and of the center coordinates of light reflections off the cornea of the eye.

Claims

1. A computerized method for locating a corneal reflection within an image of an eye comprising the steps of:

illuminating the eye with a light source to produce light reflected by the eye;

using a camera to convert the light reflected by the eye into a two-dimensional image of at least a portion of the eye, said image including a corneal reflection;

converting a segment of the image representative of the corneal reflection into a sequence of gray scale reflection data;

thresholding the gray scale reflection data by subtracting a predetermined threshold value from each element of the sequence to find the magnitudes of intensities above the threshold;

scaling the thresholded corneal reflection data according to a monotonic scale prior to finding the center of mass; and

locating the corneal reflection by finding a center of mass of the scaled thresholded data.

2. The computerized method as set forth in claim 1, further comprising the step of smoothing the collected data prior to thresholding.

3. An apparatus for locating a corneal reflection within an image of an eye comprising:

a light source illuminating the eye to produce light reflected by the eye;

a camera capturing light reflected by the eye to produce a two-dimensional image of at least a portion of the eye, said image including a corneal reflection;

means for converting a segment of the image representative of the corneal reflection into a sequence of gray scale reflection data;

means for thresholding the gray scale reflection data by subtracting a predetermined threshold value from each element of the sequence to find the magnitude of the intensities above the threshold;

means for scaling the thresholded corneal reflection data according to a monotonic scale prior to finding the center of mass; and

means for locating the corneal reflection by finding a center of mass of the scaled thresholded data.

4. The apparatus as set forth in claim 3, further comprising means for smoothing the collected sequence of gray scale data prior to thresholding.

5. A computerized method for locating an edge feature within an image of an object comprising the steps of:

illuminating the object with a light source to produce light reflected by the object;

using a camera to convert light reflected by the object into a two-dimensional image of at least a portion of the object, said image including the edge feature;

converting a segment of the image representative of the edge feature into a sequence of gray scale reflection data;

emphasizing the edge feature by spatially differentiating the sequence of gray scale reflection data to obtain a sequence of emphasized derivative data;

thresholding the emphasized derivative data by subtracting a predetermined threshold value from each element of the sequence to find the magnitudes of the data above the threshold; and

locating the edge feature by finding a center of mass of the thresholded emphasized derivative data, wherein the center of mass is obtained by the application of selected weights to the image coordinates, and wherein the selected weight at each image coordinate increases monotonically with the thresholded emphasized data for that coordinate.

6. The computerized method as set forth in claim 5, further comprising the step of smoothing the emphasized derivative data prior to thresholding.

7. The computerized method as set forth in claim 5, wherein the image is that of an eye and the edge feature is an edge feature of one of a pupil and iris of said eye.

8. The computerized method as set forth in claim 6, wherein the image is that of an eye and the edge feature is an edge feature of one of a pupil and iris of said eye.

9. A computerized method for locating a corneal reflection within an image of an eye comprising the steps of:

illuminating the eye with a light source to produce light reflected by the eye;

using a camera to convert the light reflected by the eye into a two-dimensional image of at least a portion of the eye, said image including a corneal reflection;

converting a segment of the image representative of the corneal reflection into a sequence of gray scale reflection data;

thresholding the gray scale reflection data by subtracting a predetermined threshold value from each element of the sequence to find the magnitude of intensities above the threshold; and

locating the corneal reflection by finding a center of mass of the thresholded data, wherein the center of mass is obtained by the application of selected weights to the image coordinates, and wherein the selected weight at each image coordinate increases monotonically with the thresholded data for that coordinate.

10. The computerized method as set forth in claim 9, further comprising the step of smoothing the collected data prior to thresholding.

11. The computerized method as set forth in claim 9, further comprising the step of scaling the thresholded data according to a monotonic scale prior to finding the center of mass.

12. An apparatus for locating an edge feature within an image of an object comprising:

a light source illuminating the object to produce light reflected by the object;

a camera capturing light reflected by the object to produce a two-dimensional image of at least a portion of the object, said image including the edge feature;

means for converting a segment of the image representative of the edge feature into a sequence of gray scale reflection data;

means for emphasizing the edge feature by spatially differentiating the sequence of gray reflection scale data to obtain a sequence of emphasized derivative data;

means for thresholding the emphasized derivative data by subtracting a predetermined threshold value from each element of the sequence to find the magnitudes of the emphasized data above the threshold; and

means for locating the edge feature by finding a center of mass of the thresholded emphasized data, wherein the center of mass is obtained by the application of selected weights to the image coordinates, and wherein the selected weight at each image coordinate increases monotonically with the thresholded emphasized data for that coordinate.

13. The method as set forth in claim 12, further comprising the step of smoothing the emphasized derivative data prior to thresholding.

14. The method as set forth in claim 12, wherein the image is that of an eye and the edge feature is an edge feature of one of a pupil and iris of said eye.

15. The method as set forth in claim 12, wherein the image is that of an eye and the edge feature is an edge feature of one of a pupil and iris of said eye.

16. An apparatus for locating corneal reflection within an image of an eye comprising:

a light source illuminating the eye to produce light reflected by the eye;

a camera capturing light reflected by the eye to produce a two-dimensional image of at least a portion of the eye, said image including a corneal reflection;

means for converting a segment of the image representative of the corneal reflection into a sequence of gray scale reflection data;

means for threshodling the gray scale reflection data by subtracting a predetermined threshold value from each element of the sequence to find the magnitude of the intensities above the threshold; and

means for locating the corneal reflection by finding a center of mass of the thresholded data, wherein the center of mass is obtained by the application of selected weights to the image coordinates, and wherein the selected weight at each image coordinate increases monotonically with the thresholded data for that coordinate.

17. The apparatus as set forth in claim 16, further comprising means for smoothing the collected sequence of gray scale data prior to thresholding.

18. The apparatus as set forth in claim 16, further comprising means for scaling the thresholded data according to a monotonic scale prior to finding the center of mass.