System and method for detection and segmentation of touching characters for OCR

• US 9,922,263 B2
• Filed: 03/20/2013
• Issued: 03/20/2018
• Est. Priority Date: 04/12/2012
• Status: Active Grant

• Current Assignee: Indian Statistical Institute, TATA Consultancy Services Limited (Tata Sons Pvt Ltd.)
• Original Assignee: Indian Statistical Institute, TATA Consultancy Services Limited (Tata Sons Pvt Ltd.)
• Inventors: Chattopadhyay, Tanushyam, Pal, Arpan, Sinha, Aniruddha, Garain, Utpal
• Primary Examiner(s): Carter, Aaron W
• Application Number:
  US14/387,094
• Publication Number:
  US 20150086113A1
• Time in Patent Office: 1,826 Days
• CPC Class Codes: G06V 10/267 by performing operations on regions, e.g. growing, shrinking or watersheds
• CPC Class Codes: G06V 30/153 using recognition of characters or words
• CPC Class Codes: G06V 30/158 using character size, text spacings or pitch estimation

Abstract

The present disclosure relates to a system and a method for detection of touching characters in a media, characterized by segmentation of adjoining character spaces. In the very first step, an aspect ratio is calculated for each connected component. A candidate touching position of each character is determined by calculating a threshold aspect ratio for each character. Further, a candidate cut column is determined based on a relation between column pixel densities and corresponding length thereof the column in order to segment the touching characters at the candidate cut column.

Claims

1. A method for detection of touching characters in a media by segmentation of adjoining character spaces, the method comprising:

acquiring each component of the media in a predetermined sequence, each component having at least two touching characters;

determining an aspect ratio of each component; and

performing a component investigation for each aspect ratio higher than a threshold aspect ratio, the component investigation comprising;

determining a candidate touching position of the at least two touching characters in a plurality of geometric orientations of the at least two touching characters;

computing a number of pixels representing a text of the at least two touching characters;

computing a length of a longest run of the number of pixels representing the text of the at least two touching characters for each column of the component;

determining a candidate cut column based on a relation between a column pixel density and a corresponding length of the column; and

segmenting the at least two touching characters with a referential boundary of the candidate cut column in the component.

2. The method of claim 1, wherein the media includes images captured from at least one of a camera, scanned images, and videos.

3. The method of claim 1, wherein the media is binarized before the aspect ratio of each component is determined.

4. The method of claim 1, wherein a value of the threshold aspect ratio is 1.

5. The method of claim 1, wherein the relation between the column pixel density and the corresponding length of the column includes a product of the number of pixels representing the text of the at least two touching characters and the length of longest run of the number of pixels.

6. The method of claim 1, wherein the aspect ratio is determined by dividing a width of the component by a height of the component.

7. A system for detection of touching characters in a media by segmentation of adjoining character spaces, the system comprising:

an input device configured to acquire each component of the media in a predetermined sequence, each component having at least two touching characters; and

a processor configured to determine an aspect ratio of each component, wherein the processor performs a component investigation for each aspect ratio higher than a threshold aspect ratio, wherein the processor;

determines a candidate touching position of the at least two touching characters in a plurality of geometric orientations of the characters;

computes a number of pixels representing a text of the at least two touching characters;

computes a length of a longest run of the number of pixels representing the text of the at least two touching characters for each column of the component;

determines a candidate cut column based on a column pixel density and a corresponding length of the column; and

segments the at least two touching characters with a referential boundary of the candidate cut column in the component.

8. The system of claim 7, wherein the media includes images captured from at least one of a camera, scanned images, and videos.

9. The system of claim 7, wherein a value of the threshold aspect ratio is 1.

10. The system of claim 7, wherein the processor binarizes the media before determining the aspect ratio of each component.

11. The system of claim 7, wherein the aspect ratio is determined by dividing a width of the component by a height of the component.

12. A computerized system for detection of touching characters in media by segmentation of adjoining character spaces, the system comprising:

at least one media article having a plurality of components, wherein at least a portion of the plurality of components has at least two touching characters; and

a computerized device receiving the at least one media article through at least one input device, wherein at least a portion of the plurality of components is acquired through the input device in a predetermined sequence, wherein the computerized device has a non-transient memory and a processor, the processor capable of executing a plurality of code stored on the non-transient memory, wherein the plurality of code further comprises;

code for determining an aspect ratio of each component;

code for performing a component investigation for each aspect ratio higher than a threshold aspect ratio;

code for determining a candidate touching position of the at least two touching characters in a plurality of geometric orientations of the characters;

code for computing a number of pixels representing a text of the at least two touching characters;

code for computing a length of a longest run of the number of pixels representing the text of the at least two touching characters for each column of the component;

code for determining a candidate cut column based on a column pixel density and a corresponding length of the column; and

code for segmenting the at least two touching characters with a referential boundary of the candidate cut column in the component.

13. The computerized system of claim 12, wherein the at least one media article includes images captured from at least one of a camera, scanned images, and videos.

14. The computerized system of claim 12, wherein a value of the threshold aspect ratio is 1.

15. The computerized system of claim 12, further comprising code for binarizing the at least one media article before determining the aspect ratio of each of the plurality of components acquired through the input device.

16. The computerized system of claim 12, wherein the aspect ratio is determined by dividing a width of the component by a height of the component.