Method and apparatus with liveness testing

1. A method with liveness testing, comprising:

acquiring an infrared (IR) image including an object, and a depth image including the object;

generating a first preprocessed IR image by performing first edge enhancement preprocessing on the IR image;

generating a preprocessed depth image by performing second edge enhancement preprocessing on the depth image; and

determining whether the object is a genuine object based on the first preprocessed IR image and the preprocessed depth image,

wherein the generating of the first preprocessed IR image comprises:

generating a first intermediate image based on pixel values of a current pixel and neighboring pixels in the IR image;

generating a second intermediate image by performing normalization on the IR image; and

generating the first preprocessed IR image based on the IR image, the first intermediate image, and the second intermediate image,

wherein a pixel of the first preprocessed IR image has a pixel value of a pixel at a corresponding position in the IR image, a pixel value of a pixel at a corresponding position in the first intermediate image, and a pixel value of a pixel at a corresponding position in the second intermediate image.

2. The method of claim 1 , wherein the first preprocessed IR image is an IR image in which an edge component is enhanced in comparison to the IR image.

3. The method of claim 1 , wherein the generating of the preprocessed depth image comprises:

transforming the object, in the depth image, into a frontal object; and

generating the preprocessed depth image by performing the second edge enhancement preprocessing on the depth image in which the object is transformed into the frontal object.

4. The method of claim 3 , wherein the transforming of the object, in the depth image, into the frontal object comprises:

determining feature points in the depth image;

determining a transformation matrix based on positions of the respective feature points in the depth image and positions of reference points corresponding to the respective feature points; and

transforming the object, in the depth image, into the frontal object by applying the determined transformation matrix to the object in the depth image.

5. The method of claim 1 , wherein the determining of whether the object is a genuine object comprises:

determining a first liveness score using a neural network-based first liveness test model that receives the first preprocessed IR image as an input;

determining a second liveness score using a neural network-based second liveness test model that receives the preprocessed depth image as an input; and

determining whether the object is a genuine object based on the first liveness score and the second liveness score.

6. The method of claim 1 , further comprising allowing or blocking access to a device performing the method, based on a result of the determining of whether the object is a genuine object.

7. A non-transitory computer-readable storage medium storing instructions that, when executed by a processor, cause the processor to perform the method of claim 1 .

8. An apparatus with liveness testing, comprising:

at least one image sensor configured to acquire an infrared (IR) image including an object, and a depth image including the object; and

a processor configured to:

generate a first preprocessed IR image by performing first edge enhancement preprocessing on the IR image;

generate a preprocessed depth image by performing second edge enhancement preprocessing on the depth image; and

determine whether the object is a genuine object based on the first preprocessed IR image and the preprocessed depth image;

wherein the processor is further configured to:

generate a first intermediate image based on pixel values of a current pixel and neighboring pixels in the IR image;

generate a second intermediate image by performing normalization on the IR image; and

generate the first preprocessed IR image based on the IR image, the first intermediate image, and the second intermediate image,

wherein a pixel of the first preprocessed IR image has a pixel value of a pixel at a corresponding position in the IR image, a pixel value of a pixel at a corresponding position in the first intermediate image, and a pixel value of a pixel at a corresponding position in the second intermediate image.

9. The apparatus of claim 8 , wherein the processor is further configured to:

generate a first intermediate image based on pixel values of a current pixel and neighboring pixels in the IR image;

generate a second intermediate image by performing normalization on the IR image; and

generate the first preprocessed IR image based on the IR image, the first intermediate image, and the second intermediate image.

10. The apparatus of claim 8 , wherein the processor is further configured to:

transform the object, in the depth image, into a frontal object; and

generate the preprocessed depth image by performing the second edge enhancement preprocessing on the depth image in which the object is transformed into the frontal object.

11. The apparatus of claim 10 , wherein the processor is further configured to:

determine feature points in the depth image;

determine a transformation matrix based on positions of the respective feature points in the depth image and positions of reference points corresponding to the respective feature points; and

transform the object, in the depth image, into the frontal object by applying the determined transformation matrix to the object in the depth image.

12. The apparatus of claim 8 , wherein the processor is further configured to:

determine a first liveness score using a neural network-based first liveness test model that receives the first preprocessed IR image as an input;

determine a second liveness score using a neural network-based second liveness test model that receives the preprocessed depth image as an input, and

determine whether the object is a genuine object based on the first liveness score and the second liveness score.

13. The apparatus of claim 8 , wherein the processor is further configured to allow or block access to the apparatus, based on a result of the determining of whether the object is a genuine object.

14. An electronic device comprising:

an infrared (IR) camera configured to acquire an IR image including an object;

a depth camera configured to acquire a depth image including the object; and

a processor configured to:

generate a first preprocessed IR image by performing first edge enhancement preprocessing on the IR image;

generate a preprocessed depth image by performing second edge enhancement preprocessing on the depth image; and

determine whether the object is a genuine object based on the first preprocessed IR image and the preprocessed depth image;

wherein the processor is further configured to:

generate a first intermediate image based on pixel values of a current pixel and neighboring pixels in the IR image;

generate a second intermediate image by performing normalization on the IR image; and

generate the first preprocessed IR image based on the IR image, the first intermediate image, and the second intermediate image,

wherein a pixel of the first preprocessed IR image has a pixel value of a pixel at a corresponding position in the IR image, a pixel value of a pixel at a corresponding position in the first intermediate image, and a pixel value of a pixel at a corresponding position in the second intermediate image.

15. The electronic device of claim 14 , wherein the processor is further configured to:

generate a first intermediate image based on pixel values of a current pixel and neighboring pixels in the IR image;

generate a second intermediate image by performing normalization on the IR image; and

generate the first preprocessed IR image based on the IR image, the first intermediate image, and the second intermediate image.

16. The electronic device of claim 14 , wherein the processor is further configured to:

transform the object, in the depth image, into a frontal object; and

generate the preprocessed depth image by performing the second edge enhancement preprocessing on the depth image in which the object is transformed into the frontal object.

17. The electronic device of claim 16 , wherein the processor is further configured to:

determine feature points in the depth image;

determine a transformation matrix based on positions of the respective feature points in the depth image and positions of reference points corresponding to the respective feature points; and

transform the object, in the depth image, into the frontal object by applying the determined transformation matrix to the object in the depth image.

18. The electronic device of claim 14 , wherein the processor is further configured to:

determine a first liveness score using a neural network-based first liveness test model that receives the first preprocessed IR image as an input;

determine a second liveness score using a neural network-based second liveness test model that receives the preprocessed depth image as an input, and

determine whether the object is a genuine object based on the first liveness score and the second liveness score.

19. The electronic device of claim 14 , wherein the processor is further configured to allow or block access to the electronic device, based on a result of the determining of whether the object is a genuine object.