Eye opening degree detection system, doze detection system, automatic shutter system, eye opening degree detection method, and eye opening degree detection program

Patent metadata

Field	Value
Publication number	US-9984299-B2
Application number	US-201715428680-A
Country	US
Kind code	B2
Filing date	Feb 9, 2017
Priority date	Feb 15, 2016
Publication date	May 29, 2018
Grant date	May 29, 2018

Abstract

An object of the present invention is to provide an eye opening degree detection system that can accurately and stably calculate an eye opening degree. The eye opening degree detection system includes imaging devices that generate images including regions of both eyes, a one eye opening degree calculation unit that calculates each one eye opening degree of the left and right eyes of a first image, an eye opening degree selection unit that selects a correctly-calculated one eye opening degree, an eye opening degree calculation unit that calculates an eye opening degree on the basis of the one eye opening degree and an eye opening degree determination unit that compares the eye opening degree calculated on the basis of the first image with an eye opening degree calculated on the basis of a second image prior to the first image to determine the propriety of the eye opening degree.

First claim

What is claimed is:

1. An eye opening degree detection system comprising:
a plurality of imaging devices that generates images including, at least, regions of both eyes of a target person while shifting time;
a one eye opening degree calculation unit that calculates each one eye opening degree of the left and right eyes of a first image;
an eye opening degree selection unit that creates a pixel value histogram of each region of the left and right eyes when the left and right one eye opening degrees do not substantially match each other, and selects a correctly-calculated one eye opening degree on the basis of the pixel value histogram;
an eye opening degree calculation unit that calculates an eye opening degree on the basis of the left and right one eye opening degrees when the left and right one eye opening degrees substantially match each other, and calculates an eye opening degree on the basis of the one eye opening degree selected by the eye opening degree selection unit when the left and right one eye opening degrees do not substantially match each other; and
an eye opening degree detection device having an eye opening degree determination unit that compares the eye opening degree calculated on the basis of the first image with an eye opening degree calculated on the basis of a second image prior to the first image to determine the propriety of the eye opening degree calculated on the basis of the first image.

2. The eye opening degree detection system according to claim 1 ,
wherein the eye opening degree determination unit determines that the eye opening degree calculated on the basis of the first image is valid when a difference between the eye opening degree calculated on the basis of the first image and the eye opening degree calculated on the basis of the second image is equal to or smaller than a threshold value calculated on the basis of an imaging timing difference between the first image and the second image.

3. A doze detection system comprising:
the eye opening degree detection system according to claim 1 ;
a graph creation unit that creates an eye opening degree transition graph on the basis of an eye opening degree determined by an eye opening degree determination unit; and
a doze detection unit that determines an eye opening period of time and an eye closing period of time on the basis of the eye opening degree transition graph, and detects dozing of the target person on the basis of a ratio of the eye opening period of time to the eye closing period of time in a predetermined period of time.

4. The doze detection system according to claim 3 ,
wherein when the ratio is out of a range between a predetermined lower limit threshold value excluding 0 and a predetermined upper limit threshold value excluding infinity, the doze detection unit detects dozing of the target person.

5. An automatic shutter system comprising:
the eye opening degree detection system according to claim 1 ;
a graph creation unit that creates an eye opening degree transition graph on the basis of an eye opening degree determined by an eye opening degree determination unit, and
a timing prediction unit that calculates an average eye opening period of time on the basis of the eye opening degree transition graph, and determines shutter time on the basis of the average eye opening period of time and the eye opening degree to allow the imaging devices to image.

6. An eye opening degree detection method comprising the steps of:
generating images including, at least, regions of both eyes of a target person while shifting time by a plurality of imaging devices;
calculating each one eye opening degree of the left and right eyes of a first image;
creating a pixel value histogram of each region of the left and right eyes when the left and right one eye opening degrees do not substantially match each other, and selecting a correctly-calculated one eye opening degree on the basis of the pixel value histogram;
calculating an eye opening degree on the basis of the left and right one eye opening degrees when the left and right one eye opening degrees substantially match each other, and calculating an eye opening degree on the basis of the selected one eye opening degree when the left and right one eye opening degrees do not substantially match each other, and
determining the propriety of the eye opening degree calculated on the basis of the first image by comparing the eye opening degree calculated on the basis of the first image with an eye opening degree calculated on the basis of a second image prior to the first image.

7. An eye opening degree detection program allowing a computer to execute the procedures of:
storing a plurality of images that includes, at least, regions of both eyes of a target person and is obtained by being imaged while shifting time by a plurality of imaging devices;
calculating each one eye opening degree of the left and right eyes of a first image;
creating a pixel value histogram of each region of the left and right eyes when the left and right one eye opening degrees do not substantially match each other, and selecting a correctly-calculated one eye opening degree on the basis of the pixel value histogram;
calculating an eye opening degree on the basis of the left and right one eye opening degrees when the left and right one eye opening degrees substantially match each other, and calculating an eye opening degree on the basis of the selected one eye opening degree when the left and right one eye opening degrees do not substantially match each other, and
determining the propriety of the eye opening degree calculated on the basis of the first image by comparing the eye opening degree calculated on the basis of the first image with an eye opening degree calculated on the basis of a second image prior to the first image.

Assignees

• Renesas Electronics Corp

Inventors

• Mori Yuki

Classifications

• G08B21/06 Primary

  indicating a condition of sleep, e.g. anti-dozing alarms · CPC title

• G06V20/597

  Recognising the driver's state or behaviour, e.g. attention or drowsiness · CPC title

• H04N23/611

  where the recognised objects include parts of the human body · CPC title

• H04N23/73

  by influencing the exposure time · CPC title

• G06F18/29

  Graphical models, e.g. Bayesian networks · CPC title