Recognizing text in image data

Patent metadata

Field	Value
Publication number	US-10943106-B2
Application number	US-201816139737-A
Country	US
Kind code	B2
Filing date	Sep 24, 2018
Priority date	Dec 18, 2017
Publication date	Mar 9, 2021
Grant date	Mar 9, 2021

Abstract

A device may receive image data representing a document, the document including: text, and edges. Based on the edges, the device may identify, a segment of interest within the image data and crop the segment of interest to obtain a portion of the image data. In addition, the device may perform optical character recognition on the portion of the image data, the optical character recognition producing recognized text. The device may obtain, based on the recognized text, validation data that includes verification text, and determine whether the recognized text is verified based on the verification text. Based on a result of the determination, the device may perform an action.

First claim

What is claimed is:

1. A device, comprising:
one or more memory devices; and
one or more processors, communicatively connected to the one or more memory devices, to:
receive image data representing a document, the document including:
text, and
a plurality of edges within the document;

identify, based on the plurality of edges, a segment of interest within the image data;
crop the segment of interest to obtain a portion of the image data,
the segment of interest being defined by one or more edges, of the plurality of edges, forming a rectangular shape;

determine one or more optical character recognition models to perform optical character recognition on the portion of the image data,
the one or more optical character recognition models being determined based upon a type of segment of the portion of the image data;

perform optical character recognition on the portion of the image data via the one or more optical character recognition models,
the optical character recognition producing recognized text;

obtain, based on the recognized text, validation data,
the validation data including verification text;

determine whether the recognized text is verified based on the verification text; and
perform an action based on a result of the determination of whether the recognized text is verified,
the action including retraining, based on the image data and the result, the one or more optical character recognition models to recognize text included in another segment of interest that is similar to the segment of interest.

2. The device of claim 1 , where the one or more processors, when performing the action based on the recognized text, are to:
provide the recognized text as output; and
receive, based on the output, user input indicating the validation data.

3. The device of claim 1 , where the validation data is obtained from at least one of:
an address database,
an application, or
a webpage.

4. The device of claim 1 , where the validation data includes an account identifier.

5. The device of claim 4 , where the account identifier is compared to other information to determine whether the account identifier is accurate.

6. The device of claim 1 , where the one or more processors, when performing the action based on the result of the determination of whether the recognized text is verified, are to:
identify, based on the recognized text, at least one data structure to be updated; and
update the at least one data structure using the recognized text.

7. The device of claim 1 , where the one or more processors, when performing the action based on the result of the determination of whether the recognized text is verified, are to:
cause an optical character recognition model that was used to perform optical character recognition on the segment of interest to be retrained based on the recognized text.

8. A method, comprising:
receiving, by a device, image data representing a document, the document including:
text, and
a plurality of edges within the document;

identifying, by the device and based on the plurality of edges, a segment of interest within the image data;
cropping, by the device, the segment of interest to obtain a portion of the image data, the segment of interest being defined by one or more edges, of the plurality of edges, forming a rectangular shape;
performing, by the device, optical character recognition on the portion of the image data, the optical character recognition producing recognized text;

obtaining, by the device and based on the recognized text, validation data, the validation data including verification text;
determining, by the device, whether the recognized text is verified based on the verification text; and
performing, by the device, an action based on a result of the determination of whether the recognized text is verified,
the action including retraining, based on the image data and the result, the one or more optical character recognition models to recognize text included in another segment of interest that is similar to the segment of interest.

9. The method of claim 8 , further comprising:
determining one or more optical character recognition models to perform the optical character recognition on the portion of the image data,
the one or more optical character recognition models being determined based upon a type of segment of the portion of the image data.

10. The method of claim 8 , further comprising:
determining a document type associated with the image data; and
obtaining, based on the document type, data identifying a classification for the segment of interest.

11. The method of claim 8 , where the validation data is obtained from at least one of:
an address database,
an application, or
a webpage.

12. The method of claim 8 , where the validation data includes an account identifier.

13. The method of claim 12 , where the account identifier is compared to other information to determine whether the account identifier is accurate.

14. The method of claim 8 , where performing the action based on the result of the determination of whether the recognized text is verified comprises:
providing the recognized text as output; and
receiving based on the output, user input indicating the validation data.

15. A non-transitory computer-readable medium storing instructions, the instructions comprising:
one or more instructions that, when executed by one or more processors, cause the one or more processors to:
receive image data representing a document,
the document including:
text, and
a plurality of edges within the document;

identify, based on the plurality of edges, a plurality of segments within the image data;
identify, from the plurality of segments, a segment of interest based on a location of the segment of interest within the image data;
crop the segment of interest to obtain a portion of the image data, the segment of interest being defined by one or more edges, of the plurality of edges, forming a rectangular shape;
perform optical character recognition on the portion of the image data, the optical character recognition producing recognized text;
obtaining, based on the recognized text, validation data, the validation data including verification text;
determine whether the recognized text is verified based on the verification text; and
perform an action based on a result of the determination of whether the recognized text is verified,
the action including retraining, based on the image data and the result, the one or more optical character recognition models to recognize text included in another segment of interest that is similar to the segment of interest.

16. The non-transitory computer-readable medium of claim 15 , where the one or more instructions, that cause the one or more processors to perform the action based on the result of the determination of whether the recognized text is verified, cause the one or more processors to:
identify, based on the recognized text, at least one data structure to be updated; and
update the at least one data structure using the recognized text.

17. The non-transitory computer-readable medium of claim 15 , where the one or more instructions, that cause the one or more processors to perform the action based on the result of the determination of whether the recognized text is verified, cause the one or more processors to:
provide the recognized text as output; and
receive, based on the output, user input indicating the validation data.

18. The non-transitory computer-readable medium of

Assignees

• Capital One Services Llc

Inventors

• Tripuraneni Subhashini
• Barco Jr Joseph R

Classifications

• G06V30/1473

  Recognising objects as potential recognition candidates based on visual cues, e.g. shapes · CPC title

• G06V30/12

  Detection or correction of errors, e.g. by rescanning the pattern · CPC title

• G06V30/414

  Extracting the geometrical structure, e.g. layout tree; Block segmentation, e.g. bounding boxes for graphics or text · CPC title

• G06F16/5846 Primary

  using extracted text · CPC title

• G06V10/44

  Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components · CPC title