Systems and methods for a gradient-based metric for spoof detection

1. A method for spoof detection, comprising:

receiving an input image of a biometric;

computing a gradient image of the biometric based on the input image of the biometric, wherein each pixel of the gradient image represents a value for a directional change in a characteristic at a corresponding pixel location in the input image;

dividing the gradient image into blocks of pixels;

computing, for each block of pixels, a set of gradient values, wherein each gradient value in the set of gradient values for a given block corresponds to a gradient value of a different pixel of the gradient image in the given block;

computing, for each block of pixels, a variance value, wherein the variance value for a given block corresponds to a variance of the gradient values in the set of gradient values for the given block; and

determining whether the input image is a replica of the biometric based on the variance values of the blocks of pixels of the gradient image.

2. The method of claim 1 , wherein each block corresponds to a non-overlapping set of pixels of the gradient image.

3. The method of claim 1 , wherein the input image comprises a grayscale image of a fingerprint, and computing the gradient image comprises computing a first order derivative of fingerprint ridges included in the grayscale image of the fingerprint.

4. The method of claim 1 , wherein the blocks comprise square-shaped blocks of pixels.

5. The method of claim 1 , wherein after computing, for each block of pixels, the set of gradient values, the method further comprises:

computing, for each block of pixels, a histogram of the gradient values in the set of gradient values of the pixels in the block.

6. The method of claim 5 , further comprising:

plotting the variance values as a histogram of variance.

7. The method of claim 1 , wherein a user is denied access to a device based on determining that the input image is a replica of the biometric.

8. The method of claim 1 , wherein the gradient image is divided into a fixed number of blocks regardless of dimensions of the input image.

9. The method of claim 1 , wherein the biometric comprises a fingerprint of a finger, and the replica comprises a gelatin mold, a graphite mold, or a wood glue mold of the fingerprint of the finger.

10. The method of claim 1 , wherein the biometric comprises a fingerprint of a finger, and computing the gradient image comprises computing a first order derivative of fingerprint ridges included in the fingerprint of the finger.

11. A non-transitory computer-readable storage medium storing instructions that, when executed by a processor, causes a computing device to perform spoof detection, by performing steps comprising:

receiving an input image of a biometric;

computing a gradient image of the biometric based on the input image of the biometric, wherein each pixel of the gradient image represents a value for a directional change in a characteristic at a corresponding pixel location in the input image;

dividing the gradient image into blocks of pixels;

computing, for each block of pixels, a set of gradient values, wherein each gradient value in the set of gradient values for a given block corresponds to a gradient value of a different pixel of the gradient image in the given block;

computing, for each block of pixels, a variance value, wherein the variance value for a given block corresponds to a variance of the gradient values in the set of gradient values for the given block; and

determining whether the input image is a replica of the biometric based on the variance values of the blocks of pixels of the gradient image.

12. The computer-readable storage medium of claim 11 , wherein each block corresponds to a non-overlapping set of pixels of the gradient image.

13. The computer-readable storage medium of claim 11 , wherein the input image comprises a grayscale image of a fingerprint, and computing the gradient image comprises computing a first order derivative of fingerprint ridges included in the grayscale image of the fingerprint.

14. The computer-readable storage medium of claim 11 , wherein after computing, for each block of pixels, the set of gradient values, the steps further comprise:

computing, for each block of pixels, a histogram of the gradient values in the set of gradient values of the pixels in the block.

15. The computer-readable storage medium of claim 11 , wherein a user is denied access to a device based on determining that the input image is a replica of the biometric.

16. The computer-readable storage medium of claim 11 , wherein the gradient image is divided into a fixed number of blocks regardless of dimensions of the input image.

17. The computer-readable storage medium of claim 11 , wherein the biometric comprises a fingerprint of a finger, and the replica comprises a gelatin mold, a graphite mold, or a wood glue mold of the fingerprint of the finger.

18. The computer-readable storage medium of claim 11 , wherein the biometric comprises a fingerprint of a finger, and computing the gradient image comprises computing a first order derivative of fingerprint ridges included in the fingerprint of the finger.

19. A device, comprising:

a biometric sensor; and

a processor configured to:

receive, from the biometric sensor, an input image of a biometric;

compute a gradient image of the biometric based on the input image of the biometric, wherein each pixel of the gradient image represents a value for a directional change in a characteristic at a corresponding pixel location in the input image;

divide the gradient image into blocks of pixels;

compute, for each block of pixels, a set of gradient values, wherein each gradient value in the set of gradient values for a given block corresponds to a gradient value of a different pixel of the gradient image in the given block;

compute, for each block of pixels, a variance value, wherein the variance value for a given block corresponds to a variance of the gradient values in the set of gradient values for the given block; and

determine whether the input image is a replica of the biometric based on the variance values of the blocks of pixels of the gradient image.

20. The method of claim 1 , wherein the characteristic comprises intensity or color at the corresponding pixel location in the input image.