Google Inc.

Xiaohang Wang

Henry Allan Rowley

Sanjiv Kumar

Farhan Shamsi

Yakov Okshtein

Marcus Quintana Mitchell

Debra Lin Repenning

Alessandro Bissacco

Glenn Berntson

Jose Jeronimo Moreira Rodrigues

Justin Max Scheiner

David Singleton

Lixin Zhang

Marcus Alexander Foster

Jeff Huber

Harry Lee Butler, Iv

Leon G Palm

Marria Nazif

Sam Shih

Daniel Talavera

Mark Leslie Snyder

Glenn Thrope

Douglas Alexander Gresham

Alan Newberger

Zhifei Li

Jose Moreira Rodrigues

Peter Hazlehurst

Kishore A Papineni

Mohammad Hossain Sheikh Attar

Michael James Taylor

Glen Berntson

Justin Scheiner

Leon Palm

Receiving point of interest zones and alerts on user devices comprises communicating, by a user computing device to a remote computing device, a request for point of interest data corresponding to points of interest within a proximity of the user device; presenting the received point of interest data; identifying a particular point of interest; and outputting an alert regarding the particular point of interest. Receiving point of interest zones on user devices comprises communicating a request for point of interest data; receiving the point of interest data from the remote network device wherein a size of the point of interest zone is determined based on a density of points of interest in the proximity of the user, and wherein the shape of the point of interest zone is expanded in a direction of travel and contracted in the opposite direction; and presenting the received point of interest data.

Providing points of interest to user devices in variable zones

Comparing extracted card data from a continuous scan comprises receiving, by one or more computing devices, a digital scan of a card; obtaining a plurality of images of the card from the digital scan of the physical card; performing an optical character recognition algorithm on each of the plurality of images; comparing results of the application of the optical character recognition algorithm for each of the plurality of images; determining if a configured threshold of the results for each of the plurality of images match each other; and verifying the results when the results for each of the plurality of images match each other. Threshold confidence level for the extracted card data can be employed to determine the accuracy of the extraction. Data is further extracted from blended images and three-dimensional models of the card. Embossed text and holograms in the images may be used to prevent fraud.

Comparing extracted card data using continuous scanning

Identifying the geolocation of POS terminals using non-payment events to predict when the geolocation of a computing device at a time when the device detects events corresponds to the geolocation of the terminal. The device monitors for pre-selected events and transmit data to the account system. The account system determines a frequency of the events and it reaches a pre-defined threshold, the account system identifies the location of the terminal by identifying the common geolocation of the events. The identified geolocation is saved so that when a user then enters the location and transmits event data to the account system, the system can compare the geolocation of the event data to the saved geolocation to determine whether the computing device is located at the terminal. If the computing device is located at the terminal, the account system transmits offers or other content for display and use at the identified terminal.

Point of sale terminal geolocation

Display screen with a graphical user interface of a card reader application

Extracting financial card information with relaxed alignment comprises a method to receive an image of a card, determine one or more edge finder zones in locations of the image, and identify lines in the one or more edge finder zones. The method further identifies one or more quadrilaterals formed by intersections of extrapolations of the identified lines, determines an aspect ratio of the one or more quadrilateral, and compares the determined aspect ratios of the quadrilateral to an expected aspect ratio. The method then identifies a quadrilateral that matches the expected aspect ratio and performs an optical character recognition algorithm on the rectified model. A similar method is performed on multiple cards in an image. The results of the analysis of each of the cards are compared to improve accuracy of the data.

Extracting card data from multiple cards

Extracting card data comprises receiving, by one or more computing devices, a digital image of a card; perform an image recognition process on the digital representation of the card; identifying an image in the digital representation of the card; comparing the identified image to an image database comprising a plurality of images and determining that the identified image matches a stored image in the image database; determining a card type associated with the stored image and associating the card type with the card based on the determination that the identified image matches the stored image; and performing a particular optical character recognition algorithm on the digital representation of the card, the particular optical character recognition algorithm being based on the determined card type. Another example superimposes the extracted data directly above, below, or beside the corresponding section on the displayed image.

Extracting card data for simultaneous display with an image

Extracting card information comprises a server at an optical character recognition ('OCR') system that interprets data from a card. The OCR system performs an optical character recognition algorithm an image of a card and performs a data recognition algorithm on a machine-readable code on the image of the card. The OCR system compares a series of extracted alphanumeric characters obtained via the optical character recognition process to data extracted from the machine-readable code via the data recognition process and matches the alphanumeric series of characters to a particular series of characters extracted from the machine-readable code. The OCR system determines if the alphanumeric series and the matching series of characters extracted from the machine-readable code comprise any discrepancies and corrects the alphanumeric series of characters based on the particular series of characters extracted from the machine-readable code upon a determination that a discrepancy exists.

Extracting card identification data

A geofence management system obtains location data for points of interest. The geofence management system determines, at the option of the user, the location of a user mobile computing device relative to specific points of interest and alerts the user when the user nears the points of interest. The geofence management system, however, determines relationships among the identified points of interest, and associates or 'clusters' the points of interest together based on the determined relationships. Rather than establishing separate geofences for multiple points of interest, and then alerting the user each time the user's mobile device enters each geofence boundary, the geofence management system establishes a single geofence boundary for the associated points of interest. When the user's mobile device enters the clustered geofence boundary, the geofence management system notifies the user device to alert the user of the entrance event. The user then receives the clustered, geofence-based alert.

Clustering geofence-based alerts for mobile devices

Inferring purchase intent using non-payment transaction signals predicts whether a payment transaction has been completed based on non-payment information. An account system that operates outside of the payment path does not take part in and the approval of a financial transaction between the user and the merchant system, distributes an offer to the user. The user completes a financial payment transaction with the merchant and the account system determines whether a trigger event has occurred. The user performs an action or enters information using the user computing device, and the user computing device transmits an indication of the action to the account system. In another example, the account system receives notification from another system or device. The account system determines whether the action is a trigger event and the predictive model determines whether the user completed a financial transaction and/or redeemed the distributed offer.

Inferring purchase intent using non-payment transaction events

Extracting card data comprises receiving, by one or more computing devices, a digital image of a card; perform an image recognition process on the digital representation of the card; identifying an image in the digital representation of the card; comparing the identified image to an image database comprising a plurality of images and determining that the identified image matches a stored image in the image database; determining a card type associated with the stored image and associating the card type with the card based on the determination that the identified image matches the stored image; and performing a particular optical character recognition algorithm on the digital representation of the card, the particular optical character recognition algorithm being based on the determined card type. Another example uses an issuer identification number to improve data extraction. Another example compares extracted data with user data to improve accuracy.

Comparing extracted card data with user data

Display screen with a graphical user interface of a receipt reader application

Embodiments herein provide computer-implemented techniques for allowing a user computing device to extract financial card information using optical character recognition ('OCR'). Extracting financial card information may be improved by applying various classifiers and other transformations to the image data. For example, applying a linear classifier to the image to determine digit locations before applying the OCR algorithm allows the user computing device to use less processing capacity to extract accurate card data. The OCR application may train a classifier to use the wear patterns of a card to improve OCR algorithm performance. The OCR application may apply a linear classifier and then a nonlinear classifier to improve the performance and the accuracy of the OCR algorithm. The OCR application uses the known digit patterns used by typical credit and debit cards to improve the accuracy of the OCR algorithm.

Hierarchical classification in credit card data extraction

Growing community of inventors

Jersey City, NJ, United States of America

Xiaohang Wang