Fast wake-up in gaze tracking system

Patent metadata

Field	Value
Publication number	US-10591990-B2
Application number	US-201816232591-A
Country	US
Kind code	B2
Filing date	Dec 26, 2018
Priority date	Aug 9, 2012
Publication date	Mar 17, 2020
Grant date	Mar 17, 2020

Abstract

A gaze tracking system, leaving a low power mode in response to an activation signal, initial burst of eye pictures in short time by restricting the image area of a sensor, purpose of enabling an increased frame rate. Subsequent eye pictures are captured at le rate. The first gaze point value is computed memorylessly based on the initial burst res and no additional imagery, while subsequent values may be computed recursively to account previous gaze point values or information from previous eye pictures. The restriction of the image area may be guided by a preliminary overview picture captured using the different sensor. From the gaze point values, the system may derive a control signal to a computer device with a visual display.

First claim

What is claimed is:

1. A method in a gaze tracking system for controlling a computer device associated with a visual display, wherein the gaze tracking system tracks a gaze point on said visual display, said tracking system comprising a first sensor with a light-sensitive surface, wherein:
in response to a gaze tracking activation signal from input means of said computer device at said gaze tracking system, leaving a low power mode of said gaze tracking system;
capturing, an overview picture of at least a portion of the head of the viewer, and locating an eye region of the viewer in the overview picture, wherein the overview picture is captured as a depth map by a second sensor either alone or in combination with the first sensor;
capturing, at a first frame rate, a plurality of pictures of at least a portion of the viewer's head using said first sensor;
determining a first gaze point value of said viewer based on information from said plurality of pictures and optionally on additional information from the overview picture;
providing a first control signal based on said first gaze point value;
capturing, at a second frame rate, being lower than said first frame rate, further pictures of at least a portion of the viewer's head using a smaller portion or a restricted region of said first sensor;
determining a further gaze point value of said viewer based on information from at least one of said further pictures; and
providing a further control signal based on said further gaze point.

2. The method according to claim 1 , wherein:
said first sensor includes a light-sensitive surface arranged to receive light from the head of the viewer;
said plurality of pictures is captured by reading out data from said first sensor in restricted regions of a light-sensitive surface; and
said further pictures are captured by reading out data from a less restricted or unrestricted region of the light-sensitive surface.

3. The method according to claim 1 , wherein:
said first sensor includes a light-sensitive surface arranged to receive light from the head of the viewer;
said plurality of pictures is captured by reading out data from said first sensor in restricted regions of a light-sensitive surface, which are positioned in accordance with the eye region location and the union of which occupies at most 10 percent of the area of the light-sensitive surface; and
said further pictures are captured by reading out data from a less restricted or unrestricted region of the light-sensitive surface.

4. The method according to claim 1 , wherein the determination of the first gaze point is based on information from said plurality of pictures and on additional information from the overview picture.

5. The method according to claim 1 , the determination of a further gaze point is based on information from said further picture and additionally on information from at least one previous picture.

6. The method according to claim 1 , wherein the restricted region comprises eye region location derived from the overview picture.

7. The method according to claim 1 , wherein the first sensor is CMOS-type sensor and the second sensor is CMOS-type sensor or a time of flight sensor.

8. The method according to claim 1 , wherein the depth map is created by triangulation of the outputs of the first and second sensor.

9. The method according to claim 1 , wherein the depth map is created by operating the second sensor while simultaneously illuminating the viewer's head by structured or coded light from a light source.

10. The method according to claim 9 , wherein the first sensor is a higher-resolution CMOS-type sensor and the second sensor is a lower-resolution CMOS-type sensors, wherein a depth map is created by operating the second sensor while simultaneously illuminating the viewer's head by structured or coded light from a light source.

11. The method according to claim 1 , wherein the first sensor is CMOS-type sensor and the second sensor is a time-of-flight sensor.

12. A gaze tracking system for controlling a computer device associated with a visual display, wherein the gaze tracking system tracks a gaze point on said visual display, the system comprising:
a first sensor with a light-sensitive surface arranged to receive light from the head of a viewer, the first sensor being selectively operable at a second frame rate or a first frame rate; and
a processor configured to respond to receipt, in a low power mode of the eye tracking system, of a gaze tracking activation signal by performing the following:
causing the eye tracking system to leave the low power mode;
capturing an overview picture of at least a portion of the head of the viewer and locating an eye region of the viewer in the overview picture, wherein the overview picture is captured as a depth map by a second sensor either alone or in combination with the first sensor;
capturing a plurality of pictures of at least a portion of the viewer's head using said first sensor, from which data are read out at a first frame rate;
determining a first gaze point value of said viewer based on information from said plurality of pictures;
providing a first control signal based on said first gaze point value;
capturing further pictures of at least a portion of the viewer's head using a smaller portion of the or a restricted region of said first sensor, from which data are read out at a second frame rate, being lower than said first frame rate;
determining, for each further picture, a further gaze point value of said viewer on the basis of at least one of said further pictures; and
providing a further control signal based on said further gaze point.

13. A gaze tracking system according to claim 12 , wherein:
the first sensor receives light from the head of a viewer of the visual display, and the first sensor is selectively operable to read out the light-sensitive surface at a second frame rate or a restricted region at a first frame rate, being higher than said second frame rate; and
wherein the capturing of the plurality of pictures from which data are read out at a first frame rate and in restricted regions;
wherein the determining a first gaze point value of said viewer on the visual display based on information from said plurality of pictures and optionally on additional information from the overview picture; and
determining, for each further picture, a further gaze point value of said viewer on the basis said further picture and additionally of at least one previous picture.

14. The system according to claim 12 , wherein the first sensor is CMOS-type sensor and the second sensor is CMOS-type sensor or a time of flight sensor.

15. The system according to claim 12 , wherein the depth map is configured to be created by triangulation of the outputs of the first and second sensor.

16. The method according to claim 15 , wherein the first sensor is a higher-resolution CMOS-type sensor and the second sensor is a lower-resolution CMOS-type sensors, wherein a depth map is configured to be created by operating the second sensor while simultaneously illuminating the viewer's head by structured or coded light from a light source.

17. The method according to claim 12 , wherein the depth map is configured to be created by operating the second sensor while simultaneously illuminating the viewer's head by structured or coded light from a light source.

18. The method according to claim 12 , wherein the first sensor is CMOS-type sensor and the second sensor is a time-of-flight sensor.

19. A non-transitory computer rea

Assignees

• Tobii Ab

Inventors

• Skogö Mårten
• Olsson Anders
• Elvesjö John Mikael
• Yu Aron

Classifications

• G06F1/325

  Power saving in peripheral device · CPC title

• G06F3/013 Primary

  Eye tracking input arrangements (G06F3/015 takes precedence) · CPC title