Eye open or closed state detection method and electronic device

1. An eye open or closed state detection method, comprising:

obtaining a grayscale image of an eye on a human face in a to-be-processed face image;

extracting an eye feature from the grayscale image, wherein the eye feature comprises an eye opening feature, wherein the eye feature further comprises an iris shape feature or a vertical direction feature, wherein the eye opening feature represents an eye opening degree, wherein the iris shape feature represents a shape of the iris, and wherein the vertical direction feature represents a change degree of an eyelid curve;

obtaining a plurality of smallest pixels in a first plurality of pixels comprised in the grayscale image, wherein a smallest pixel of the smallest pixels comprises a smallest pixel value in a second plurality of pixels with a same horizontal coordinate in an image coordinate system, and wherein a quantity of the smallest pixels is equal to a quantity of selected horizontal coordinates;

obtaining the vertical direction feature based on vertical coordinates of the smallest pixels; and

identifying an eye open state or an eye closed state based on the eye feature.

2. The eye open or closed state detection method of claim 1 , further comprising:

performing binarization processing on the grayscale image to obtain a binary image of the grayscale image;

obtaining a convex set in a largest connected region of the binary image;

obtaining distances, in a horizontal direction and a vertical direction of an image coordinate system, between every two pixels comprised in the convex set;

determining a first largest distance of the distances in the horizontal direction;

determining a second largest distance of the distances in the vertical direction; and

determining the iris shape feature based on the first largest distance and the second largest distance, wherein the iris shape feature is a first ratio of the first largest distance to the second largest distance, a second ratio of the second largest distance to the first largest distance, or a smaller value of the first ratio and the second ratio.

3. The eye open or closed state detection method of claim 2 , further comprising:

traversing the grayscale image based on a specified rectangular window and a specified sliding step;

determining an average pixel value of pixels in the specified rectangular window to obtain a plurality of average pixel values; and

performing binarization processing on the grayscale image using a smallest pixel value in the average pixel values as a threshold.

4. The eye open or closed state detection method of claim 1 , further comprising:

obtaining a plurality of absolute values of differences between vertical coordinates of every two pixels with adjacent horizontal coordinates in the smallest pixels;

determining a plurality of feature scores corresponding to each of the absolute values; and

using a sum of the feature scores as the vertical direction feature.

5. The eye open or closed state detection method of claim 4 , wherein a feature score scoreV j corresponding to an absolute value V j is based on the following:

if A<V j ≤B , score V j =V j −A;

if B<V j ≤C , score V j =k 1*( V j −A );

if C<V j , score V j =k 2*( V j −A ),

wherein A, B, C, k1 and k2 are positive numbers, wherein k1<k2, and wherein A<B<C.

6. The eye open or closed state detection method of claim 1 , further comprising:

obtaining a plurality of grayscale images of eyes on human faces in a plurality of face images in a face image training library;

extracting eye features corresponding to the grayscale images;

constructing a classification model based on the eye features and pre-marked eye open states or pre-marked eye closed states corresponding to the grayscale images; and

identifying the eye open state or the eye closed state based on the eye features and the classification model.

7. An electronic device, comprising:

a processor; and

a memory coupled to the processor and configured to store instructions that, when executed by the processor, cause the electronic device to be configured to:

obtain a grayscale image of an eye on a human face in a to-be-processed face image;

extract an eye feature from the grayscale image, wherein the eye feature comprises an eye opening feature, wherein the eye feature further comprises an iris shape feature or a vertical direction feature, wherein the eye opening feature represents an eye opening degree, wherein the iris shape feature represents a shape of the iris, and wherein the vertical direction feature represents a change degree of an eyelid curve;

obtain a plurality of smallest pixels in a first plurality of pixels comprised in the grayscale image, wherein a smallest pixel of the smallest pixels comprises a smallest pixel value in a second plurality of pixels with a same horizontal coordinate in an image coordinate system, and wherein a quantity of the smallest pixels is equal to a quantity of selected horizontal coordinates;

obtain the vertical direction feature based on vertical coordinates of the smallest pixels; and

identify an eye open state or an eye closed state based on the eye feature.

8. The electronic device of claim 7 , wherein the instructions further cause the electronic device to be configured to:

perform binarization processing on the grayscale image to obtain a binary image of the grayscale image;

obtain a convex set in a largest connected region of the binary image;

obtain distances, in a horizontal direction and a vertical direction of an image coordinate system, between every two pixels comprised in the convex set;

determine a first largest distance of the distances in the horizontal direction;

determine a second largest distance of the distances in the vertical direction; and

determine the iris shape feature based on the first largest distance and the second largest distance, wherein the iris shape feature is a first ratio of the first largest distance to the second largest distance, a second ratio of the second largest distance to the first largest distance, or a smaller value of the first ratio and the second ratio.

9. The electronic device of claim 8 , wherein the instructions further cause the electronic device to be configured to:

traverse the grayscale image based on a specified rectangular window and a specified sliding step;

determine an average pixel value of pixels in the specified rectangular window to obtain a plurality of average pixel values; and

perform binarization processing on the grayscale image using a smallest pixel value in the average pixel values as a threshold.

10. The electronic device of claim 7 , wherein the instructions further cause the electronic device to be configured to:

obtain a plurality of absolute values of differences between vertical coordinates of every two pixels with adjacent horizontal coordinates in the smallest pixels;

determine a plurality of feature scores corresponding to each of the absolute values; and

use a sum of the feature scores as the vertical direction feature.

11. The electronic device of claim 10 , wherein a feature score scoreV j corresponding to an absolute value V j is based on the following:

if A<V j ≤B , score V j =V j −A;

if B<V j ≤C , score V j =k 1*( V j −A );

if C<V j , score V j =k 2*( V j −A ),

wherein A, B, C, k1 and k2 are positive numbers, wherein k1<k2, and wherein A<B<C.

12. The electronic device of claim 7 , wherein the instructions further cause the electronic device to be configured to:

obtain a plurality of grayscale images of eyes on human faces in a plurality of face images in a face image training library;

extract eye features corresponding to the grayscale images;

construct a classification model based on the eye features and pre-marked eye open states or pre-marked eye closed states corresponding to the grayscale images; and

identify the eye open state or the eye closed state based on the eye features and the classification model.

13. A computer program product comprising computer-executable instructions stored on a non-transitory computer-readable medium that, when executed by a processor, cause an electronic device to:

obtain a grayscale image of an eye on a human face in a to-be-processed face image;

extract an eye feature from the grayscale image, wherein the eye feature comprises an eye opening feature, wherein the eye feature further comprises an iris shape feature or a vertical direction feature, wherein the eye opening feature represents an eye opening degree, wherein the iris shape feature represents a shape of the iris, and wherein the vertical direction feature represents a change degree of an eyelid curve;

obtain a plurality of smallest pixels in a first plurality of pixels comprised in the grayscale image, wherein a smallest pixel of the smallest pixels comprises a smallest pixel value in a second plurality of pixels with a same horizontal coordinate in an image coordinate system, and wherein a quantity of the smallest pixels is equal to a quantity of selected horizontal coordinates;

obtain the vertical direction feature based on vertical coordinates of the smallest pixels; and

identify an eye open state or an eye closed state based on the eye feature.

14. The computer program product of claim 13 , wherein the instructions further cause the electronic device to:

perform binarization processing on the grayscale image to obtain a binary image of the grayscale image;

obtain a convex set in a largest connected region of the binary image;

obtain distances, in a horizontal direction and a vertical direction of an image coordinate system, between every two pixels comprised in the convex set;

determine a first largest distance of the distances in the horizontal direction;

determine a second largest distance of the distances in the vertical direction; and

determine the iris shape feature based on the first largest distance and the second largest distance, wherein the iris shape feature is a first ratio of the first largest distance to the second largest distance, a second ratio of the second largest distance to the first largest distance, or a smaller value of the first ratio and the second ratio.

15. The computer program product of claim 14 , wherein the instructions further cause the electronic device to:

traverse the grayscale image based on a specified rectangular window and a specified sliding step;

determine an average pixel value of pixels in the specified rectangular window to obtain a plurality of average pixel values; and

perform binarization processing on the grayscale image by using a smallest pixel value in the average pixel values as a threshold.

16. The computer program product of claim 13 , wherein the instructions further cause the electronic device to:

obtain a plurality of absolute values of differences between vertical coordinates of every two pixels with adjacent horizontal coordinates in the smallest pixels;

determine a plurality of feature scores corresponding to each of the absolute values; and

use a sum of the feature scores as the vertical direction feature.

17. The computer program product of claim 13 , wherein the instructions further cause the electronic device to:

obtain a plurality of grayscale images of eyes on human faces in a plurality of face images in a face image training library;

extract eye features corresponding to the grayscale images;

construct a classification model based on the eye features and pre-marked eye open or pre-marked eye closed states corresponding to the grayscale images; and

identify the eye open state or the eye closed state based on the eye features and the classification model.