Image processing apparatus and image processing method

• US 8,326,085 B2
• Filed: 06/08/2009
• Issued: 12/04/2012
• Est. Priority Date: 06/19/2008
• Status: Expired due to Fees

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: Yokoi, Masanori
• Primary Examiner(s): COUSO, YON JUNG
• Application Number:
  US12/480,610
• Publication Number:
  US 20090316996A1
• Time in Patent Office: 1,275 Days
• Field of Search: 382/289, 382/199, 382/260, 382/263, 382/281, 382/298
• US Class Current: 382/289
• CPC Class Codes: G06T 2207/10008 from scanner, fax or copier
• CPC Class Codes: G06T 2207/20061 Hough transform
• CPC Class Codes: G06T 2207/30176 Document
• CPC Class Codes: G06T 7/12 Edge-based segmentation
• CPC Class Codes: G06T 7/70 Determining position or orientation of objects or cameras camera calibration G06T7/80
• CPC Class Codes: G06V 30/1478 of characters or characters lines

Abstract

The present invention enables inclination detection by detecting an inclination of a document image based on a feature of a document area. In order to achieve this, reduction processing is performed on document image data including the document area so as to generate a reduced document image corresponding to the document area that has been extracted. Thereafter, filter processing is performed on the reduced document image so as to generate an edge image by extracting at least one edge of the document area. Then, a straight line adjoining an edge of the edge image is detected using a Hough transformation so that the inclination of the document image is determined based on an inclination of the straight line. Consequently, inclination detection with high accuracy can be performed based on the feature of the document area.

Claims

1. An image processing apparatus, comprising:

an image input unit configured to input a document image;

a reducing unit configured to generate a reduced document image by reducing the document image;

an edge image generating unit configured to generate an edge image by extracting at least one edge of the reduced document image; and

an inclination detecting unit configured to detect an inclination of the document image based on an inclination of a straight line representing an edge of the edge image, wherein said inclination detecting unit detects the straight line representing the edge of the edge image by performing a Hough transformation, and detects the inclination of the document image based on an inclination of the straight line, and

wherein said inclination detecting unit is operable, if a plurality of straight lines that have different inclinations are detected, to determine a smallest inclination among the inclinations of the plurality of straight lines to be the inclination of the document image.

2. The image processing apparatus according to claim 1, wherein said reducing unit includes:

a first reducing unit configured to generate a first reduced image from the document image, wherein the document image comprises at least one reduction unit area, which contains a plurality of pixels, and the first reducing unit is configured to determine a value of each pixel of the first reduced image based on values of said plurality of pixels in a corresponding such reduction unit area; and

a second reducing unit configured to generate the reduced document image from the first reduced image by allocating the same pixel value to a predetermined range of pixels within the first reduced image, the allocated pixel value being based on at least one of the values of the pixels in the predetermined range.

3. The image processing apparatus according to claim 1, wherein said edge image generating unit performs edge extraction using a Laplacian filter.

4. The image processing apparatus according to claim 1, wherein said edge image generating unit performs edge extraction using a high-pass filter.

5. The image processing apparatus according to claim 1, wherein the reducing unit generates the reduced document image by:

determining, if even one black pixel exists in a reduction unit area in accordance with the reduction ratio R, a corresponding pixel in a simple reduced image as being a black pixel, and

filling up, with black pixels, pixels in a range “

L”

to connect characters or text lines in the simple reduced image if even one black pixel exists in the range L,

wherein R and L are selected so as to satisfy expressions (1) and (2);

L×

1/√

R>

cs 

 

(1)

L×

1/√

R>

ls 

 

(2)

where cs indicates the space between characters and ls indicates the space between text lines.

6. An image processing apparatus, comprising:

an image input unit configured to input a document image;

a reducing unit configured to generate a reduced document image by reducing the document image;

an edge image generating unit configured to generate an edge image by extracting at least one edge of the reduced document image; and

an inclination detecting unit configured to detect an inclination of the document image based on an inclination of a straight line representing an edge of the edge image, wherein said inclination detecting unit detects the straight line representing the edge of the edge image by performing a Hough transformation and detects the inclination of the document image based on an inclination of the straight line, and

wherein said inclination detecting unit is operable, if a plurality of straight lines that have different inclinations are detected, to select two orthogonal lines, and to determine a smaller inclination between the inclinations of the two straight lines to be the inclination of the document image.

7. The image processing apparatus according to claim 6, wherein the reducing unit generates the reduced document image by:

determining, if even one black pixel exists in a reduction unit area in accordance with the reduction ratio R, a corresponding pixel in a simple reduced image as being a black pixel, and

filling up, with black pixels, pixels in a range “

L”

to connect characters or text lines in the simple reduced image if even one black pixel exists in the range L,

wherein R and L are selected so as to satisfy expressions (1) and (2);

L×

1/√

R>

cs 

 

(1)

L×

1/√

R>

ls 

 

(2)

where cs indicates the space between characters and ls indicates the space between text lines.

8. An image processing method comprising:

inputting a document image;

generating a reduced document image by reducing the document image;

generating an edge image by extracting at least one edge of the reduced document image; and

detecting an inclination of the document image based on an inclination of a straight line representing an edge of the edge image, wherein the straight line representing the edge of the edge image is detected by performing a Hough transformation, and the inclination of the document image is detected based on an inclination of the straight line,

wherein if a plurality of straight lines that have different inclinations are detected, a smallest inclination among the inclinations of the plurality of straight lines is determined to be the inclination of the document image, and

wherein at least one processor executes steps stored in a memory to perform at least one of the inputting, generating, and detecting.

9. The image processing method according to claim 8, wherein the reducing of the document image further includes:

generating a first reduced image from the document image, wherein the document image comprises at least one reduction unit area, which contains a plurality of pixels, and the first reducing step determines a value of each pixel of the first reduced image based on values of the pixels in a corresponding such reduction unit area; and

generating the reduced document image from the first reduced image by allocating a same pixel value to a predetermined range of pixels within the first reduced image, the allocated pixel value being based on at least one of the values of the pixels in the predetermined range.

10. The method according to claim 8, wherein in the reduced document image generating step, the reduced document image is generated by:

determining, if even one black pixel exists in a reduction unit area in accordance with the reduction ratio R, a corresponding pixel in a simple reduced image as being a black pixel, and

filling up, with black pixels, pixels in a range “

L”

to connect characters or text lines in the simple reduced image if even one black pixel exists in the range L,

wherein R and L are selected so as to satisfy expressions (1) and (2);

L×

1/√

R>

cs 

 

(1)

L×

1/√

R>

ls 

 

(2)

where cs indicates the space between characters and is indicates the space between text lines.

11. A computer-readable storage medium storing a computer program configured, when run on a computer, to cause said computer to:

input a document image;

generate a reduced document image by reducing the document image;

generate an edge image by extracting at least one edge of the reduced document image; and

detect an inclination of the document image based on an inclination of a straight line representing an edge of the edge image, wherein the straight line representing the edge of the edge image is detected by performing a Hough transformation, and the inclination of the document image is detected based on an inclination of the straight line,

wherein if a plurality of straight lines that have different inclinations are detected, a smallest inclination among the inclinations of the plurality of straight lines is determined to be the inclination of the document image.

12. An image processing method, comprising:

inputting a document image;

generating a reduced document image by reducing the document image;

generating an edge image by extracting at least one edge of the reduced document image; and

detecting an inclination of the document image based on an inclination of a straight line representing an edge of the edge image, wherein the straight line representing the edge of the edge image is detected by performing a Hough transformation, and the inclination of the document image is detected based on an inclination of the straight line,

wherein if a plurality of straight lines that have different inclinations are detected, two orthogonal lines are selected, and a smaller inclination between the inclinations of the two straight lines is determined to be the inclination of the document image, and

wherein at least one processor executes steps stored in a memory to perform at least one of the inputting, generating, and detecting.

13. The method according to claim 12, wherein in the reduced document image generating step, the reduced document image is generated by:

determining, if even one black pixel exists in a reduction unit area in accordance with the reduction ratio R, a corresponding pixel in a simple reduced image as being a black pixel, and

filling up, with black pixels, pixels in a range “

L”

to connect characters or text lines in the simple reduced image if even one black pixel exists in the range L,

wherein R and L are selected so as to satisfy expressions (1) and (2);

L×

1/√

R>

cs 

 

(1)

L×

1/√

R>

ls 

 

(2)

where cs indicates the space between characters and ls indicates the space between text lines.

14. A non-transitory computer-readable storage medium storing a computer program configured, when run on a computer, to cause said computer to:

input a document image;

generate a reduced document image by reducing the document image;

generate an edge image by extracting at least one edge of the reduced document image; and

detecting an inclination of the document image based on an inclination of a straight line representing an edge of the edge image, wherein the straight line representing the edge of the edge image is detected by performing a Hough transformation, and the inclination of the document image is detected based on an inclination of the straight line, and

wherein if a plurality of straight lines that have different inclinations are detected, two orthogonal lines are selected, and a smaller inclination between the inclinations of the two straight lines is determined to be the inclination of the document image.