JP2009290438A - Image processing apparatus, image processing method, program, and recording medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processing apparatus which produces less incorrect separation between characters on a dot matrix and a dot matrix pattern by separating characters on the dot matrix in a legible manner and reproducing a dot matrix pattern without uncomfortable feeling during operation in economy mode. <P>SOLUTION: A means of detecting edge pixel area on dot matrix detects a target pixel area as a character edge pixel area on the dot matrix, in the case when the target pixel area is detected as a character edge pixel area by an edge pixel area detection means, and pixel areas that are away by a predetermined distance from the target pixel area and are opposite to each other with the target pixel area therebetween are detected as a dot matrix area of character background by the dot matrix pixel area detection means. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to an image processing apparatus that makes it easy to read characters on an image and can reduce consumption of toner, ink, and the like.

In the current copying machine, when a document area is mixed with a character area and a pattern (printing, photographic paper) area, the resolution is given priority to the character area and the gradation is given priority to the pattern area. In order to switch between them adaptively, an area identification unit, so-called image area separation technology is installed. In color copiers, in addition to the image area separation technique described above, processing for black character areas and processing for color character areas and design areas are performed using local color information to reproduce the character areas in black. And an image area separation technique (hereinafter, image area separation technique A) for adaptively switching between the two.
FIG. 13 is an overall flowchart of the image area separation technique A.
In FIG. 13, a scanner 101 acquires image data composed of RGB signals by reading a document. The scanner γ correction unit 102 corrects the γ characteristic of the scanner and converts the linear reflectance signal into a linear density signal. The scanner color correction unit 103 converts the RGB signal depending on the scanner characteristics into a device-independent R′G′B ′ signal that does not depend on the scanner.

The filter 104 performs a smoothing process that suppresses the undulation of the halftone dot portion and suppresses moire, and an edge enhancement process that increases the sharpness of the character portion. The edge enhancement processing is switched depending on the type of document based on the determination result (character or picture) from the image attribute determination A. For example, edge emphasis processing is applied with emphasis on character sharpness for character originals, and smoothing processing is applied with emphasis on moire suppression for pattern originals. The determination result of the image attribute determination A is applied to the filter 104 and the black processing unit 106.
The printer color correction unit 105 performs conversion from a device-independent R′G′B ′ signal to a C′M′Y ′ signal depending on printer characteristics. The black processing means 106 generates a K signal according to the C′M′Y ′ data (black generation), and subtracts the amount corresponding to K from the C′M′Y ′ (under color removal (UCR)). Do. CMYK corresponds to the color material color of the printer. As a method of black generation and UCR processing, there are a method based on an expression, a method using a LUT, and the like.
When the user designates the toner save mode, the toner save processing unit 107 responds to a signal indicating the type of document from the operation panel (not shown) and the edge detection result from the edge detection unit (not shown). Switch toner save processing. In the processing flow of FIG. 13, toner save is OFF.
The image after the toner saving process is processed by the printer γ correction unit 108 and the pseudo halftone processing unit 109, and the image is output on the recording medium by the printer 110. The printer γ correction unit 108 performs conversion processing using a density conversion table in accordance with the density characteristics of the printer. The pseudo halftone processing unit 109 performs pseudo halftone processing such as dithering and error diffusion.

The above description is an example in the case of improving the image quality on the premise of faithful reproduction of the original image. However, the use of the image area separation technique is not limited to the above-described use examples. For example, the image area separation technique is also used when it is desired to form an image so that characters can be read while saving toner and ink. In this case, it is required to reproduce the picture area thinly and reproduce the character portion itself in a state close to the original image while maintaining or enhancing the contrast with the background as well as the characters on the white background as well as the characters on the white background. .
When such adaptive image processing is performed (hereinafter referred to as eco mode), both the fine characters on the white background and the characters on the halftone dots are separated with high accuracy, and the pattern edges are erroneously separated. An image area separation technique (hereinafter referred to as image area separation technique B) is required.

FIG. 14 is an overall flowchart when the image area separation technique B is used. Each processing block is the same as that shown in FIG. Of course, in the toner save processing means 107, if the pixel of interest or the block of interest is a picture portion, processing for saving toner is performed, and if it is a character portion, no special processing is performed. Is done.
Various methods have been proposed as the image area separation technique B.
For example, in the invention described in Patent Document 1, a character is detected by paying attention to or using a white background with a wide character background. Further, in the inventions described in Patent Documents 2 to 4, a design is detected by paying attention to or using the presence of dots of halftone design. In the invention described in Patent Document 5, a character on a halftone dot is positively determined as a character.
Japanese Patent No. 3153221 Japanese Patent No. 2972171 Japanese Patent No. 3862374 JP 2002-262077 A Patent 2002-199210

However, the image area separation technique B that separates both the fine characters on the white ground and the characters on the halftone dots with high accuracy and does not erroneously separate the pattern edges is still inadequate in performance and has not yet been put into practical use. It is.
In the invention of Patent Document 1, white ground characters are separated, but there is a problem that characters on a halftone dot are determined as a picture. More specifically, the character area on the halftone dot is determined as a pattern except when the halftone dot background is very thin, that is, when it is determined as a white background. On the other hand, with respect to the halftone dot pattern, there are very few places that are erroneously detected as characters.
Further, in the inventions described in Patent Documents 2 to 4, there is a mechanism (such as a dot counting unit and an integration unit) that expands a halftone dot region from the vicinity of a dot. There is a problem that it is judged as a design.
Further, in the invention described in Patent Document 5, even if characters on a halftone dot can be separated, a region having a structure like a character in a halftone dot pattern, for example, a contour portion of an image (contour of a person or an object), etc. There is a problem such as misjudging.

As can be understood from the above, when the image area separation method described in Patent Documents 1 to 4 is selected in the eco mode, the character on the halftone dot is determined as a picture, and as a result, the character on the halftone dot and its character In the worst case, the characters cannot be read. In addition, when the image area separation method described in Patent Document 5 is selected, an area frequently determined as a character portion appears in the design. As a result, in the same region, not only the eco-mode effect of saving the coloring material is diminished, but also the density of the pattern fluctuates drastically, causing a sense of incongruity in the reproduced image. As a result, the mis-separated part is noticeable abnormally and often causes discomfort to the user. When eco-mode is offered as a sales point, a replay image with a halftone dot pattern may not be accepted (although a uniform thin pattern image is acceptable), it may be a negative point.
The present invention has been made in view of the above-described circumstances, and when using the eco mode, the characters on the white ground are separated from the pattern with high accuracy using a conventional technique such as Patent Document 1, for example, while using halftone dots. When reconstructing an image using the separation results for the upper character area, the image should be separated to a legible level. On the other hand, in the halftone dot pattern area, it is possible to reproduce the pattern with less discomfort for the user. An object is to provide a small number of image processing apparatuses.

In order to solve the above problem, the invention according to claim 1 is an image data acquisition unit that converts a document into a digital signal and acquires the image data, and an edge pixel that detects a character edge pixel region from the image data. Based on an area detection means, a halftone pixel area detection means for detecting a halftone pixel area of a character background from the image data, a detection result by the edge pixel area detection means, and a detection result by the halftone pixel area detection means Halftone dot edge pixel area detecting means for detecting a character edge pixel area on the halftone dot, and the halftone dot edge pixel area detecting means is configured such that the target pixel area is a character edge pixel by the edge pixel area detecting means. A pixel area that is detected as an area and has a predetermined distance from the target pixel area and is opposed to the target pixel area is detected as the halftone dot pixel area. When it is detected as a halftone pixel region of the character background by stage, and wherein the image processing device for detecting the target pixel area as a character edge pixel region on the dot.
According to a second aspect of the present invention, in the image processing apparatus according to the first aspect, the halftone pixel area detecting means includes a flat pixel area detecting means for detecting a flat halftone pixel area. To do.
According to a third aspect of the present invention, in the image processing device according to the second aspect, the flat pixel region detecting means has a predetermined distance from the target pixel region and faces the target pixel region across the target pixel region. It is characterized by detecting that the halftone dot density is below a certain level difference with respect to the background halftone pixel region.
According to a fourth aspect of the present invention, in the image processing apparatus according to the second aspect, the flat pixel region detecting means has a predetermined distance from the target pixel region and faces the target pixel region across the target pixel region. It is characterized in that it is detected that a color difference with respect to a background halftone pixel region is not more than a certain value.

According to a fifth aspect of the present invention, the image data acquisition means converts the document into a digital signal and acquires it as image data, and the edge pixel area detection means detects the character edge pixel from the image data. An edge pixel area detecting step for detecting an area; a halftone pixel area detecting unit for detecting a halftone pixel area of a character background from the image data; and a detection result in the edge pixel area detecting step. And a halftone dot edge pixel area detecting means for detecting a character edge pixel area on the halftone dot based on the detection result in the halftone dot pixel area detecting step. In the halftone dot edge pixel region detecting step, the pixel region of interest is a character edge pixel region in the edge pixel region detecting step. When a pixel area that is detected and has a predetermined distance from the target pixel area and that faces the target pixel area across the target pixel area is detected as a halftone pixel area of a character background by the halftone pixel area detection step, The image processing method is a step of detecting a target pixel area as a character edge pixel area on a halftone dot.
The invention according to claim 6 is characterized by a program for causing a computer to execute the image processing method according to claim 5.
The invention according to claim 7 is characterized by a computer-readable recording medium on which the program according to claim 6 is recorded.

  According to the present invention, after detecting a character edge pixel region and a character dot halftone pixel region in a document, a pixel region having a predetermined distance from the target pixel region and facing the target pixel region on the halftone dot is detected. Since the character area on the halftone dot is readable and separable even when using the eco mode, there is little misseparation that allows the user to reproduce the pattern with less discomfort for the user in the halftone dot picture area. An image processing apparatus can be provided.

Hereinafter, embodiments of the present invention will be described in detail.
FIG. 1 is a processing block diagram of the entire image area separation processing according to the present invention. Hereinafter, the configuration and operation of each block will be described.
[Scanner 10] (Image data acquisition means)
The present invention has been made on the assumption that it is used in a color copying machine or the like. Therefore, the scanner 10 is a 400 dpi RGB 8-bit output color scanner device. The output signal from the scanner 10 is a substantially linear signal, but after γ conversion is performed as necessary, the signal is detected by the white ground / character area detection processing means 20 or the halftone dot / character area detection processing means 30. Each is entered.
[White ground / character area detection processing means 20]
This block performs processing for interpreting white ground characters as characters. Since this processing block is not the essence of the present invention, the contents will be briefly described below. The required function is to detect a character area having a white background from two-dimensional image data and not to erroneously detect a pattern (printing paper, printing halftone) area as a character area. Specifically, the conventional processing described in Patent Document 1 or the like may be performed. A brief description is given below.

FIG. 2 is a diagram for explaining the white ground / character area detection processing means 20. In the figure, a log conversion circuit 21 converts a linear reflectance signal from the scanner 10 into a linear density signal, and C1 represents the amount of each color ink of C (cyan), M (magenta), and Y (yellow). M1 and Y1 signals are output. The masking circuit 22 converts the signal from the log conversion circuit 21 into a printer recording signal (C2, M2, Y2). The character detection circuit 23 detects a character area from the printer recording signal and outputs the detection result as a binary signal. For example, 1 is output if the target pixel is a character area, and 0 is output if it is a picture area. The white background detection circuit 24 detects an area having a white background from the printer recording signal and outputs the detection result as a binary signal. For example, 1 is output if the background of the pixel of interest is white, and 0 otherwise. The AND circuit 25 takes these ands as an output signal of the white ground / character area detection processing means 20.
In this processing block, any device that can detect characters on the white ground can be used as long as it has been proposed in the past. It is important and necessary to have no performance.
The output signal of this processing block is a binary signal of 1 [on] if the target pixel is a white ground character region, and 0 [off] otherwise.

[On-dot / character area detection processing means 30] (FIG. 1)
It is a processing block which is the characteristic structure of this invention. Details of the processing will be described in an embodiment described later. Here, the required functions and features will be described. The function of this processing block is to detect characters on halftone dots from two-dimensional image data, especially when the original is printed, while the edge of the picture, especially the edge of the halftone picture. It is important not to misdetect parts as much as possible. The present invention uses the following image features in which characters on halftone dots are present in an actual document, and in particular, devise to reduce misjudgments in halftone patterns. That is,
(1) In the printed matter, the characters on the halftone dot are generally sandwiched between the halftone dot regions.
(2) In addition to the above (1), the halftone dots of the character background are often flat (uniform) with little change in density and color.
In other words, the edge portion in the pattern that seems to be mistaken for the characters on the halftone dot and the background thereof often do not have the features (1) and (2). We propose a processing algorithm using the features. Although specific processing contents will be described in the embodiment, the principle of the processing algorithm will be briefly described below.

FIG. 3 is a diagram for explaining image characteristics of characters on halftone dots. For example, it can be seen that the horizontal bar of the letter “R” indicated by reference numeral 1 has halftone dot regions above and below the bar and satisfies the requirement (1).
On the other hand, “−” indicated by reference numeral 2 in FIG. 3 has a halftone dot region at the top, but does not have a halftone dot region at the bottom, so half can be said to be a halftone character, but half is white ground. It can be said that it is a character. In such a case, the process proposed this time is not determined as a character on a halftone dot and cannot be detected as a character. However, even if it is thinly reproduced, it is not a letter in the first place (it is a part of the periphery of the edge) and does not become a big problem.
The output signal of this processing block is a binary signal of 1 [on] if the pixel of interest is a character area on a halftone dot, and 0 [off] otherwise.

[OR processing means 40]
This processing block performs an OR operation on the output signal from the above-described white ground / character area detection processing means 20 and the output signal from the halftone dot / character area detection processing means 30 and outputs the result. The reason for applying this processing is that the same reproduction processing (for example, processing corresponding to color material saving is made through) is applied to pixels existing in the character area on the white background and pixels existing on the halftone dots. is there. However, if you want to perform different playback processing on the pixels existing in the character area on the white background and the pixels existing on the halftone dots, it is not necessary to perform such a logical OR. The output signal from the processing means 20 and the output signal from the halftone dot / character area detection processing means 30 may be used independently of each other.
An example of the merit of processing independently without performing logical sum processing is shown. For example, when the character itself has the same density, the white ground character has a feature that the contrast is larger by the amount of the lighter background than the halftone character. Therefore, when an image is reproduced in the eco mode, the former is less readable than the latter, and for example, even if γ is laid and reproduced, the legibility is guaranteed.

[First embodiment]
A first embodiment of the present invention will be described. FIG. 4 is a block diagram showing halftone dot / character area detection processing means 30a according to the first embodiment of the present invention. An original signal is read by the scanner 10, a color image signal composed of RGB is acquired, and after various image recognition processing described later, an attribute signal (binary signal) indicating whether the character area is on a halftone dot or not. ) Is output.
[Edge Pixel Area Detection Processing Unit 31]
The function of this processing block is to positively extract character edges from image data regardless of whether they are on the white ground or on halftone dots. Here, the method or apparatus described in Japanese Patent No. 3260815, which is an existing technology, is used. A brief description is given below.
Here, the 8-bit signal of the input signal is γ-converted to 6 bits (0 to 63: 0 on the white side of the document and 63 on the black side) and used for edge detection processing.

FIG. 5 is a diagram for explaining edge separation. Edge separation is performed as follows. If at least one of the following conditions is satisfied in the 3 × 3 mask, the center pixel E is determined as an edge pixel. That is, for the four masks shown in FIGS. 5A to 5D (A to I are pixel values),
E−B ≧ Cont (threshold) &
E ≧ Bk (threshold) &
G ≧ Bk &
H ≧ Bk &
I ≧ Bk
Or
E−I ≧ Cont &
E ≧ Bk &
B ≧ Bk &
D ≧ Bk &
If is established, the center pixel E is determined as an edge pixel.
[Halftone Pixel Area Detection Processing Unit 32]
The function of this processing block is to extract a halftone dot region from the image data. Here, the method or apparatus described in Patent Document 2 which is an existing technique is used.

FIG. 6 is a diagram showing a processing flow of the halftone pixel area detection processing means 32.
The input image signal unit 321 γ-converts an 8-bit signal, which is an input signal, to 6 bits (0 to 63: 0 on the white side of the document and 63 on the black side), and uses it for the pole detection process.
For example, in a 3 × 3 pixel mask (not shown), the pole detection unit 322 detects a pole (peak) when the center pixel value is the maximum or the minimum compared to the other eight pixel values around it. ) Pixel candidate. If the pixel value is maximum, it is a peak peak pixel candidate, and if the pixel value is minimum, it is a valley peak pixel candidate. Further, as shown in the following formula (1), the absolute value | Δmi | of the difference between the central pixel value and the surrounding eight pixel values is calculated, respectively, and when each absolute value is not less than a predetermined threshold value Δmth, Are determined to be extreme pixels. The output value is a 2-bit signal indicating the peak peak pixel [10], valley peak pixel [01], or other pixel [00]. Note that the mask is moved in units of pixels when performing the detection process.
| Δmi | ≧ Δmth (where i is 1 to 8)
(1)

The halftone dot area detection unit 323 will be described. FIG. 7 is a diagram illustrating a pixel block for explaining detection of a halftone dot region. For example, in the target block B0 (N = 9) composed of 9 × 9 pixels, the above-described peak peak pixel and valley peak pixel are counted, and the larger value is used as the peak pixel number of the target block B0. When the number is larger than a predetermined threshold, the block of interest B0 is detected as a halftone dot region. In addition, the movement of the block in the case of processing is performed in units of blocks (9 × 9 pixels). The output value is a binary signal [1] in the case of the halftone dot region, and [0] if it is not the halftone dot region.
The region determination output signal output unit 324 calculates the number of blocks determined as the halftone dot block by the halftone dot region detection unit 323 for the eight blocks (B1 to B8) surrounding the target block B0 shown in FIG. If the calculated value is larger than the predetermined threshold value, the target block B0 is determined as a halftone dot region, and a halftone dot determination signal is output. The movement of the block when processing is performed is performed in units of blocks (9 × 9 pixels). The output value is a binary signal [1] when the halftone dot is determined and [0] when the halftone dot is not determined.
In detecting the halftone area, the method or apparatus described in Patent Document 2 is used as an example of this time, but the halftone pixel area detection processing means 32 is of course limited to this method or apparatus. The present invention is not limited to this, and any existing halftone dot area detection method or apparatus disclosed in Patent Document 3 may be used.
[Correction processing means 33] (Detection of halftone dot edge pixel area)
The function of this processing block is to use the result of the edge pixel area detection processing means 31 and the result of the halftone pixel area detection processing means 32 to obtain an edge pixel area (a character edge on a halftone dot) between the halftone dot areas. Pixel area) is detected.

FIG. 8 is a diagram for explaining the processing contents of the correction processing means. Respective pixels existing at a predetermined distance away from the target pixel (P0) in a predetermined direction are defined as P1 to P8. Here, since the distance L has an influence on the line width of the character to be detected, it can be determined experimentally while looking at the final output image in the adaptive eco mode using this image area separation technique. That's fine.
At this time, P0 is determined to be an edge pixel by the edge pixel area detection processing unit 31, and the signal [1 indicating that the correction pixel is a halftone dot character edge from the correction processing unit 33 to the target pixel P0 under the following conditions. ] Is output. If the character edge is not a halftone dot, signal [0] is output.
P1 = halftone dot region & P5 = halftone dot region or P2 = halftone dot region & P6 = halftone dot region or P3 = halftone dot region & P7 = halftone dot region or P4 = halftone dot region & P8 = halftone dot region P1 and P5, P2 and P6, P3 and P7, and P4 and P8 are pairs of pixels that exist with the target pixel P0 interposed therebetween, as can be seen from FIG.

In other words, P0 is determined to be an edge pixel by the edge pixel region detection processing unit 31, and from two pixels that are separated from the target pixel P0 by a predetermined distance L and are opposed to each other with the target pixel P0 interposed therebetween. If any one of the pixel pairs (P1 and P5, P2 and P6, P3 and P7, or P4 and P8) is a halftone dot region, it is determined that the target pixel P0 is a halftone dot character edge.
As described above, according to the present invention, a character edge pixel (for example, FIG. 3) can be detected with high accuracy. On the other hand, the portion indicated by reference numeral 2 in FIG. 3 or a thick character on a halftone dot is not determined as a character edge depending on the distance parameter L described above.

[Second Embodiment]
A second embodiment of the present invention will be described. The same parts as those in the first embodiment are denoted by the same reference numerals, and the description thereof is omitted. FIG. 9 is a block diagram showing halftone dot / character area detection processing means 30b according to the second embodiment of the present invention. As in the case of the first embodiment, the document is read by the scanner 10, a color image signal composed of RGB is acquired, and after performing various image recognition processes described later, whether or not the character is on a halftone dot. An attribute signal (binary signal) is output.
[Pole pixel density detection processing means 34]
The function of this processing block is to calculate the dot density for each local region using image data. Here, a description will be mainly given of a different part of the processing content from the halftone pixel region detection processing means 32 described above. FIG. 10 is a diagram showing a processing flow of the extreme pixel density detection processing means 34.
The pole pixel density calculation unit 341 will be described. For example, in the target block B0 (N = 9) composed of 9 × 9 pixels in FIG. 7, the above-described peak peak pixel and valley peak pixel are counted, and the larger value is used as the peak pixel number of the target block B0. . If the number of peak pixels is equal to or less than a predetermined threshold, “0” is output as the output value. Otherwise, the number of peak pixels is output as the pixel density value of the central pixel of the block. The movement of the block when processing is performed is performed in units of pixels. That is, a density value is calculated for each pixel.
Note that the method of pole detection is not limited to the above-described method. For example, the isolated point determination disclosed in Patent Document 4 may be replaced with the pole detection unit 322 and used.

[Correction processing means 35]
This processing block uses the result of the edge pixel region detection processing unit 31 and the result of the extreme pixel density detection processing unit 34 to obtain an edge pixel region sandwiched between flat (uniform density) halftone dot regions. Detect (flat pixel area detection). Specific processing contents will be described with reference to FIG.
Respective pixels existing at a predetermined distance away from the target pixel (P0) in a predetermined direction are defined as P1 to P8. Here, since the distance L has an influence on the width of the character to be detected, it can be determined experimentally while looking at the final output image in the adaptive eco mode using this image area separation technique. Good.
At this time, P0 is determined to be an edge pixel by the edge pixel region detection processing unit 31, and the correction processing unit 35 determines that the pixel of interest P0 is an edge pixel region sandwiched between flat halftone dot regions under the following conditions. A signal [1] indicating the presence is output. If the edge pixel region is not sandwiched between halftone dot regions having a uniform density, the signal [0] is output.
(P1 ≧ threshold_dot & P5 ≧ threshold_dot & | P1-P5 | <threshold_flat)
Or (P2 ≧ threshold_dot & P6 ≧ threshold_dot & | P2-P6 | <threshold_flat)
Or (P3 ≧ threshold_dot & P7 ≧ threshold_dot & | P3-P7 | <threshold_flat)
Or (P4 ≧ threshold_halftone dot & P8 ≧ threshold_halftone dot & | P4-P8 | <threshold_flat)
As described above, the present invention uses the feature that a flat halftone pixel area exists in the background of the halftone character in the document. Therefore, according to the present invention, the character edge pixel area on the halftone dot is used. Corresponding pixels can be detected with high accuracy. In addition, since the present invention utilizes the feature that there is a halftone dot area with the same halftone dot ratio in both directions across the edge in the background of the halftone dot character in the document, It can be detected with high accuracy.

[Third embodiment]
A third embodiment of the present invention will be described. The same parts as those in the first and second embodiments are denoted by the same reference numerals, and the description thereof is omitted. FIG. 11 is a block diagram showing halftone dot / character area detection processing means 30c according to the third embodiment of the present invention. As in the case of the first embodiment, the document is read by the scanner 10, a color image signal composed of RGB is acquired, and after performing various image recognition processes described later, whether or not the character is on a halftone dot. An attribute signal (binary signal) is output.
[Masking processing means 36]
The function of this processing block is to convert the RGB signals of the scanner into CMY signals that are recording signals. As a result, the halftone dots of the original can be more accurately determined for each color plate. That is, a color original is made up of cyan (C), magenta (M), yellow (Y), and black (K) plates. The halftone dots of the magenta plate generate other signals in the subsequent pole detection by generating M signals. This makes it possible to extract halftone dots only for the magenta version without extracting halftone dots for the plate.
The masking process is performed by the following linear linear expression, and each coefficient parameter [aij] may be determined experimentally.
C = a11 * R + a12 * G + a13 * B + a14
M = a21 * R + a22 * G + a23 * B + a24
Y = a31 * R + a32 * G + a33 * B + a34

[Pole pixel density detection processing means 37]
The function of this processing block is to calculate a halftone dot density for each local region by using each CMY image data after the masking process. Basically, three processes in FIG. 10 are performed in parallel. Here, a description will be mainly given of a different part of the processing content from the halftone pixel region detection processing means 32 described above. FIG. 12 is a diagram illustrating a processing flow of the extreme pixel density detection processing 37.
For example, in the attention block B0 (N = 9) including 9 × 9 pixels in FIG. 7, the pole pixel density calculation unit 371 counts the above-described peak peak pixels, and calculates the value as the number of peak pixels of the attention block B0. And A method of detecting a valley peak as in the above-described extreme pixel density calculation unit 341 may be used, but a color halftone dot on which characters are present is generally a highlight, that is, a halftone dot on a mountain peak side with a low dot percentage. In most cases, the peak peak density is calculated in this embodiment. If the number of peak pixels is equal to or less than a predetermined threshold, “0” is output as the output value. Otherwise, the number of peak pixels is output as the pixel density value of the central pixel of the block. The movement of the block when processing is performed is performed in units of pixels. That is, a density value is calculated for each pixel.

[Correction processing means 38]
The function of this processing block is to use the result of the edge pixel area detection processing unit 31 and the result of the extreme pixel density detection processing unit 37 to obtain an edge pixel sandwiched between flat (uniform color) halftone dot areas. An area is detected (flat pixel area detection). Specific processing contents will be described with reference to FIG. Respective pixels existing at a predetermined distance away from the target pixel (P0_c, m, y) in a predetermined direction are defined as P1_c, m, y to P8_c, m, y. Here, since the distance L has an influence on the width of the character to be detected, it can be determined experimentally while looking at the final output image in the adaptive eco mode using this image area separation technique. Good.
At this time, P0 is determined to be an edge pixel by the edge pixel area detection processing unit 31, and an active signal is output from the correction processing unit 38 to the target pixel under the following conditions.

(((P1_c ≧ threshold_dot & P5_c ≧ threshold_dot) or (P1_m ≧ threshold_dot & P5_m ≧ threshold_dot) or (P1_y ≧ threshold_dot & P5_y ≧ threshold_dot)) And ((| P1_c-P5_c | + | P1_m-P5_m | + | P1_y-P5_y |) <threshold_flat)) ⇒A
Or (((P2_c ≧ threshold_dots & P6_c ≧ threshold_dots)) or (P2_m ≧ threshold_dots & P6_m ≧ threshold_dots) or (P2_y ≧ threshold_dots & P6_y ≧ threshold_dots) ) And ((| P2_c-P6_c | + | P2_m-P6_m | + | P2_y-P6_y |) <threshold_flat)) ⇒B
Or (((P3_c ≧ threshold_halftone dot & P7_c ≧ threshold_halftone dot) or (P3_m ≧ threshold value_halftone dot & P7_m ≧ threshold_halftone dot) or (P3_y ≧ threshold value_halftone dot & P7_y ≧ threshold value_dot) ) And ((| P3_c-P7_c | + | P3_m-P7_m | + | P3_y-P7_y |) <threshold_flat)) ⇒C
Or (((P4_c ≧ threshold_halftone dot & P8_c ≧ threshold_halftone dot) or (P4_m ≧ threshold value_halftone dot & P8_m ≧ threshold_halftone dot) or (P4_y ≧ threshold value_halftone dot & P8_y ≧ threshold value_dot) ) And ((| P4_c-P8_c | + | P4_m-P8_m | + | P4_y-P8_y |) <threshold_flat)) ⇒D
The output signal is input to the logical sum processing means 40, and the logical sum of [ABCD] is taken and output as a binary signal.

In this case, the threshold is determined by [ABCD] in each of the four directions across the target pixel, and the logical sum of [ABCD] is obtained by the logical sum processing means 40. However, the condition may be made stricter. it can. For example, if the logical product of [ABCD] is taken, the flatness (uniformity of color) will be judged, and it will be difficult to detect characters existing on flat halftone dots. It is possible to reduce the mis-separation in the middle, and it is possible to change the processing at the adjustment level.
Further, this embodiment has described an image area separation, particularly a halftone dot character detection device, when an eco mode is selected in a color copying machine. For example, when you want to apply different processing to the halftone dot character edge and the middle edge of the picture, such as strong edge enhancement (spatial filtering), weak edge enhancement on the edge of the picture, etc. It is also possible to employ the present invention as the recognition system processing.
As described above, the present invention uses the feature that a flat halftone area exists in the background of the character on the halftone dot in the document. Therefore, according to the present invention, the character edge pixel area on the halftone dot has a high accuracy. Can be detected. In addition, since the present invention uses the feature that there is a halftone dot area of approximately the same color in both directions across the edge in the background of the halftone character in the document, according to the present invention, the character edge on the halftone dot is used. The pixel area can be detected with high accuracy.

As described above, when the image processing apparatus described in the first to third embodiments is installed in an image forming apparatus such as a copying machine or a facsimile, a character area on a halftone dot can be read even when the eco mode is used. It is possible to provide a copying machine, a facsimile machine, or the like that can perform image processing with less misseparation and can reproduce a pattern with less discomfort for the user in the halftone dot pattern area while being separated.
In addition, the present invention is applied to various image processing apparatuses by including the program having the image processing blocks described in the first to third embodiments in a computer or by recording the program on a computer-readable recording medium. It becomes possible.

It is a processing block diagram of the entire image area separation processing according to the present invention. It is a figure explaining a white ground and character area detection processing means. It is a figure explaining the image feature of the character on a halftone dot. FIG. 3 is a block diagram showing halftone dot / character area detection processing means according to the first embodiment of the present invention. (A) thru | or (d) is a figure explaining edge isolation | separation. It is a figure which shows the processing flow of a halftone pixel area | region detection process means. It is a figure which shows the pixel block for demonstrating the detection of a halftone dot area | region. It is a figure explaining the processing content of a correction process means. It is a block diagram which shows the halftone dot and character area detection process means based on 2nd embodiment of this invention. It is a figure which shows the processing flow of a pole pixel density detection process means. It is a block diagram which shows the halftone dot and character area detection process means based on 3rd embodiment of this invention. FIG. 10 is a diagram showing a processing flow of pole pixel density detection processing 37. It is a whole flowchart of the image area separation technique A which is a prior art. It is a whole flowchart of the image area separation technique B which is a prior art.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 ... Scanner, 20 ... White ground and character area detection processing means, 21 ... Log conversion circuit, 22 ... Masking circuit, 23 ... Character detection circuit, 24 ... White background detection circuit, 25 ... AND circuit, 30, 30a, 30b, 30c ... halftone dot / character area detection processing means, 31 ... edge pixel area detection processing means, 32 ... halftone pixel area detection processing means, 321 ... input image signal section, 322 ... extreme point detection section, 323 ... halftone dot area detection 324 ... Area determination output signal output unit 33 ... Correction processing means 34 ... Pole pixel density detection processing means 341 ... Pole pixel density calculation part 35 ... Correction processing means 36 ... Mask processing means 37 ... Pole pixel Density detection processing, 371 ... pole pixel density calculation unit, 38 ... correction processing means, 40 ... logical sum processing means, 101 ... scanner, 102 ... correction means, 103 ... scanner Correcting means 104 ... filter, 105 ... printer color correction unit, 106 ... ink processing means, 107 ... toner save processing unit, 108 ... correction means, 109 ... halftoning unit, 110 ... printer

Claims (7)

Image data acquisition means for converting an original into a digital signal and acquiring it as image data; edge pixel area detection means for detecting a character edge pixel area from the image data; and detecting a dot pixel area of a character background from the image data Halftone dot pixel area detection means for detecting a character edge pixel area on a halftone dot based on a detection result by the edge pixel area detection means and a detection result by the halftone pixel area detection means Detecting means,
The halftone dot edge pixel area detecting means detects the target pixel area as a character edge pixel area by the edge pixel area detecting means, has a predetermined distance from the target pixel area, and opposes the target pixel area. An image processing apparatus, wherein when the pixel area is detected as a halftone pixel area of a character background by the halftone pixel area detection means, the target pixel area is detected as a character edge pixel area on a halftone dot.   2. The image processing apparatus according to claim 1, wherein the halftone pixel area detecting means includes a flat pixel area detecting means for detecting a flat halftone pixel area.   3. The image processing apparatus according to claim 2, wherein the flat pixel region detecting means has a halftone dot with respect to a halftone pixel region of a character background having a predetermined distance from the target pixel region and facing the target pixel region. An image processing apparatus for detecting that the density is equal to or less than a certain level difference.   3. The image processing apparatus according to claim 2, wherein the flat pixel region detecting means has a predetermined distance from the target pixel region and has a color with respect to a halftone pixel region of a character background opposed to the target pixel region. An image processing apparatus that detects that a difference is equal to or less than a certain value. An image data acquisition means for converting the document into a digital signal and acquiring it as image data;
An edge pixel area detecting means for detecting a character edge pixel area from the image data;
A halftone pixel region detecting means for detecting a halftone pixel region of a character background from the image data;
A halftone dot edge detection unit that detects a halftone dot edge pixel region based on a detection result in the edge pixel region detection step and a detection result in the halftone pixel region detection step. A pixel area detecting step,
In the halftone dot edge pixel region detecting step, the target pixel region is detected as a character edge pixel region in the edge pixel region detecting step, and has a predetermined distance from the target pixel region and is opposed to the target pixel region. When the pixel area to be detected is detected as the halftone dot pixel area of the character background by the halftone pixel area detecting step, the target pixel area is detected as a character edge pixel area on the halftone dot. Image processing method.   A program for causing a computer to execute the image processing method according to claim 5.   A computer-readable recording medium on which the program according to claim 6 is recorded.

Images