JP6263951B2 - Image processing apparatus and program - Google Patents

Description

  The present invention relates to an image processing apparatus, an image reading apparatus, an image forming apparatus, and a program.

  2. Description of the Related Art Conventionally, an apparatus that binarizes an image obtained by reading with a scanner or the like to obtain binary image information is known.

In US Pat. No. 6,057,033, scanned color image data is obtained from at least two color channels, at least one region of interest containing foreground content and background content is identified in the scanned color image data, and at least one threshold value is identified. The data values are obtained according to image attributes that differ between the foreground content and the background content within the region of interest, and the scanned color image data of the document is determined according to at least one threshold data value obtained from the region of interest. A method for obtaining bi-tonal image data from a document converted to data is disclosed.
Patent Document 2 discloses a plurality of edge portions that detect edge information of an image that has passed through filters of each color, a logical sum circuit that performs a logical sum of outputs from the edge detection portion, and an image obtained by converting a color image into black and white. A detecting unit for detecting, an exclusive OR circuit for obtaining an exclusive OR of the output of the OR circuit and the output of the edge detecting unit, and a multi-value binary converting unit for binarizing an image obtained by converting a color image into black and white And an exclusive OR circuit that takes the exclusive OR of the outputs of the XOR and the exclusive OR circuit is disclosed.

JP 2009-535899 A Japanese Patent Laid-Open No. 11-205617

  In the present invention, it is desirable to obtain a binary image with a clearer boundary between characters and background when binarizing input image information.

The invention according to claim 1 is based on an input image information acquisition unit that acquires input image information that is image information in a predetermined color space, and the input image information acquired by the input image information acquisition unit. A target pixel selection unit that selects a target pixel, a local region setting unit that sets an image region in a predetermined range including the target pixel selected by the target pixel selection unit as a local region, and the local region setting unit An average color value calculation unit that calculates an average color value that is an average value of color values of pixels included in the local region set by the average color value, and the average color value calculated by the average color value calculation unit and the target pixel A color displacement calculation unit that calculates a color displacement in which a difference from the color value is represented by a vector quantity in the color space; and binary image information of the pixel of interest based on the color displacement calculated by the color displacement calculation unit Includes a binarization decision unit that, the, the local region setting unit sets a plurality of local regions having different sizes of the image area, the binarization decision unit, the size of the color displacement An image characterized in that the image information of the pixel of interest is binarized based on the color displacement calculated for the smallest image region of the plurality of local regions that is equal to or greater than a predetermined value. It is a processing device.
The invention according to claim 2 is based on the input image information acquisition unit that acquires input image information that is image information in a predetermined color space, and the input image information acquired by the input image information acquisition unit. A target pixel selection unit that selects a target pixel, a local region setting unit that sets an image region in a predetermined range including the target pixel selected by the target pixel selection unit as a local region, and the local region setting unit An average color value calculation unit that calculates an average color value that is an average value of color values of pixels included in the local region set by the average color value, and the average color value calculated by the average color value calculation unit and the target pixel A color displacement calculation unit that calculates a color displacement in which a difference from the color value is represented by a vector quantity in the color space; and binary image information of the pixel of interest based on the color displacement calculated by the color displacement calculation unit A binarization determining unit that sets a plurality of local regions having different sizes of the image region, and the binarization determining unit has a similar color displacement direction. The image processing apparatus is characterized in that the image information of the pixel of interest is binarized based on the color displacement calculated for the largest image region from among the local regions.
The invention according to claim 3, wherein the binarization decision unit according to claim 1 or 2, characterized in that binarizing the image information of the pixel of interest based on the size and / or direction of the color displacement The image processing apparatus described in the above.
According to a fourth aspect of the present invention, the binarization determination unit selects a specific color axis from the color displacement calculated by the color displacement calculation unit, and the attention is based on the color displacement with respect to the specific color axis. the image processing apparatus according to any one of claims 1 to 3, characterized in that binarizing the image information of the pixel.
According to a fifth aspect of the present invention, the computer has a function of acquiring input image information that is image information in a predetermined color space, a function of selecting a target pixel based on the acquired input image information, A function of setting an image area in a predetermined range including the selected pixel of interest as a local area, and a function of calculating an average color value that is an average value of color values of pixels included in the set local area And a function of calculating a color displacement in which a difference between the calculated average color value and the color value of the target pixel is represented by a vector amount in the color space, and an image of the target pixel based on the calculated color displacement A function for binarizing information, and the function for setting sets a plurality of the local areas having different sizes of the image area, and the function for binarizing is a magnitude of the color displacement. Is a predetermined value Is a program for binarizing the image information of the pixel of interest based on the color displacement calculated for the smallest image area from the plurality of local regions to be the upper.
According to a sixth aspect of the present invention, a computer has a function of acquiring input image information that is image information in a predetermined color space, a function of selecting a target pixel based on the acquired input image information, A function of setting an image area in a predetermined range including the selected pixel of interest as a local area, and a function of calculating an average color value that is an average value of color values of pixels included in the set local area And a function of calculating a color displacement in which a difference between the calculated average color value and the color value of the target pixel is represented by a vector amount in the color space, and an image of the target pixel based on the calculated color displacement A function for binarizing information, and the function for setting sets a plurality of the local areas having different sizes of the image area, and the function for binarizing has a direction of the color displacement. Similar local regions Program for binarizing the image information of the pixel of interest based on the color displacement calculated for the largest image area from the

According to the first aspect of the present invention, it is possible to provide an image processing apparatus that can obtain a binary image with a clearer boundary between the character and the background than when the present configuration is not adopted. Further, the accuracy of the binarization process is further improved as compared with the case where this configuration is not adopted.
According to the second aspect of the present invention, it is possible to provide an image processing apparatus capable of obtaining a binary image with a clearer boundary between the character and the background than when this configuration is not employed. Further, the accuracy of the binarization process is further improved as compared with the case where this configuration is not adopted.
According to the third aspect of the present invention, the binarization processing of the target pixel becomes easier as compared with the case where this configuration is not adopted.
According to the fourth aspect of the present invention, the processing time required for the binarization process is shorter than when this configuration is not adopted.
According to the fifth aspect of the present invention, a process for obtaining a binary image with a clearer boundary between the character and the background can be realized by a computer as compared with the case where this configuration is not adopted. Further, the accuracy of the binarization process is further improved as compared with the case where this configuration is not adopted.
According to the sixth aspect of the present invention, a process for obtaining a binary image with a clearer boundary between the character and the background can be realized by a computer as compared with the case where this configuration is not adopted. Further, the accuracy of the binarization process is further improved as compared with the case where this configuration is not adopted.

It is the block diagram explaining the functional structure of the image reading apparatus with which the image processing apparatus of this Embodiment is applied. It is a block diagram explaining the functional structure of the image forming apparatus to which the image processing apparatus of this Embodiment is applied. It is the block diagram explaining the functional structure of the binarization process part of this Embodiment. It is the flowchart explaining operation | movement of the binarization process part of this Embodiment. It is the figure which demonstrated an example of the method in which the binarization determination part binarizes the image information of a focused pixel based on a color displacement. (A)-(d) is the figure explaining the position of the attention pixel in a local region. (A)-(c) is the figure explaining the change of the color displacement when the several local area | region from which the magnitude | size of an image area differs is set. (A)-(b) is the figure which compared the binary image binarized by this Embodiment with the binary image binarized by the prior art.

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

<Description of Image Reading Device>
FIG. 1 is a block diagram illustrating a functional configuration of an image reading apparatus 1 to which the image processing apparatus of the present embodiment is applied.
An image reading apparatus 1 shown in the figure binarizes image information obtained by reading an image of a document to create binary image information. The image reading apparatus 1 performs a predetermined process using the binary image information in the post-processing unit.

  FIG. 1 shows a control unit 10, an image reading unit 20, a signal processing unit 30, and a post-processing unit 40 in the image reading apparatus 1.

  The control unit 10 controls each unit in the image reading operation of the image reading apparatus 1. In the illustrated example, the control unit 10 controls the image reading unit 20, the signal processing unit 30, and the post-processing unit 40.

The image reading unit 20 is an example of an image reading unit that reads an image of a document. The image reading unit 20 includes a drive motor 21, a light emitting unit 22, and a light receiving unit 23.
The light emitting unit 22 is configured by arranging a plurality of light emitting elements (for example, LEDs (Light Emitting Diodes)) side by side in the main scanning direction. The light receiving unit 23 may be a CCD type that receives light with a CCD (Charge Coupled Devices), or a CIS type that receives reflected light with respect to the light sequentially irradiated from the light emitting unit 22 to the document with a CIS (Contact Image Sensor). it can.

  The drive motor 21 moves the image reading unit 20 in the sub-scanning direction and the opposite direction. Then, the light emitting unit 22 irradiates the original with light with a predetermined amount of light on the original. Further, the light receiving unit 23 transmits the R signal, the G signal, and the B signal, which are image data generated by receiving the light reflected from the document, to the signal processing unit 30.

  The signal processing unit 30 includes a sample hold unit 31, a gain adjustment unit 32, an offset adjustment unit 33, an A / D conversion unit 34, a shading correction unit 35, and a binarization processing unit 36.

The sample hold unit 31 samples the R signal, G signal, and B signal, which are analog image signals output from the light receiving unit 23.
The gain adjusting unit 32 adjusts the voltage output from the light receiving unit 23 when the light emitting unit 22 is lit so that the voltage is close to the maximum input voltage of the A / D conversion unit 34.
The offset adjustment unit 33 adjusts the voltage output from the light receiving unit 23 when the light emitting unit 22 is turned off to a value close to the minimum input voltage of the A / D conversion unit 34.

The A / D conversion unit 34 converts the analog image signal output from the offset adjustment unit 33 into a digital image signal.
The shading correction unit 35 corrects the reading density difference in the main scanning direction on the digital image signal output from the A / D conversion unit 34 using the shading data.
The binarization processing unit 36 functions as an image processing unit that performs processing for converting image information of a document read by the image reading unit 20 from image information in a predetermined color space to binary image information. Here, the RGB signal, which is image information in the RGB color space, is binarized to perform binary image information processing. The binarization processing unit 36 will be described in more detail later.

  Each function performed by the signal processing unit 30 can be realized by providing an application specific integrated circuit (ASIC) or the like.

The post-processing unit 40 performs processing using the binary image information created by the binarization processing unit 36.
The post-processing unit 40 is, for example, an image information transmission unit for directly transmitting binary image information for forming an image in an image forming unit 70 (see FIG. 2) of the image forming apparatus 2 described later. As a result, the image forming apparatus 2 forms a binary image. In this case, the image reading apparatus 1 can also be regarded as a scanner unit that reads an image of a document in an image forming unit that reads the image of the document and forms an image based on the obtained binary image information.
The post-processing unit 40 is a functional unit that performs facsimile transmission, for example. That is, the binary image information is transmitted to a facsimile apparatus as an external device via a public telephone line, a LAN (Local Area Network), a network such as the Internet. In this case, the image reading apparatus 1 can also be regarded as a facsimile apparatus.
Further, the post-processing unit 40 is a functional unit that performs document archiving that stores image information of a document, for example. That is, the obtained binary image information is stored in an HDD (Hard Disk Drive, etc.), and necessary document information is downloaded from the HDD by an operation from another personal computer, portable terminal, etc., and the personal computer, portable terminal, etc. In this case, the image reading apparatus 1 can be regarded as a document archive system.
Further, the post-processing unit 40 is a functional unit that performs character recognition, for example. That is, based on the obtained binary image information, a character included in the binary image information is specified by collation with a previously stored pattern. In this case, the image reading apparatus 1 can also be regarded as an OCR (Optical Character Reader).

<Description of Image Forming Apparatus>
FIG. 2 is a block diagram illustrating a functional configuration of the image forming apparatus 2 to which the image processing apparatus of the present embodiment is applied.
An image forming apparatus 2 shown in the figure binarizes image information sent from an external device such as a personal computer or a portable terminal, and creates binary image information. In the image forming apparatus 2, a binary image is formed in the image forming unit based on the obtained binary image information.

  As illustrated, the image forming apparatus 2 includes a communication I / F 50, a control unit 60, and an image forming unit 70.

  The communication I / F 50 transmits and receives various types of information to and from a personal computer or portable terminal that is an external device via a network.

  The control unit 60 receives a print data and converts it into a page description language (PDL), a PDL generation unit 61, and a rasterize unit 62 that generates a raster image from the PDL generated by the PDL generation unit 61. The color conversion processing unit 63 that converts RGB data into YMCK data, the raster image adjustment unit 64 that adjusts the raster image converted by the color conversion processing unit 63, and binarizes the YMCK data into binary image information. And a binarization processing unit 65.

  In the present embodiment, first, the PDL generation unit 61 receives print data from the PC. This image data is image information that a user using a personal computer or a portable terminal wants to print using the image forming apparatus 2. The PDL generation unit 61 that has received the image information converts it into code data described in PDL and outputs it.

  The rasterizing unit 62 converts the code data described in PDL output from the PDL generating unit 61 into raster data for each pixel, and forms a raster image. The rasterizing unit 62 outputs the converted raster data as RGB (Red, Green, Blue) video data (RGB video data). At this time, the rasterizing unit 62 outputs RGB data for each page.

The color conversion processing unit 63 converts the RGB data input from the rasterizing unit 62 into device-independent color values such as [XYZ], [L ^* a ^* b ^* ], and [L ^* u ^* v ^* ]. After that, it is converted into YMCK data that is a reproduction color (yellow, magenta, cyan, black) of the image forming apparatus 2 and output. This YMCK data is composed of Y color data, M color data, C color data, and K color data separated for each color, and can be regarded as an example of image information here.

  The raster image adjustment unit 64 performs γ conversion, definition processing, halftone processing, and the like on the YMCK data input from the color conversion processing unit 63 so that a better image quality can be obtained by the image forming apparatus 2. Make various adjustments.

The binarization processing unit 65 is an example of an image processing unit that performs processing for converting image information in a predetermined color space into binary image information. Here, YMCK data, which is image information in the YMCK color space, is binarized to perform binary image information.
The binarization processing unit 65 will be described in more detail later.

  The image forming unit 70 is an example of an image forming unit, and forms an image on a sheet or the like as a recording medium based on the binary image information generated by the binarization processing unit 65. The image forming unit 70 is, for example, a printer, and is an electrophotographic system that forms an image by transferring toner attached to a photoreceptor to a recording medium, or an ink jet system that forms an image by ejecting ink onto the recording medium. Things can be used.

Heretofore, when the image information is binarized in the image reading apparatus 1 or the image forming apparatus 2, there are the following problems.
(1) When the background of image information is determined, and binarization is performed with, for example, characters set to ON and background set to OFF, the correction accuracy depends on the determination accuracy.
(2) When binarization is performed by superimposing edge information or correcting adjacent neighboring areas, for example, in the case of a small character, it is difficult to obtain a clear boundary between the character and the background, and the line width is In the case of wide characters, omission (white omission) may occur near the center of the line.
(3) When performing edge detection, frequency change analysis, histogram determination, etc., the processing time becomes long.

<Description of functional configuration of binarization processing unit>
In the present embodiment, in order to suppress the above problems, the above-described binarization processing unit that binarizes image information has the following configuration. In the following description, the binarization processing unit 65 in FIG. 2 will be described as an example.

FIG. 3 is a block diagram illustrating a functional configuration of the binarization processing unit 65 of the present embodiment.
The binarization processing unit 65 of this embodiment includes an input image information acquisition unit 651, a target pixel selection unit 652, a local region setting unit 653, a local region size storage unit 654, an average color value calculation unit 655, A color displacement calculation unit 656, a binarization determination unit 657, and a binary image information output unit 658.

  The input image information acquisition unit 651 acquires input image information that is image information in a predetermined color space. In the present embodiment, the input image information acquisition unit 651 acquires image information in the YMCK color space from the raster image adjustment unit 64 (see FIG. 2) of the control unit 60 as input image information.

  The target pixel selection unit 652 selects a target pixel based on the input image information acquired by the input image information acquisition unit 651. As will be described in detail later, the target pixel is one pixel selected from the pixels constituting the input image. The pixels constituting the input image are sequentially selected.

  The local region setting unit 653 sets an image region in a predetermined range including the target pixel selected by the target pixel selection unit 652 as a local region. As will be described in detail later, the local region is an image region in a predetermined range in the input image, and the pixel of interest is located, for example, near the center thereof.

  The local area size storage unit 654 stores the size of the local area. The local area is, for example, a square. In this case, the local area size storage unit 654 stores the size of the square. As will be described in detail later, there are cases where a plurality of local areas having different sizes are prepared. In this case, the local area size storage unit 654 stores the size of each local area. The shape of the local region is not limited to a square, but may be other shapes such as a rectangle, a hexagon, and a circle.

  The average color value calculation unit 655 calculates an average color value that is an average value of the color values of pixels included in the local region set by the local region setting unit 653. That is, when the color space is the YMCK color space, the average color value calculation unit 655 calculates the total value of the Y, M, C, and K color values for all the pixels included in the local area. And the average value is calculated by dividing by the number of pixels. The average value is set as the average color value for each of the Y, M, C, and K colors.

  The color displacement calculation unit 656 calculates a color displacement represented by a vector amount in a color space using the difference between the average color value calculated by the average color value calculation unit 655 and the color value of the target pixel. That is, in the color space to be used, it is possible to consider a vector quantity starting from the average color value and ending at the color value of the pixel of interest. This vector quantity has a magnitude and direction. Here, the vector amount is a color displacement, and the color displacement calculation unit 656 calculates this color displacement.

  The binarization determination unit 657 binarizes the image information of the target pixel based on the color displacement calculated by the color displacement calculation unit 656. As will be described in detail later, the binarization determination unit 657 binarizes the image information of the target pixel based on the magnitude and / or direction of the color displacement that is the vector amount.

  The binary image information output unit 658 outputs the binary image information generated by the binarization determination unit 657 to the image forming unit 70 (see FIG. 2).

<Description of operation of binarization processing unit>
FIG. 4 is a flowchart for explaining the operation of the binarization processing unit 65 of the present embodiment.
Hereinafter, the operation of the binarization processing unit of the present embodiment will be described with reference to FIGS. 3 and 4.

  First, the input image information acquisition unit 651 acquires input image information (step 101). In the present embodiment, the input image information acquisition unit 651 acquires YMCK data in the YMCK color space as input image information.

  Next, the pixel-of-interest selection unit 652 selects a pixel of interest based on the input image information (step 102). Here, the target pixel is one pixel selected from the input image and is a pixel to be binarized. The target pixel is selected from, for example, a pixel located in the first column of the first row in the input image, and the pixels in the first row are selected in order. Further, when all the pixels in one row have been selected, the pixels in the next row are similarly selected in order. This is continued until the pixel in the last column of the last row is selected.

  Then, the local area setting unit 653 sets an image area in a predetermined range including the target pixel as a local area (step 103). The size of the local area is determined by the local area setting unit 653 with reference to the local area size storage unit 654. The local area is an image area obtained by cutting out a predetermined range in the input image. And it is preferable to set so that the pixel of interest is located near the center of the local region. Although details will be described later, a plurality of local areas may be selected at this time.

  Next, the average color value calculation unit 655 calculates an average color value that is an average value of the color values of the pixels included in the local region (step 104). In this embodiment, since the color space to be used is the YMCK color space, the average color value of the color values of the pixels included in the local area for each of the four colors of Y, M, C, and K colors. Is calculated. When a local region having a plurality of sizes is selected by the local region setting unit 653, an average color value is further calculated for each local region having a different size. For example, when three types of local regions are selected, the average color value calculated by the average color value calculation unit 655 is 4 colors × 3 types = 12.

Next, the color displacement calculation unit 656 calculates a color displacement represented by a vector quantity in a color space using the difference between the average color value and the color value of the target pixel (step 105). The vector quantity at this time is
In the color space to be used, this is a vector quantity starting from the average color value and ending at the color value of the pixel of interest. For example, the average color values of Y color, M color, C color, and K color are Yave, Mave, Cave, and Kave, and the color values of Y color, M color, C color, and K color of the pixel of interest are It is assumed that they are C0, M0, Y0, K0. At this time, as a coordinate in the YMCK color space, a vector amount starting from (Yave, Mave, Cave, Kave) and ending at (Y0, M0, C0, K0) is a color displacement. If the coordinates (Yave, Mave, Cave, Kave) in the YMCK color space are point A and the coordinates (Y0, M0, C0, K0) are point B, the AB vector is the color displacement.
At this time, the magnitude of the color displacement is the distance between A (Yave, Mave, Cave, Kave) and B (Y0, M0, C0, K0), and the direction of the color displacement is A (Yave, Mave, Cave, The direction is from Kave) to B (Y0, M0, C0, K0).

Then, the binarization determination unit 657 binarizes the image information of the target pixel based on the color displacement (step 106).
Further, the pixel-of-interest selecting unit 652 determines whether or not there are further pixels of interest to be selected (step 107).
If there is a target pixel to be selected (Yes in Step 107), the process returns to Step 102. On the other hand, when there is no target pixel to be selected (No in Step 107), the binary image information output unit 658 outputs the generated binary image information to the image forming unit 70 (Step 108).

<Description of method for binarizing image information of target pixel>
There are various methods for the binarization determining unit 657 to binarize the image information of the target pixel based on the color displacement.

FIG. 5 is a diagram illustrating an example of a method in which the binarization determination unit 657 binarizes image information of a target pixel based on color displacement.
FIG. 5 conceptually illustrates color displacement in the YMCK color space.
In the figure, the color displacement is shown as an AB vector. At this time, when the magnitude of the color displacement (the magnitude of the AB vector) is larger than this threshold value, the binarization determining unit 657 turns the binary image information of the target pixel ON (black). On the other hand, when the magnitude of the color displacement is equal to or smaller than the threshold value, the binarization determination unit 657 sets the binary image information of the target pixel to OFF (white).
That is, in the example described with reference to FIG. 5, it can be said that the binarization determination unit 657 binarizes the image information of the pixel of interest based on the magnitude of the color displacement.

Further, the binarization determination unit 657 may binarize the image information of the target pixel in consideration of the color displacement direction in addition to the color displacement magnitude.
In this case, for example, it is conceivable to consider a direction along the lightness axis of the color displacement (lightness axis direction). That is, when the magnitude of the color displacement is larger than the threshold and the direction of the color displacement is a direction in which the brightness is reduced, as described above, the binary image information of the target pixel is set to ON (black). That is, in the input image, characters are usually drawn as an image having a low brightness. Therefore, by doing in this way, the character part can be turned ON (black) when the image information is binarized. On the other hand, when the magnitude of the color displacement is larger than the threshold value and the direction of the color displacement is a direction in which the brightness increases, the binary image information of the target pixel is set to OFF (white). The lightness axis direction of the color displacement can be derived by taking the inner product of the lightness vector and the color displacement.

  The direction of color displacement is not limited to the lightness axis direction, and other color axes may be employed. For example, when the inner product of the brightness vector and the color displacement does not reach a predetermined value, the direction along the Y color axis (Y color axis direction) is considered as the other color axis, and the direction of the color displacement When the Y value is in the increasing direction, the binary image information of the target pixel is set to ON (black). When the Y value is in the decreasing direction, the binary image information of the target pixel is set to OFF (white). Also good. A direction along the C color axis (C color axis direction) and a direction along the M color axis (M color axis direction) may be employed.

Further, the binarization determination unit 657 selects a specific color axis from the color displacement calculated by the color displacement calculation unit 656, and binarizes the image information of the target pixel based on the color displacement with respect to the specific color axis. Good.
That is, in the example of FIG. 5, the size of the AB vector is not calculated, but, for example, the Y color axis is selected as the color axis. Then, the value of Y0−Yave, which is the difference between the average color value for Y color and the color value of the target pixel, is compared with a predetermined threshold value. When the value of Y0-Yave is larger than the threshold value, the binarization determining unit 657 sets the binary image information of the target pixel to ON (black). On the other hand, when the magnitude of the color displacement is equal to or smaller than the threshold value, the binarization determination unit 657 sets the binary image information of the target pixel to OFF (white).
This method has the advantage that less processing is performed in the binarization determination unit 657 than the size of the AB vector is obtained.

  At this time, there is a problem of which color axis to select from among Y color, M color, C color, and K color, but of each color axis, the difference between the average color value and the color value of the target pixel is the largest. A color axis may be selected. That is, the values of Y0-Yave, M0-Mave, C0-Cave, and K0-Kave may be calculated, and the maximum value may be selected from these color value differences. Alternatively, YMCK having the largest color value of the target pixel may be selected. That is, the color axis for the largest one of Y0, M0, C0, and K0 is selected.

  When a specific color axis is selected in this way and the image information of the pixel of interest is binarized based on the color displacement about the specific color axis, the binarization determining unit 657 selects the AB vector that is the color displacement. In other words, the image information of the pixel of interest can be binarized based on the size when a specific color axis is projected.

In the above-described example, the color axes for the Y color, the M color, the C color, and the K color are considered. However, the present invention is not limited to this, and other color axes may be used. For example, the lightness axis described above or the saturation axis representing saturation may be selected.
In selecting the color axis, the color displacement of a pixel located in the vicinity of the target pixel may be referred to. In other words, by selecting a color axis that is the same as the color axis that is selected when the surrounding pixel is the pixel of interest, the color axis selection does not vary greatly, and determination accuracy when binarization is performed Will be improved.

FIGS. 6A to 6D are diagrams illustrating the position of the target pixel in the local region.
FIG. 6A shows a case where the local region R is only a background that does not include characters, and the pixel of interest Q is also a background pixel. FIG. 6B shows a case where the local region R includes a part of characters, but the target pixel Q is a background pixel. Further, FIG. 6C shows a case where the local region R includes a part of a character and the target pixel Q is a character pixel. Further, FIG. 6D shows a case where the local region R is only a character that does not include a background, and the pixel of interest Q is also a character pixel.

  In this case, in the case of FIG. 6A, if the local region R is an appropriate size (not so large), the background is formed by one color, and therefore the average color value and the color value of the pixel of interest Q There is almost no difference. Therefore, in this case, if the threshold value is set to an appropriate value, the difference between the color values becomes equal to or less than the threshold value. Therefore, the target pixel Q is determined as a background pixel, and the binary image information can be turned off (white).

  When a character is included in the local region R as shown in FIG. 6B, the difference between the average color value and the color value of the pixel of interest Q increases as the region including the character increases. Go. That is, the average color value is affected by the character color value, and the average color value is between the background color value and the character color value. As a result, a difference from the color value of the target pixel Q indicating the background color value occurs. However, if the threshold value is set to an appropriate value, in the case of FIG. 6B, the color value difference is equal to or smaller than the threshold value, the target pixel Q is determined to be a background pixel, and the binary image information is OFF (white). Become.

  On the other hand, in the case of FIG. 6C, the color value of the pixel of interest Q is the color value of the character. The average color value is between the background color value and the character color value, as in FIG. In such a case, since the difference between the average color value and the color value of the target pixel Q is generally large, if the threshold is set to an appropriate value, the difference between the average color value and the color value of the target pixel Q is It becomes larger than the threshold value. Accordingly, the target pixel Q is determined as a character pixel, and the binary image information is turned ON (black).

  Therefore, in the case of FIGS. 6A to 6C, if the threshold value is set to an appropriate value, it can be determined whether the target pixel Q is a background pixel or a character pixel. In this case, the background portion is OFF (white) and the character portion is ON (black). According to this method, it is possible to distinguish between the case of FIG. 6B and the case of FIG.

  That is, in this embodiment, first, it is assumed that the input image is locally composed of one color or two colors. If the input image is composed of one color, the color displacement is small, but if the input image is composed of two colors, the fact that the color displacement is large is utilized. That is, a threshold is set for the magnitude of color displacement, and the pixel of interest Q is a pixel constituting a character or a pixel constituting a background depending on whether the magnitude of the color displacement exceeds this threshold or is less than or equal to the threshold. Judging whether there is. When the target pixel Q is determined to be a pixel constituting a character, the binary image information is set to ON (black), for example. When the target pixel Q is determined to be a pixel constituting a background, the binary image information is For example, it is set to OFF (white).

  However, in the case of FIG. 6D, the difference between the color values of the average color value and the color value of the target pixel Q is almost the same as in FIG. 6A. This pixel is determined. As a result, the binary image information is turned off (white), and erroneous determination occurs. In this case, for example, a phenomenon occurs in which an omission (white omission) occurs near the center of a character line.

  Therefore, in the present embodiment, the binarization determination unit 657 performs determination based on the color displacement of the local region R having an appropriate size when binarizing the image information of the target pixel Q. And it prevents that the local area | region R at the time of determining becomes a case like FIG.6 (d), and avoids a misjudgment.

FIGS. 7A to 7C are diagrams illustrating changes in color displacement when a plurality of local regions R having different image region sizes are set.
FIG. 7A shows a case where three types (small, medium, large) having different sizes are selected as the local region R. FIG. 7B conceptually shows the color displacement in each local region R shown in FIG. 7A for the target pixel Q1. Further, FIG. 7C conceptually shows the color displacement in each local region R shown in FIG. 7A for the target pixel Q2.

  In the case of the pixel of interest Q1 shown in FIG. 7B, when the local region R is “small”, the magnitude of the color displacement is small and the direction is upward in the drawing. When the local region R is “medium”, the magnitude of the color displacement is large and the direction is substantially upward. Further, when the local region R is “large”, the magnitude of the color displacement is the same as when the local region R is “medium”, and the direction is the upper left direction in the figure.

  In the case of the pixel of interest Q2 shown in FIG. 7C, when the local region R is “small”, the magnitude of the color displacement is large and the direction is in the upper right direction in the figure. When the local region R is “medium”, the magnitude and direction of the color displacement is the same as when the local region R is “small”. Further, when the local region R is “large”, the magnitude of the color displacement is the same as when the local region R is “medium”, and the direction is substantially upward in the figure.

  At this time, in this embodiment, the binarization determination unit 657 selects a local region R according to the following criteria (I) and (II), for example, and based on the color displacement calculated for the local region R: The image information of the target pixel Q is binarized.

(I) The thing of the minimum image area is selected from the several local area | region R from which the magnitude | size of a color displacement becomes more than a predetermined value.
(II) From the local region R having the similar color displacement direction, the one with the largest image region is selected.

  According to the reference (I), in the case of the target pixel Q1, if the local region R is “medium” or “large” and the magnitude of the color displacement is greater than or equal to a predetermined value, the minimum value is determined from these The “medium” local region R is selected as the image region. In the case of the pixel of interest Q2, if the local region R is all of “small”, “medium”, and “large” and the magnitude of the color displacement is greater than or equal to a predetermined value, As the object, the “small” local region R is selected.

  That is, in the case of the pixel of interest Q1, when the local region R is “small”, the magnitude of the color displacement is small, which corresponds to the case of FIG. 6A, and the local region R is “medium” and “large”. This case corresponds to the case of FIG. 6B or 6C because the magnitude of the color displacement becomes large. Therefore, in this case, the local region R may be selected as “medium” or “large”. However, when the local region R is “large”, the direction of color displacement is changed with respect to the local region R being “small” and “medium”. This is considered to be due to the fact that the local region R becomes larger, so that the color constituting the background is mixed with other colors instead of one color. The color constituting the background is more preferably one color (that is, the local region R is composed of two colors in combination with the character color) from the viewpoint of determination accuracy when binarizing. Therefore, when there are a plurality of candidates as the local region R in this way, a smaller image region R is selected in order to suppress mixing of other colors. Therefore, in this case, the local region R is selected as “medium”.

  In the case of the pixel of interest Q2, it means that the local region R is all “small”, “medium”, and “large” and corresponds to the case of FIG. 6B or FIG. 6C. Of these, the one with the local region R “small” is selected as the smallest image region.

  Further, according to the standard (II), in the case of the pixel of interest Q1, the direction of color displacement is similar when the local region R is “small” and “medium”. The “medium” local region R is selected. In the case of the target pixel Q2, the direction of color displacement is similar when the local region R is “small” and “medium”, but the “middle” local region R is the largest image region R among them. Select.

  That is, as described above, if the direction of the color displacement is substantially the same, there is a high possibility that the color constituting the background is one color (that is, the local region R is composed of two colors together with the character color). Therefore, a local region R having a similar color displacement direction is first selected from the viewpoint of determination accuracy when binarizing. The larger local region R is more likely to correspond to the case of FIG. 6B or FIG. Therefore, the “medium” local region R is selected in both the target pixels Q1 and Q2.

Although the two criteria (I) and (II) have been described above, either of these criteria may be adopted, or both may be used in combination.
By setting a plurality of local areas R having different image area sizes and adopting the references (I) and (II), the local area R as shown in FIG. It is possible to select a local region R as shown in 6 (b) and FIG. 6 (c).
As for the size of the local region R, the determination accuracy when binarizing is larger is higher unless another color is mixed. Therefore, when the largest local region R is selected, a larger local region R may be added and set, and the criteria (I) and (II) may be applied.

According to the present embodiment, it is difficult to depend on the accuracy of background determination, and a binary image with a clearer boundary between characters and the background can be obtained.
Further, according to the present embodiment, it is possible to obtain a binary image that is less dependent on the size of the character and the combination of the character and the background color and has a clearer boundary between the character and the background.
Furthermore, even in the case of an input image in which characters are present on a gentle pattern, by setting a local area of an appropriate size, the image in the local area can be configured with one or two colors. . Therefore, even in such a case, it is possible to obtain a binary image in which the boundary between the character and the background is clearer.
Since edge detection, frequency change analysis, histogram determination, etc. need not be performed, the processing is light and the processing time of the binarization processing can be shortened.
In addition, by setting a local area of an appropriate size, it is difficult for missing (whiteout) to occur near the center of a line even in the case of a character with a wide line width.

FIGS. 8A to 8B are diagrams comparing a binary image binarized by the present embodiment and a binary image binarized by the conventional technique. Here, the result of binarization processing is shown for an input image in which the background is blue and the character “long” portion is green. Moreover, as a prior art, the case where the edge of a character is extracted and the boundary of a character and a background is reproduced by superimposing the extracted edge is shown.
As shown in FIG. 8A, in the binarization process according to the present embodiment, it can be seen that the character portion indicated by “long” is separated from the background with a clear boundary. On the other hand, as shown in FIG. 8B, in the binarization processing according to the prior art, the boundary between the character portion indicated by “long” and the background is unclear, and further, a “long” character is formed. There is a dropout (white dropout) near the center of the line.

In the example described in detail above, when the input image information is binarized, white (white pixel) is set to ON and black (black pixel) is set to OFF. However, the present invention is not limited to this. Other colors may be set.
In the example described in detail above, the case where the YMCK color space is used has been described. However, the present invention is not limited to this, and other color spaces such as an RGB color space may be used.
In the example described in detail above, the difference between the average color value, which is the color displacement, and the color value of the pixel of interest is simply considered as a difference in color value. However, the present invention is not limited to this. A change rate, a differential value, or the like when the color value of the pixel is compared may be obtained and used as a difference between the average color value and the color value of the target pixel.
Further, in the example described in detail above, the average color value is derived by calculating a simple average value of the color values of the pixels included in the local region, but is not limited to this, and is not limited to each pixel. You may give weight.

  In the example described above in detail, a threshold value is set, and a fixed binarization method for determining ON / OFF of the pixel of interest depending on whether the magnitude of the color displacement exceeds the threshold value is used. It is not limited. For example, for each target pixel, for example, a floating binarization method may be adopted in which the threshold value is adjusted according to the density of the peripheral region to determine ON / OFF of the target pixel. In this case, the accuracy of the binarization process depends on the size of the local area. Therefore, it is preferable that the size of the local region is variable and the size of the local region is determined according to the color displacement.

  In the example described above in detail, the binarization processing unit 36 (see FIG. 1) is a part of the image reading apparatus 1 (see FIG. 1), and the binarization processing unit 65 (see FIG. 2) Although it was a part of the forming apparatus 2 (see FIG. 2), the present invention is not limited to this and can be used alone as an image processing apparatus.

<Description of the program>
The processing performed by the binarization processing units 36 and 65 is realized by cooperation of software and hardware resources. That is, a program (not shown) in the control computer provided in the signal processing unit 30 (see FIG. 1) or the control unit 60 (see FIG. 2) implements a program for realizing the functions of the binarization processing units 36 and 65. Each of these functions is realized by executing and performing the control described above.

  Accordingly, the processing performed by the binarization processing units 36 and 65 selects a target pixel based on the function of acquiring input image information, which is image information in a predetermined color space, and the acquired input image information. A function to set a predetermined range of image area including the selected pixel of interest as a local area, and to calculate an average color value which is an average value of the color values of pixels included in the set local area A function for calculating a color displacement in which a difference between the calculated average color value and the color value of the target pixel is expressed by a vector quantity in a color space, and image information of the target pixel based on the calculated color displacement. It can be understood as a program that realizes the function to be valued.

  The program for realizing the present embodiment can be provided not only by communication means but also by storing it in a recording medium such as a CD-ROM.

DESCRIPTION OF SYMBOLS 1 ... Image reading apparatus, 2 ... Image forming apparatus, 10 ... Control part, 20 ... Image reading unit, 30 ... Signal processing part, 36 ... Binarization processing part, 40 ... Post-processing part, 50 ... Communication I / F, 60 ... Control unit, 65 ... Binarization processing unit, 70 ... Image forming unit, 651 ... Input image information acquisition unit, 652 ... Target pixel selection unit, 653 ... Local region setting unit, 654 ... Local region size storage unit, 655 ... average color value calculation unit, 656 ... color displacement calculation unit, 657 ... binarization determination unit, 658 ... binary image information output unit

Claims (6)

An input image information acquisition unit that acquires input image information that is image information in a predetermined color space;
A target pixel selection unit that selects a target pixel based on the input image information acquired by the input image information acquisition unit;
A local region setting unit that sets, as a local region, an image region of a predetermined range including the target pixel selected by the target pixel selection unit;
An average color value calculation unit that calculates an average color value that is an average value of color values of pixels included in the local region set by the local region setting unit;
A color displacement calculation unit that calculates a color displacement in which a difference between the average color value calculated by the average color value calculation unit and the color value of the target pixel is represented by a vector amount in the color space;
A binarization determining unit that binarizes image information of the target pixel based on the color displacement calculated by the color displacement calculating unit;
Equipped with a,
The local region setting unit sets a plurality of the local regions having different sizes of the image region,
The binarization determining unit is configured to perform the attention based on the color displacement calculated for the smallest image region from among the plurality of local regions where the color displacement is equal to or greater than a predetermined value. An image processing apparatus that binarizes image information of pixels . An input image information acquisition unit that acquires input image information that is image information in a predetermined color space;
A target pixel selection unit that selects a target pixel based on the input image information acquired by the input image information acquisition unit;
A local region setting unit that sets, as a local region, an image region of a predetermined range including the target pixel selected by the target pixel selection unit;
An average color value calculation unit that calculates an average color value that is an average value of color values of pixels included in the local region set by the local region setting unit;
A color displacement calculation unit that calculates a color displacement in which a difference between the average color value calculated by the average color value calculation unit and the color value of the target pixel is represented by a vector amount in the color space;
A binarization determining unit that binarizes image information of the target pixel based on the color displacement calculated by the color displacement calculating unit;
With
The local region setting unit sets a plurality of the local regions having different sizes of the image region,
The binarization determination unit binarizes the image information of the target pixel based on the color displacement calculated for the largest image region from among the local regions having similar color displacement directions. An image processing apparatus. The binarization decision unit, image processing apparatus according to claim 1 or 2, characterized in that binarizing the image information of the pixel of interest based on the size and / or direction of the color displacement. The binarization determination unit selects a specific color axis from the color displacement calculated by the color displacement calculation unit, and binarizes the image information of the target pixel based on the color displacement about the specific color axis. the image processing apparatus according to any one of claims 1 to 3, characterized in that. On the computer,
A function of acquiring input image information which is image information in a predetermined color space;
A function of selecting a pixel of interest based on the acquired input image information;
A function of setting an image area of a predetermined range including the selected target pixel as a local area;
A function of calculating an average color value that is an average value of color values of pixels included in the set local region;
A function of calculating a color displacement in which a difference between the calculated average color value and the color value of the target pixel is represented by a vector amount in the color space;
A function of binarizing the image information of the target pixel based on the calculated color displacement;
To achieve,
The setting function sets a plurality of the local regions having different sizes of the image region,
The binarizing function is based on the color displacement calculated based on the color displacement calculated for the smallest image region from among the plurality of local regions where the color displacement is greater than or equal to a predetermined value. A program that binarizes pixel image information . On the computer,
A function of acquiring input image information which is image information in a predetermined color space;
A function of selecting a pixel of interest based on the acquired input image information;
A function of setting an image area of a predetermined range including the selected target pixel as a local area;
A function of calculating an average color value that is an average value of color values of pixels included in the set local region;
A function of calculating a color displacement in which a difference between the calculated average color value and the color value of the target pixel is represented by a vector amount in the color space;
A function of binarizing the image information of the target pixel based on the calculated color displacement;
Realized
The setting function sets a plurality of the local regions having different sizes of the image region,
The binarization function binarizes image information of the pixel of interest based on the color displacement calculated for the largest image region from among the local regions having similar color displacement directions. program.

Images