Document Processing Apparatus, Document Processing Method and Scanner

• US 20130322757A1
• Filed: 05/29/2013
• Published: 12/05/2013
• Est. Priority Date: 05/31/2012
• Status: Active Grant

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: PAN, Yifeng, SUN, Jun, HE, Yuan, NAOI, Satoshi
• Granted Patent: US 9,070,035 B2
• Time in Patent Office: Days
• US Class Current: 382/182
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/1463 Orientation detection or correction, e.g. rotation of multiples of 90 degrees
• CPC Class Codes: G06V 30/224 of printed characters having additional code marks or containing code marks
• CPC Class Codes: G06V 30/40 Document-oriented image-based pattern recognition

Abstract

The disclosure provides a document processing apparatus, method and a scanner. The document processing apparatus includes: a text line extraction unit extracting a text line from an input document; a language classification unit determining whether an OCR process is necessary for a language of the input document; an OCR unit determining, by performing the OCR process, an OCR confidence in the case that it is determined that the OCR process is necessary; an graphic feature recognition unit determining an graphic feature recognition confidence; and a determination unit determining a combination confidence based on at least one of the determined graphic feature recognition confidences and the determined OCR confidences, and determining an orientation of the input document based on the combination confidences. This technical solution can determine better an orientation of the document, and is especially applicable when the quality of the image of the document is deteriorated.

Claims

1. A document processing apparatus, comprising:

a text line extraction unit configured to extract at least one text line from an input document;

a language classification unit configured to determine, by performing an graphic feature recognition process, whether or not an optical character recognition process is necessary for a language of the input document;

an optical character recognition unit configured to determine, by performing the optical character recognition process, an optical character recognition confidence with respect to each candidate direction for each of at least some of the text lines, in the case that it is determined that the optical character recognition process is necessary for the language of the input document;

an graphic feature recognition unit configured to determine, by performing the graphic feature recognition processing, an graphic feature recognition confidence with respect to each candidate direction for each text lines; and

a determination unit configured to determine, based on at least one of the determined graphic feature recognition confidences and the determined optical character recognition confidences, a combination confidence with respect to each candidate direction for each of the at least some of the text lines, and determine, based on the combination confidences, an orientation of the input document.

2. The document processing apparatus according to claim 1, wherein:

the language classification unit is further configured to determine that the optical character recognition process is necessary for the language of the input document, in the case that the language of the input document is determined to be Chinese, Japanese, Korean or a language in Latin alphabet.

3. The document processing apparatus according to claim 1, wherein:

the text line extraction unit is further configured to calculate, for each of the text lines, the possibility of being a line of characters in the document as a text line confidence; and

the optical character recognition unit is further configured to determine, according to an descending order of the text line confidences, the optical character recognition confidence for each text line one by one, until the orientation of the document is determined by the determination unit.

4. The document processing apparatus according to claim 3, wherein:

the determination unit is further configured to determine, according to the descending order of the text line confidences, the combination confidence for each of the at least some of the text lines, determine, based on the determined combination confidences, a current confidence with respect to each candidate direction, and determine, in the case that a maximum value among the current confidences complies with s a first predetermined condition, the candidate direction corresponding to the current confidence of the maximum value to be the orientation of the input document.

5. The document processing apparatus according to claim 4, wherein:

the determination unit is further configured to determine, based on the determined combination confidences and the determined optical character recognition confidences, the current confidence with respect to each candidate direction.

6. The document processing apparatus according to claim 1, wherein:

the text line extraction unit is further configured to calculate, for each of the text lines, the possibility of being a line of characters in the document as a text line confidence; and

the language classification unit is further configured to determine, according to a descending order of the text line confidences, a text line language confidence with respect to each candidate language for each of the text lines one by one, determine, based on the determined text line language confidences, a current language confidence with respect to each candidate language, and determine, in the case that a maximum value among the current language confidences complies with a second predetermined condition, the candidate language corresponding to the language confidence of the maximum value as the language of the input document.

7. The document processing apparatus according to claim 1, further comprising:

a resize unit configured to resize the input document and provide the resized document to the text line extraction unit for extracting the text lines.

8. The document processing apparatus according to claim 1, further comprising:

a binarization unit configured to convert the input document into a binarized image document, and provide the binarized image document to the text line extraction unit for extracting the text lines.

9. A scanner, comprising the document processing apparatus according to claim 1 for determining the orientation of the input document.

10. A document processing method, comprising:

extracting at least one text line from an input document;

determining, by performing an graphic feature recognition process, whether or not an optical character recognition process is necessary for a language of the input document;

determining, by performing the optical character recognition process, an optical character recognition confidence with respect to each candidate direction for each of at least some of the text lines, in the case that it is determined that the optical character recognition process is not necessary for the language of the input document;

determining, by performing the graphic feature recognition processing, an graphic feature recognition confidence with respect to each candidate direction for each text lines; and

determining, based on at least one of the determined graphic feature recognition confidences and the determined optical character recognition confidences, a combination confidence with respect to each candidate direction for each of the at least some of the text lines, and determining, based on the combination confidences, an orientation of the input document.

11. The document processing method according to claim 10, wherein in the language classification step, it is determined that the optical character recognition process is necessary for the language of the input document, in the case that the language of the input document is determined to be Chinese, Japanese, Korean or a language in Latin alphabet.

12. The document processing method according to claim 10, wherein

in the text line extraction step, for each of the text lines, the possibility of being a line of the characters in the document is further calculated as text line confidences;

and

in the optical character recognition step, the optical character recognition confidence for each text line is further determined one by one according to an descending order of the text line confidences, until the orientation of the document is determined by the determination step.

13. The document processing method according to claim 12, wherein

in the determination step, the combination confidence for each of the at least some of the text line is determined according to the descending order of the text line confidences, a current confidence with respect to each candidate direction is determined based on the determined combination confidences, and in the case that a maximum value among the current confidences meets a first predetermined condition, the candidate direction corresponding to the current confidence of the maximum value is determined to be the orientation of the input document.

14. The document processing method according to claim 13, wherein

in the determination step, the current confidence with respect to each candidate direction is determined based on the determined combination confidences and the determined optical character recognition confidences.

15. The document processing method according to claim 10, wherein

in the text line extraction step, for each of the text lines, the possibility of being a line of the characters in the document is further calculated as text line confidences;

and

in the language classification step, a text line language confidence with respect to each candidate language for each of the text lines is determined one by one according to a descending order of the text line confidences, a current language confidence with respect to each candidate language is determined based on the determined text line language confidences, and in the case that a maximum value among the current language confidences complies with a second predetermined condition, the candidate language corresponding to the language confidence of the maximum value is determined as the language of the input document.

16. The document processing method according to claim 10, further comprising:

a resizing step for resizing the input document and applying the resized document for extracting the text lines in the text line extraction step.

17. The document processing method according to claim 10, further comprising:

a binarization step for converting the input document into a binarized image document, wherein the language of the binarized image document is used for extracting the text lines in the text line extraction step.

18. A program causing a computer device to execute the document processing method according to claim 10, for determining an orientation of the input document.

19. A computer readable storage medium on which a program executable by a computer is stored, and when being executed, the program causes the computer device to execute the document processing method according to claim 10.