Image recognition apparatus, image recognition method and computer-readable medium

1. An image recognition apparatus comprising:

a memory storing instructions; and

at least one processor configured to process the instructions to:

determine an occlusion pattern of a face in a facial image representing the face by comparing the facial image with a standard facial image representing a standard face; and

identify the facial image using a region determined by the occlusion pattern in the facial image;

wherein the at least one processor determines the occlusion pattern further on a basis of:

a distribution of a reliability on a basis of comparison of small regions included in the facial image with small regions in the standard facial image corresponding to the small regions in the facial image, and

a medium region including a group of the small regions within a predetermined range, when a distribution of a reliability of the group of the small regions satisfies a predetermined occlusion condition.

2. The image recognition apparatus according to claim 1 , wherein

the medium region is an upper face region including a group of the small regions above a nose in the facial image.

3. The image recognition apparatus according to claim 2 , wherein the at least one processor is further configured to process the instructions to:

determine that the occlusion pattern is an occlusion pattern with sunglasses, when a distribution of a reliability in the upper face region satisfies the occlusion condition.

4. The image recognition apparatus according to claim 1 , wherein

the medium region is a lower face region including a group of the small regions below a nose in the facial image.

5. The image recognition apparatus according to claim 4 , wherein the at least one processor is further configured to process the instructions to:

determine that the occlusion pattern is an occlusion pattern with a mask, when a distribution of a reliability in the lower face region satisfies the occlusion condition.

6. The image recognition apparatus according to claim 1 , wherein the at least one processor is further configured to process the instructions to:

divide the facial image into the small regions, and

calculate the reliability by comparing the corresponding small regions of the facial image and the standard facial image with each other.

7. The image recognition apparatus according to claim 1 , wherein the at least one processor is further configured to process the instructions to:

identify for each of the small regions determined by the occlusion pattern in the facial image, and

identify the facial image on a basis of a result of identifying for each of the small regions.

8. The image recognition apparatus according to claim 1 , wherein the at least one processor is further configured to process the instructions to:

identify the facial image by using information on an entire area of the face, and

determine whether to use a result of identifying the facial image by using information on the region determined by the occlusion pattern in the facial image or a result of identifying the facial image by using information on an entire area of the face.

9. The image recognition apparatus according to claim 1 , wherein the at least one processor is further configured to process the instructions to:

the region determined by the occlusion pattern in the facial image is obtained by excluding a region based on the occlusion pattern in the facial image.

10. The image recognition apparatus according to claim 7 , wherein the at least one processor is further configured to process the instructions to:

identify for each of the small regions other than a region based on the occlusion pattern in the facial image.

11. The image recognition apparatus according to claim 1 , wherein

the standard facial image is obtained by averaging the facial images of a plurality of persons.

12. The image recognition apparatus according to claim 1 , wherein

the standard facial image is an image whose similarity to facial images of a plurality of persons satisfies a predetermined condition.

13. The image recognition apparatus according to claim 1 , wherein

the standard facial image is a facial image representing an artificial object of a standard face.

14. The image recognition apparatus according to claim 1 , wherein

the facial image and the standard facial image include information indicated by feature vectors.

15. An image recognition method comprising:

determining an occlusion pattern of a face in a facial image representing the face by comparing the facial image with a standard facial image representing a standard face; and

identifying the facial image using a region determined by the occlusion pattern in the facial image,

wherein the occlusion pattern is determined further on a basis of:

a distribution of a reliability on a basis of comparison of small regions included in the facial image with small regions in the standard facial image corresponding to the small regions in the facial image, and

a medium region including a group of the small regions within a predetermined range, when a distribution of a reliability of the group of the small regions satisfies a predetermined occlusion condition.

16. A non-transitory computer-readable storage medium storing instructions that when executed by a computer enable the computer to implement a method comprising:

determining an occlusion pattern of a face in a facial image representing the face by comparing the facial image with a standard facial image representing a standard face; and

identifying the facial image using a region determined by the occlusion pattern in the facial image,

wherein the occlusion pattern is determined further on a basis of:

a distribution of a reliability on a basis of comparison of small regions included in the facial image with small regions in the standard facial image corresponding to the small regions in the facial image, and

a medium region including a group of the small regions within a predetermined range, when a distribution of a reliability of the group of the small regions satisfies a predetermined occlusion condition.