Methods and systems for determining the authenticity of an identity document

1. A method for determining authenticity of a document comprising the steps of:

capturing, by an electronic device, an image of a document, the document including image data of a biometric modality of a user and other data;

cropping the captured image to include an image of the document only;

removing the other data from the cropped image;

generating a gray scale image from the cropped image, the gray scale image and the cropped image being a pair of images; and

simultaneously processing, by a trained machine learning model (MLM), each image in the pair of images to determine whether the identity document in the captured image is authentic.

2. The method according to claim 1 said simultaneously processing step comprising the steps of:

calculating, by the trained MLM using the pair of images, a genuineness score;

comparing the genuineness score against a threshold score; and

in response to determining the genuineness score is greater than the threshold score, determining the identity document in the captured image is authentic; and

in response to determining the genuineness score is less than the threshold score, determining the document is fraudulent.

3. The method according to claim 1 said determining step comprising the steps of:

calculating, by the trained MLM using the pair of images a first score representing the likelihood that the document in the captured image data is fraudulent and a second score representing the likelihood that the document is authentic;

comparing the first score against the second score;

in response to determining the first score is greater than the second score, determining the captured image requires manual review; and

in response to determining the first score is less than the second score, determining the document is authentic.

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

obtaining images from an image database, the images being images of identity documents;

cropping each obtained image to include an image of the respective identity document only;

generating a gray scale image from each cropped image, the gray scale image and the cropped image being a pair of images; and

training the MLM using the pairs of images, wherein the MLM simultaneously processes the images in each pair of images.

5. The method according to claim 4 , said generating a gray scale image step comprising the steps of:

determining the location of the biometric modality image data in the cropped image; and

detecting the edges of each obtained image.

6. An electronic device for determining authenticity of a document comprising:

a processor; and

a memory configured to store data, said electronic device being associated with a network and said memory being in communication with said processor and having instructions stored thereon which, when read and executed by said processor, cause said electronic device to:

capture an image of a document, the document including image data of a biometric modality of a user and other data;

crop the captured image to include an image of the document only;

remove the other data from the cropped image;

generate a gray scale image from the cropped image, the gray scale image and the cropped image being a pair of images; and

simultaneously process, by a trained machine learning model (MLM), each image in the pair of images to determine whether the identity document in the captured image is authentic.

7. The electronic device according to claim 6 , wherein the instructions when read and executed by said processor, cause said electronic device to:

calculate using the pair of images, a genuineness score;

compare the genuineness score against a threshold score;

in response to determining the genuineness score is greater than the threshold score, determine the identity document in the captured image is authentic; and

in response to determining the genuineness score is less than the threshold score, determine the document is fraudulent.

8. The electronic device according to claim 6 , wherein the instructions when read and executed by said processor, cause said electronic device to:

calculate by the trained MLM using the pair of images a first score representing the likelihood that the document in the captured image data is fraudulent and a second score representing the likelihood that the document is authentic;

compare the first score against the second score;

in response to determining the first score is greater than the second score, determine the captured image requires manual review; and

in response to determining the first score is less than the second score, determine the document is authentic.

9. The electronic device according to claim 6 , wherein the instructions when read and executed by said processor, cause said electronic device to:

obtain images from an image database, the images being images of documents;

crop each obtained image to include an image of the respective document only;

generate a gray scale image from each cropped image, the gray scale image and the cropped image being a pair of images; and

train the MLM using the pairs of images, wherein the MLM simultaneously processes the images in each pair of images during said training step.

10. The electronic device according to claim 9 , wherein the instructions when read and executed by said processor, cause said electronic device to:

determine the location of the biometric modality image data in the cropped image; and

detect the edges of each obtained image.

11. A non-transitory computer-readable recording medium in an electronic device for determining authenticity of a document, the non-transitory computer-readable recording medium storing instructions which when executed by a hardware processor cause the non-transitory recording medium to perform steps comprising:

capturing an image of a document, the document including an image of a biometric modality of a user and other data;

cropping the captured image to include an image of the document only;

removing the other data from the cropped image;

generating a gray scale image from the cropped image, the gray scale image and the cropped image being a pair of images; and

simultaneously processing, by a trained machine learning model (MLM), each image in the pair of images to determine whether the identity document in the captured image is authentic.

12. The non-transitory computer-readable recording medium according to claim 11 , wherein the instructions when read and executed by said processor, cause said non-transitory computer-readable recording medium to perform the steps of:

calculating, by the trained MLM using the pair of images, a genuineness score;

comparing the genuineness score against a threshold score;

in response to determining the genuineness score is greater than the threshold score, determining the identity document in the captured image is authentic; and

in response to determining the genuineness score is less than the threshold score, determining the document is fraudulent.

13. The non-transitory computer-readable recording medium according to claim 11 , wherein the instructions when read and executed by said processor, cause said non-transitory computer-readable recording medium to perform the steps of:

calculating, by the trained MLM using the pair of images a first score representing the likelihood that the document in the captured image data is fraudulent and a second score representing the likelihood that the document is authentic;

comparing the first score against the second score;

in response to determining the first score is greater than the second score, determining the captured image requires manual review; and

in response to determining the first score is less than the second score, determining the document is authentic.

14. The non-transitory computer-readable recording medium according to claim 11 , wherein the instructions when read and executed by said processor, cause said non-transitory computer-readable recording medium to perform the steps of:

obtaining images from an image database, the images being images of documents;

cropping each obtained image to include an image of the respective document only;

generating a gray scale image from each cropped image, the gray scale image and the cropped image being a pair of images; and

training the MLM using the pairs of images, wherein the MLM simultaneously processes the images in each pair of images.

15. The non-transitory computer-readable recording medium according to claim 14 , wherein the instructions when read and executed by said processor, cause said non-transitory computer-readable recording medium to perform the steps of:

determining the location of the biometric modality image data in the cropped image; and

detecting the edges of each obtained image.