PERFORMING OPTICAL CHARACTER RECOGNITION USING SPATIAL INFORMATION OF REGIONS WITHIN A STRUCTURED DOCUMENT

• US 20180032842A1
• Filed: 07/26/2016
• Published: 02/01/2018
• Est. Priority Date: 07/26/2016
• Status: Active Grant

• Current Assignee: Intuit, Inc.
• Original Assignee: Intuit, Inc.
• Inventors: YELLAPRAGADA, Vijay, CHIANG, Peijun, MADDIKA, Sreeneel K.
• Granted Patent: US 10,013,643 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 18/214 Generating training patterns; Bootstrap methods, e.g. bagging or boosting
• CPC Class Codes: G06F 18/24 Classification techniques
• CPC Class Codes: G06Q 40/123 Tax preparation or submission
• CPC Class Codes: G06T 2207/20081 Training; Learning
• CPC Class Codes: G06T 2207/30176 Document
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/19147 Obtaining sets of training patterns; Bootstrap methods, e.g. bagging or boosting
• CPC Class Codes: G06V 30/412 Layout analysis of documents structured with printed lines or input boxes, e.g. business forms or tables
• CPC Class Codes: G06V 30/414 Extracting the geometrical structure, e.g. layout tree; Block segmentation, e.g. bounding boxes for graphics or text

Abstract

Techniques are disclosed for facilitating optical character recognition (OCR) by identifying one or more regions in an electronic document to perform the OCR. For example a method for identifying information in an electronic document includes obtaining a set of training documents for each template of a plurality of templates for the electronic document, extracting spatial attributes for at least a first label region and at least a first corresponding value region from the set, and training a classifier model based on the extracted spatial attributes, wherein the classifier model is used to identify the information in the electronic document. The spatial attributes represent a position of at least the first label region and at least the first value region within the electronic document.

Claims

1. A computer-implemented method for identifying information in an electronic document, comprising:

obtaining a set of training documents for each template of a plurality of templates for the electronic document;

extracting spatial attributes for at least a first label region and at least a first corresponding value region from the set, the spatial attributes representing a position of at least the first label region and at least the first value region within the electronic document; and

training a classifier model based on the extracted spatial attributes, wherein the classifier model is used to identify the information in the electronic document.

2. The method of claim 1, wherein the information comprises at least one label and at least one corresponding value in the electronic document.

3. The method of claim 1, further comprising:

segmenting an image of the electronic document to obtain spatial attributes of a plurality of regions in the electronic document, the regions including at least a first label and at least a first corresponding value; and

identifying at least the first label based on the trained classifier model.

4. The method of claim 3, further comprising:

designating one or more of the regions that are not identified as labels, as value regions; and

performing Optical Character Recognition in the one or more value regions to obtain at least one value corresponding to the identified at least one label.

5. The method of claim 4, further comprising identifying the designated one or more value regions as corresponding to the identified at least one label based on the classifier model.

6. The method of claim 5, wherein the designated one or more value regions are identified as corresponding to the identified at least one label based at least on a position of the designated one or more value regions relative to the at least one label.

7. The method of claim 3, further comprising obtaining the image of the electronic document by capturing the image using a camera device of a mobile device.

8. The method of claim 1, wherein the spatial attributes comprises at least one of dimensions of each of the at least one label region and the at least one corresponding value region, position of each of the at least one label region and the at least one corresponding value region within the electronic document, or position of the at least one label region and the at least one corresponding value region relative to other regions in the electronic document.

9. The method of claim 1, wherein the electronic document comprises a semi-structured document.