Background OCR during card data entry

Patent metadata

Field	Value
Publication number	US-9483760-B2
Application number	US-201615008177-A
Country	US
Kind code	B2
Filing date	Jan 27, 2016
Priority date	Nov 3, 2014
Publication date	Nov 1, 2016
Grant date	Nov 1, 2016

Abstract

Financial transaction card data can be entered by providing a picture of the card to a server programmed with a text recognition algorithm. The server can perform text recognition on the image at the same time that a consumer enters additional required data, such as a zip code. The server can perform as much text recognition processing as possible in the time the consumer is entering the additional data. Once the additional data is received, a signal can be provided to the server indicating that the user is now waiting for results of the text recognition process, meaning the server should provide them as quickly as possible. Once text recognition results are received, a consumer can make a selection to identify a character which the text recognition algorithm did not sufficiently identify. Based on known account number constraints, the user selection can cause multiple characters to be identified.

First claim

The invention claimed is:

1. A method, performed on a mobile computing device, for reducing a number of user corrections entered to obtain a correct account number for a financial transaction card, the method comprising: obtaining, at the mobile computing device, multiple versions of text recognition results, wherein each version of the text recognition results is associated with a total confidence score; identifying, as a best guess, a version of the text recognition results that is associated with a highest total confidence score; selecting, based on individual confidence scores of characters of the best guess, characters to verify from the best guess; receiving a user selection, as a selected character, for one of the characters to verify; and updating the multiple versions of the text recognition results by performing one or more of: eliminating, from the multiple versions of text recognition results, one or more of the multiple versions of text recognition results that do not match the selected character; eliminating, from the multiple versions of text recognition results, one or more of the multiple versions of text recognition results that do not satisfy the Luhn algorithm; eliminating, from the multiple versions of text recognition results, one or more of the multiple versions of text recognition results that do not match any of multiple known issuer identification numbers; updating the total confidence score for the multiple versions of text recognition results; or any combination thereof; and identifying a new best guess from the remaining versions of the text recognition results, wherein the new best guess comprises at least a first difference from the best guess that is a change of a first character in the best guess to the selected character, and a second difference from the best guess that is a change of a second character in the best guess other than to the selected character.

2. The method of claim 1 further comprising: verifying that the new best guess includes the correct account number by: sending a proposed account number associated with the new best guess to a financial institution; and receiving, from the financial institution, a confirmation that the proposed account number associated with the new best guess indicates valid account data.

3. The method of claim 2 further comprising: sending the verified correct number associated with the new best guess to a server, wherein the server is configured to link the financial transaction card to a service provided by a payment service system by storing the verified correct account number associated with the new best guess, in a database, with an association to a user and the service provided by the payment service system.

4. The method of claim 2 further comprising: sending the verified correct account number associated with the new best guess to a server, wherein the server is configured to perform a transaction using the verified correct account number associated with the new best guess.

5. The method of claim 1 wherein the total confidence score for each selected version of the multiple versions of text recognition results is based on a combination of individual confidence scores assigned to recognized characters of that selected version.

6. The method of claim 1 , wherein selecting the best guess comprises selecting a version of the text recognition results that is associated with a highest total confidence score; and wherein identifying the new best guess comprises selecting a version of the text recognition results with an updated highest total confidence score.

7. The method of claim 1 , wherein the updated highest total confidence score is based on a combination of individual confidence scores assigned to recognized characters of the new best guess.

8. The method of claim 1 , wherein updating the visual representation with the new best guess comprises showing at least: a first difference indicating the correction character, and a second difference indicating a difference between the best guess and the correct character set other than the correction character.

9. The method of claim 1 , wherein the method is performed on a mobile computing device with a display; wherein displaying the visual representation of the best guess comprises displaying the visual representation on the display of the mobile computing device with one or more input areas, each input area indicating a character of the best guess which is a candidate for receiving the correction character; and wherein the correction character is received through a user selection to one of the input areas.

10. The method of claim 1 , wherein: the characters of the best guess which are candidates for receiving the correction character are selected based on the candidates for receiving the correction character having individual confidence scores below a defined threshold value.

11. The method of claim 1 , wherein at least one of the input areas for a particular character position is configured to include options for the correction character comprising less than ten different number characters; and wherein the options for the correction character are selected based on a union of the characters in the particular character position from two or more of the multiple versions of text recognition results.

12. The method of claim 1 , wherein at least one of the input areas is for a particular character position to receive the correction character; and wherein the particular character position is selected based on a computation that receiving a correction character for the particular character position is most likely to narrow possibilities for character positions other than the particular character position.

13. The method of claim 1 , wherein the second difference is based on the elimination of the one or more versions of text recognition results that: do not satisfy the Luhn algorithm, do not match any of multiple known issuer identification numbers, or both.

14. A non-transitory computer-readable storage medium storing instructions that, when executed by a computing system, cause the computing system to perform operations for reducing a number of user corrections entered to obtain a correct character set, the operations comprising: selecting a best guess from multiple versions of text recognition results; displaying a first visual representation of the best guess; receiving a correction character corresponding to a character displayed in the first visual representation, the correction character indicating a difference between the best guess and the correct character set; identifying, based on the correction character, a new best guess from the multiple versions of text recognition results; and updating the first visual representation to show a second visual representation with the new best guess, wherein the new best guess comprises at least a first difference from the best guess that is a change of a first character in the best guess to the selected character, and a second difference from the best guess that is a change of a second character in the best guess other than to the selected character.

15. The computer-readable storage medium of claim 14 , wherein: identifying a new best guess is accomplished by: eliminating any of the multiple versions that do not match the correction character; eliminating any of the multiple versions that, based on the correction character, either do not satisfy the Luhn algorithm or do not match any of multiple known issuer identification numbers; or any combination thereof.

16. The computer-readable storage medium of claim 14 , wherein selecting the best guess comprises s

Assignees

• Square Inc

Inventors

• Bekmann Joachim
• Guo Fei

Classifications

• G06T11/60

  Creating or editing images; Combining images with text · CPC title

• G06Q20/3276

  using a pictured code, e.g. barcode or QR-code, being read by the M-device · CPC title

• G07F7/122

  Online card verification · CPC title

• G06Q20/34 Primary

  using cards, e.g. integrated circuit [IC] cards or magnetic cards · CPC title

• G06T7/74

  involving reference images or patches · CPC title