JP4039911B2 - Image processing apparatus and image processing method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention performs processing for determining image areas such as characters, halftone dots, photographs (printing paper), solids, etc., and document types for input image data obtained by scanning a document, and processing according to the determination result. The present invention relates to an image processing apparatus that performs the above, an image forming apparatus including the same, an image processing method, a program that can execute the method, and a recording medium that records the program.
[0002]
[Prior art]
In addition to the conventional analog type, digital type is widely used as an image forming apparatus such as an electronic copying machine, and with the advancement of digital image processing technology, a full-color digital copying machine that reproduces color images with high image quality is available. It has been commercialized.
Document images copied by the above-described image forming apparatus include characters, line drawings, printed photographs (halftone dots), and photographic paper photographs in combination of these. It is necessary to perform image processing suitable for the original. In order to efficiently perform the above image processing, a character mode, a character / photo (printed photo) mode, a photo (printed photo) mode, etc. are selected as the operation mode of the image forming apparatus. ing.
[0003]
However, it is very troublesome for the user to switch the mode for each document. In addition, when an inappropriate mode is selected, the image is often considerably deteriorated, and unnecessary copying is performed. Therefore, in order to solve such a problem, it has been proposed to automatically determine the type of document.
[0004]
For example, Japanese Patent Laid-Open No. 9-261460 discloses an image processing method for automatically identifying an image area of a document image in which a plurality of images are mixed.
In the image processing method disclosed in the above publication, parameters for performing image region separation at least twice and determining a document based on the first image region separation determination, and parameters for determining a region locally during the main scan, change. When determining a document, only highly accurate information is used. For example, when a parameter that makes it difficult to detect an edge is set during pre-scanning and the image region separation criterion (threshold value) is increased, the inside of a character or the like is not detected as an edge, and the halftone dot The area where an image is erroneously detected as a character is reduced.
[0005]
Japanese Patent Application Laid-Open No. 8-251402 discloses a technique for automatically determining the type of document and automatically switching the operation mode. According to the technique described in the above publication, a brightness histogram is created from data read in the preliminary scan. The document type is automatically determined from this histogram. Then, an operation mode suitable for each document type is selected, and image processing (background adjustment, presence / absence of black character discrimination, gradation reproduction switching, etc.) according to the document type is executed.
That is, in this technique, R, G, B signals of three primary colors obtained by reading a document (including a color document and a monochrome document) by a photoelectric conversion element are converted into a brightness signal, a histogram is created from the brightness signal, The document type is determined based on the shape of the created histogram, and the subsequent image processing contents are switched based on the determination result.
[0006]
In Japanese Patent Laid-Open No. 8-251406, chromatic / achromatic determination of a document is performed, and an image separation function, that is, edge separation / halftone dot determination, is performed, whereby a document is converted into (1) line drawing and (2). ▼ Halftone, ▲ 3 ▼ Continuous, ④ Line drawing and halftone, ▲ 5 ▼ Halftone and continuous, ６ Six continuation and line, ▲ Seven ”Line drawing, halftone, and continuous It is shown that one of seven types of images is determined and a processing mode corresponding to this is automatically given to the filter, color correction, selector, and gradation processing.
[0007]
[Problems to be solved by the invention]
However, the image processing method described in the above publication has the following problems.
[0008]
That is, in Japanese Patent Application Laid-Open No. 9-261460, only the threshold value is adjusted, and the processing contents are the same. Therefore, the discrimination accuracy is not sufficiently high, and the fixed threshold value is used regardless of the document type. Since the processing is performed using, the discrimination accuracy is not sufficiently high.
[0009]
In Japanese Patent Application Laid-Open No. 8-251402, since determination is performed based on global information (histogram), it is impossible to distinguish between a photographic paper photograph and a printed photograph even in a photograph, and the image area in the document is small. It is difficult to make a determination at the time or when a plurality of images are mixed. Accordingly, erroneous determination may occur, but in this case, image processing that does not conform to the original document type is performed, which causes a problem of adversely affecting the image quality of the output image. This increases misprinting and wastes paper, which is not preferable from the viewpoint of resource saving.
[0010]
Also, the procedure for manually inputting the processing mode is not preferable because the output image is not preferable. When the image is output again, there is a risk that it will not be understood how to switch the operation mode if there is no previous image information. There is. That is, it is insufficient as a user interface.
[0011]
In Japanese Patent Application Laid-Open No. 8-251406, it is very complicated to perform each region separation (image separation) for each element for each pixel to determine the type of document. There is a problem that a circuit for processing becomes very large.
[0012]
In other words, in order to determine the type of document, in order to obtain some determination result for each pixel, an attempt is made to determine the pixel that can be determined either way. There is a problem that a circuit for this is required, each of which requires very complicated processing contents, and a circuit for processing becomes very large. In addition, since a single color signal is used, even if the same pixel as in a color original, the density varies greatly depending on the signal, or by performing ternarization for determining characters, high density and low density Although character determination is performed for each pixel, intermediate density character determination such as color characters is out of scope, so erroneous determination for each pixel will result in erroneous document determination results. There is a problem that may occur. In addition, the paper is not halftone dots / characters, so it lacks accuracy. There is also a problem that it is not possible to cope with pre-scan data whose resolution in the sub-scanning direction is reduced.
[0013]
The present invention has been made in view of the above-described conventional problems, and an object of the present invention is to accurately determine an image area and a document type by a simple method when determining an image area and an image type of an input image. An image processing apparatus, an image forming apparatus including the image processing apparatus, an image processing method, a program capable of executing the method, and a recording medium on which the program is recorded.
[0014]
[Means for Solving the Problems]
  An image processing apparatus according to the present invention includes document type determination means for determining a document type from input image data read by scanning a document with an image input means, and performing image processing based on the determination result In the deviceSaidThe document type determination means includes first density determination means for determining the presence or absence of the background of the document based on the density histogram data of the input image data and threshold data for determining the background of the document input as image data from the density histogram data. ,SaidSecond determination for determining the type of document input as image data from the determination result of the background of the document determined by the first determination means and the type of input image obtained by determination in units of pixels of the input image data Means, andThe first determination means extracts a histogram of density from the input image data to create a density histogram indicating the relationship between the density of each pixel and its frequency, and extracts a maximum frequency value in the density histogram as a first maximum frequency value. In addition, the concentration category to which the first maximum power value belongs is extracted as the first maximum power concentration region, and the second maximum power value is extracted except for the concentration region adjacent to the first maximum power concentration region. 2) The maximum power density class extracting means for extracting the density class to which the maximum power value belongs as the second maximum power density class, the first threshold value for determining the solid background, and the background consisting of solid and halftone dots. A threshold value setting means for setting a second threshold value, and the first maximum power value / second maximum power value and the first / second threshold value are compared, and a first maximum power value < If 1 threshold value or 1st maximum power value + second maximum power value <second threshold value, it is determined that there is no background, and if 1st maximum power value ≧ first threshold value or 1st maximum power value + second maximum value In the case of the frequency value ≧ the second threshold value, it is determined that the background is present, and further, the background determination means for calculating the background density is provided.It is characterized by that.
[0015]
  An image processing apparatus according to the present invention includes:In an image processing apparatus that includes a document type determination unit that determines a document type from input image data read by scanning a document with an image input unit, and that performs image processing based on the determination result, the document type determination unit includes: First determination means for determining presence / absence of a document background based on density histogram data of input image data and threshold data for determining a document background input as image data from the density histogram data; A second determination unit that determines the type of the document input as image data from the determination result of the background of the document determined by the unit and the type of input image obtained by determining in units of pixels of the input image data; And the second determination means applies to a target pixel having a block made up of a plurality of pixels extracted from the input image data. The feature amount is extracted from the image data in the main scanning direction and the background determination result of the document determined by the first determination unit, and whether the pixel of interest belongs to the solid, photograph, character, or halftone dot type The target amount is determined from the target pixel determination unit that performs the determination, and the image data in the main scanning direction and the background determination result of the document determined by the first determination unit for the peripheral pixels that exist around the target pixel. Peripheral pixel determining means for extracting and determining whether the peripheral pixel belongs to a solid, photo, character, or halftone area; and peripheral pixel determining means for the same line existing in the preceding stage in the sub-scanning direction of the target pixel An upper line determining means for determining whether a line existing in the preceding stage in the sub-scanning direction belongs to a solid, a photo, a character, or a halftone dot, and a subsequent stage in the sub-scanning direction of the target pixel. Exist Lower line determination means for determining whether a line existing in the subsequent stage in the sub-scanning direction belongs to a solid, photo, character, or halftone dot based on the determination result of the peripheral pixel determination means for the same line, And determining whether the target pixel belongs to a solid, photo, character, or halftone dot based on the determination results of the target pixel determination unit, the upper line determination unit, and the lower line determination unit It is characterized by that.
[0016]
  An image processing apparatus according to the present invention includes:In an image processing apparatus that includes a document type determination unit that determines a document type from input image data read by scanning a document with an image input unit, and that performs image processing based on the determination result, the document type determination unit includes: First determination means for determining presence / absence of a document background based on density histogram data of input image data and threshold data for determining a document background input as image data from the density histogram data; A second determination unit that determines the type of the document input as image data from the determination result of the background of the document determined by the unit and the type of input image obtained by determining in units of pixels of the input image data; The first determination means includes a density histogram indicating a relationship between the density of each pixel and the frequency from the input image data. A histogram generating means configured to extract the maximum frequency value in the density histogram as a first maximum frequency value, and extract a density category to which the first maximum frequency value belongs as a first maximum frequency density category; A maximum frequency concentration category extracting means for extracting the second maximum frequency value other than the density category adjacent to the maximum frequency concentration category and extracting the concentration category to which the second maximum frequency value belongs as the second maximum frequency concentration category; Threshold setting means for setting a first threshold value for discriminating the background consisting of the above and a second threshold value for discriminating the background consisting of solid and halftone dots, and the first maximum power value and the second maximum power value described above Are compared with the first and second threshold values, and it is determined that there is no background if the first maximum degree value <the first threshold value or the first maximum degree value + the second maximum degree value <the second threshold value. , 1st maximum In the case of numerical value ≧ first threshold value or in the case of the first maximum degree value + second maximum degree value ≧ second threshold value, it is determined that there is a background, and further, background determination means for calculating the background density, The second determination means determines the image data in the main scanning direction and the background of the original document determined by the first determination means for a target pixel having a block composed of a plurality of pixels extracted from the input image data. A feature amount is extracted from the result, and the target pixel determination means for determining whether the target pixel belongs to a solid, a photograph, a character, or a halftone dot, and a peripheral pixel existing around the target pixel, A feature amount is extracted from the image data in the main scanning direction and the background determination result of the document determined by the first determination unit, and whether the surrounding pixel belongs to a solid, photograph, character, or halftone dot region Surrounding pixel determination means for performing determination; Whether the line existing in the preceding stage in the sub-scanning direction belongs to the solid, photograph, character, or halftone dot type based on the determination result of the surrounding pixel determining means for the same line existing in the preceding stage in the sub-scanning direction Based on the determination results of the upper line determination means for determining the above and the surrounding pixel determination means for the same line existing in the subsequent stage of the target pixel, the photograph, character Or a lower line determination unit that determines which type of halftone dot belongs, and based on the determination results of the target pixel determination unit, the upper line determination unit, and the lower line determination unit, It is characterized in that it is determined whether the image belongs to solid, photo, character or halftone dot.
[0043]
In the present invention, the document type determination means for automatically determining the document type from the input image data can use the result based on the presence or absence of the background to determine the document type, so that the accuracy of determining the document type is improved.
In addition, instead of fixing the threshold value, it is possible to extract only specific image data having a predetermined frequency or more when determining the document type by setting the threshold value according to the type of each document. The discrimination accuracy can be improved.
In addition, since only the pixels that are clear to which area belong are extracted except for pixels having uncertain elements, and the document type is determined, the accuracy of determining the document type can be improved.
Further, when the document type discrimination means cannot discriminate the original, instead of performing processing in the default image mode, the fact that the document type is uncertain is displayed on the display means, and the user can be prompted to enter the processing mode. . Thus, even when the document type determination unit cannot determine the document, the user selects the processing mode via the input unit, so that image processing according to the original document type can be executed.
[0044]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of the present invention will be described below with reference to the drawings.
[0045]
As shown in FIG. 1, an image forming apparatus 10 according to the present embodiment includes a color image input device (image input means) 2, an image processing device 1, a color image output device 3, and an operation panel 4. .
For example, the color image input device 2 includes a scanner unit, and reads a reflected light image from a document as an RGB (R: red, G: green, B: blue) analog signal by a CCD (Charge Coupled Device). Is.
The color image output device 3 is a device that performs predetermined image processing in the image processing device 1 and outputs the result.
[0046]
The image processing apparatus 1 includes an A / D (analog / digital) conversion unit 11, a shading correction unit 12, a document type automatic discrimination unit 13, an input tone correction unit 14, a color correction unit 15, a black generation and under color removal unit 16, It comprises a spatial filter processing unit 17, an output tone correction unit 18, a tone reproduction processing unit 19, and a region separation processing unit 20.
[0047]
The A / D converter 11 converts an analog signal read by the image input device 2 into a digital signal.
The shading correction unit 12 performs shading correction for removing various distortions generated in the illumination system, the imaging system, and the imaging system of the image input apparatus 2.
[0048]
The automatic document type discrimination unit 13 is an image processing system employed in a color image processing apparatus such as a density signal for an RGB signal (RGB reflectance signal) from which various distortions have been removed by the shading correction unit 12. Converts the signal into an easy-to-use signal and determines the document type such as whether the input document image is a text document, a printed photo document, a photographic paper photo, or a combined character / printed photo document It is. Details will be described later.
[0049]
The input tone correction unit 14 adjusts the color balance and simultaneously performs image quality adjustment processing such as removal of background area density and contrast based on the determination result of the document type automatic determination unit 13.
[0050]
The color correction unit 15 performs color correction processing to remove color turbidity based on spectral characteristics of CMY (C: cyan, M: magenta, Y: yellow) color materials including unnecessary absorption components in order to realize color reproduction faithfully. Is to do. The color-corrected image signal is separated into a character, halftone dot, or photograph (other) region for each pixel by the region separation processing unit 20 based on the determination result of the document type automatic determination unit 13. . This region separation processing unit 20 sends a region identification signal indicating which region a pixel belongs to to the black generation and under color removal unit 16, the spatial filter processing unit 17, and the gradation reproduction processing unit 19 based on the separation result. Is output.
[0051]
The black generation and under color removal unit 16 performs black generation processing for generating a black (K) signal from the CMY three-color signals after color correction, while subtracting the K signal obtained by black generation from the original CMY signal to newly The undercolor removal process for generating a simple CMY signal is performed. As a result of these processes (black generation process / under color removal process), the CMY three-color signal is converted into a CMYK four-color signal.
[0052]
The spatial filter processing unit 17 performs spatial filter processing using a digital filter and corrects spatial frequency characteristics, thereby preventing blurring and graininess deterioration of the output image.
The output tone correction unit 18 performs output tone correction processing for converting a signal such as a density signal into a halftone dot area ratio that is a characteristic value of the image output device 3.
The gradation reproduction processing unit 19 performs gradation reproduction processing (halftone generation processing) in which an image is finally divided into pixels and processed so that each gradation can be reproduced.
[0053]
It should be noted that the image region extracted as a black character or possibly a color character by the region separation processing unit 20 is subjected to a high frequency in the sharpness enhancement processing in the spatial filter processing unit 17 in order to improve the reproducibility of the black character or the color character. The amount of frequency enhancement is increased. At the same time, in the halftone generation process, a binarization process or a multi-value process on a high-resolution screen suitable for high-frequency reproduction is selected.
[0054]
On the other hand, with respect to the region determined as a halftone dot by the region separation processing unit 20, the spatial filter processing unit 17 performs a low pass filter process for removing the input halftone dot component. At the same time, in the halftone generation process, a binarization process or a multi-value process is performed on the screen with an emphasis on gradation reproducibility.
[0055]
The operation panel 4 includes a liquid crystal display as a display unit, and setting buttons as input units for setting a numeric keypad, a copy button, a processing mode, and the like. As will be described later, in the present embodiment, the determination result in the automatic document type determination unit 13 is displayed to the user via the liquid crystal display. The user can select and input a processing mode via a setting button or the like based on the determination result.
[0056]
As described above, the image data subjected to the above-described processes is temporarily stored in a storage unit (not shown), read at a predetermined timing, and input to the color image output device 3. The above processing is performed by a CPU (Central Processing Unit).
[0057]
The image output device 3 outputs image data onto a recording medium (for example, paper). Examples of the image output device 3 include a color image output device using an electrophotographic method or an ink jet method. It is not limited.
[0058]
The present invention is characterized in the image processing apparatus 1 of the image forming apparatus 10, and particularly in the document type automatic discrimination unit 13. Therefore, an embodiment of the automatic document type determination unit (document type determination unit) 13 will be described in detail below.
[0059]
<First Embodiment>
FIG. 2 is a block diagram showing the document type automatic discrimination unit 13-1 in the first embodiment.
First, as shown in FIG. 2, the document type automatic determination unit 13-1 of the first embodiment includes a signal conversion unit 21, a first determination unit (first determination unit) 31, and a second determination unit (second Determination means) 41.
[0060]
The signal converter 21 converts the RGB reflectance signal into a density signal and converts the RGB density signal into a complementary CMY signal.
The first determination unit 31 includes a histogram generation unit (histogram generation unit) 32, a maximum frequency density category extraction unit (maximum frequency density category extraction unit) 33, a threshold setting unit (threshold setting unit) 34, and a background determination unit (background determination). Means) 35. Then, the first determination unit 31 determines whether or not the document has a background, and sets the density category determined to have the background as the background density category.
[0061]
The histogram creation unit 32 creates a density histogram indicating the relationship between the density of each pixel of the document image and its frequency. Specifically, as shown in FIG. 4, for example, image data in the main scanning direction (see FIG. 4B) read during the forward path (feed) of the scanner (image input apparatus 1), a plurality of pixels are recorded. An average value (for example, 7 × 7 pixels: see FIG. 4B) is calculated, and a density histogram (see FIG. 4A) is created. For example, as shown in FIG. 3B, when the background is formed of halftone dots (background (halftone dots)), a density histogram is created for each pixel as shown in FIG. A characteristic having two maximum power values is shown, and the maximum power value may become small and may not be detected as a background. On the other hand, when an average value of a plurality of pixels is obtained as shown in FIG. 4B and a density histogram is created, as shown in FIG. It is possible to detect as a background (in a solid background, the average value of a plurality of pixels and the density value of the pixels in the background area are almost the same, and the density category having the maximum power value does not change).
[0062]
As shown in FIGS. 5 to 7, the maximum power density category extraction unit 33 extracts the maximum power value in the density histogram created by the histogram creation unit 32 as a first maximum power value (MAX1) and MAX1. Is extracted as the first maximum power density category.
Further, the maximum power density category extraction unit 33 extracts the second maximum power value (MAX2) other than the density zone adjacent to the first maximum power density zone, and sets the density zone to which MAX2 belongs to the second maximum power frequency. Extract as concentration category. That is, the second largest power value in the density section that is not adjacent to at least the density section having MAX1 is extracted as MAX2.
[0063]
The threshold value setting unit 34 sets a first threshold value for discriminating from a solid background and a second threshold value for discriminating from a solid and halftone background.
The background determination unit 35 compares MAX1 and MAX2 with the first and second threshold values and performs the following determination.
[0064]
When MAX1 ≧ first threshold value, it is determined that there is a background, and the density category to which MAX1 belongs is set as the background density category. For example, as shown in FIG. 5, this determination uses the fact that the first maximum power density classification (so-called white background) in a text document composed of a solid background is higher than other density values.
[0065]
When MAX1 + MAX2 ≧ second threshold value, it is determined that there is a background, and the density classification to which MAX1 and MAX2 belong is set as the background density classification. For example, as shown in FIG. 6, this determination is performed by using a first original power density class indicating a white background and a second maximum power density class indicating a halftone dot background in a character document including a solid / halftone background. However, it uses that it is higher than other density values.
[0066]
  MAX1<In the case of the first threshold or MAX1 + MAX2<In the case of the second threshold, it is determined that there is no background. For example, as shown in FIG. 7, this determination utilizes the fact that the first maximum power density classification is not significantly higher than other density values in a photographic document made of photographic paper.
[0067]
Next, the second determination unit 41 will be described.
First, as shown in FIG. 2, the second determination unit 41, in the signal of each color component converted by the signal conversion unit 21, is a region where the processing target pixel is solid, that is, a region where the density is substantially constant, photographic paper (photograph ), A signal processing unit 42 (42c, 42m, 42y) provided for each color component signal (C, M, Y) for determining whether the region belongs to a character or a halftone dot, and each of these signal processing Based on the result of the signal processing for each color component signal by the units 42c, 42m, and 42y, a determination unit for determining whether the pixel of interest finally belongs to a type of photographic paper, character, or halftone dot As a general determination unit 47.
[0068]
As shown in FIG. 2, each of the signal processing units 42c, 42m, and 42y includes an attention pixel determination unit 45, a surrounding pixel determination unit 48 (48a, 48b, 48c, 48d, 48e, and 48f), and an upper line determination unit. 43, a lower line determination unit 44, and a signal-specific determination unit 46.
[0069]
The pixel-of-interest determination unit 45 extracts the feature amount of the pixel of interest from image data in the main scanning direction in an n × n block such as 3 × 3 centered on the pixel of interest. This is a pixel-of-interest determining means for determining whether it belongs to a solid, photographic paper, character or halftone dot.
[0070]
The peripheral pixel determination unit 48 performs the same processing as the target pixel determination unit 45 on the target pixel determination unit 45 and the peripheral pixels a, b, c, d, e, and f that exist around the target pixel, This is a peripheral pixel determination means for determining the type of peripheral pixels a, b, c, d, e, and f.
[0071]
The upper line determination unit 43 determines an upper line for the target pixel X based on the peripheral pixel determination results of the peripheral pixels a, b, and c above the target pixel X, that is, in the preceding stage in the sub-scanning direction. It is.
The lower line determination unit 44 is a lower line determination unit that determines the lower line for the target pixel X based on the peripheral pixel determination result of the peripheral pixels d, e, and f below the target pixel X, that is, downstream of the sub-scanning direction. .
[0072]
The signal determination unit 46 is a signal determination unit that performs determination for each signal based on the determination results of the target pixel determination unit 45, the upper line determination unit 43, and the lower line determination unit 44. Then, based on the determination result of the signal-specific determination unit 46 in each of the signal processing units 42c, 42m, and 42y, the above-described comprehensive determination unit 47 converts the pixel of interest X into any type of photographic paper, text, or halftone dot. A final determination is made as to whether it belongs.
[0073]
In the above example, when the types of the peripheral pixels a, b, c, d, e, and f are determined, the peripheral pixel determination units 48 a, 48 b, 48c, 48d, 48e, and 48f are provided.
[0074]
However, the present invention is not necessarily limited to this, and the peripheral pixel determination units 48a, 48b, 48c, 48d, 48e, and 48f are combined into one, and the determination results of the peripheral pixels a, b, c, d, e, and f are displayed as upper line determination units. 43 and the lower line determination unit 44 may be sequentially stored, and the determination of the upper line and the lower line may be performed based on the result. That is, the peripheral pixel determination units 48a, 48b, 48c, 48d, 48e, and 48f all perform the same processing, and therefore, a single peripheral pixel determination unit is shared and processed in time series. Is possible.
At this time, the peripheral pixel a, b, c, d, e, and f are first subjected to the determination process of the peripheral pixel a, and then the peripheral pixel b, the peripheral pixel c,. Judgment processing is performed.
[0075]
For the peripheral pixels a, b, and c existing in the previous stage of the target pixel X, the respective determination results are stored in the upper line determination unit 32, while the peripheral pixels d, e, and f existing in the subsequent stage of the target pixel X are stored. Is stored in the lower line determination unit 44. Then, after a predetermined number of discrimination results are stored, the upper line and the lower line are judged based on the respective results.
[0076]
Further, each determination process in the upper line determination unit 43 and the lower line determination unit 44, and each determination result of the upper line determination unit 43, the target pixel determination unit 45, and the lower line determination unit 44 input to the signal-specific determination unit 46 Alternatively, a priority order is determined for each determination result of each signal for each color component from each signal determination unit 46, 46, 46 input to the overall determination unit 47. When the determination results of the pixels entering the signal determination unit 46 and the comprehensive determination unit 47 are different from each other, the determination is performed according to the priority order of the signal determination unit 46 and the comprehensive determination unit 47. For example, when the determination result differs for each color component, halftone dot> character> printing paper> background. This is because if the result of halftone dots and characters is obtained for each color component, it is determined that the number of inversions is large, and this is a feature that cannot be seen in characters, so that halftone dots are given priority.
[0077]
Regarding the priority order, since the reliability of the determination result varies depending on the first and second threshold values used in the first determination unit 31, it is desirable to determine the priority order based on the reliability of the determination result.
[0078]
In addition, regarding the priority of each determination process in the upper line determination unit 43 and the lower line determination unit 44, each of the upper line determination unit 43 and the lower line determination unit 44 has a plurality of peripheral pixels a, b, c and peripheral pixels. Based on the result of d · e · f, the upper line and the lower line are determined. If the results of each peripheral pixel a · b · c and each peripheral pixel d · e · f are different, the final order is determined by the priority order. Make a good judgment.
The priority order is determined in advance based on, for example, setting of a threshold value used for pixel discrimination processing.
[0079]
Assume that it is determined that the peripheral pixel a is solid and the peripheral pixel b is a halftone dot. In this case, the threshold value for extracting halftone dots is set to a level at which it is difficult to extract halftone dots, and the threshold values for other areas are set to be loose compared to the extraction level for halftone dots. Is determined, the reliability of the result is determined to be high. Priorities are determined based on such information.
[0080]
Next, the configuration of the target pixel determination unit 45 will be described with reference to FIG. In the present embodiment, each of the surrounding pixel determination units 48a, 48b, 48c, 48d, 48e, and 48f has the same determination mechanism as the target pixel determination unit 45. Only for the unit 45.
[0081]
The pixel-of-interest determination unit 45 includes an adjacent pixel density difference calculation unit 51, a rising / falling determination unit 52, an inversion number calculation unit 53, a pixel determination unit 54, a solid pixel determination threshold setting unit 55, a rising edge, as shown in FIG. A falling judgment threshold setting unit 56 and an inversion number threshold setting unit 57 are included.
[0082]
The adjacent pixel density difference calculation unit 51 calculates a density difference (adjacent pixel density difference) between adjacent pixels in the main scanning direction including the target pixel X.
As illustrated in FIGS. 13 to 16, the rising / falling determination unit 52 determines whether the pixel of interest X belongs to a solid photographic paper based on the calculated adjacent pixel density difference, or whether the pixel is a rising pixel. A falling pixel is discriminated. The details of FIGS. 13 to 16 will be described later.
[0083]
The inversion number calculation unit 53 does not count even if a plurality of rising pixels or falling pixels continue in a certain region in the main scanning direction, that is, M pixels (M is an integer of 2 or more) with respect to the target pixel X, and from the rising It counts the number of falling and the number of times of reversal from falling to rising.
The pixel determination unit 54 determines whether the target pixel X is a character or a halftone dot based on the calculated inversion number, and further, the background determination unit 35 detects the density value for the pixel determined to be a halftone dot. In comparison with the density of the background (halftone dot), it is determined whether it is a halftone dot or a background (halftone dot).
[0084]
The solid pixel determination threshold value setting unit 55 is provided in the rising / falling determination unit 52 described above, and the pixel of interest X is solid with respect to the result of the pixel of interest X calculated by the adjacent pixel density difference calculation unit 51. A solid pixel determination threshold value for determining whether or not the pixel belongs is set. The solid pixel determination threshold setting unit 55 can separate a solid background region even when the background is small and the background determination unit 35 determines that there is no background region.
[0085]
Like the above-described pixel determination threshold setting unit 55, the rising / falling determination threshold setting unit 56 is provided in the rising / falling determination unit 52, and the target pixel X calculated by the adjacent pixel density difference calculation unit 51 is provided. The rising / falling determination threshold value for determining whether the target pixel X is a rising pixel or a falling pixel is set for the above result.
[0086]
The inversion number threshold setting unit 57 is provided in the pixel determination unit 54, and based on the inversion number calculated by the inversion number calculation unit 53, whether the target pixel X belongs to a character or a halftone dot. Is set as a third threshold value for discriminating.
[0087]
Next, an image determination method in the signal conversion unit 21, the first determination unit 31, and the second determination unit 41 described above will be described with reference to the flowcharts shown in FIGS. Further, the symbol indicating the step is “S”.
[0088]
First, the signal converter 21 converts the RGB reflectance signal into a density signal and converts the RGB density signal into a complementary CMY signal (S1). Next, the histogram creation unit 32 creates a density histogram indicating the relationship between the density of each pixel of the document image and its frequency (S2). Then, the maximum power density category extraction unit 33 extracts the first and second maximum power density categories (S3). Thereafter, the background determination unit 35 determines the presence / absence of the background area, and if there is a background area, sets the background density classification (S4). The image processing from S1 to S4 corresponds to the first determination step.
[0089]
Next, the signal conversion unit 21 converts the RGB reflectance signal into a density signal again, converts the RGB density signal into a complementary CMY signal (S5), and outputs the converted signal to the second determination unit 41. The converted CMY signal is input to the target pixel determination unit 45 of the second determination unit 41, and the adjacent pixel density difference calculation unit 51 calculates the density difference with the adjacent pixels in the main scanning direction including the target pixel X ( S6).
[0090]
Then, the rising / falling determination unit 52 compares the calculated absolute value of the adjacent pixel density difference with the solid pixel determination threshold (S7). When the calculated absolute value of the adjacent pixel density difference is smaller than the solid pixel determination threshold, the rising / falling determination unit 52 has a relationship of “maximum background density> pixel density> minimum background density”. Whether or not the above is satisfied is confirmed (S9). If it is satisfied, it is determined that the background is solid (S10). If it is not satisfied, it is determined to be a photographic paper photograph (S11). Note that the background determination unit 35 obtains the minimum and maximum values of the background density. The minimum and maximum values of the background density are the minimum value and the maximum value of the density classification determined as the background.
[0091]
On the other hand, in S7, for the pixel having the larger absolute value of the adjacent pixel density difference, the rise / fall determination unit 52 compares the calculated absolute value of the adjacent pixel density difference with the rise / fall determination threshold value. (S8).
[0092]
A pixel having an absolute value of the adjacent pixel density difference larger than the rising / falling determination threshold and having a positive adjacent pixel density difference is determined as a falling pixel. A pixel having an absolute value of the adjacent pixel density difference larger than the rising / falling determination threshold and having a negative adjacent pixel density difference is determined as a rising pixel.
[0093]
In this example, the adjacent pixel density difference is obtained by subtracting the density value of the right pixel from the density value of the left pixel, and positive / negative is determined accordingly. Also, the rising pixel / falling pixel is determined for the left pixel based on the density difference.
If the rising / falling determination threshold value is larger than the absolute value of the adjacent pixel density difference in S8, it is determined as photographic paper (S11).
[0094]
Next, when it is determined in S8 that the absolute value of the adjacent pixel density difference is larger than the rising / falling determination threshold, the inversion number calculation unit 53 calculates the inversion number (S12).
[0095]
In this inversion number calculation, as shown in FIG. 11, even if a plurality of rising pixels or falling pixels continue in a certain region composed of M pixels such as 30 pixels in the main scanning direction, the rising pixels are not counted. Are detected and counted from the falling pixel to the falling pixel and from the falling pixel to the rising pixel.
[0096]
Specifically, as shown in FIG. 11, both the fifth and sixth pixels from the left are determined as rising pixels “1”, but when obtaining the inversion number, only the fifth pixel is counted, The sixth pixel is not counted. In this way, as shown in FIG. 12, for each n × n pixel, the rising pixel / falling pixel in a certain area composed of M pixels such as 30 pixels in the main scanning direction is determined. The fixed area shown in FIG. 12 indicates the range of the fixed area in the main scanning direction when the rising pixel / falling pixel is determined, and indicates that a line memory is provided for each peripheral pixel. It is not a thing.
[0097]
Next, when the above-described inversion number is calculated in S12, the pixel determination unit 54 compares the counted inversion number with an inversion number threshold (S13). If the inversion number is smaller than the inversion number threshold, it is determined that the character is a character (S14).
[0098]
On the other hand, if the inversion number is larger than the inversion number threshold, the pixel determination unit 54 determines whether or not the pixel density satisfies the relationship of “maximum background density> pixel density> minimum background density”. Confirm (S15). If it is satisfied, it is determined to be a background (halftone dot) (S17), and if not satisfied, it is determined to be a halftone dot (S16). The minimum and maximum values of the background density are obtained by the background determination unit 35 as in the above case.
[0099]
In this way, the determination process is performed, and the final determination result for the pixel of interest is obtained in consideration of the priorities described above. The determination result for each pixel is counted by the comprehensive determination unit 47 and compared with a predetermined threshold value to determine the document type. For example, if the characters and halftone dots are equal to or greater than the threshold, the document is determined to be a character / printed photo document. If the character and background (halftone dot) are equal to or greater than the threshold value, the background (halftone dot) may be determined to be a text document, for example.
As described above, in the image processing according to the present invention, the document type can be determined using the minimum and maximum values of the background density obtained by the background determination unit 35. The processing from S5 to S17 corresponds to the second determination step.
[0100]
As described above, the process for determining the presence / absence of the background by the first determination unit (first determination step) and the process by the second determination unit for extracting the feature amount based on the presence / absence of the background and determining the document type (first) (2 determination step) may be performed by scanning the document once each (prescanning is performed twice before performing the main scan), but the forward path when the image input device 2 prescans the document image. It is desirable to perform the first determination step during (feed) and to perform the second determination step during return (return). By doing so, it is possible to effectively use the pre-scan return path and to improve the document type discrimination accuracy.
[0101]
Here, a representative example of the density distribution according to the type of each image will be described based on the distribution charts shown in FIGS. In addition, the vertical axis | shaft of the figure shows a density value.
Normally, as shown in FIG. 13, the density distribution diagram of a solid image has a very small adjacent pixel density difference because the density change is small. Accordingly, it is possible to determine that a pixel having a smaller adjacent pixel density difference than the solid pixel determination threshold is solid.
[0102]
Next, as shown in FIG. 14, the density distribution diagram of an image of a photographic paper photograph has a smooth density change, and the adjacent pixel density difference is small and slightly larger than solid. Also, it is possible to determine that the adjacent pixel density difference is large and smaller than the rising / falling threshold is a photographic paper photograph. Also, since there are solid areas in which the adjacent pixel density difference is very small even in photographic paper photographs, the pixel density of such areas is “maximum background density> pixel density> background density as described above. By confirming whether or not the relationship of “minimum value” is satisfied, it is possible to distinguish from the solid region of the background.
[0103]
Further, as shown in FIG. 15, the density distribution diagram of a halftone dot image has a large adjacent pixel density difference, and there are as many rising and falling pixels as the number of halftone dots, and the inversion number becomes very large. It is possible to determine that a pixel density difference is larger than the rising / falling determination threshold and the inversion number is larger than the inversion number threshold as a halftone dot. In addition, as described above, it is possible to distinguish the pixel density from the background (halftone dot) by checking whether or not the pixel density satisfies the relationship of “maximum background density> pixel density> minimum background density”. It is.
[0104]
Finally, as shown in FIG. 16, the density distribution diagram of the character image has a very large adjacent pixel density difference, but the number of rising pixels and falling pixels is present at the edge portion of the character. Since the number of inversions decreases, it is possible to determine that a character whose inversion number is smaller than the inversion number threshold is a character.
[0105]
In addition, it is possible to perform a wider range of processing by arbitrarily adjusting each of the set threshold values. For these threshold values, a plurality of expected values are stored in advance in a ROM (Read Only Memory) or the like, and values stored in a storage means (not shown) such as a memory are set by a switch or the like as necessary. You should be able to do it.
[0106]
By the processing operation described above, the type of document (type of image) is determined, and based on the result, as shown in FIG. 1, the subsequent input tone correction / color correction / black generation under color removal processing / space is performed. Filter processing, gradation reproduction processing, and area separation processing are switched.
[0107]
An example of the switching will be shown below.
[0108]
For example, when it is determined that the document is a character document, the area separation processing unit 20 validates the area separated as a character or line drawing, and erroneously identifies a continuous tone such as halftone dot or photographic paper. It is regarded as separation (because it is a character document, it may be misidentified depending on the type of document) and is not reflected. In the input tone correction unit 14 and the tone reproduction processing unit 19, a large amount of highlight is removed or a correction curve that increases the contrast is used. The color correction unit 15 and the black generation and under color removal unit 16 perform color correction processing with an emphasis on saturation on color characters, and set a large amount of black generation for black characters in the black generation and under color removal processing. The spatial filter processing unit 17 switches parameters such as emphasizing edges with spatial filter processing and setting filter coefficients so as to weaken smoothing processing.
[0109]
If it is determined that the document is a character / photographic paper photographic original, processing using an intermediate parameter between the character original processing and the photographic paper photographic original processing is performed in each processing. Then, the area separation processing unit 20 validates the area separated as a character, line drawing, or photographic paper, and regards the area separated as a halftone dot as erroneous separation (even if it is a character / photographic paper original). However, depending on the type of document, there is a case where it is erroneously determined), so that it is not reflected. The input tone correction unit 14 and tone reproduction processing unit 19 use the intermediate parameters between the photographic paper photo manuscript processing and the character manuscript processing depending on whether the text manuscript or the photographic paper photo manuscript is important. Remove and adjust contrast. The color correction unit 15 and the black generation and under color removal unit 16 perform color correction processing so that the intensity of saturation and the balance of gradation are not excessive. In the black generation and under color removal processing, the black generation amount is adjusted to such an extent that the photographic paper photographic image is not affected.
[0110]
Second Embodiment
Next, the document type automatic determination unit 13-2 according to the second embodiment of the present invention will be described.
In the first embodiment, the type of document is determined by the document type automatic determination unit 13 including the first determination unit 31 and the second determination unit 41. However, in the present invention, the document type can also be determined by the document type automatic determination unit 13-2 including only the second determination unit 41.
[0111]
FIG. 17 is a block diagram illustrating the document type automatic determination unit 13-2 according to the second embodiment. In addition, about the part which has the same function as the part used in 1st Embodiment, the same code | symbol is attached and the description is abbreviate | omitted. Note that the present invention is not limited to this.
[0112]
As shown in this figure, the configuration is substantially the same as the second determination unit 41 in FIG. However, the configuration of the target pixel determination unit 145 and the peripheral pixel determination unit 148 (148a, 148b, 148c, 148d, 148e, and 148f) is different from the target pixel determination unit 45 and the peripheral pixel determination unit 48 of the first embodiment. Therefore, the pixel-of-interest determination unit 145 and the surrounding pixel determination unit 148 will be described. Note that the surrounding pixel determination unit 148 has the same determination mechanism as that of the target pixel determination unit 145, and therefore the description will be given only for the target pixel determination unit 145.
[0113]
As illustrated in FIG. 18, the target pixel determination unit 145 includes an adjacent pixel density difference calculation unit 51, an inversion number calculation unit 53, a pixel determination unit 154, and a solid pixel determination threshold that are the same as the target pixel determination unit 45 of the first embodiment. In addition to the setting unit 55, the rising / falling determination threshold setting unit 56, and the inversion number threshold setting unit 57, a comparison threshold setting unit 61 is provided. The rising / falling determination unit 52 is not provided.
[0114]
The comparison threshold value setting unit 61 creates a three-dimensional frequency distribution diagram composed of the inversion number, adjacent pixel density difference, and frequency, which are image feature amounts, and in order to accurately determine the type of document, for each document to be read Set an appropriate threshold. The pixel determination unit 154 uses the threshold set by the comparison threshold setting unit 61 to (1) compare and determine the inversion number threshold and the inversion number, and (2) absolute rise / fall determination threshold and the adjacent pixel density difference. (3) A comparison determination between the solid pixel determination threshold value and the absolute value of the adjacent pixel density difference is performed for each pixel.
[0115]
Here, the relationship between the adjacent pixel density difference, the inversion number, and the frequency of each document type (base / printing paper, characters, halftone dots) will be described with reference to FIGS. 19 and 20. In these drawings, the X axis indicates the number of inversions, the Y axis indicates the adjacent pixel density difference, and the Z axis indicates the frequency, and the frequency distribution of each normal region is shown. In FIG. 19 and FIG. 20, for the sake of simplification, the distribution on the XY plane excluding the frequency distribution on the Z axis is shown.
[0116]
As shown in FIG. 19, a base / printing paper (printing paper photograph), characters, and halftone dots usually appear. In the background / photographic paper photograph, the difference between the number of inversions and the adjacent pixel density tends to appear in a small area. For characters and halftone dots, the adjacent pixel density difference appears over a wide range, but the inversion number tends to appear more in the halftone dots.
[0117]
By the way, as shown in this figure, there is a region where characters and halftone dots intersect. Therefore, as shown in FIG. 20, a document is determined by providing a threshold at a location where characters / halftone dots appear prominently. For example, a three-dimensional frequency distribution diagram composed of the inversion number, the adjacent pixel density difference, and the frequency is created, and the threshold value is determined so that the character and halftone dot regions do not overlap.
[0118]
Specifically, a threshold value is set at which the frequency of the character area / halftone dot area is equal to or greater than a predetermined value.
[0119]
The character area is defined by a rising / falling determination threshold 1, a rising / falling determination threshold 2, an inversion number threshold 1, and an inversion number threshold 2. The rising / falling determination threshold 1 and the rising / falling determination threshold 2 are determined based on the adjacent pixel density difference. As shown in FIG. 19, since a part where a character appears to some extent can be specified by the adjacent pixel density difference, an upper limit and a lower limit of the part are set as threshold values. The inversion number threshold value 1 and the inversion number threshold value 2 are the upper limit (inversion number threshold value 2) and lower limit (inversion number) of the inversion number frequently appearing in the rising / falling determination threshold value 1 and rising / falling determination threshold value 2 described above. It is defined as the number threshold 1).
[0120]
The halftone dot region is defined by an inversion number threshold 3, an inversion number threshold 4, a rising / falling determination threshold 3, and a rising / falling determination threshold 4. As shown in FIG. 19, the reversal number threshold value 3 and the reversal number threshold value 4 can specify a portion where a halftone dot appears to some extent from the reversal number, so the upper limit and the lower limit of the portion are set as threshold values. The rising / falling determination threshold 3 and the rising / falling determination threshold 4 are the upper limit (rising / falling determination threshold 4) of adjacent pixel density differences that frequently occur within the range of the inversion number threshold 3 and the inversion number threshold 4. It is determined as the lower limit (rising / falling judgment threshold 3).
As described above, the present invention extracts only the pixels with high accuracy determined to belong to the character area or the dot area, and improves the accuracy of document type discrimination.
[0121]
In addition, as shown in FIG. 21, in order to reduce the memory area, an area to be counted for characters / halftone dots may be determined. That is, for characters, the number of inversions changes for one original, but the adjacent pixel density difference does not change much. For this reason, the setting is made so that the count width of the inversion number is narrow and the adjacent pixel density difference is widely counted. On the other hand, in the halftone dot, the adjacent pixel density difference changes with respect to one original, but the inversion number does not change so much.
For this reason, the count width of the inversion number is set to be wide and the adjacent pixel density difference is set to be narrowly counted.
[0122]
Next, a processing method by the pixel-of-interest determination unit 145 shown in FIGS. 22 and 23 will be described with reference to flowcharts.
First, the signal conversion unit 21 converts an RGB reflectance signal into a density signal and converts the RGB density signal into a complementary CMY signal (S21). Next, the inversion number calculation unit 53 calculates the inversion number (S22), and the adjacent pixel density difference calculation unit 51 calculates the density difference between adjacent pixels in the main scanning direction (S23).
[0123]
Then, the comparison threshold setting unit 61 creates a three-dimensional frequency distribution diagram including the inversion number, the adjacent pixel density difference, and the frequency (S24). Then, from the distribution diagram, the rising / falling determination threshold 1, the rising / falling determination threshold 2, the inversion number threshold 1, the inversion number threshold 2, and the halftone dot are reliably detected. The falling judgment threshold 3, rising / falling judgment threshold 4, inversion number threshold 3, and inversion number threshold 4 are set (S25). The rising / falling determination thresholds 1 to 4 are stored in the rising / falling determination threshold setting unit 56, and the inversion number thresholds 1 to 4 are stored in the inversion number threshold setting unit 57. The processing so far is performed only by the pixel-of-interest determination unit 145 of the automatic document type determination unit 13-2 shown in FIG. 17, and in the processing described below, all the processing units of the automatic document type determination unit 13-2. Works.
[0124]
Next, the original is read again, the signal conversion unit 21 converts the RGB reflectance signal into a density signal and converts the RGB density signal into a complementary CMY signal (S26), and the inversion number calculation unit 53 calculates the inversion number. Calculate (S27). The adjacent pixel density difference calculation unit 51 calculates the adjacent pixel density difference (S28).
[0125]
Thereafter, the pixel determination unit 154 determines a target pixel using the inversion number / adjacent pixel density difference calculated in S27 and S28 and each threshold value set so far.
First, the pixel determination unit 154 determines that “inversion number threshold 3 <calculated inversion number <inversion number threshold 4 and rise / fall determination threshold 3 <absolute value of calculated adjacent pixel density difference <rise / fall determination”. It is determined whether or not the relationship of “threshold value 4” is satisfied (S29). If it is satisfied, it is regarded as a halftone dot (S30).
[0126]
On the other hand, when the relationship in S29 is not satisfied, the pixel determination unit 154 determines that “the inversion number threshold 1 <the calculated inversion number <the inversion number threshold 2 and the rising / falling determination threshold 1 <the calculated adjacent pixel density difference calculation”. It is determined whether or not the relationship of absolute value <rising / falling threshold value 2 ”is satisfied (S31). If it is satisfied, it is regarded as a character (S32).
[0127]
On the other hand, when the relationship in S31 is not satisfied, the pixel determination unit 154 determines whether or not the relationship of “absolute pixel density difference <absolute value <solid pixel determination threshold” is satisfied (S33). If it is satisfied, it is regarded as a ground (S34). If not, it is regarded as photographic paper (S35).
[0128]
As in the case of the first embodiment, the determination result is determined in consideration of the priority order, and the result for the target pixel is determined. The determination result for each pixel is counted by the comprehensive determination unit 47 and determined in advance. The document type is determined by comparing with a threshold value.
As described above, the process of setting an appropriate threshold value from S21 to S25, that is, the feature amount of the inversion number and the adjacent pixel density difference in order to extract pixels that are determined to belong to the character area or the halftone dot area reliably. A process for setting an appropriate threshold based on this and a process from S26 to S35 for determining the document type using these thresholds each scan the document once (pre-scan twice before performing the main scan). However, the threshold value is set during the forward path (feed) when the document image is pre-scanned by the image input device 2, and the document type is determined using the threshold value during the return path (return). It is desirable to do. By doing so, it is possible to effectively use the pre-scan return path and to improve the document type discrimination accuracy.
As described above, the image processing apparatus can perform document type determination even using only the second determination unit 41.
[0129]
<Embodiment 3>
Next, the document type automatic discrimination unit 13-3 according to the third embodiment of the present invention will be described.
FIG. 24 is a block diagram illustrating the document type automatic determination unit 13-3 according to the third embodiment. In addition, about the part which has the same function as the part used in 1st Embodiment, the same code | symbol is attached and the description is abbreviate | omitted. Note that the present invention is not limited to this.
[0130]
First, as shown in FIG. 24, the automatic document type discrimination unit 13-3 includes a signal conversion unit 21, a judgment block unit 71, a main scanning direction judgment unit 72 (72c, 72m, 72y; area judgment unit), and a sub scanning direction. The determination unit 73 (73c / 73m / 73y; area determination unit), the signal-specific determination unit 74 (74c / 74m / 74y), and the comprehensive determination unit (original determination unit) 75 are configured. Note that the main scanning direction determination unit 72, the sub-scanning direction determination unit 73, and the signal-specific determination unit 74 are arranged for each CMY signal.
[0131]
The signal conversion unit 21 converts the RGB reflectance signal input from the shading correction unit 12 into a density signal, converts the RGB density signal into a complementary CMY signal, and outputs the converted signal to the determination block unit 71. .
[0132]
The determination block unit 71 is for storing image data of a block composed of a plurality of n × m pixels such as 15 × 7 for the converted CMY signal. The main scanning direction determination unit 72 extracts, from the image data of the determination block unit 71, image data in the main scanning direction including the target pixel, that is, in the direction orthogonal to the scanning direction of the scanner (image input device). Each pixel is separated into a character, halftone dot, character or halftone dot, background, photographic paper photograph, background or photographic paper photograph area.
[0133]
The sub-scanning direction determining unit 73 extracts image data in the sub-scanning direction including the target pixel, that is, the same direction as the scanning direction of the scanner, from each image data of the determination block unit 71, , Character or halftone dot, background, photographic paper photograph, background or photographic paper photograph area.
[0134]
The signal discriminating unit 74 discriminates each signal based on the determination result and priority of the main scanning direction determination unit 72 and the sub-scanning direction determination unit 73 (priority with respect to the results in the main scanning direction and the sub-scanning direction). It is.
[0135]
The overall determination unit 75 further performs final pixel determination based on the result of the signal-specific determination unit for each signal. Priorities are determined for the signal (region identification signal) of each pixel that enters the signal-specific determination unit and the comprehensive determination unit. If the pixel determination results that enter the respective determination units differ, the priority order is determined. Make a decision.
[0136]
As for the priority order, the data size of the image in the main scanning direction and the sub scanning direction stored in the determination block unit 71 (because the data (number of pixels) in the main scanning direction and the sub scanning direction differs depending on the block size) Since the reliability of the determination result varies depending on the threshold value used in the resolution and area separation processing, it is desirable to determine the priority order based on the reliability of the determination result. For example, if even one image area is included, it is validated. In the case of a color manuscript, all black characters have a characteristic amount as a character having the same color component signal. For example, in a red character, the MY signal has a characteristic amount as a character, but does not have a C signal. In this way, it is possible to cope with a case where the pixel density value largely changes depending on the color component signal.
[0137]
Next, specific configurations of the main scanning direction determination unit 72 and the sub scanning direction determination unit 73 will be described with reference to FIG. That is, both the main scanning direction determination unit 72 and the sub scanning direction determination unit 73 have the same configuration. Further, the pixels used in the main scanning direction determination unit 72 are pixels only in the main scanning direction of the n × m determination block including the target pixel stored in the determination block unit 71, while the sub scanning direction determination unit 73 The pixels to be used are pixels only in the sub-scanning direction of the n × m determination block including the target pixel.
[0138]
First, as shown in FIG. 25, the main scanning direction determination unit 72 or the sub scanning direction determination unit 73 includes a minimum density value calculation unit 81, a maximum density value calculation unit 82, a maximum density difference calculation unit 83, and a total difference value calculation. A section 84, a determination area setting section 85, a character / halftone determination section 86, and a background / photographic paper determination section 87.
[0139]
The minimum density value calculation unit 81 and the maximum density value calculation unit 82 are for calculating the minimum density value and the maximum density value in the pixel in the main scanning direction or the sub scanning direction.
The maximum density difference calculator 83 is for calculating the maximum density difference using the minimum density value and the maximum density value calculated by the minimum density value calculator 81 and the maximum density value calculator 82.
The sum difference value calculation unit 84 is for calculating the sum of absolute values of density differences between adjacent pixels.
[0140]
The determination area setting unit 85 compares the maximum density difference calculated by the maximum density difference calculation unit 83 and the total difference value calculated by the total difference value calculation unit 84 with each threshold value, thereby comparing the background area. -It is for separating into a photographic paper (photo) area and a character area / halftone dot area.
[0141]
The character / halftone dot determination unit 86 determines that the pixel of interest is a character region, a halftone dot region, a character, or a halftone dot region (which is unknown) with respect to the pixels determined as the character / halftone dot region by the determination region setting unit 85. It is for determining to which one of these.
[0142]
The background / photographic paper determination unit 87 selects a pixel of interest as a background region, a photographic paper region, a background, or a photographic paper region (which is unknown) with respect to the pixels determined as the background / photographic paper region by the determination region setting unit 85. It is for determining to which one of these.
[0143]
The determination area setting unit 85 separates the result calculated by the maximum density difference calculation unit 83 of the target pixel into a background / printing paper (photo) area and a character / halftone dot area. A maximum density difference threshold value setting unit 88 for setting a maximum density difference threshold value as a threshold value of the image, and a background / printing paper region and a character / halftone dot region for the result calculated by the total difference value calculation unit 84 of the target pixel And a total difference value setting unit 89 for setting a total difference value threshold value as a second threshold value.
[0144]
Further, the character / halftone dot determination unit 86 includes a character determination threshold value setting unit 90 for setting a character determination threshold value as a fifth threshold value for determining that the target pixel belongs to the character region, and the target pixel is a halftone dot. And a halftone dot threshold value setting unit 91 for setting a halftone dot threshold value as a sixth threshold value for determining that the image belongs to the area. Further, the background / photographic paper determination unit 87 determines a photographic paper determination threshold value setting unit 92 as a third threshold value for determining that the target pixel belongs to the photographic paper region, and determines that the target pixel belongs to the background region. A background determination threshold value setting unit 93 serving as a fourth threshold value;
[0145]
Here, an example of pixel density distribution in the character / halftone dot region and the photographic paper photograph / background region will be described with reference to FIGS. Further, a distribution diagram based on the maximum density difference and the total difference value for each region will be described with reference to FIGS. 30 and 31 (FIG. 30 shows an actual distribution of each pixel, and FIG. 31 shows this distribution. ) And a threshold value for separating each pixel is entered). An area where the maximum density difference is equal to or less than the total difference value indicates an area where the total difference value does not become less than the maximum density difference and no pixel exists.
[0146]
First, as shown in FIG. 26, the density distribution of the base region is usually very small because the change in density is small, and both the maximum density difference and the total difference value are very small and are distributed in the region A shown in FIG. That is, as shown in FIG. 31, when the maximum density difference is smaller than the background determination threshold for the pixel determined as the background / printing paper area, it can be determined as the background pixel.
[0147]
Next, as shown in FIG. 27, the density distribution of the photographic paper region normally changes smoothly and the maximum density difference and the total difference value are both small and slightly larger than the background area. , Distributed in a region B shown in FIG. Therefore, as shown in FIG. 31, when the maximum density difference is larger than the photographic paper determination threshold for the pixels determined to be the background area and the photographic paper area, it can be determined to be the photographic paper area.
[0148]
Furthermore, for a pixel determined to be a background / photographic paper area, if the maximum density difference is larger than the background determination threshold value but smaller than the photographic paper determination threshold value, it is unclear to which area it belongs. It can be determined that it belongs to the area.
[0149]
In addition, as shown in FIG. 28, the density distribution in the halftone dot region varies depending on the halftone dot, but the total difference value has a density change corresponding to the number of halftone dots. The ratio of the difference value increases. For this reason, it becomes distribution like the area | region D shown in FIG. Therefore, as shown in FIG. 31, if the sum difference value is larger than the product of the maximum density difference and the halftone dot determination threshold for the pixel determined as the character / halftone dot region, it can be determined that the pixel belongs to the halftone dot region. .
[0150]
Next, as shown in FIG. 29, the density distribution of the character area has a maximum density difference and a total difference value that increases. However, since the density change is smaller than that of the halftone area, the summation is smaller than that of the halftone area. The difference value becomes smaller. In particular, since the ratio of the total difference value with respect to the maximum density difference is small, the distribution is as in region C shown in FIG. Therefore, as shown in FIG. 31, when a pixel determined to be a character / halftone dot region has a total difference value smaller than the product of the maximum density difference and the character determination threshold, it can be determined that the pixel belongs to the character region.
[0151]
Furthermore, for pixels determined to be a character / halftone area, if the sum total difference value is smaller than the product of the maximum density difference and the halftone dot determination threshold, but greater than the product of the maximum density difference and the character determination threshold, Therefore, it can be determined that it belongs to the character or halftone dot region.
[0152]
Next, the document type discrimination operation will be described below. In this embodiment, the document type is determined based on image data obtained by pre-scanning. As a result, it is possible to automatically determine the document type, and when the document type is uncertain, it is possible to easily determine the document type by the user by setting the priority order. Strategy can be achieved.
[0153]
First, the main scanning direction determination unit 72 and the sub-scanning direction determination unit 73 determine in which region each pixel belongs to the image data read from the document to be determined. This determination procedure will be described based on the flowchart shown in FIG.
[0154]
First, it is automatically detected that a document is placed on the document table, and pre-scanning is started. Then, as shown in the figure, the maximum density value in the n × m block including the target pixel is calculated from the image data obtained by the prescan (S41), and the minimum density value is calculated (S42). Next, the maximum density difference is calculated using the calculated maximum density value and minimum density value (S43), and the sum of absolute values of density differences between adjacent pixels, that is, the total difference value is calculated (S44).
[0155]
Next, the calculated maximum density difference and the maximum density difference threshold value are compared, and the calculated total difference value and the supply sum difference value threshold value are compared (S45). When it is determined that the maximum density difference is smaller than the maximum density difference threshold value and the total difference value is smaller than the total difference value threshold value, it is determined that the target pixel belongs to the background / printing paper region (S46). .
On the other hand, when the above condition is not satisfied in S45, it is determined that the target pixel belongs to the character / halftone dot region (S52).
[0156]
In the background / printing paper area shown in S46, the calculated maximum density difference is compared with the background determination threshold (S47). If the maximum density difference is smaller, the target pixel belongs to the background area. It is determined (S48). If the maximum density difference is larger in S47, the maximum density difference is compared with the photographic paper determination threshold value (S49). Here, when the maximum density difference is larger, it is determined that the target pixel belongs to the photographic paper area, but when the maximum density difference is smaller, it is determined to belong to the background or the photographic paper area (which is unknown). (S51).
[0157]
On the other hand, in the character / halftone dot area shown in S52, the calculated sum difference value is compared with the value obtained by multiplying the maximum density difference by the character determination threshold value (S53), and if the sum difference value is smaller. The target pixel is determined to belong to the character area (S54). If the sum difference value is larger in S53, a comparison is further made between the sum difference value and a value obtained by multiplying the maximum density difference by a halftone threshold value (S55). Here, when the total difference value is larger, it is determined that the target pixel belongs to the halftone dot region (S56), and when the maximum density difference is smaller, it is determined that it belongs to the character or halftone dot region (which is unknown). (S57).
[0158]
After determining which region each pixel belongs to, the following calculation is performed in the signal determining unit 74. First, the number of pixels classified into each region is counted. Then, the count value of each area is compared with a threshold value set in advance for each area, and the document type is determined.
[0159]
If the determination results are different between the main scanning direction and the sub-scanning direction, the data size of the image in the main scanning direction and the sub-scanning direction stored in the determination block unit 71 (the main scanning direction / sub-scanning depends on the block size). Because the reliability of the determination result varies depending on the direction data (number of pixels)), the resolution, and the threshold value used in each scanning direction determination unit, the reliability of the determination result (the larger the block, the more pixels used for the determination) Therefore, the amount of information increases, and when the threshold value used for the determination is determined to be strict, the reliability is higher), and the determination result is adopted based on that. .
[0160]
Further, when image data in the sub-scanning direction is thinned out during pre-scanning, only the image data in the main scanning direction is used.
For example, when the count value of the character area exceeds a predetermined threshold value of the character area, it is determined that the document is a character document. If the count value of the area other than the character area is close to the threshold value that does not exceed the threshold value of the area, it is determined that the document type is unknown, and the user's own determination is prompted by the following procedure.
[0161]
First, the count value of each area is defined as follows. The count values of these areas may be displayed to the user via the operation panel 4.
Ct: Count value of character area Ch: Count value of halftone dot area
Cth: Count value of character or halftone dot area
Cb: Count value of background area Cp: Count value of photographic paper photograph area
Cbp: Count value of the background or photographic paper photographic area
[0162]
Further, the threshold value is also defined as follows. These threshold values may also be displayed to the user via the display means.
THt: threshold value for the count value of the character area
THt1: Threshold value for the count value of the character area and the character or halftone dot area
THh1: threshold for halftone dot area and count value of character or halftone dot area
THb1: Threshold value for the count value of the background area and the background or photographic paper photographic area
THp1: Threshold value for the count value of the photographic paper simple area and the background or photographic paper photographic area
These threshold values are determined in advance so that a preferable result can be obtained based on a large number of documents.
[0163]
Here, a case where Ct> THt will be described. In this case, although it is confirmed that the document includes a character area, it is unknown whether it is a character document, a character / photo document, or a character / photographic paper photo document. Therefore, according to the present invention, when the count value of the other area is close to the threshold value even if it does not exceed the threshold value, the document type is predicted, and the priority order of the document type is determined and displayed as follows. Yes.
[0164]
1) When Ct> THt, Ct + Cth> THt1, and Ch + Cth <THh1
(1) Text manuscript
(2) Text / photo (printed photo) manuscript
(Numbers in circles are priorities)
2) When Ct> THt and Ct + Cth <THt1, Ch + Cth> THh1, or when Ct> THt and Ct + Cth <THt1, Ch + Cth <THh1
(1) Text / Photo manuscript
(2) Text manuscript
3) When Ct> THt, Cb + Cbp> THb1, and Cp + Cbp <THp1
(1) Text manuscript
(2) Text / photographic paper photo manuscript
4) When Ct> THt and Cb + Cbp <THb1, Cp + Cbp> THp1, or when Ct> THt and Cb + Cbp <THb1, Cp + Cbp <THp1
(1) Text / photographic paper photo manuscript
(2) Text manuscript
[0165]
As described above, for example, when the count value of the area other than the character area does not exceed the threshold value but is close to the threshold value, the user can be prompted to select the document type by determining that the document type is unknown. Thereby, erroneous determination of the document type can be prevented. Also, misprinting can be reduced because various correction processes are performed in a processing mode corresponding to the document type.
[0166]
Furthermore, even when the document type is unknown, the user can easily determine the document type by displaying the expected document types with priorities, and select a processing mode according to the priority. Can do.
[0167]
Further, the image processing apparatus 1 includes a CPU (control means) and a storage means (not shown), and this CPU stores the document type discrimination result, the total number of pixels, and the count value of each area in a storage means (not shown). Thereby, the user can know the determination result at the time of previous output when the output image is not preferable and the image is output again. In other words, in order to know the determination result of the previous document type, place the document on the document table, perform a pre-scan, obtain the total number of pixels and the count value of each area, and compare the result with the stored data. Is done. Then, by providing display means for displaying the determination result for the user and input means for allowing the user to select and input the processing mode, the user can select and input the processing mode according to the determination result. In this case, the display means and the input means can be substituted by the operation panel 4.
The document type determination result thus performed is fed back to each processing unit after the document type automatic determination unit 13-3 as described above.
[0168]
In this way, the image processing apparatus 1 can switch the processing mode corresponding to the type of document automatically or by the user's own selection.
[0169]
Note that the image processing apparatus 1 of the present invention can also be applied only to the image reading apparatus (image input apparatus) 2. In this case, the image processing apparatus 1 includes the A / D conversion unit 11, the shading correction unit 12, the document type automatic determination unit 13, and the input tone correction unit 14 illustrated in FIG. 1.
[0170]
<Fourth embodiment>
Next, the automatic document type discrimination unit 13-4 according to the fourth embodiment of the present invention will be described.
FIG. 33 is a block diagram illustrating the document type automatic determination unit 13-4 according to the fourth embodiment. In addition, about the member which has the same function as the member used in 3rd Embodiment, the same code | symbol is attached and the description is abbreviate | omitted. Note that the present invention is not limited to this.
[0171]
The document type automatic determination unit 13-4 is different from the third embodiment in a main scanning direction determination unit 172 (172c, 172m, 172y) and a sub-scanning direction determination unit 173 (173c, 173m, 173y). Both the main scanning direction determination unit 172 and the sub-scanning direction determination unit 173 have the same configuration. FIG. 34 is a block diagram of the main scanning direction determination unit 172. The same parts as those in the main scanning direction determination unit 72 of the third embodiment shown in FIG.
[0172]
The pixels used in the main scanning direction determination unit 172 are pixels only in the main scanning direction of the n × m determination block including the target pixel stored in the determination block unit 71, and are used in the sub-scanning direction determination unit 173. The pixel is a pixel only in the sub-scanning direction of the n × m determination block including the target pixel.
[0173]
The main scanning direction determination unit 172 and the sub scanning direction determination unit 173 include a minimum density value calculation unit 81 that calculates a minimum density value, a maximum density value calculation unit 82 that calculates a maximum density value, and the calculated minimum density value and The maximum density difference calculation unit 83 that calculates the maximum density difference using the maximum density value, the total difference value calculation unit 84 that calculates the sum of absolute values of the density differences of adjacent pixels, and the maximum density difference calculation unit The maximum density difference and the calculated total difference value are compared with respective predetermined thresholds, a background determination unit 101 that determines a background pixel, a photographic paper determination unit 102 that determines a photographic paper pixel, and a character pixel A character determination unit 103 and a halftone dot determination unit 104 that determines halftone pixels are configured.
[0174]
Further, the background determination unit 101 includes a background maximum density difference threshold setting unit 111 for setting a threshold for determining a background pixel for the result calculated by the maximum density difference calculation unit 83, and a total difference value calculation unit. A background total difference value threshold setting unit 112 for setting a threshold for determining a background pixel with respect to the total difference value calculated in 84, and the photographic paper determination unit 102 includes a maximum density difference calculation unit 83. Is calculated by the photographic paper minimum density difference threshold value setting unit 113, the photographic paper maximum density difference threshold value setting unit 114, and the total difference value calculating unit 84 for setting a threshold value for determining a photographic paper pixel. The photographic paper determination threshold value setting unit 115 for setting a threshold value for determining the photographic paper pixel as a result of the determination, and the character determination unit 103 perform the character calculation for the result calculated by the maximum density difference calculation unit 83. Picture A character determination threshold for setting a threshold for determining a character pixel for the result calculated by the character minimum density difference threshold setting unit 116 and the sum difference value calculation unit 84. Similarly to the setting unit 117, the halftone dot determination unit 104 is a halftone dot total difference value threshold setting unit for setting a threshold for determining a halftone pixel for the result calculated by the total difference calculation unit 84. 118 and a halftone dot determination threshold value setting unit 119.
[0175]
FIG. 35 is a flowchart showing the flow of region separation processing for performing document type determination in each of the main scanning direction determination unit 72 and the sub-scanning direction determination unit 73 in the n × m block including the target pixel in the present invention. It is.
[0176]
First, at step 61 (S61), the maximum density value is calculated, and at S62, the minimum density value is calculated. Then, the maximum density difference is calculated using the minimum density value and the maximum density value calculated in S61 and S62 (S63). Further, in S64, the sum of absolute values of density differences between adjacent pixels is calculated. In S65, the maximum density difference calculated in S63 and the background maximum density difference threshold value are compared, and the total difference value calculated in S63 and the background total difference value threshold value are compared. When it is determined that the maximum density difference is smaller than the background maximum density difference threshold and the total difference value calculated in S64 is smaller than the background total difference value threshold (“YES” in S65), the target pixel is the background pixel. It is determined that there is (S73).
[0177]
On the other hand, when the above condition is not satisfied in S65, it is determined in S66 whether the pixel is a photographic paper pixel. The maximum density difference calculated in S63 is larger than the photographic paper minimum density difference threshold and smaller than the photographic paper maximum density difference threshold, and the product of the maximum density difference calculated in S63 and the photographic paper determination threshold is calculated in S64. If it is greater than the sum total difference value, it is determined that the pixel is a photographic paper photographic pixel (S72). Further, when the above condition is not satisfied in S66, the maximum density difference calculated in S63 in S67 is larger than the character minimum density difference threshold, and the product of the maximum density difference calculated in S63 and the character determination threshold is S64. If it is larger than the total difference value calculated in step S1, it is determined that the target pixel is a character pixel (S71).
[0178]
Next, when the above condition is not satisfied in S67, in S68, the total sum difference value calculated in S64 is larger than the halftone dot total difference value threshold value, and the maximum density difference calculated in S63 and the halftone dot determination threshold value are calculated. Is larger than the sum difference value calculated in S64, it is determined that the pixel is a halftone dot pixel (S70), and otherwise, it is determined that it is an indeterminate pixel (S69).
[0179]
In this way, the determination process is performed, and the final determination result for the target pixel is obtained. The determination result for each pixel is counted by the comprehensive determination unit 75 and compared with a predetermined threshold value to determine the document type. For example, if the characters and halftone dots are equal to or greater than the threshold, the document is determined to be a character / printed photo document.
[0180]
Here, an example of pixel density distribution in the character / halftone dot region, photographic paper photograph / background region will be described with reference to FIGS. 26 to 29 of the third embodiment. In addition, a distribution diagram based on the maximum density difference and the total difference value for each region will be described with reference to FIGS. 36 and 37. FIG. (FIG. 36 shows an actual distribution of each pixel, and FIG. 37 shows this distribution in a simulated manner and a threshold value for separating each pixel is entered. Maximum density difference = total difference The region below the value indicates a region where the sum difference value does not become the maximum density difference or less and no pixel exists.
[0181]
First, as shown in FIG. 26, the density distribution of the base region is usually very small because the density change is small, and both the maximum density difference and the total difference value are very small and are distributed in the region A shown in FIG. That is, as shown in FIG. 37, when the maximum density difference is smaller than the background maximum density difference threshold and the total difference value is smaller than the background total difference value threshold, it can be determined that the pixel is a background pixel.
[0182]
Next, as shown in FIG. 27, the density distribution of the photographic paper region normally changes smoothly, and the maximum density difference is small but slightly larger than the background region. In addition, since the density change is smooth, the density change of adjacent pixels is also small, and the maximum density difference and the total difference value are very close to each other, and are therefore distributed in the region B shown in FIG. Therefore, as shown in FIG. 37, the maximum density difference is larger than the photographic paper minimum density difference threshold, the maximum density difference is smaller than the photographic paper maximum density difference threshold, and the product of the maximum density difference and the photographic paper determination threshold. If the total difference value is small, it can be determined that the area is a photographic paper photograph area.
[0183]
In addition, as shown in FIG. 28, the density distribution in the halftone dot region varies depending on the halftone dot, but the total difference value has a density change corresponding to the number of halftone dots. The ratio of the difference value increases. For this reason, it becomes distribution like the area | region D shown in FIG. Therefore, as shown in FIG. 37, when the total sum difference value is larger than the halftone dot total difference value threshold value and the total sum difference value is larger than the product of the maximum density difference and the halftone dot determination threshold value, it belongs to the halftone dot region. Can be determined.
[0184]
Next, as shown in FIG. 29, the density distribution of the character area has a maximum density difference and a total difference value that increases. However, since the density change is smaller than that of the halftone area, the summation is smaller than that of the halftone area. The difference value becomes smaller. In particular, since the ratio of the total difference value with respect to the maximum density difference is small, the distribution is as in region C shown in FIG. Therefore, as shown in FIG. 37, when the maximum density difference is larger than the character minimum density difference threshold and the total difference value is smaller than the product of the maximum density difference and the character determination threshold, it is determined to belong to the character area. it can.
[0185]
Further, since it is not clear which areas other than those determined as the background area, the photographic paper photograph area, the halftone dot area, and the text area belong to the indeterminate area.
As described above, in the present embodiment, except for pixels having uncertain elements, only the pixels that are clear to which region belong are extracted and the document type is determined. Can be raised. Further, when determining the area of the input image data, only a specific feature amount is extracted, so that the circuit can be small.
[0186]
Note that a wider range of processing can be performed by arbitrarily adjusting the threshold values set above. For these threshold values, a plurality of expected values are stored in advance in a ROM (Read Only Memory) or the like, and values stored in a storage means (not shown) such as a memory are set by a switch or the like as necessary. You should be able to do it.
[0187]
By the processing operation described above, the document type (pixel type) is determined, and based on the result, as described above, the subsequent input tone correction, color correction, black generation and under color removal processing, and spatial filter processing are performed.・ Tone reproduction processing and area separation processing are switched.
[0188]
In the present invention, the image processing method may be recorded on a computer-readable recording medium in which a program to be executed by a computer is recorded. As a result, it is possible to provide a portable recording medium on which a program for performing the image processing method is recorded.
In the present embodiment, as the recording medium, a memory (not shown) such as a ROM itself may be a program medium because processing is performed by a microcomputer. However, it may be a program medium provided with a program reading device as an external storage device and readable by inserting a recording medium therein.
[0189]
In any case, the stored program may be configured to be accessed and executed by the microprocessor, or in any case, the program is read and the read program is illustrated in the microcomputer. The program may be downloaded to a non-program storage area and executed. It is assumed that this download program is stored in the main device in advance.
[0190]
Here, the program medium is a recording medium configured to be separable from the main body, such as a tape system such as a magnetic tape or a cassette tape, a magnetic disk such as a floppy (R) disk or a hard disk, or a CD-ROM / MO / MD. / Disk system of optical disks such as DVD, IC card (including memory card) / card system such as optical card, or mask ROM, EPROM (Erasable Programmable Read Only Memory), EEPROM (Electrically Erasable Programmable Read Only Memory) flash It may be a medium that carries a fixed program including a semiconductor memory.
[0191]
Further, in the present embodiment, since it is a system configuration capable of connecting a communication network including the Internet, it may be a medium that fluidly carries the program so as to download the program from the communication network. When the program is downloaded from the communication network in this way, the download program may be stored in the main device in advance or may be installed from another recording medium.
[0192]
The recording medium is read by a program reading device provided in the digital color image forming apparatus, whereby the above-described image processing method is executed.
[0193]
In the above, a digital copying machine using an electrophotographic process has been described as an example. However, the present invention is not limited to the above example, and predetermined image processing is performed by inputting information from an image input device. Needless to say, the present invention can be applied to an image forming apparatus that performs and outputs the result, for example, an image forming apparatus that uses an ink jet recording method or a sublimation recording method.
[0194]
【The invention's effect】
According to the present invention, the document type determining means for automatically determining the document type from the input image data first determines the presence or absence of the background, and determines the document type based on the result. Therefore, the accuracy of determining the document type is improved.
[0195]
According to the present invention, when the first maximum numerical value ≧ the first threshold, for example, the first maximum power density classification (so-called white background) in a solid text document is higher than other density values. Can be used to determine whether or not there is a groundwork.
In the case of the first maximum frequency value + the second maximum frequency value ≧ the second threshold value, for example, in a text document composed of solid and halftone dots, the first maximum frequency density classification indicating a white background and the first background indicating a halftone dot background. It is possible to determine whether or not there is a ground by utilizing the fact that the 2 maximum power density categories are higher than other density values.
In the case of the first maximum power value ≦ the first threshold value, for example, in the case of a photographic document composed of a photographic paper photograph, the first maximum power power density classification is utilized which is not significantly higher than other density values. , It can be determined whether there is a ground. Therefore, the document type determination accuracy is further improved.
[0196]
According to the present invention, since the histogram creation unit creates a histogram based on the average value of the density values of a plurality of input image data, the extraction accuracy of the background (halftone dot) can be improved.
[0197]
According to the present invention, based on the determination result of the upper pixel determination and the lower line determination based on the determination result of the target pixel and the determination result of the peripheral pixel in the front stage in the sub-scanning direction and the rear stage in the sub-scanning direction, It is determined whether it belongs to a photographic paper photograph, a character or a halftone dot. For this reason, information on peripheral pixels before and after the target pixel is used as line information, and thus the amount of information increases. Therefore, it is possible to obtain an accurate determination result.
In addition, since the determination process is performed using the image data in the main scanning direction for the target pixel, the upper line, and the lower line, for example, the waste of the image data deteriorated in the sub-scanning direction, which is data read by the prescan, It is easy to remove the determination result of the image data in the sub-scanning direction, which is a simple information, and the determination result of the image data in the main scanning direction, which is effective information, can be extracted and applied to the type determination. .
Therefore, when determining the image area and image type of an input image, an image processing apparatus capable of accurately determining the image type by a simple method regardless of black and white or color originals and preventing an increase in circuit scale is provided. be able to.
[0198]
According to the present invention, the determination by the first determination unit is performed during the forward pass when reading the document with the image input unit, and the determination by the second determination unit is performed during the return pass when reading the document with the image input unit. The accuracy of document type discrimination can be improved by effectively utilizing the prescan return path.
[0199]
According to the present invention, instead of fixing the threshold value, the threshold value is set according to the type of each document, and only specific image data of a predetermined frequency or more is extracted when determining the document type. Therefore, the discrimination accuracy can be improved.
Further, by extracting image data of a specific area with respect to the inversion number and adjacent pixel density difference, which is a feature amount of the type of document (image), it is possible to reduce a memory area for storing image data.
[0200]
According to the present invention, the optimum threshold value is set during the forward pass when reading the document with the image input unit, and the document type is determined during the return pass when reading the document with the image input unit. By making effective use of the return path, the document type discrimination accuracy can be increased.
[0201]
According to the present invention, except for pixels having uncertain elements, only the pixels that are clear to which region belong are extracted and the document type is determined, so that the accuracy of determining the document type can be improved. Further, when determining the area of the input image data, only a specific feature amount is extracted, so that the circuit can be small.
[0202]
According to the present invention, the document discriminating unit counts the number of pixels separated into each area by the area discriminating unit, and compares the count number with a predetermined threshold value to discriminate the document. The scale can be reduced, and the document type can be determined by a simple method. Since uncertain elements are excluded, it is possible to determine the document type using a simple method, and the accuracy does not deteriorate much.
[0203]
According to the present invention, when the document type determination unit cannot determine the document (when the document type is uncertain), the display unit displays that the document type is uncertain rather than performing processing in the default image mode. The user can be prompted to enter the processing mode. Thus, even when the document type determination unit cannot determine the document, the user selects the processing mode via the input unit, so that image processing according to the original document type can be executed. Therefore, misprinting can be suppressed and wasteful use of paper can be reduced. Further, it is possible to avoid a situation in which the determination result of the document type is unknown.
[0204]
According to the present invention, since the control unit prioritizes candidate document types and displays them on the display unit, the user can refer to the priorities and easily select an image mode (processing mode). it can.
[0205]
According to the present invention, when the determination result of the document type determination unit is uncertain, the control unit stores the document type selected by the user in the storage unit. Therefore, when the output image is not preferable, the process is performed again. In addition, since the previous result can be referred to, the operation mode can be easily switched.
[0206]
According to the present invention, since the control unit stores at least a part of the processing result of the document type determination unit together with the document type in the storage unit, if the output image is not preferable, the processing result of the document type determination unit (for example, The previous image mode can be found based on the total number of pixels and the count value of each area), and can be easily redone even when the time has elapsed after the image is output.
[0207]
According to the present invention, since the image forming apparatus includes these image processing apparatuses, the accuracy of document type discrimination is improved, image processing according to the accuracy can be performed, and an image of high quality is output. Can be provided.
[0208]
According to the present invention, since the program is a program for causing a computer to execute the image processing method, the computer can read and execute the image processing method for accurately determining the type of document, which can be general purpose. it can.
[0209]
According to the present invention, since these programs are stored in a computer-executable manner, an image processing method for accurately determining the document type can be easily supplied to the computer.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating an image forming apparatus according to the present invention.
FIG. 2 is a block diagram illustrating a document type automatic determination unit in the first embodiment.
FIG. 3 is a diagram illustrating an example of image data of a halftone dot region and a density histogram created based on the image data.
4 is a diagram showing a density histogram created by using an average value of a plurality of pixels obtained from the image data shown in FIG. 3; FIG.
FIG. 5 is a diagram illustrating an example of a density histogram of a character document including a solid background.
FIG. 6 is a diagram showing an example of a density histogram of a character document composed of solid and halftone backgrounds.
FIG. 7 is a diagram showing an example of a density histogram of a photographic paper photographic document.
FIG. 8 is a block diagram showing a target pixel determination unit in the first embodiment.
FIG. 9 is a flowchart showing a processing procedure from step 1 to step 6 of image determination in the first embodiment.
FIG. 10 is a flowchart showing a processing procedure from step 7 to step 17 of image determination in the first embodiment following FIG. 9;
FIG. 11 is an explanatory diagram illustrating calculation of the number of inversions of rising pixels and falling pixels in a fixed area composed of M pixels.
FIG. 12 is an explanatory diagram showing calculation of inversion numbers of rising pixels and falling pixels in a fixed region composed of M pixels for each of n × n pixels.
FIG. 13 is a density distribution diagram of a solid image.
FIG. 14 is a density distribution diagram of a photographic paper photograph.
FIG. 15 is a density distribution diagram of a halftone image.
FIG. 16 is a density distribution diagram of a character image.
FIG. 17 is a block diagram illustrating a document type automatic determination unit according to the second embodiment.
FIG. 18 is a block diagram illustrating a pixel-of-interest determination unit in the second embodiment.
FIG. 19 is a diagram showing the relationship between the adjacent pixel density difference, the inversion number, and the frequency in a character area, a halftone dot area, and a background / photographic paper photograph area.
FIG. 20 is a diagram illustrating a relationship among adjacent pixel density difference, inversion number, frequency, and threshold value in a character area, a dot area, and a background / photographic paper photograph area.
FIG. 21 is a diagram showing areas for counting in adjacent pixel density difference, inversion number, and frequency with respect to characters and halftone dots;
FIG. 22 is a flowchart showing a processing procedure from step 21 to step 28 of image determination in the second embodiment.
FIG. 23 is a flowchart showing a processing procedure from step 29 to step 35 of image determination in the second embodiment following FIG. 22;
FIG. 24 is a block diagram illustrating a document type automatic determination unit according to a third embodiment.
FIG. 25 is a block diagram illustrating a scanning direction determination unit and a sub-scanning direction determination unit in the third embodiment.
FIG. 26 is a density distribution diagram of a base region.
FIG. 27 is a density distribution diagram of a photographic paper photograph region.
FIG. 28 is a density distribution diagram of a halftone dot region.
FIG. 29 is a density distribution diagram of a character region.
30 is a diagram showing a distribution of a character area, a character area or a halftone dot area, a halftone dot area, a photographic paper photograph area, a ground or photographic paper photograph area, and a ground area with respect to the maximum density difference and the total difference value.
31 is a diagram showing thresholds for separating each region based on FIG. 30. FIG.
FIG. 32 is a flowchart illustrating an image determination processing procedure according to the third embodiment.
FIG. 33 is a block diagram illustrating a document type automatic determination unit according to a fourth embodiment.
FIG. 34 is a block diagram illustrating a main scanning direction determination unit in the fourth embodiment.
FIG. 35 is a flowchart illustrating an image determination processing procedure according to the fourth embodiment.
FIG. 36 is a diagram showing a distribution of a character area, a halftone dot area, a photographic paper photograph area, and a background area with respect to the maximum density difference and the total difference value.
FIG. 37 is a diagram showing threshold values for separating each region including an indeterminate region based on FIG.
[Explanation of symbols]
1 Image processing device
2 Image input device
3 Image output device
4 Operation panel
10 Image forming apparatus
11 A / D converter
12 Shading correction part
13 Document type automatic discrimination section
14 Input tone correction unit
15 color correction section
16 Black generation under color removal part
17 Spatial filter processing section
18 Output tone correction unit
19 Tone reproduction processing section
20 Region separation processing unit
21 Signal converter
31 1st determination part
32 Histogram creation part
33 Maximum frequency concentration classification extraction unit
34 Threshold setting unit
35 Ground determination unit
41 Second determination unit
42 Signal processor
43 Upper line judgment part
44 Lower line judgment part
45 Target pixel determination unit
46 Signal-specific judgment section
47 Comprehensive judgment part
48 Peripheral pixel determination unit
51 Adjacent pixel density difference calculation unit
52 Rising / falling judgment part
53 Inversion number calculation unit
54 Pixel determination unit
55 Solid Pixel Determination Threshold Setting Unit
56 Rising / falling threshold setting unit
57 Inversion number threshold value setting unit
61 Comparison threshold setting unit
71 Judgment block
72 Main scanning direction determination unit
73 Sub-scanning direction determination unit
74 Signal-specific judgment section
75 Comprehensive judgment part
81 Minimum density value calculator
82 Maximum concentration value calculator
83 Maximum density difference calculator
84 Total difference calculation unit
85 Judgment area setting section
86 Character / half-tone dot decision part
87 Substrate / photographic paper judgment section
88 Maximum density difference threshold setting section
89 Sum total difference value threshold setting section
90 character judgment threshold setting part
91 Halftone determination threshold setting unit
92 photographic paper determination threshold value setting section
93 Background judgment threshold value setting unit
101 Background judgment unit
102 Photographic paper judgment unit
103 Character judgment part
104 Halftone determination unit
111 Base maximum density difference threshold setting unit
112 Background total difference value threshold value setting unit
113 photographic paper minimum density difference threshold setting section
114 Maximum density difference threshold setting section for photographic paper
115 photographic paper determination threshold value setting unit
116 character minimum density difference threshold setting section
117 Character judgment threshold setting part
118 Halftone dot total difference value threshold value setting unit
119 Halftone dot threshold setting unit
145 Attention pixel determination unit
148 Peripheral pixel determination unit
154 Pixel determination unit
172 Main scanning direction determination unit
173 Sub-scanning direction determination unit

Claims (3)

In an image processing apparatus that includes a document type determination unit that determines a document type from input image data read by scanning a document with an image input unit, and performs image processing based on the determination result.
The document type determining means includes
First determination means for determining the presence or absence of the background of the document based on density histogram data of the input image data and threshold data for determining the background of the document input as image data from the density histogram data;
A second type for determining the type of document input as image data from the determination result of the background of the document determined by the first determination unit and the type of input image obtained by determining in units of pixels of the input image data. A determination means;
With
The first determination means includes
Histogram creation means for creating a density histogram showing the relationship between the density of each pixel and its frequency from input image data;
The maximum frequency value in the density histogram is extracted as the first maximum frequency value, the density category to which the first maximum frequency value belongs is extracted as the first maximum frequency density category, and further adjacent to the first maximum frequency density category. A maximum frequency density category extracting means for extracting the second maximum power value other than the density zone and extracting the density zone to which the second maximum power value belongs as the second maximum power density zone;
Threshold setting means for setting a first threshold value for determining a solid background and a second threshold value for determining a solid and halftone background;
The first maximum power value / second maximum power value and the first / second threshold values are compared, and if the first maximum power value <the first threshold value or the first maximum power value + the second maximum power If numerical value <second threshold, it is determined that there is no background, and if first maximum numerical value ≧ first threshold or if first maximum numerical value + second maximum numerical value ≧ second threshold, it is determined that there is background. In addition, background determination means for calculating the background density,
An image processing apparatus comprising:   In an image processing apparatus that includes a document type determination unit that determines the type of a document from input image data read by scanning the document with an image input unit, and performs image processing based on the determination result.
  The document type determining means includes
  First determination means for determining presence / absence of a document background based on density histogram data of input image data and threshold data for determining a document background input as image data from the density histogram data;
  A second type for determining the type of document input as image data from the determination result of the background of the document determined by the first determination unit and the type of input image obtained by determining in units of pixels of the input image data. A determination means;
  With
  The second determination means includes
  For a target pixel having a block composed of a plurality of pixels extracted from input image data, feature amounts are extracted from the image data in the main scanning direction and the background determination result of the document determined by the first determination unit. Pixel-of-interest determination means for determining whether the pixel of interest belongs to a solid, photo, character, or halftone dot;
  For a peripheral pixel existing around the pixel of interest, a feature amount is extracted from the image data in the main scanning direction and the determination result of the background of the document determined by the first determination unit, and the peripheral pixel is solid. Peripheral pixel determination means for determining whether the region belongs to a character or a halftone dot;
  Based on the determination result of the peripheral pixel determination means for the same line existing in the previous stage of the target pixel in the sub-scanning direction, any line of the photograph, text, or halftone dot in the previous stage in the sub-scanning direction is solid Upper line determination means for determining whether it belongs to a type,
  Based on the determination result of the surrounding pixel determination means for the same line existing in the subsequent stage of the target pixel in the sub-scanning direction, whether the line existing in the subsequent stage in the sub-scanning direction belongs to a solid, photograph, character, or halftone dot type A lower line determining means for determining
  Based on the determination results of the target pixel determination unit, the upper line determination unit, and the lower line determination unit, it is determined whether the target pixel belongs to a solid, photo, character, or halftone dot type. An image processing apparatus.   In an image processing apparatus that includes a document type determination unit that determines the type of a document from input image data read by scanning the document with an image input unit, and performs image processing based on the determination result.
  The document type determining means includes
  First determination means for determining presence / absence of a document background based on density histogram data of input image data and threshold data for determining a document background input as image data from the density histogram data;
  A second type for determining the type of document input as image data from the determination result of the background of the document determined by the first determination unit and the type of input image obtained by determining in units of pixels of the input image data. A determination means;
  With
  The first determination means includes
  Histogram creation means for creating a density histogram showing the relationship between the density of each pixel and its frequency from input image data;
  The maximum frequency value in the density histogram is extracted as the first maximum frequency value, the density category to which the first maximum frequency value belongs is extracted as the first maximum frequency density category, and further adjacent to the first maximum frequency density category. A maximum frequency density category extracting means for extracting the second maximum power value other than the density zone and extracting the density zone to which the second maximum power value belongs as the second maximum power density zone;
  Threshold setting means for setting a first threshold value for discriminating from a solid background and a second threshold value for discriminating from a solid and halftone background;
  The first maximum power value / second maximum power value and the first / second threshold values are compared, and the first maximum power value <the first threshold value or the first maximum power value + the second maximum power If numerical value <second threshold, it is determined that there is no background, and if first maximum numerical value ≧ first threshold or if first maximum numerical value + second maximum numerical value ≧ second threshold, it is determined that there is background. In addition, background determination means for calculating the background density,
  Have
  The second determination means includes
  For a target pixel having a block composed of a plurality of pixels extracted from input image data, feature amounts are extracted from the image data in the main scanning direction and the background determination result of the document determined by the first determination unit. Pixel-of-interest determination means for determining whether the pixel of interest belongs to a solid, photo, character, or halftone dot;
  For a peripheral pixel existing around the pixel of interest, a feature amount is extracted from the image data in the main scanning direction and the determination result of the background of the document determined by the first determination unit, and the peripheral pixel is solid. Peripheral pixel determination means for determining whether the region belongs to a character or a halftone dot;
  Whether the line existing in the preceding stage in the sub-scanning direction belongs to the solid, photograph, character, or halftone dot type based on the determination result of the surrounding pixel determining means for the same line existing in the preceding stage in the sub-scanning direction Upper line determination means for determining
  Based on the determination result of the surrounding pixel determination means for the same line existing in the subsequent stage of the target pixel in the sub-scanning direction, whether the line existing in the subsequent stage in the sub-scanning direction belongs to solid, photograph, character, or halftone dot A lower line determining means for determining
  Based on the determination results of the target pixel determination unit, the upper line determination unit, and the lower line determination unit, a determination is made as to whether the target pixel belongs to a solid, photo, character, or halftone dot type. An image processing apparatus characterized by that.

Images