Character recognition systems and methods
Techniques described herein are directed to generation and use of a system that allows a payment service to read payment objects, onboard users, and detect fraud using data from the payment objects. The systems and methods may include receiving instructions to obtain images of a payment instrument and generating data representing those images, which may include a 3D model of the payment instrument. A comparison of the generated data and previously stored data may be performed and the results of this comparison may be utilized to determine a likelihood that a fraudulent event is occurring with respect to the payment instrument.
1 . A computer-implemented method comprising:
receiving, by one or more computing devices, first data representing an instruction to obtain images of a payment instrument in two or more perspectives, wherein the two or more perspectives are determined based at least in part on at least one of a type of the payment instrument or a risk metric associated with a user account and wherein the first data is configured to cause a template to be displayed on a screen of an individual computing device of the one or more computing devices, the template indicating how the payment instrument is to be moved to obtain the images of the payment instrument in the two or more perspectives;
generating, by the one or more computing devices and based at least in part on the instruction, second data representing images of the payment instrument taken by a camera in the two or more perspectives, wherein the first data is configured to cause an application associated with the camera to be displayed on a foreground of the screen;
generating, by the one or more computing devices and utilizing the second data, a three-dimensional model of the payment instrument, the three-dimensional model indicating information present on the payment instrument and physical attributes of the payment instrument,
wherein the information present on the payment instrument includes one or more selected from a Quick Response (QR) code, a barcode, a photograph of a cardholder, a symbol associated with tap-to-pay functionality, a signature of a cardholder, design or artwork, and combinations thereof, and
wherein the physical attributes include at least physical dimensions of the payment instrument selected from a group of a thickness of the payment instrument, a degree of embossing, a degree of concavity of text on the payment instrument, and combinations thereof;
determining, by the one or more computing devices, that the information and the physical attributes differ from third data associated with the payment instrument as stored by a payment server; and
generating fourth data indicating a likelihood of a fraudulent event based at least in part on the information and the physical attributes differing from the third data.
2 . The computer-implemented method of claim 1 , wherein the images are captured utilizing a digital scan of the payment instrument.
3 . The computer-implemented method of claim 1 , wherein the payment instrument includes a near-field communication object, and the second data represents images of the near-field communication object.
4 . The computer-implemented method of claim 1 , further comprising determining the risk metric by:
retrieving stored data known about the user account associated with the payment instrument; and
outputting, with a machine-learning model and based on the stored data, the risk metric.
5 . The computer-implemented method of claim 1 , further comprising causing output of a request to modify the information as determined from the three-dimensional model by user input.
6 . The computer-implemented method of claim 1 , wherein determining that the information differs from the third data is based at least in part on analysis of the three-dimensional model utilizing optical character recognition processing and computer vision processing.
7 . A system comprising:
one or more processors; and
non-transitory computer-readable media storing instructions that, when executed by the one or more processors, cause the one or more processors to perform operations comprising:
receiving first data representing an instruction to obtain images of a payment instrument in two or more perspectives, wherein the two or more perspectives are determined based at least in part on at least one of a type of the payment instrument or a risk metric associated with a user account and wherein the first data is configured to cause a template to be displayed on a screen of an individual computing device, the template indicating how the payment instrument is to be moved to obtain the images of the payment instrument in the two or more perspectives;
generating, based at least in part on the instruction, second data representing images of the payment instrument taken by a camera, wherein the first data is configured to cause an application associated with the camera to be displayed on a foreground of the screen;
generating, utilizing the second data, a three-dimensional model of the payment instrument, the three-dimensional model indicating information present on the payment instrument and physical attributes of the payment instrument,
wherein the information present on the payment instrument includes one or more selected from a Quick Response (QR) code, a barcode, a photograph of a cardholder, a symbol associated with tap-to-pay functionality, a signature of a cardholder, design or artwork, and combinations thereof, and
wherein the physical attributes include at least physical dimensions of the payment instrument selected from a group of a thickness of the payment instrument, a degree of embossing, a degree of concavity of text on the payment instrument, and combinations thereof;
determining that the information and the physical attributes differ from third data associated with the payment instrument as stored by a payment server; and
generating fourth data indicating a likelihood of a fraudulent event based at least in part on the information and the physical attributes differing from the third data.
8 . The system of claim 7 , wherein
the images are captured utilizing a digital scan of the payment instrument.
9 . The system of claim 7 , wherein the operations further comprise determining the risk metric by:
retrieving stored data known about the user account; and
outputting, with a machine-learning model and based on the stored data, the risk metric.
10 . The system of claim 7 , the operations further comprising:
receiving an indication that the payment instrument is requested to be used in association with a transaction;
determining a risk metric associated with the transaction, wherein receiving the first data is based at least in part on the risk metric associated with the transaction satisfying a threshold risk metric; and
receiving, based at least in part on the risk metric satisfying the threshold risk metric, the third data associated with the payment instrument as previously stored from the payment server.
11 . The system of claim 7 , the operations further comprising:
determining a first perspective of a first image of the payment instrument to be obtained;
determining a second perspective of a second image of the payment instrument to be obtained;
determining when a field of view of the camera depicts the payment instrument pursuant to the first perspective;
causing the camera to generate first image data when the field of view of the camera depicts the payment instrument pursuant to the first perspective;
determining when the field of view of the camera transitions to depicting the payment instrument pursuant to the second perspective;
causing the camera to generate second image data when the field of view of the camera transitions to depicting the payment instrument pursuant to the second perspective; and
wherein the second data comprises the first image data and the second image data.
12 . The system of claim 7 , wherein the payment instrument includes a near-field communication object, and the second data represents images of the near-field communication object.
13 . The system of claim 7 , the operations further comprising:
performing optical character recognition on the second data such that text data is identified from the payment instrument;
performing computer vision processing on the second data such that object attributes are identified from the payment instrument; and
determining the information from the text data and from the object attributes.
14 . A non-transitory computer-readable medium with instructions stored thereon that, when executed by one or more computers, cause the one or more computers to perform operations, the operations comprising:
receiving, by the one or more computers, first data representing an instruction to obtain images of a payment instrument in two or more perspectives, wherein the two or more perspectives are determined based at least in part on at least one of a type of the payment instrument or a risk metric associated with a user account and wherein the first data is configured to cause a template to be displayed on a screen of an individual computing device of the one or more computing devices, the template indicating how the payment instrument is to be moved to obtain the images of the payment instrument in the two or more perspectives;
generating, by the one or more computers and based at least in part on the instruction, second data representing images of the payment instrument taken by a camera in the two or more perspectives, wherein the first data is configured to cause an application associated with the camera to be displayed on a foreground of the screen;
generating, by the one or more computers and utilizing the second data, a three-dimensional model of the payment instrument, the three-dimensional model indicating information present on the payment instrument and physical attributes of the payment instrument,
wherein the information present on the payment instrument includes one or more selected from a Quick Response (QR) code, a barcode, a photograph of a cardholder, a symbol associated with tap-to-pay functionality, a signature of a cardholder, design or artwork, and combinations thereof, and
wherein the physical attributes include at least physical dimensions of the payment instrument selected from a group of a thickness of the payment instrument, a degree of embossing, a degree of concavity of text on the payment instrument, and combinations thereof;
determining, by the one or more computers, that the information and the physical attributes differ from third data associated with the payment instrument as stored by a payment server; and
generating fourth data indicating a likelihood of a fraudulent event based at least in part on the information and the physical attributes differing from the third data.
15 . The non-transitory computer-readable medium of claim 14 , wherein the images are captured utilizing a digital scan of the payment instrument.
16 . The non-transitory computer-readable medium of claim 14 , wherein the payment instrument includes a near-field communication object, and the second data represents images of the near-field communication object.
17 . The non-transitory computer-readable medium of claim 14 , wherein the operations further include determining the risk metric by:
retrieving stored data known about the user account associated with the payment instrument; and
outputting, with a machine-learning model and based on the stored data, the risk metric.
18 . The non-transitory computer-readable medium of claim 14 , wherein the operations further include causing output of a request to modify the information as determined from the three-dimensional model by user input.
19 . The non-transitory computer-readable medium of claim 14 , wherein determining that the information differs from the third data is based at least in part on analysis of the three-dimensional model utilizing optical character recognition processing and computer vision processing.