Fingerprint anti-counterfeiting method and electronic device

1. A method, comprising:

obtaining, by an electronic device, a fingerprint image and a multi-frame touch signal; and

determining, by the electronic device based on the fingerprint image and the multi-frame touch signal, whether the fingerprint image is of a real finger or of a fake finger;

wherein the electronic device stores a multi-modal network, and the multi-modal network comprises a fingerprint image subnetwork and a touch signal subnetwork; and

wherein determining, by the electronic device, whether the fingerprint image is of a real finger or a fake finger comprises:

inputting, by the electronic device, the fingerprint image into the fingerprint image subnetwork for processing, to obtain a first eigenvector;

inputting, by the electronic device, the multi-frame touch signal into the touch signal subnetwork for processing, to obtain a second eigenvector;

fusing, by the electronic device, the first eigenvector with the second eigenvector to obtain a third eigenvector;

classifying, by the electronic device, the third eigenvector to obtain a classification result; and

determining, by the electronic device based on the classification result, whether the fingerprint image is of a real finger or of a fake finger.

2. The method according to claim 1 , wherein determining, by the electronic device based on the classification result, whether the fingerprint image is of a real finger or a fake finger comprises:

when the classification result comprises a first confidence value, and the first confidence value is greater than or equal to a first threshold, determining, by the electronic device, that the fingerprint image is of a real finger, wherein the first confidence value is a confidence value at which the fingerprint image is of a real finger; or

when the classification result comprises a second confidence value, and the second confidence value is greater than or equal to a second threshold, determining, by the electronic device, that the fingerprint image is of a fake finger, wherein the second confidence value is a confidence value at which the fingerprint image is of a fake finger; and

wherein the first threshold and the second threshold are adjustable.

3. The method according to claim 1 , further comprising:

before the electronic device inputs the fingerprint image into the fingerprint image subnetwork for processing, and the inputs the multi-frame touch signal into the multi-modal network for processing, fusing the fingerprint image with the multi-frame touch signal; and

after the electronic device inputs the fingerprint image into the fingerprint image subnetwork for processing, and inputs the multi-frame touch signal into the multi-modal network for processing, separating the fingerprint image and the multi-frame touch signal.

4. The method according to claim 1 , further comprising:

after obtaining, by the electronic device, the fingerprint image and the multi-frame touch signal, performing, by the electronic device, normalization processing on the fingerprint image and the multi-frame touch signal.

5. The method according to claim 1 , further comprising:

when the electronic device determines that the fingerprint image is of a fake finger, performing the following:

sending, by the electronic device, fake fingerprint prompt information to a user; or

sending, by the electronic device, operation failure prompt information to a user.

6. The method according to claim 1 , further comprising:

when the electronic device determines that the fingerprint image is of a real finger, performing the following:

determining, by the electronic device, that the fingerprint image matches a fingerprint image preset by a user; and

determining, by the electronic device, that user identity authentication succeeds.

7. The method according to claim 1 , wherein the multi-frame touch signal represents one or more of a touch strength, a touch strength change trend, a touch shape, or a touch shape change trend.

8. The method according to claim 1 , wherein the multi-frame touch signal comprises:

a multi-frame touch signal synchronously collected when the fingerprint image is collected; or

a multi-frame touch signal synchronously collected when the fingerprint image is collected, and a multi-frame touch signal buffered before the fingerprint image is collected.

9. The method according to claim 1 , wherein the fingerprint image is other than a first fingerprint image, and the multi-frame touch signal is a multi-frame touch signal synchronously collected when the first fingerprint image is collected.

10. The method according to claim 1 , further comprising:

after obtaining, by the electronic device, the fingerprint image and the multi-frame touch signal, preprocessing, by the electronic device, the multi-frame touch signal based on a screen status or a touch collection frequency of the electronic device, wherein the preprocessing comprises one or more of interpolation processing, zero padding processing, or downsampling processing.

11. The method according to claim 1 , further comprising:

after obtaining, by the electronic device, the fingerprint image and the multi-frame touch signal, obtaining, by the electronic device through truncation, a touch signal with a predetermined frame quantity or a predetermined size.

12. An electronic device, comprising:

a fingerprint sensor;

a touch sensor;

one or more processors; and

one or more memories, wherein the one or more memories store one or more computer programs that are executable by the one or more processors to cause the electronic device to:

detect a touch operation in a fingerprint unlocking area;

collect, by the fingerprint sensor, a fingerprint image;

collect, by the touch sensor, a multi-frame touch signal; and

determine, based on the fingerprint image and the multi-frame touch signal, whether the fingerprint image is of a real finger or of a fake finger;

wherein the electronic device stores a multi-modal network, and the multi-modal network comprises a fingerprint image subnetwork and a touch signal subnetwork; and

wherein the one or more computer programs are further executable by the one or more processors to cause the electronic device to:

input the fingerprint image into the fingerprint image subnetwork for processing, to obtain a first eigenvector;

input the multi-frame touch signal into the touch signal subnetwork for processing, to obtain a second eigenvector;

fuse the first eigenvector with the second eigenvector to obtain a third eigenvector;

classify the third eigenvector to obtain a classification result; and

determine, based on the classification result, whether the fingerprint image is a of a real finger or of a fake finger.

13. The electronic device according to claim 12 , wherein the one or more computer programs are further executable by the one or more processors to cause the electronic device to:

when the classification result comprises a first confidence value, and the first confidence value is greater than or equal to a first threshold, determining that the fingerprint image is a fingerprint image of a real finger, wherein the first confidence value is a confidence value at which the fingerprint image is of a real finger; or

when the classification result comprises a second confidence value, and the second confidence value is greater than or equal to a second threshold, determining that the fingerprint image is of a fake finger, wherein the second confidence value is a confidence value at which the fingerprint image is of a fake finger; and

wherein the first threshold and the second threshold are adjustable.

14. The electronic device according to claim 12 , wherein the one or more computer programs are further executable by the one or more processors to cause the electronic device to:

before the fingerprint image is input into the fingerprint image subnetwork for processing and the multi-frame touch signal is input into the multi-modal network for processing, fuse the fingerprint image with the multi-frame touch signal; and

after the fingerprint image is input into the fingerprint image subnetwork for processing and the multi-frame touch signal is input into the multi-modal network for processing, separate the fingerprint image and the multi-frame touch signal.

15. The electronic device according to claim 12 , wherein the one or more computer programs are further executable by the one or more processors to cause the electronic device to:

when it is determined that the fingerprint image is of a fake finger, send fake fingerprint prompt information to a user, or send operation failure prompt information to a user.

16. The electronic device according to claim 12 , wherein the one or more computer programs are further executable by the one or more processors to cause the electronic device to:

when it is determined that the fingerprint image is of a real finger, determine that the fingerprint image matches a fingerprint image preset by a user; and

determine that user identity authentication succeeds.

17. The electronic device according to claim 12 , wherein the multi-frame touch signal comprises:

a multi-frame touch signal synchronously collected when the fingerprint image is collected; or

a multi-frame touch signal synchronously collected when the fingerprint image is collected, and a multi-frame touch signal buffered before the fingerprint image is collected.

18. The electronic device according to claim 12 , wherein the one or more computer programs are further executable by the one or more processors to cause the electronic device to:

after the fingerprint image and the multi-frame touch signal are obtained, preprocessing the multi-frame touch signal based on a screen status or a touch collection frequency of the electronic device, wherein the preprocessing comprises one or more of interpolation processing, zero padding processing, or downsampling processing.