JP2008109361A - Image processor and processing method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processor and an image processing method for accurately deciding and removing various punch holes. <P>SOLUTION: This image processor includes a noise removing means 11 for removing segments being equal to or shorter than length preliminarily set for the width direction and height direction of a binary image; a punch hole candidate detecting means 12 for detecting a region isolated in an image acquired by the noise removing means 11 as a punch hole candidate; and a size deciding means 13 for deciding the punch hole candidate as a punch hole when the width and height of the circumscribed rectangle of the punch hole candidate is not less than the minimum value and not more than the maximum value of the preliminarily set size of the punch hole to be removed. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an image processing apparatus and an image processing method for processing a document image in which punch holes are processed.

  When a document is read by the image reading apparatus, if a punch hole is processed in the document, the background color (usually black) of the punch hole is read. Therefore, there is a problem that the punched hole is included as a shadow in the read document image, although the punched hole is originally an unnecessary part.

Therefore, as a technique for erasing the shadow of the punch hole, for example, in (Patent Document 1), a punch hole pattern is stored in a storage unit in advance, and the original reading data is collated with the punch hole pattern to determine the punch hole pattern. It is described that the image data of punch holes is erased or hollowed out.
JP-A-9-9041

  However, in the method described in (Patent Document 1), punch holes cannot be determined unless the punch holes completely coincide with the punch hole pattern. For this reason, when the punch hole has been deformed due to the use of a document or the like, there is a problem that the punch hole cannot be accurately determined and erased or the like cannot be performed.

  Accordingly, an object of the present invention is to provide an image processing apparatus and an image processing method capable of accurately determining and removing various punch holes.

  In order to solve the above-described problem, the present invention removes a line segment having a length equal to or less than a preset length in the width direction and the height direction of a binary image, and punches an isolated region in the obtained image. A configuration in which a punch hole is detected when the width and height of the circumscribed rectangle of the punch hole candidate are both greater than the preset minimum value and less than the maximum value of the punch hole size to be removed. It is.

  According to the present invention, by removing a line segment that is equal to or less than a preset length in the width direction and the height direction of the binary image, characters in the binary image, ruled lines hanging on punch holes, and the like are removed. In addition, since punch holes of a predetermined size are determined, various punch holes can be accurately determined and removed.

  According to a first aspect of the present application, a noise removing unit that removes a line segment having a length equal to or less than a preset length in the width direction and the height direction of a binary image, and an isolated image in the image obtained by the noise removing unit. The punch hole candidate detecting means for detecting the punched area as a punch hole candidate, and the width and height of the circumscribed rectangle of the punch hole candidate are both equal to or larger than the preset minimum value and the maximum value of the punch hole size to be removed. In this case, the image processing apparatus includes a size determination unit that determines a punch hole. According to the present invention, by removing a line segment that is equal to or less than a preset length in the width direction and the height direction of the binary image, characters in the binary image, ruled lines hanging on punch holes, and the like are removed. In addition, since punch holes of a predetermined size are determined, various punch holes can be accurately determined and removed.

  The second invention of the present application further determines a punch hole when the number of pixels having different values in the circumscribed rectangle is smaller than a predetermined ratio between the punch hole candidate and an ellipse inscribed in the circumscribed rectangle of the punch hole candidate. This is an image processing apparatus having pattern determining means. According to the present invention, even if the punch hole is an ellipse broken from a perfect circle, it can be determined, and even when the punch hole is deformed, it can be accurately determined and removed.

  According to a third aspect of the present invention, there is provided an image processing apparatus, wherein the punch hole candidate detecting means detects a punch hole candidate within a predetermined distance from the end of the binary image and within a predetermined width from the predetermined distance. It is a thing. According to the present invention, punch hole candidates are detected only in a portion where a punch hole may exist, so that high-speed processing is possible.

  The fourth invention of the present application is an image processing apparatus having a binarizing means for converting a multi-valued image into a binary image. The multi-valued image is also converted into a binary image, so that the same as described above. Various punch holes can be accurately determined and removed.

  According to a fifth aspect of the present invention, there is provided an image processing apparatus having punch hole removing means for removing a punch hole by performing linear interpolation using pixels in a peripheral region of a circumscribed rectangle of the punch hole candidate. The contour part of the circumscribed rectangle cannot be used for interpolation because of the large difference from the surrounding pixel values due to the binarization. However, by using the pixels in the peripheral area of the circumscribed rectangle of the punch hole candidate, With respect to the value image, punch holes can be removed using pixels in the peripheral area.

  According to a sixth aspect of the present invention, a noise removal step for removing a line segment having a length equal to or less than a preset length in the width direction and the height direction of a binary image, and an image obtained through the noise removal step The punch hole candidate detecting step for detecting an isolated area as a punch hole candidate, and the width and height of the circumscribed rectangle of the punch hole candidate are both equal to or larger than a preset minimum value of the punch hole size to be removed. The image processing method includes a size determination step of determining a punch hole when the value is equal to or less than the value. According to the present invention, by removing a line segment that is equal to or less than a preset length in the width direction and the height direction of the binary image, characters in the binary image, ruled lines hanging on punch holes, and the like are removed. In addition, since punch holes of a predetermined size are determined, various punch holes can be accurately determined and removed.

  In the seventh invention of the present application, a punch hole is determined when the number of pixels having different values in the circumscribed rectangle is smaller than a predetermined ratio between the punch hole candidate and an ellipse inscribed in the circumscribed rectangle of the punch hole candidate. The image processing method includes a pattern determination step. According to the present invention, even if the punch hole is an ellipse broken from a perfect circle, it can be determined, and even when the punch hole is deformed, it can be accurately determined and removed.

  According to an eighth aspect of the present invention, there is provided an image processing method in which the punch hole candidate detecting step detects a punch hole candidate within a predetermined distance from the end of the binary image and within a predetermined width from the predetermined distance. It is a thing. According to the present invention, punch hole candidates are detected only in a portion where a punch hole may exist, so that high-speed processing is possible.

  A ninth invention of the present application is an image processing method including a binarization processing step for converting a multi-value image into a binary image. By converting a multi-value image into a binary image as well, Similarly, various punch holes can be accurately determined and removed.

  A tenth invention of the present application is an image processing method including a punch hole removing step for removing a punch hole by performing linear interpolation using pixels in a peripheral region of a circumscribed rectangle of the punch hole candidate. The contour part of the circumscribed rectangle cannot be used for interpolation because of the large difference from the surrounding pixel values due to the binarization. However, by using the pixels in the peripheral area of the circumscribed rectangle of the punch hole candidate, With respect to the value image, punch holes can be removed using pixels in the peripheral area.

  Embodiments of the present invention will be described below with reference to the drawings.

  FIG. 1 is a block diagram showing a configuration of an image processing apparatus according to an embodiment of the present invention.

  In FIG. 1, an image processing apparatus 1 according to an embodiment of the present invention is connected to an image reading apparatus 2 that reads an original image in color, and binarizing means 10 that converts a multi-valued image into a binary image; Noise removing means 11 for removing noise from the image significantly, punch hole candidate detecting means 12 for detecting an isolated region as a punch hole candidate, and size judging means 13 for judging a punch hole based on a preset size The pattern determining means 14 for determining a punch hole based on a difference from a predetermined pattern, and the punch hole removing means 15 for removing a portion determined to be a punch hole.

  The binarizing means 10 binarizes the color original image (multi-valued image) read by the image reading device 2 with a predetermined threshold value, for example. Note that the image reading apparatus 2 may read a monochrome image, and in this case, the binarizing means 10 may be omitted. In the following, an example in which the background color of the document image is white and the punched hole portion is black will be described, but the same process can be performed in the opposite case.

  The noise removing unit 11 sets a preset length (for example, 0.7 mm in both the width direction and the height direction) in the width direction and the height direction of the binary image obtained by the binarization unit 10. ) Noise is removed by removing the following line segments.

  In this way, by removing line segments that are less than or equal to a preset length in the width direction and height direction of the binary image, characters in the binary image, ruled lines that are applied to punch holes, and the like are removed. .

  The punch hole candidate detecting means 12 detects an isolated region in the image obtained by the noise removing means 11 as a punch hole candidate. Here, the punch hole candidate detecting means 12 is separated from the end of the binary image by a predetermined distance PD (for example, 2 mm) or more and within a predetermined width HD (for example, 20 mm) from the predetermined distance PD. A punch hole candidate is detected in the region.

  FIG. 2 is a diagram showing an area where punch hole candidate detection means detects a punch hole candidate, and an area where punch hole candidate detection means 12 detects a punch hole candidate (hereinafter referred to as “detection area”) 20a. , 20b, 20c, 20d.

  FIG. 3 is a diagram illustrating an example in which the punch hole candidate is not processed. As shown in FIG. 3, the punch hole candidate detecting means 12 is used when the punch hole candidate 30 is in contact with the detection areas 20a, 20b, 20c and 20d (in the case shown in FIG. 3A), or when the detection area 20a, When the punch hole candidate 30 straddles 20b, 20c, and 20d (as shown in FIG. 5B), the punch hole candidate 30 is not processed.

  FIG. 4 is a diagram showing the overlapping of the detection areas at the corners. There is a possibility that the connecting portions of the detection areas 20a, 20b, 20c, and 20d at the corners of the two-dimensional image may not be accurately processed as the punch hole candidates 30 corresponding to the case where the punch hole candidates 30 contact and straddle. Therefore, as shown in FIG. 4, a region corresponding to a maximum value Smax of the punch hole size described later is detected in an overlapping manner.

  The size determination means 13 is such that the width and height of the circumscribed rectangle of the punch hole candidate 30 are both equal to or larger than the preset minimum value Smin (for example, 4 mm) of the punch hole size to be removed and the maximum value Smax ( For example, it is determined to be 12 mm.) In the following cases, it is determined as a punch hole, and in the case of being out of any one of the conditions, it is not determined as a punch hole.

  FIG. 5 is a conceptual diagram of punch hole determination by the size determination means. As shown in FIG. 5 (a), the size determining means 13 determines whether the width w and the height h of the circumscribed rectangle 31 of the punch hole candidate 30 are Smin ≦ w ≦ Smax and Smin ≦ h ≦ Smax. Is determined. Further, as shown in FIG. 5B, the size determining means 13 does not determine that the hole is a punch hole when w <Smin or h <Smin. Further, as shown in FIG. 5 (c), the size determining means 13 does not determine a punch hole even when Smax <w or Smin <h.

  The pattern determination means 14 is configured such that the number of pixels having different values in the circumscribed rectangle 31 is smaller than a predetermined ratio between the punch hole candidate 30 and a predetermined pattern (for example, an ellipse) inscribed in the circumscribed rectangle 31 of the punch hole candidate 30. In this case, the punch hole is determined.

  FIG. 6 is a conceptual diagram of punch hole determination by the pattern determination means. As shown in FIG. 6, the pattern determination unit 14 divides each pixel in the circumscribed rectangle 31 of the punch hole candidate 30 into three regions (white, black, white) in the horizontal direction one line at a time. The number of pixels different from the color (i.e., black if inside the ellipse inscribed in the circumscribed rectangle 31 of the punch hole candidate 30 and white if outside) is counted.

  FIG. 7 is a diagram showing an example of counting the number of pixels by the pattern determining means. As shown in FIG. 7, the pattern determining means 14 counts diffCnt = 3 as the number of different pixels when the punch candidate 30 is applied to the originally white area on the left side and black is included. Further, when white is included in the originally black region in the center, diffCnt = 2 is counted as a different number of pixels.

  Then, the pattern determination unit 14 performs such determination for the entire area, and the number of pixels of the color different from the original color (diffCnt) is a predetermined ratio (HOLE_RATIO, for example, the number of pixels (W0 × H0) of the entire area) If it is smaller than 15 (%), it is determined as a punch hole if it is smaller (diffCnt / (W0 × H0) × 100 <HOLD_RATIO), and if it is larger, it is not determined as a punch hole.

  The punch hole removing means 15 removes punch holes by performing linear interpolation using pixels in the peripheral area of the circumscribed rectangle 31 of the punch hole candidate 30.

  FIG. 8 is a conceptual diagram of punch hole removal by the punch hole removing means. As shown in FIG. 8A, the punch hole removing means 15 performs linear interpolation using pixels in a rectangle 32 in a peripheral area that is widened by a predetermined distance d outward from the circumscribed rectangle 31 of the punch hole candidate 30. Is used to remove punch holes.

  For example, as shown in FIG. 8B, the punch hole removing unit 15 weights four pixels P0, P1, P2, and P3 that are in contact with the circumscribed rectangle 31 in the vertical and horizontal directions from the target pixel V according to the distances. And interpolate to determine the pixel value after removal. Each pixel value V after removal is

となる。例えば、Ｐ０＝２１０，Ｐ１＝２０５，Ｐ２＝２０７，Ｐ３＝２１２，Ｘ＝１０，Ｙ＝５０，Ｗ＝１００，Ｈ＝１５０のときは、

It becomes. For example, when P0 = 210, P1 = 205, P2 = 207, P3 = 212, X = 10, Y = 50, W = 100, H = 150,

となる。

It becomes.

  Next, punch hole removal by the image processing apparatus 1 configured as described above will be described with reference to the flowcharts of FIGS. FIG. 9 is a flowchart showing a processing procedure of the entire image processing. FIG. 10 is a flowchart showing details of the punch hole removing process of FIG.

  As shown in FIG. 9, the image processing apparatus 1 determines the image type of the original image read by the image reading apparatus 2 (S101). If the image type is other than monochrome binary, the binarizing means 10 Binarization processing is performed (S102). Next, the image processing apparatus 1 performs noise removal processing by the noise removing unit 11 (S103). Then, the image processing apparatus 1 obtains a detection region from the left end of the binary image (S104), and performs punch hole removal processing on this detection region (S105).

  As shown in FIG. 10, in the punch hole removal process, the punch hole candidate detecting means 12 detects an isolated area as a punch hole candidate by extracting an isolated point in the detection area (S201). Next, the size determination means 13 determines the size of the punch hole candidate (S202), and as a result, if it is determined that the punch hole is determined, it determines whether the isolated area is within the detection area (S203).

  Subsequently, it is determined whether or not the shape of the punch hole is rectangular by the pattern determining means 14 (S204), and if it is not rectangular, it is determined whether or not it is a round punch hole shape (S205). If the punch hole has a round shape, the punch hole removing means 15 removes the punch hole (S206).

  At this time, when the type of the original image read by the image reading device 2 is monochrome binary, the punch hole is filled with white, and when it is multivalued, it is filled with the value interpolated using the pixels in the peripheral area. Then, the process proceeds to the next punch hole candidate process (S207).

  When the punch hole removal process is completed for the left detection area, the process returns to FIG. 9 to obtain the next right detection area (S106), and the punch hole removal process is similarly performed for this detection area (S107). The same processing is performed for the upper end and the lower end (S108, S109, S110, S111), and the punch hole removal is completed.

  As described above, in the image processing apparatus 1 according to the present embodiment, characters in a binary image are removed by removing line segments that are less than or equal to a preset length in the width direction and the height direction of the binary image. And the ruled lines or the like applied to the punch holes are removed, and a punch hole of a predetermined size is determined. Therefore, various punch holes can be accurately determined and removed.

  In addition, since the punch hole candidate 30 and an ellipse inscribed in the circumscribed rectangle 31 of the punch hole candidate 30 have a number of pixels having different values in the circumscribed rectangle 31 are smaller than a predetermined ratio, the punch hole is determined. Can be determined even if it is an ellipse broken from a perfect circle, and even when the punch hole is deformed, it can be accurately determined and removed.

  Further, by detecting the punch hole candidate 30 within a predetermined distance or more from the edge of the binary image and within the predetermined width from the predetermined distance, the punch hole is limited to a portion where the punch hole may exist. Candidates can be detected, and high-speed processing is possible.

  Also, by converting a multi-value image into a binary image, various punch holes can be accurately determined and removed as described above. Further, in this case, the punch holes are removed by linear interpolation using the pixels of the rectangle 32 in the peripheral area of the circumscribed rectangle 31 of the punch hole candidate 30. Therefore, the punch holes are also used for the multi-value image using the pixels in the peripheral area. Can be removed.

  The present invention is useful as an image processing apparatus and an image processing method for processing a document image in which punch holes are processed. In particular, the present invention is suitable as an image processing apparatus and an image processing method capable of accurately determining and removing various punch holes.

1 is a block diagram illustrating a configuration of an image processing apparatus according to an embodiment of the present invention. The figure which shows the area | region which detects a punch hole candidate by a punch hole candidate detection means The figure which shows the example which is not processed as a punch hole candidate The figure which shows the overlap of each detection area of a corner Conceptual diagram of punch hole determination by size determination means Conceptual diagram of punch hole judgment by pattern judgment means The figure which shows the example of the count of the number of pixels by a pattern determination means Conceptual diagram of punch hole removal by punch hole removal means Flowchart showing processing procedure of overall image processing The flowchart which shows the detail of the punch hole removal process of FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Image processing apparatus 2 Image reading apparatus 10 Binarization means 11 Noise removal means 12 Punch hole candidate detection means 13 Size determination means 14 Pattern determination means 15 Punch hole removal means 20a, 20b, 20c, 20d Detection area 30 Punch hole candidates 31 Circumscribed rectangle 32 surrounding rectangle

Claims (10)

Noise removing means for removing a line segment having a length equal to or less than a preset length in the width direction and the height direction of the binary image;
Punch hole candidate detecting means for detecting an isolated region in the image obtained by the noise removing means as a punch hole candidate;
Image processing having size determination means for determining a punch hole when the width and height of the circumscribed rectangle of the punch hole candidate are both set to a preset minimum value and a maximum value of the punch hole size to be removed apparatus. Furthermore, pattern determining means for determining a punch hole when the number of pixels having different values in the circumscribed rectangle between the punch hole candidate and an ellipse inscribed in the circumscribed rectangle of the punch hole candidate is smaller than a predetermined ratio is provided. The image processing apparatus according to claim 1. 2. The punch hole candidate detecting means detects the punch hole candidate within an area that is a predetermined distance or more away from an end of the binary image and within a predetermined width from the predetermined distance. Or the image processing apparatus of 2. 4. The image processing apparatus according to claim 1, further comprising binarization means for converting a multi-value image into the binary image. 5. The image processing apparatus according to claim 4, further comprising punch hole removing means for removing the punch hole by performing linear interpolation using pixels in a peripheral region of a circumscribed rectangle of the punch hole candidate. A noise removal step of removing line segments of a length equal to or less than a preset length in the width direction and the height direction of the binary image;
Punch hole candidate detection step for detecting an isolated region in the image obtained through this noise removal step as a punch hole candidate;
Image processing including a size determination step for determining a punch hole when the width and height of a circumscribed rectangle of the punch hole candidate are both preset minimum and maximum values of the punch hole size to be removed Method. And a pattern determination step for determining a punch hole when the number of pixels having different values in the circumscribed rectangle is smaller than a predetermined ratio between the punch hole candidate and an ellipse inscribed in the circumscribed rectangle of the punch hole candidate. The image processing method according to claim 6. The punch hole candidate detecting step detects the punch hole candidate within an area that is a predetermined distance or more away from an end of the binary image and within a predetermined width from the predetermined distance. Or the image processing method of 7. 9. The image processing method according to claim 6, further comprising a binarization processing step of converting a multi-value image into the binary image. The image processing method according to claim 9, further comprising a punch hole removing step of removing the punch hole by performing linear interpolation using pixels in a peripheral region of a circumscribed rectangle of the punch hole candidate.

Images