JP5975945B2 - Image processing device - Google Patents

Description

  The present invention relates to an image processing apparatus.

  In some image processing apparatuses, a good image is obtained by distinguishing an image area as a “halftone dot area”, “photo area”, or “character area” according to the characteristics of a color image, and applying different image processing to each. ing. For example, gradation is emphasized in halftone dots and photographic areas, and sharpness and discrimination are emphasized in character areas. Therefore, it is desirable to switch and output image processing according to the purpose.

  In the case of the “character area”, different processing is further performed for the black character area or the color character area. For example, in the image processing apparatus described in Japanese Patent Application Laid-Open No. H10-228867, when the color is an achromatic color and a character area, the output of color (CMY) is stopped and only black is output. Further, in the image processing apparatus described in Patent Document 2, different image processing is applied to black characters, color characters, continuous tone images, and halftone images.

  In order to perform such processing for each area, it is necessary to accurately recognize and separate each area, and various proposals have been made.

  For example, in the image processing apparatus described in Patent Document 3, the color discrimination unit determines whether the color is black or white, and in the case of a color character attribute, the brightness of the character color is determined, and a very dark color and a very bright color are determined. Only the pixels other than the color are output as color characters.

  However, in the above-described method, the boundary between the area determined to be a black character and the area determined as a color character may be emphasized unnecessarily, and different processing is performed for each part as a character. Visibility may be reduced.

  Further, in the image processing apparatus described in Patent Document 4, correction of the color character area is performed by connecting the cutout of the color character by the expansion process to the color pixel area and removing the color pixel of the black character part by the contraction process. Processing is performed to improve the determination accuracy.

JP-A 64-072662 Japanese Patent Laid-Open No. 04-160981 JP 2000-069309 A JP-A-11-004344

  However, in the above-described method, due to the characteristics of the expansion / reduction process, if the color pixel area has a backlash, the expansion / contraction process is executed with this backlash, and there is a possibility that sufficient correction is not performed. In addition, in the case of a character on a halftone dot region or a character on a non-white background region, there is a concern that the black character may disappear due to the color pixel expansion process. Furthermore, since two expansion processing circuits and one contraction processing circuit are required for the correction process of the color character area, the scale of the arithmetic circuit becomes enormous and the cost of the apparatus increases.

  The present invention has been made in view of the above-described problems, and an image processing apparatus having good determination accuracy for black characters, color characters, and image portions (gradation portions) without increasing the circuit scale of image processing. The purpose is to obtain.

The image processing apparatus according to the present invention specifies the area attribute of each pixel in the input color image data as one of a character area, a white background area, a non-white background area, and a halftone area, and the predetermined peripheral pixels of the pixel. Based on the region attribute, an image region determination unit that specifies a background attribute of the pixel as one of the white background region, the non-white background region, and the halftone dot region; and a target pixel whose region attribute is the character region On the other hand, a region for correcting the region attribute of the target pixel based on an attribute combination of the region attribute and the background attribute for a plurality of reference pixels existing in a reference window of a predetermined size including the target pixel. An attribute correction unit and correction data are provided. The region attribute correction unit counts, for each attribute combination, the reference pixels having the attribute combination for a plurality of reference pixels existing in the reference window, and sets the count result and the attribute combination. The region attribute of the pixel of interest is corrected in accordance with the attribute combination whose comparison result exceeds the threshold. The correction data indicates, for each attribute combination, the threshold value and the region attribute after correction when the count result exceeds the threshold value, and the region attribute correction unit includes the region attribute of the target pixel. Is corrected, the corrected region attribute is determined based on the correction data. The correction data has priority information in association with the corrected area attribute when the count result exceeds the threshold, and the area attribute correction unit has the count result exceeds the threshold When there are a plurality of attribute combinations, the corrected region attribute is determined based on the priority indicated by the priority information.

  According to the present invention, it is possible to obtain an image processing apparatus having good determination accuracy for black characters, color characters, and image portions (gradation portions) without increasing the circuit scale of image processing.

FIG. 1 is a block diagram showing a configuration of an image processing apparatus according to an embodiment of the present invention. FIG. 2 is a flowchart for explaining the operation of the image processing apparatus shown in FIG. FIG. 3 is a diagram showing an example of correction data used by the region attribute correction unit 12 in FIG. FIG. 4 is a diagram showing a specific example of region attribute correction by the image processing apparatus shown in FIG.

  Hereinafter, embodiments of the present invention will be described 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.

  The image processing apparatus shown in FIG. 1 includes an image input unit 1, an image processing unit 2, and an image output unit 3.

  The image input unit 1 includes, for example, a scanner, optically reads a document image, generates image data of the document image, and outputs the image data to the image processing unit 2.

  The image processing unit 2 performs various image processes on the image data input from the image input unit 1, and outputs the image data after the image processing to the image output unit 3.

  The image output unit 3 includes, for example, a print engine, and prints an image based on the image data from the image processing unit 2 on printing paper.

  The image processing unit 2 includes an image region determination unit 11, a region attribute correction unit 12, and an output image processing unit 13.

  The image area determination unit 11 performs image area determination for each pixel in the input image data. Specifically, the image area determination unit 11 specifies the area attribute of each pixel in the input color image data as one of a character area, a white background area, a non-white background area, and a halftone dot area, Based on the area attribute of the peripheral pixel, the background attribute of the pixel is specified as one of a white area, a non-white area, and a halftone area.

  In the image region determination unit 11, the character region determination unit 21 determines whether each pixel is a pixel in the character region and whether it is a character edge, and the halftone region determination unit 22 determines whether each pixel is a pixel edge. It is determined whether or not the pixel is in a halftone dot region, and the ground region determination unit 23 determines whether each pixel is a pixel in the ground region (that is, a plain region) and Specify the background color for the pixel. The background area determination unit 23 determines that the background color is white when the background color is white, and determines that the background color is a non-white background (hereinafter also referred to as a color background) when the background color is a color other than white.

  The character area determination unit 21 extracts a character structure by edge extraction, for example. Further, the halftone dot region determination unit 22 extracts a halftone dot structure by extracting density change patterns in vertical, horizontal, and diagonal directions, for example. Further, the background color area determination 23 determines whether the area is a white area or a non-white area based on, for example, luminance information or density information.

  In addition, the color determination unit 24 in the image region determination unit 11 determines whether the pixel determined as the character region by the character region determination unit 121 is a chromatic color or an achromatic color. Thereby, the character region is further divided into a black character region and a color character region.

  For example, the color determination unit 24 performs threshold processing on the U value and V value of the YUV data converted from the RGB data of the input image, and specifies whether the character is a black character or a color character.

  Note that the region extraction method and the chromatic pixel determination method described in this embodiment are examples, and other methods may be used.

  The region attribute correction unit 12 corrects the region attribute of the target pixel obtained by the image region determination unit 11. Specifically, the region attribute correction unit 12 performs the region attribute and the background attribute for a plurality of reference pixels existing in a reference window of a predetermined size including the target pixel with respect to the target pixel whose region attribute is a character region. The region attribute of the target pixel is corrected based on the attribute combination.

  The reference window is an m × n pixel range around the target pixel.

  Further, in this embodiment, the region attribute correction unit 12 counts the reference pixels having the attribute combination for each attribute combination for a plurality of reference pixels existing in the reference window, and the result of the count (count value) Ni) is compared with the threshold value THi set for each attribute combination, and the region attribute of the target pixel is corrected according to the attribute combination whose count value Ni exceeds the threshold value THi.

 Furthermore, in this embodiment, the region attribute correction unit 12 has correction data indicating the region attribute after correction when the threshold value THi and the count value Ni exceed the threshold value THi for each attribute combination, When correcting the region attribute of the target pixel, the corrected region attribute is determined based on the correction data.

  In addition, the correction data has priority information in association with the corrected region attribute when the count value exceeds the threshold value. Then, when there are a plurality of attribute combinations in which the count value Ni exceeds the threshold value THi, the region attribute correction unit 12 determines the corrected region attribute based on the priority indicated by the priority information.

  Note that the region attribute correction unit 12 may perform the above-described region attribute correction only on the pixel of interest on the character edge where the region attribute is a character region.

  The output image processing unit 13 performs image processing necessary for printing based on the image data. For example, the output image processing unit 13 performs edge enhancement processing using a differential filter or the like for a character region, color correction processing, black generation / UCR (Under Color Removal) processing, halftone processing, and the like adapted to the character region. Then, smoothing processing using an integration filter or the like, color correction processing, black generation / UCR processing, halftone processing, etc. adapted to the halftone region or background region are performed on the halftone region or background region. As a result, the character portion in the printed image is sharpened, and moire and noise in the halftone dot portion are suppressed.

  Next, the operation of the image processing apparatus will be described. FIG. 2 is a flowchart for explaining the operation of the image processing apparatus shown in FIG.

  First, the image input unit 1 generates bitmap image data of a document image read by an image scanner and stores it in a memory (not shown) or the like (step S1).

  Then, the image processing unit 2 selects a pixel of interest in a predetermined scanning order for each page of the image data in each page (step S2).

  The image area determination unit 11 determines the image area for the target pixel as described above (step S3). Thereby, the region attribute and the background attribute for the target pixel are specified.

  Then, the image area determination unit 11 determines whether or not the area attribute of the target pixel is a character area (step S4). If the region attribute of the target pixel is not a character region, the region attribute of the target pixel is not corrected.

  On the other hand, when the region attribute of the target pixel is a character region, the image region determination unit 11 determines whether the target pixel is achromatic, and based on the determination result, the character region is a black character region or a color character region. To identify.

  Then, the region attribute correction unit 12 specifies the region attribute and background attribute of the target pixel in the image region determination unit 11, and further specifies the region attribute and background attribute of each reference pixel in the reference window of the target pixel. For each attribute combination, reference pixels are counted (step S5).

  FIG. 3 is a diagram showing an example of correction data used by the region attribute correction unit 12 in FIG.

  As shown in FIG. 3, there are a plurality of attribute combinations for the region attribute and the background attribute, and the region attribute correction unit 12 has a plurality of counters for each of the plurality of attribute combinations. The attribute combination of the reference pixel is specified, and the count value i of the counter for the specified attribute combination is increased by 1.

  Then, after the count of the attribute combinations for all the reference pixels in the reference window is completed, the region attribute correction unit 12 specifies the threshold value THi for each attribute combination from the correction data, and the count value Ni for each attribute combination. It is determined whether or not exceeds a threshold value THi (step S6).

  When the count value Ni does not exceed the threshold value THi for all the attribute combinations, the region attribute of the target pixel is not corrected.

  On the other hand, when the count value Ni exceeds the threshold value THi for one attribute combination, the region attribute correction unit 12 specifies the corrected region attribute and background attribute for the attribute combination with the correction data (step S7). ).

  When the correction data shown in FIG. 3 is used, for example, when the region attribute of the target pixel is determined to be a black character region (black character region), the region attribute is a color character region (chromatic character region) for the reference pixel. When the number N5 of reference pixels whose background attributes are color character areas exceeds the threshold TH5 of the attribute combination, the area attribute of the target pixel is changed from the black character area to the color character area, and the background attribute of the target pixel Is a color background area. As shown in FIG. 3, in some cases, even when the threshold value THi is exceeded, correction may not be performed. Further, a different threshold THi can be used for each attribute combination.

  On the other hand, when the count value Ni exceeds the threshold value THi for a plurality of attribute combinations, the region attribute correction unit 12 specifies the priority information for the plurality of attribute combinations with the correction data. In FIG. 3, a numerical value with parentheses described after the corrected area attribute indicates the priority information. Here, the priority is the highest when the priority indicated by the priority information is 1, and the lower the priority value, the lower the priority. Then, the region attribute correction unit 12 identifies the attribute combination having the highest priority among the plurality of attribute combinations, and sets the corrected region attribute for the attribute combination as the region attribute of the target pixel.

  Thereafter, the output image processing unit 13 performs the above-described image adaptation processing (spatial filter processing, color correction processing, black generation / UCR processing, halftone processing, etc.) for the target pixel (step S8).

  When the above-described processing (steps S2 to S8) for the target pixel is completed, the image processing unit 2 determines whether the target pixel has reached the end of the page (step S9), and the target pixel is the end of the page. If not, the process returns to step S2, selects the next target pixel, and executes the above-described processing (steps S2 to S8) for the target pixel.

  On the other hand, if the target pixel has reached the end of the page, the image processing unit 2 ends the image processing of this page, and causes the image output unit 3 to print this page (step S10).

  FIG. 4 is a diagram showing a specific example of region attribute correction by the image processing apparatus shown in FIG.

  FIG. 4A is a diagram illustrating an example of a document. The document shown in FIG. 4A has a black character “1”.

  FIG. 4B shows an example of region attributes obtained by region attribute determination processing in the image region determination unit 11. In the region attribute shown in FIG. 4B, a black character pixel is erroneously determined to be a color character region in the character edge portion. In FIG. 4, the hatched pixels are pixels that are determined to be color character areas.

  FIG. 4C shows an example of a reference window used by the region attribute correction unit 12. As shown in FIG. 4C, a pixel erroneously determined as a color character region by correcting the region attribute of the target pixel based on the attribute combination of the reference pixel in the reference window centered on the target pixel. The area attribute is corrected to the correct area attribute. FIG. 4D is a diagram illustrating the region attributes after correction.

  In the reference window of FIG. 4C, there are a relatively large number of pixels (N10 = 65) determined as a black character region and a relatively small number of pixels (N11 = 2) erroneously determined as a color character region. include. Therefore, in FIG. 3, since the condition “N11> TH11” is not satisfied, “Uncorrected (1)” with a priority of 1 is not selected, and the condition “N10> TH10” is satisfied, so the priority is 2. “Black character on white (2)” is selected. Thereby, a relatively small number of pixels erroneously determined to be color character areas are corrected to black.

  As described above, according to the above embodiment, the image region determination unit 11 specifies the region attribute of each pixel in the input color image data as one of a character region, a white background region, a non-white background region, and a halftone dot region. In addition, based on the region attribute of a predetermined peripheral pixel of the pixel, the background attribute of the pixel is specified as one of a white region, a non-white region, and a halftone region, and the region attribute correction unit 12 For a pixel of interest that is a character region, correct the region attribute of the pixel of interest based on the attribute combination of the region attribute and the background attribute for multiple reference pixels that exist within a reference window of a predetermined size that includes the pixel of interest. To do.

  As a result, correction is made even if an erroneous determination occurs when determining the area of the black character, the color character, and the image portion. Therefore, the determination accuracy of black characters, color characters, and image portions (gradation portions) is improved without increasing the circuit scale of image processing.

  The above-described embodiments are preferred examples of the present invention, but the present invention is not limited to these, and various modifications and changes can be made without departing from the scope of the present invention. is there.

  The present invention is applicable to an image processing apparatus such as a digital multifunction peripheral or an image scanner.

11 Image region discriminating unit 12 Region attribute correcting unit

Claims (2)

The area attribute of each pixel in the input color image data is specified as one of a character area, a white background area, a non-white background area, and a halftone dot area, and based on the area attribute of a predetermined peripheral pixel of the pixel, An image area determination unit that specifies a background attribute as one of the white area, the non-white area, and the halftone area;
Based on an attribute combination of the region attribute and the background attribute for a plurality of reference pixels existing in a reference window of a predetermined size including the target pixel for the target pixel whose region attribute is the character region A region attribute correction unit that corrects the region attribute of the pixel of interest;
Correction data ,
The region attribute correction unit counts, for each attribute combination, the reference pixels having the attribute combination for a plurality of reference pixels existing in the reference window, and sets the count result and the attribute combination. Is compared with a threshold value, and the region attribute of the target pixel is corrected according to the attribute combination in which the count result exceeds the threshold value,
The correction data indicates, for each attribute combination, the threshold value, and the region attribute after correction when the count result exceeds the threshold value,
When correcting the region attribute of the target pixel, the region attribute correction unit determines the corrected region attribute based on the correction data,
The correction data has priority information associated with the region attribute after correction when the count result exceeds the threshold,
The region attribute correction unit, when there are a plurality of the attribute combinations whose count result exceeds the threshold, determining the corrected region attribute based on the priority indicated by the priority information;
An image processing apparatus. The region attribute correction unit includes the region attribute and the background attribute for a plurality of reference pixels existing in the reference window for only the pixel of interest on the character edge where the region attribute is the character region. attribute on the basis of a combination, the target pixel of the area attribute image processing apparatus according to claim 1 Symbol placement and corrects the.

Images