Image estimation apparatus, image processing apparatus and computer readable medium

1. An image estimation apparatus comprising:

a forming unit that forms a reference frame with a size determined uniquely in accordance with a size of a recording region where an image is recorded; and

an estimating unit that estimates a type of a single image based on a size of a bounding frame around the single image and the size of the reference frame formed by the forming unit, the single image being a target of type inference in an image contained in a designation region in a recording region image which is an image indicating the recording region where an image is recorded,

wherein the estimating unit estimates that the single image as the target of type inference contained in the bounding frame is a character image expressing a character when the bounding frame is included in the reference frame formed by the forming unit, and estimates that the single image as the target of type inference contained in the bounding frame is another image than the character image when the bounding frame is not included in the reference frame formed by the forming unit,

when the estimating unit estimates that the single image is the character image, the estimating unit further estimates that the designation region is a character image region containing one or more character images when an occupied quantity of character images in the designation region is not smaller than an occupied quantity of one or more images other than the character images in the designation region, and

when the estimating unit estimates that the single image is said another image than the character image, the estimating unit further estimates that the designation region is a non-character image region containing one or more images other than character images when an occupied quantity of character images in the designation region is smaller than an occupied quantity of one or more images other than the character images in the designation region.

2. The image estimation apparatus according to claim 1 , wherein:

the forming unit further forms a table inference reference frame with a larger size than the size of the reference frame;

the image estimation apparatus further comprises a grating image detecting unit which detects a specific color line segment from the non-character image region estimated by the estimating unit and which detects a grating image including the detected line segment as a constituent member when the line segment is detected; and

the estimating unit further estimates that the non-character image region is a table image region containing a table image indicating a table when both a first condition and a second condition are satisfied, and estimates that the non-character image region is a general image region containing a general image expressing another image than the character image and the table image when at least one of the first and second conditions is not satisfied, the first condition being a condition that a size of the grating image detected by the grating image detecting unit is substantially equal to a size of the table inference reference frame formed by the forming unit, the second condition being a condition that character images are contained in a region surrounded by a contour of the grating image detected by the grating image detecting unit.

3. The image estimation apparatus according to claim 2 , wherein:

the table inference reference frame is a rectangular frame with a size determined uniquely in accordance with a vertical length and a lateral length in the recording region.

4. The image estimation apparatus according to claim 1 , wherein:

the single image is an image formed from a set of continuous specific color pixels.

5. The image estimation apparatus according to claim 1 , wherein:

the reference frame is a rectangular frame with a size determined uniquely in accordance with a vertical length and a lateral length in the recording region.

6. The image estimation apparatus according to claim 1 , wherein:

the bounding frame is a bounding rectangular frame around the single image which is the target of type inference.

7. An image processing apparatus comprising:

a forming unit which forms a reference frame with a size determined uniquely in accordance with a size of a recording region where an image is recorded;

an estimating unit which estimates a type of a single image based on a size of a bounding frame around the single image and the size of the reference frame formed by the forming unit, the single image being a target of type inference in an image contained in a designation region in a recording region image which is an image indicating the recording region where an image is formed; and

a hiding unit which hides a specific region contained in the designation region by a hiding method determined uniquely in accordance with the type estimated by the estimating unit,

wherein the estimating unit estimates that the single image as the target of type inference contained in the bounding frame is a character image expressing a character when the bounding frame is included in the reference frame formed by the forming unit, and estimates that the single image as the target of type inference contained in the bounding frame is another image than the character image when the bounding frame is not included in the reference frame formed by the forming unit,

when the estimating unit estimates that the single image is the character image, the estimating unit further estimates that the designation region is a character image region containing one or more character images when an occupied quantity of character images in the designation region is not smaller than an occupied quantity of one or more images other than the character images in the designation region, and

when the estimating unit estimates that the single image is said another image than the character image, the estimating unit further estimates that the designation region is a non-character image region containing one or more images other than character images when an occupied quantity of character images in the designation region is smaller than an occupied quantity of one or more images other than the character images in the designation region.

8. The image processing apparatus according to claim 7 , wherein:

the hiding unit hides the character image region estimated by the estimating unit after shaping the character image region into a predetermined shape to embrace the character image region in the specific region, and hides the specific region contained in the non-character image region estimated by the estimating unit by a hiding method corresponding to the type of each image contained in the specific region.

9. The image processing apparatus according to claim 8 , wherein:

the forming unit further forms a table inference reference frame with a larger size than the size of the reference frame;

the image processing apparatus further comprises a grating image detecting unit which detects a specific color line segment from the non-character image region estimated by the estimating unit and which detects a grating image including the detected line segment as a constituent member when the line segment is detected;

the estimating unit further estimates that the non-character image region is a table image region containing a table image indicating a table when both a first condition and a second condition are satisfied, and estimates that the non-character image region is a general image region containing a general image indicating another image than the character image and the table image when at least one of the first and second conditions is not satisfied, the first condition being a condition that a size of the grating image detected by the grating image detecting unit is substantially equal to a size of the table inference reference frame formed by the forming unit, the second condition being a condition that character images are contained in a region surrounded by a contour of the grating image detected by the grating image detecting unit; and

the hiding unit hides the character image region estimated by the estimating unit after shaping the character image region into a predetermined shape to embrace the character image region in the specific region, hides an entire region surrounded by a contour of the table image region estimated by the estimating unit, and hides an entire region surrounded by a contour of the general image region estimated by the estimating unit.

10. The image processing apparatus according to claim 8 , wherein:

the forming unit further forms a table inference reference frame with a larger size than the size of the reference frame;

the image processing apparatus further comprises a grating image detecting unit which detects a specific color line segment from the non-character image region estimated by the estimating unit and which detects a grating image including the detected line segment as a constituent member when the line segment is detected;

the estimating unit further estimates that the non-character image region is a table image region containing a table image indicating a table when both a first condition and a second condition are satisfied, and estimates that the non-character image region is a general image region containing a general image indicating another image than the character image and the table image when at least one of the first and second conditions is not satisfied, the first condition being a condition that a size of the grating image detected by the grating image detecting unit is substantially equal to a size of the table inference reference frame formed by the forming unit, the second condition being a condition that character images are contained in a region surrounded by a contour of the grating image detected by the grating image detecting unit; and

the hiding unit hides the character image region estimated by the estimating unit after shaping the character image region into a predetermined shape to embrace the character image region in the specific region, hides an entire region surrounded by the contour of the grating image detected by the grating image detecting unit, and hides an entire region surrounded by a contour of the general image region estimated by the estimating unit.

11. The image processing apparatus according to claim 8 , wherein:

the forming unit further forms a table inference reference frame with a larger size than the size of the reference frame;

the image processing apparatus further comprises a grating image detecting unit which detects a specific color line segment from the non-character image region estimated by the estimating unit and which detects a grating image including the detected line segment as a constituent member when the line segment is detected;

the estimating unit further estimates that the non-character image region is a table image region containing a table image indicating a table when both a first condition and a second condition are satisfied, and estimates that the non-character image region is a general image region containing a general image indicating another image than the character image and the table image when at least one of the first and second conditions is not satisfied, the first condition being a condition that a size of the grating image detected by the grating image detecting unit is substantially equal to a size of the table inference reference frame formed by the forming unit, the second condition being a condition that character images are contained in a region surrounded by a contour of the grating image detected by the grating image detecting unit; and

the hiding unit hides the character image region estimated by the estimating unit after shaping the character image region into a predetermined shape to embrace the character image region in the specific region, hides an entire region surrounded by a contour of a cell contained in the grating image detected by the grating image detecting unit and contained in the designation region, and hides an entire region surrounded by a contour of the general image region estimated by the estimating unit.

12. The image processing apparatus according to claim 9 , wherein:

the table inference reference frame is a rectangular frame with a size determined uniquely in accordance with a vertical length and a lateral length in the recording region.

13. The image processing apparatus according to claim 7 , wherein:

the single image is an image formed from a set of continuous specific color pixels.

14. The image processing apparatus according to claim 7 , wherein:

the reference frame is a rectangular frame with a size determined uniquely in accordance with a vertical length and a lateral length in the recording region.

15. The image processing apparatus according to claim 7 , wherein:

the bounding frame is a bounding rectangular frame around the single image which is the target of type inference.

16. A non-transitory computer readable medium storing a program causing a computer to execute a process for an image estimating method, the process comprising:

forming a reference frame with a size determined uniquely in accordance with a size of a recording region where an image is recorded; and

estimating a type of a single image based on a size of a bounding frame around the single image and the size of the formed reference frame, the single image being a target of type inference in an image contained in a designation region in a recording region image which is an image indicating the recording region where an image is recorded,

wherein the estimating step includes estimating that the single image as the target of type inference contained in the bounding frame is a character image expressing a character when the bounding frame is included in the formed reference frame, and estimating that the single image as the target of type inference contained in the bounding frame is another image than the character image when the bounding frame is not included in the formed reference frame,

when the single image is estimated as the character image, the process further comprises estimating that the designation region is a character image region containing one or more character images when an occupied quantity of character images in the designation region is not smaller than an occupied quantity of one or more images other than the character images in the designation region, and

when the single image is estimated as said another image than the character image, the process further comprises estimating that the designation region is a non-character image region containing one or more images other than character images when an occupied quantity of character images in the designation region is smaller than an occupied quantity of one or more images other than the character images in the designation region.

17. A non-transitory computer readable medium storing a program causing a computer to execute a process for an image processing method, the process comprising:

forming a reference frame with a size determined uniquely in accordance with a size of a recording region where an image is recorded;

estimating a type of a single image based on a size of a bounding frame around the single image and the size of the formed reference frame, the single image being a target of type inference in an image contained in a designation region in a recording region image which is an image indicating the recording region where an image is recorded; and

hiding a specific region contained in the designation region by a hiding method determined uniquely in accordance with the type estimated by the estimating unit,

wherein the estimating step includes estimating that the single image as the target of type inference contained in the bounding frame is a character image expressing a character when the bounding frame is included in the formed reference frame, and estimating that the single image as the target of type inference contained in the bounding frame is another image than the character image when the bounding frame is not included in the formed reference frame,

when the single image is estimated as the character image, the process further comprises estimating that the designation region is a character image region containing one or more character images when an occupied quantity of character images in the designation region is not smaller than an occupied quantity of one or more images other than the character images in the designation region, and

when the single image is estimated as said another image than the character image, the process further comprises estimating that the designation region is a non-character image region containing one or more images other than character images when an occupied quantity of character images in the designation region is smaller than an occupied quantity of one or more images other than the character images in the designation region.