JP5299334B2 - Image processing apparatus, image processing method, and program - Google Patents

Description

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

  The line scanning type image reading apparatus has a one-dimensional image sensor composed of a large number of light receiving elements arranged in the main scanning direction, and the image sensor or an optical system that guides scanning light to the image sensor is relative to the document in the sub scanning direction. The original is subdivided into pixels and read by moving the target.

 2. Description of the Related Art Conventionally, in this type of image reading apparatus, a plurality of image sensors are arranged parallel to each other in the sub-scanning direction. That is, in a multi-color original reading and color discrimination including a full color, a three-color integrated sensor in which image sensors of each color of R (red), G (green), and B (blue) are integrated into one chip is used. ing.

When a plurality of image sensors are arranged in the sub-scanning direction as described above, a reading time difference depending on a distance (gap) between the image sensors is generated. That is, the image sensor on one end side in the arrangement direction reads the document in advance of the image sensor on the other end side, and the reading position of the document at the same time of each image sensor is different.
Therefore, in order to simultaneously output image data of the same line in the document read by each image sensor (a signal obtained by quantizing the photoelectric conversion output of the image sensor), it is necessary to perform correction corresponding to the difference in reading position. .

In a conventional image reading apparatus, a process called line-to-line correction is performed in which transmission of image data corresponding to an image sensor that reads a document in advance is delayed in line units by a delay memory (FIFO or the like).
That is, if the gap between the image sensors is g, the magnification of the optical system (document projection magnification) is F, and the read pixel density (number of lines / unit length) in the sub-scanning direction on the document is Dy, the image sensor. There is a gap in the reading position corresponding to the number n of lines represented by the formula (1).
n = (g / F) × Dy (1)
Therefore, the conventional image reading apparatus is configured to sequentially read image data for n lines of an image sensor that reads an original document in advance after it is temporarily stored in a delay memory.

For example, in an image reading apparatus incorporated in a copying machine, the reading magnification M is changed according to the copying magnification. That is, the reading pixel density Dy in the sub-scanning direction is changed to a value (Dy = Ds × M) obtained by multiplying the reference reading pixel density Ds when the reading magnification is “1” (100%) by the reading magnification M. . As a method for changing the reading pixel density, the reading cycle (image sensor driving timing) is fixed, and the relative movement speed (sub-scanning speed) between the sensor module (image sensor or optical system) and the original is set to 1 of the reference speed. A method of changing to a speed of / M is common.
Usually, the reference reading pixel density Ds and the magnification F of the optical system are set so that the number of lines n (hereinafter referred to as the number of different lines) corresponding to the deviation of the reading position becomes an integer. Therefore, when the reading magnification M is an integer (excluding 0), the number of different lines n is also an integer. Therefore, in principle, the period corresponding to the n-line scanning is delayed as described above. Thus, the inconvenience due to the gap between the image sensors is completely eliminated.

  However, in actuality, mechanical vibrations or unevenness in conveyance of the sensor module (difference between the target speed and the actual speed) occurs, which may cause the number n of different lines to be reduced. This means that a positional deviation on the document occurs in addition to the temporal deviation of the reading for each image sensor reading line. In an automatic document reader of the type that keeps the sensor stationary at the time of reading and conveys the document, unevenness in the conveyance speed occurs such as the document passes between multiple rollers or the friction is not constant due to the surface condition of the document It's easy to do.

Therefore, in this type of color image reading apparatus, when a color original image is read due to a slight shift of each reading position of the three-line color sensor or mechanical vibration, R (red), Color data different from the original color may be output at an edge portion of G (green) and B (blue) image data, and color misregistration may occur.
Another problem is that color misregistration occurs not only in the sub-scanning direction but also in the main scanning direction due to chromatic aberration and the like.

Therefore, for the purpose of such color misregistration correction, for example, as described in Patent Documents 1 and 2, the amount of color misregistration between the R, G, and B sensors is detected from the image data of the read document, and the subordinates are detected. Techniques for correcting color misregistration in the scanning direction have been proposed.
In addition, as described in Patent Document 3, a reference chart is arranged outside the original to detect color misregistration due to misalignment of reading positions of the R, G, and B sensors caused by mechanical vibrations. As described in Japanese Patent Application Laid-Open No. 2004-26853, a correction technique for correcting color misregistration in the sub-scanning direction, and a document conveying speed is detected, a deviation from a predetermined speed is obtained, and a sensor reading speed is constant based on the deviation. As described above, there has been proposed a color misregistration correction technique that variably controls the position of a reading cylindrical lens so that the reading speed is uneven.
Further, as described in Patent Document 5, there is a technique in which a monochrome edge portion is detected from a change in sensor output and three colors are aligned with the data of the sensor arranged at the center.
Further, Patent Document 6 describes a technique for correcting a color shift caused by mechanical vibration or chromatic aberration by detecting an edge portion and bringing the R and B values of the edge portion close to the G value.

JP-A-6-334814 JP-A-8-139949 JP-A-9-98256 JP-A-8-107472 Japanese Patent Laid-Open No. 10-215381 JP 2001-203900 A

However, in the conventional color misregistration correction technique as described above, for example, in the techniques of Patent Document 1 and Patent Document 3, due to the vibration of the machine in the sub-scanning direction of the R, G, B sensor, etc. Since the color shift that changes every moment in the sub-scanning direction is obtained and corrected from the original image or the reference image outside the original, there is a problem that the color shift in one main scanning line cannot be handled.
In the technique described in Patent Document 2, interpolation data and correlation functions are obtained for each of R, G, and B in order to detect a color misregistration amount from document data, so that the processing becomes complicated and the circuit becomes complicated. Another problem is that only the color shift in the sub-scanning direction is corrected.
In addition, the technique described in Patent Document 4 has a problem in that color misregistration in the main scanning direction due to vibrations and aberrations cannot be corrected because color misregistration is corrected from the amount of change in reading speed.
Further, with the technique described in Patent Document 5, the sensor detection value does not exceed the threshold value for detecting the black-and-white edge portion for edges of colors other than black and white, and correction is not performed. There is.

  In the technique described in Patent Document 6, only RH1 to RH15 (R interpolation pixels), BH1 to BH15 (B interpolation pixels), and 15 interpolation points (R and B) are obtained only by correction in the main scanning direction. Since each of the points is an input pixel) and Gn (G input value of the target pixel), if the sub-scanning direction is performed in addition to the main scanning direction, the circuit scale is increased. There's a problem. In addition, the interpolation points of R and G do not always coincide with Gn, and there is usually a slight difference from Gn. For this reason, in order to improve the color misregistration correction performance, it is necessary to increase the number of interpolation points, leading to an increase in circuit scale. Furthermore, in the technique described in Patent Document 6, color misregistration correction is performed on the detected edge portion, and there is a problem that the color misregistration correction is difficult to be performed depending on the strength of the edge.

  The present invention has been made in order to solve the above-described problems, and performs simple processing for correcting color misregistration in the vicinity of black characters in the main scanning direction and the sub-scanning direction caused by chromatic aberration, mechanical vibration, and the like. It is to be realized by.

In order to solve the above-mentioned problem, the invention described in claim 1
An image processing apparatus that performs image processing on an image obtained by reading a document using a plurality of image sensors arranged in parallel to each other and separated into a plurality of colors,
One of the color-separated colors is set as a reference color, the other at least one color is set as a correction target color, a virtual pixel is set between a target pixel of the correction target color and an adjacent pixel adjacent thereto, The gradation value of the virtual pixel is calculated based on the gradation value of the target pixel and the gradation value of the adjacent pixel, and the reference pixel gradation value of the target pixel is the target pixel of the correction target color and the virtual pixel. A color misregistration correction unit that converts the gradation value of the target pixel of the correction target color into the gradation value of the reference color when the gradation value of the virtual pixel group consisting of

The invention according to claim 2 is the invention according to claim 1,
When the gradation value of the reference color of the pixel of interest exceeds the maximum value of the gradation value of the virtual pixel group, the color misregistration correction unit determines the gradation value of the pixel of interest of the correction target color as the virtual pixel group. When the gradation value of the reference color of the target pixel is less than the minimum value of the virtual pixel group, the gradation value of the target pixel of the correction target color is The value is converted into the minimum gradation value of the virtual pixel group.

The invention according to claim 3 is the invention according to claim 1 or 2,
The interval between the target pixel and the virtual pixel can be set to different values in the main scanning direction and the sub-scanning direction.

The invention according to claim 4 is the invention according to any one of claims 1 to 3,
An operation unit is provided for setting an interval between the target pixel and the virtual pixel in each of the main scanning direction and the sub-scanning direction.

The invention according to claim 5 is the invention according to any one of claims 1 to 3,
The interval between the pixel of interest and the virtual pixel is set to a different value depending on whether the reading device reads the document in a stationary state or moves and moves the document.

The invention according to claim 6 is the invention according to any one of claims 1 to 5,
The color misregistration correction unit determines whether the pixel of interest and the surrounding pixels existing within a predetermined range from the pixel of interest are black pixels or color pixels, and the pixels determined to be black pixels The gradation value of the target pixel of the color to be corrected is converted only when the number of pixels exceeds a predetermined threshold.

The invention according to claim 7 is the invention according to any one of claims 1 to 5,
A weighting factor is associated with the target pixel and the peripheral pixel according to a positional relationship with the target pixel, and the color misregistration correction unit multiplies the pixel determined to be the black pixel with the weighting factor. Only when the integrated value exceeds a predetermined threshold value, the gradation value of the target pixel of the correction target color is converted.

The invention according to claim 8 is the invention according to claim 6 or 7,
The color misregistration correction unit determines whether each of the target pixel and the peripheral pixel is a black pixel or a color pixel based on the saturation and brightness of the target pixel and the peripheral pixel.

The invention according to claim 9 is:
An image processing method for performing image processing on an image obtained by reading a document using a plurality of image sensors arranged in parallel to each other and separated into a plurality of colors,
A virtual pixel is set between a target pixel of the correction target color and an adjacent pixel adjacent to the target pixel of the correction target color, with one of the color-separated colors as a reference color and at least one other color as a correction target color, The gradation value of the virtual pixel is calculated based on the gradation value of the target pixel and the gradation value of the adjacent pixel, and the reference pixel gradation value of the target pixel is the target pixel of the correction target color and the virtual pixel. Converting the gradation value of the target pixel of the color to be corrected into the gradation value of the reference color when the gradation value of the virtual pixel group consisting of is not less than the minimum value and not more than the maximum value.

The program of the invention according to claim 10 is:
A computer used in an image processing apparatus that performs image processing on an image obtained by reading a document using a plurality of image sensors arranged in parallel to each other and separated into a plurality of colors,
One of the color-separated colors is set as a reference color, the other at least one color is set as a correction target color, a virtual pixel is set between a target pixel of the correction target color and an adjacent pixel adjacent thereto, The gradation value of the virtual pixel is calculated based on the gradation value of the target pixel and the gradation value of the adjacent pixel, and the reference pixel gradation value of the target pixel is the target pixel of the correction target color and the virtual pixel. A color misregistration correction unit that converts a gradation value of a pixel of interest of the correction target color into a gradation value of the reference color when the gradation value of the virtual pixel group consisting of
To function as.

  According to the present invention, it is possible to realize color misregistration correction around the black character in the main scanning direction and the sub-scanning direction caused by chromatic aberration, mechanical vibration, and the like by a simple process.

3 is a block diagram illustrating a functional configuration of the image reading apparatus according to the present embodiment. It is a figure which shows the main structures of the reader of FIG. It is a figure which shows schematic structure of the image sensor which the reader of FIG. 1 has. It is the elements on larger scale of the image sensor shown to FIG. 3A. It is a block diagram which shows the image processing circuit which processes the image signal of R, G, and B read with the image sensor. It is a figure which shows the other structural example of the image processing apparatus of FIG. It is a figure which shows the state which color misregistration generate | occur | produced in the achromatic pixel. 5 is a flowchart showing a color misregistration correction process A executed in the color misregistration correction unit in FIG. 4 in the first embodiment. It is a flowchart which shows the R-color-shift correction process performed in step S1 of FIG. It is a figure which shows an example of the adjacent pixel window for performing color shift correction | amendment of the main scanning direction and a subscanning direction. It is a figure which shows an example of the adjacent pixel window for performing color misregistration correction only in the main scanning direction. It is a figure which shows an example of the adjacent pixel window for performing color misregistration correction only in the sub-scanning direction. It is a figure for demonstrating a virtual pixel. It is a figure which shows typically the result of the color misregistration correction process in the case of Rmax ≧ G (X, Y) ≧ Rmin and Bmax ≧ G (X, Y) ≧ Bmin. It is a figure which shows typically the result of the color shift correction process in the case of Rmax <G (X, Y). It is a figure which shows typically the result of the color shift correction process in the case of Rmin> G (X, Y). 6 is a flowchart showing a color misregistration correction process B executed in a color misregistration correction unit in FIG. 4 in the second embodiment. It is a flowchart which shows the black area determination process performed in FIG.12 S11. It is a figure which shows an example of a black area | region determination window. It is a figure which shows the other example of a black area | region determination window. It is a figure which shows the other example of a black area | region determination window. It is a flowchart which shows the black pixel / color pixel discrimination | determination process performed in step S205 of FIG. It is a figure which shows an example of saturation reference Wref (V). It is a figure which shows an example of the weighting filter of the black area | region determination window shown to FIG. 14A. It is a flowchart which shows the black area determination process B of the modification in 2nd Embodiment. 12 is a flowchart illustrating a scan control process executed by the control unit of FIG. 1 in the third embodiment.

[First Embodiment]
Hereinafter, an example of the first embodiment of the present invention will be described in detail with reference to the drawings.
FIG. 1 shows a functional configuration example of the image reading apparatus 1 according to the present embodiment. As shown in FIG. 1, the image reading apparatus 1 includes a control unit 10, a reading device 20, an image processing device 30, a communication I / F 40, an operation unit 50, a display unit 60, and the like. ing.

  The controller 10 includes a CPU (Central Processing Unit), a RAM (Random Access Memory), and a ROM (Read Only Memory). The CPU of the control unit 10 reads out the system program, the overall control program, and various processing programs stored in the ROM in accordance with the operation of the operation unit 50 and the like, expands them in the RAM, and performs image reading according to the expanded programs. The operation of each part of the apparatus 1 is centrally controlled.

As illustrated in FIG. 2, the reading device 20 includes an ADF (Auto Document Feeder) unit 20 a and a reading unit 20 b.
The ADF unit 20a is an automatic document feeder. As shown in FIG. 2, the ADF unit 20 a is configured to bring a document into close contact with a sheet feeding roller 22 that sequentially conveys documents stacked on the document placing tray 21 from the top and a platen glass 201 that is a document reading portion. A contact roller 23 for passing the paper and a guide roller 24 for guiding the document conveyed by the paper feed roller 22 along the contact roller 23, and reading the document loaded on the document placement tray 21 on the platen glass 201. Convey one sheet at a time. The ADF unit 20a also includes a switching claw 25 for switching the advancing direction of the document that has passed through the platen glass 201, a reversing roller 26 for reversing the front and back of the document, and a discharge tray 27 for discharging the document that has been read. When the document is a double-sided document, the front and back sides of the document are reversed after one side of the document is read, and the document is conveyed again to the reading position of the platen glass 201.

  The document placement tray 21 of the ADF unit 20a is provided with a document presence / absence detection sensor (not shown). The document presence / absence detection sensor detects the presence / absence of a document on the document placing tray 21 and outputs the detection signal to the control unit 10. An opening / closing sensor (not shown) is provided on the surface of the document cover 28 of the ADF unit 20a on the platen glass 201 side. The open / close sensor detects the open / close state of the document cover 28 and outputs the detection signal to the control unit 10.

  The reading unit 20b is an exposure scanning unit including a light source and a mirror, an optical system that guides scanning light to the image sensor 202, and an image that receives reflected light from a document and outputs an electrical signal (image signal) corresponding to the light intensity. A sensor module including the sensor 202 is provided.

  FIG. 3A is a diagram schematically illustrating the configuration of the image sensor 202. FIG. 3B is a partial enlarged view of a part of FIG. 3A. As shown in FIGS. 3A and 3B, the image sensor 202 is a substrate of a three-line CCD array (R sensor 202R, G sensor 202G, B sensor 202B) in which a plurality of light receiving elements are arranged in the main scanning direction of document scanning. This is a line image sensor integrally formed therein. The R sensor 202R, the G sensor 202G, and the B sensor 202B are arranged in parallel in the direction orthogonal to the main scanning direction (sub scanning direction) with an interval of integer lines nd (n is an integer). On the light receiving surfaces of the R sensor 202R, the G sensor 202G, and the B sensor 202B, the original image is separated into three primary colors R (red), G (green), and B (blue), and read in order to read the original image. A spectral filter that transmits the B light is provided.

  In accordance with a scan instruction from the control unit 10, the reading unit 20 b relatively moves the document transported or placed on the reading position of the platen glass 201 and the sensor module in the sub-scanning direction orthogonal to the main scanning direction. The original is read, and the read R, G, and B image signals are converted into image data via the analog processing unit 203 and the A / D converter 204 and output to the image processing apparatus 30. In the image data, for example, the gradation value of each color pixel is represented by 0 to 255.

  The reading unit 20b can read an original by two types of reading methods, an ADF original reading method and an FB original reading method. In the ADF document reading method, a document conveyed from the ADF unit 20a is read. In this case, the sensor module of the reading unit 20b is kept stationary, and the document is moved in the sub-scanning direction with respect to the sensor module. In the FB document reading system, the document cover 28 is opened and the document placed on the platen glass 201 is read. In this case, the document is kept stationary, and the sensor module of the reading unit 20b is moved in the sub-scanning direction with respect to the document.

The image processing apparatus 30 performs various image processing on each of the R, G, and B image data output from the reading apparatus 20 and outputs the processed image data to the communication I / F 40.
FIG. 4 is a block diagram illustrating a processing circuit that processes R, G, and B image signals read by the image sensor 202. As shown in FIG. 4, the image processing apparatus 30 includes a shading correction unit 31, a γ correction unit 32, an interline correction unit 33, a color misregistration correction unit 34, an enlargement / reduction processing unit 35, a color conversion processing unit 36, and an image correction. A unit 37, an image area identification processing unit 38, and the like.

The shading correction unit 31 corrects non-uniform light quantity of the light source system and variation in output characteristics of each pixel of the image sensor 202, and ensures uniformity in the main scanning direction.
The γ correction unit 32 adjusts the gradation characteristics so as to be suitable for the subsequent image processing, and absorbs individual differences between devices by using a look-up table (LUT) independent for each RGB. Perform the conversion.

  The inter-line correction unit 33 performs correction to absorb a difference in reading position distance caused by a physical interval (e.g., 8 lines) between the color lines of the image sensor 202. For example, correction is performed using a delay memory such as a FIFO.

  The color misregistration correction unit 34 performs main scanning direction and sub-scanning caused by mechanical vibration, conveyance unevenness (original conveyance unevenness, sensor module conveyance unevenness in the reading unit 20b), chromatic aberration, manufacturing error of each color of the image sensor 202, and the like. Correct the color shift in the direction. Details will be described later.

The enlargement / reduction processing unit 35 performs image enlargement / reduction processing by interpolation processing or the like when conversion of the read image and the output resolution is necessary.
The color conversion processing unit 36 performs conversion into image data that matches the configuration of the color channel requested by the output device. For example, when an image is output to an image output device of a copying machine using CMYK toner as a color material, conversion from RGB to CMYK is performed. In addition, when outputting an image to be displayed on a color monitor such as a personal computer, color conversion processing that matches the characteristics of the monitor such as sRGB is performed.

The image correction unit 37 corrects deterioration of MTF (Modulation Transfer Function) generated in the reading unit, and corrects an image optimal for image characteristics such as a character region and a halftone dot.
The image area identification processing unit 38 analyzes the characteristics of each pixel, and identifies and outputs various attributes necessary for correction of the image correction unit 37, such as edge regions, color pixels / black pixels, and halftone dots.

  Each unit of the image processing device 30 is configured by a dedicated electric circuit (hardware). However, as shown in FIG. 5, the image processing device 30 executes processing of the CPU, RAM, and each unit. Are stored in the ROM (shading correction program P1, γ correction program P2, inter-line correction program P3, color misregistration correction processing program P4,...), And the cooperation between the CPU and the program stored in the ROM is provided. It is also possible to adopt a configuration in which the function of each unit is realized by software processing according to the above. In this case, for example, the color misregistration correction processing program P4 includes a program for executing a color misregistration correction processing A (see FIGS. 7 and 8) described later.

  The communication I / F 40 is a communication interface for transmitting image processed image data to an image output device such as a printer or a monitor (image display device).

The operation unit 50 includes a touch panel, hard keys, and the like.
The touch panel is a pressure-sensitive (resistive film pressure type) touch panel in which transparent electrodes are arranged in a grid pattern so as to cover the upper surface of an LCD (Liquid Crystal Display) of the display unit 60. The coordinates are detected by a voltage value, and the detected position signal is output to the control unit 10 as an operation signal. Note that the touch panel is not limited to a pressure-sensitive type, and may be another electrostatic type, an optical type, or the like.
The hard key includes various operation buttons such as a numeric button and a start button (not shown) for instructing start of scanning, and outputs an operation signal generated by the button operation to the control unit 10.

  The display unit 60 is configured by an LCD or the like, and displays various operation buttons, the status of the apparatus, the operation status of each function, and the like on the display screen in accordance with an instruction of a display signal input from the control unit 10.

Next, the operation of the color misregistration correction unit 34 will be described in detail.
FIG. 6 is a diagram illustrating a state in which color misregistration has occurred when an achromatic edge portion of a document is read by the image sensor 202. The vertical axis in FIG. 6 indicates the gradation value (signal value) of the input pixel, and the horizontal axis indicates the pixel position in the sub-scanning direction. When an achromatic edge portion is read, the gradation value of R should be equal to the gradation value of G = the gradation value of B. However, if the reading position between the three colors R, G, and B, which should be the same pixel position, is shifted due to mechanical vibration, conveyance unevenness, manufacturing error of each color sensor, etc., as shown in FIG. The gradation values of G and B are different values. That is, a quality problem called color misregistration occurs in which a color different from the original color is output. In particular, in the case of a black character, if the outline is colored, it is visually noticeable, so it tends to be a quality problem.
Therefore, the color misregistration correction unit 34 efficiently and accurately corrects the color misregistration in the peripheral portion of the black character in the main scanning direction and the sub scanning direction caused by mechanical vibration, conveyance unevenness, manufacturing error of each color sensor, chromatic aberration, and the like. Do.

In the following description, the coordinates of the target pixel to be corrected are represented by (X, Y). Here, X represents coordinates in the main scanning direction, and Y represents coordinates in the sub-scanning direction.
Further, the R gradation value before correction in the target pixel (correlated with the physical quantity of luminance input to the R sensor 202R) and the coordinate of the target pixel in the R input image are represented by R (X, Y). Similarly, the G tone value before correction at the target pixel and the coordinate of the target pixel in the G input image are represented by G (X, Y), and the B tone value before correction at the target pixel and the B input image are corrected. The coordinates of the pixel of interest in are represented by B (X, Y).
The output result of the color misregistration correction unit 34 at the target pixel is represented as Rout (X, Y), Gout (X, Y), and Bout (X, Y).

  In this embodiment, G is a reference color, R and B are correction target colors, and an R input image and a B input image are corrected using a G input image input from the interline correction unit 33 as a reference. The G input image and the B input image may be corrected using the R input image as a reference, or the R input image and the G input image are corrected using the B input image as a reference. It is good as well.

FIG. 7 shows a flowchart of the color misregistration correction process A executed by the color misregistration correction unit 34. In the color misregistration correction unit 34, for each of the R, G, and B input images, the color misregistration correction processing A shown in FIG. 7 is performed on each pixel while sequentially moving the target pixel (X, Y). .
As shown in FIG. 7, in color misregistration correction processing A, R-color misregistration correction processing is executed for each pixel of R (step S1), and Gout (X, A process of outputting as Y) ← G (X, Y) is performed (step S2), and a B-color misregistration correction process is executed for each pixel of B (step S3).

FIG. 8 shows a flowchart of the R-color misregistration correction process.
In the R-color misregistration correction process, first, an adjacent pixel window is generated (step S101).
FIG. 9A shows an example (R example) of the adjacent pixel window. In step S101 of FIG. 8, as shown in FIG. 9A, the target pixel R (X, Y) and the eight pixels adjacent thereto (adjacent pixels R (X−1, Y−1) to R (X + 1, An adjacent pixel window consisting of 9 pixels with Y + 1) is generated, where the vertical direction in Fig. 9A is the main scanning direction and the horizontal direction is the sub-scanning direction (Figs. 9B, 9C, and 10). The same applies to FIGS. 14A to 14C and FIG.

Next, a virtual pixel is provided between the target pixel and the adjacent pixel, and the gradation value is calculated (step 102).
Here, the virtual pixel will be described with reference to FIG. When color misregistration occurs due to the above-described factors, the R sensor 202R of the image sensor 202 reads a position that is deviated relative to the G sensor 202G. Assuming that the color misregistration is within a range of one pixel, the pixel when no color misregistration is present exists between the target pixel and the adjacent pixel. Therefore, a virtual pixel is provided at a position where the target pixel is virtually shifted around one pixel, a gradation value of the virtual pixel is calculated, and the gradation value of the virtual pixel group including the virtual pixel and the target pixel is calculated. Based on this, color misregistration correction is performed.

  Specifically, as shown in FIG. 10, positions (eight locations) separated from the target pixel R (X, Y) = r22 by a distance Kh in the main scanning direction and Kv in the sub-scanning direction are set as virtual pixel positions. The gradation values r11, r12, r13, r21, r23, r31, r32, r33 of the eight virtual pixels are calculated based on the gradation values of the pixels in the neighboring pixel window around the eight virtual pixels. Then, 9 pixels obtained by adding the target pixel to 8 virtual pixels are set as a virtual pixel group.

As a method for calculating the gradation value of the virtual pixel, since simple and stable performance can be obtained, bi-linear interpolation based on the gradation values of the pixel of interest and the adjacent pixels can be used. The interpolation formula is shown below.
The operation can be easily implemented by replacing Kh ′ = 256 × Kh and Kv ′ = 256 × Kv with a power of 2 and replacing it with a fixed-point operation.
In the following interpolation formula, the portion divided by 256 or 65536 is replaced with a shift operation, thereby eliminating a costly division operation.

(Interpolation formula)
r11 = (Kh '× Kv' × R (X-1, Y-1) + (256-Kh ') × Kv' × R (X, Y-1) +
Kh '× (256-Kv') × R (X-1, Y) + (256-Kh ') × (256-Kv') × R (X, Y)} / 65536
r12 = {Kv '× R (X, Y-1) + (256-Kv') × R (X, Y)} / 256

r13 = (Kh '× Kv' × R (X + 1, Y-1) + (256-Kh ') × Kv' × R (X, Y-1) +
Kh '× (256-Kv') × R (X + 1, Y) + (256-Kh ') × (256-Kv') × R (X, Y)} / 65536
r21 = {Kh '× R (X-1, Y) + (256-Kh') × R (X, Y)} / 256
r22 = R (X, Y)

r23 = {Kh '× R (X + 1, Y) + (256-Kh') × R (X, Y)} / 256

r31 = (Kh '× Kv' × R (X-1, Y + 1) + (256-Kh ') × Kv' × R (X, Y + 1) +
Kh '× (256-Kv') × R (X-1, Y) + (256-Kh ') × (256-Kv') × R (X, Y)} / 65536

r32 = (Kv '× R (X, Y + 1) + (256-Kv') × R (X, Y)) / 256

r33 = (Kh '× Kv' × R (X + 1, Y + 1) + (256-Kh ') × Kv' × R (X, Y + 1) +
Kh '× (256-Kv') × R (X + 1, Y) + (256-Kh ') × (256-Kv') × R (X, Y)} / 65536

In the above example, fixed-point arithmetic is performed at 256 times. However, if more precision is required, for example, 512 times or 1024 times may be used. If the circuit scale is to be reduced, 128 times or 64 times is used. And so on.
In the case of 256 times, the position of Kh and Kv can be adjusted with an accuracy of a distance of 0.00390625 = ^1/256 of one pixel.
As a method for calculating the virtual pixel, a known interpolation method other than bi-linear interpolation can be used.

  As the values of the distances Kv and Kh from the target pixel to the virtual pixel, values that are appropriate experimentally and empirically based on individual conveyance unevenness, mechanical vibration, chromatic aberration, and the like of the image reading apparatus 1 may be set. Good. Since the amount of color misregistration caused by conveyance unevenness, mechanical vibration, chromatic aberration, etc. differs between the main scanning direction and the sub scanning direction, the value of the interval Kh between the target pixel and the virtual pixel in the main scanning direction, the target pixel and the virtual pixel The value of the interval Kv in the sub-scanning direction can be set to a different value. In addition, the amount of color misregistration caused by conveyance unevenness, mechanical vibration, chromatic aberration, and the like varies depending on the reading resolution, the conveyance speed, the degree of wear of the apparatus, and the like. Therefore, the values of Kv and Kh can be individually set by an operation from the operation unit 50.

Alternatively, the values of Kv and Kh may be automatically set instead of being input from the operation unit 50. For example, while changing the value of Kh within the range of the pixel interval, sequentially read a document containing a plurality of black lines in the line direction (main scanning direction) of the image sensor 202 and integrate the number of color pixels (chromatic color pixels). However, among the Kh that minimizes the total number of color pixels, a value that increases the Kh value as much as possible may be automatically set as the Kh value (for example, JP 2000-022964, JP 2000-022965, JP 2000-022966, JP 2000-022967, JP 2000-022968, 2005-184156). As for the value of Kv, while changing the value of Kv within the range of the pixel interval, a document containing a plurality of black lines in the direction perpendicular to the line direction of the image sensor 202 is read, the number of color pixels is added, and the color pixels Of the Kvs with the smallest total number, the value with the largest Kv value can be automatically set as the Kv value.
Since it is assumed that the color misregistration amounts are different between R and B, it is preferable that the values of Kv and Kh can be set separately for R and B.

  When the calculation of the gradation value of the virtual pixel is completed, the maximum value Rmax of the gradation values (r11 to r33) of the R virtual pixel group is calculated in step S103 of FIG. 8 (step S103). Then, the calculated maximum value Rmax is compared with the G gradation value G (X, Y) of the target pixel (step S104).

If it is determined that Rmax <G (X, Y) as a result of the comparison (step S104; Yes), Rmax is the correction result of R, which is the value closest to G (X, Y) among the virtual pixels. (Rout) is output (step S105). (That is, Rout (X, Y) = Rmax.)
On the other hand, if it is determined that Rmax <G (X, Y) is not satisfied, that is, Rmax ≧ G (X, Y) (step S104; No), the process proceeds to step S106.

  In step S106, the minimum value Rmin of the gradation values (r11 to r33) of the R virtual pixel group is calculated (step S106). Then, the calculated minimum value Rmin is compared with the G gradation value G (X, Y) of the target pixel (step S107).

  If it is determined that Rmin> G (X, Y) as a result of the comparison (step S107; Yes), Rmin, which is the value closest to G (X, Y) among the virtual pixels, is determined. The correction result (Rout) of R is output (step S108). (In other words, Rout (X, Y) = Rmin.)

  On the other hand, if it is determined as a result of the comparison that Rmin> G (X, Y) is not satisfied, that is, Rmax ≧ G (X, Y) ≧ Rmin (step S107; No), G (X , Y) is output as an R correction result (Rout) (step S109). (That is, Rout (X, Y) = G (X, Y).)

  Since the B-color misregistration correction process is the same as the R-color misregistration correction process, the description is cited. In other words, in the B-color misregistration correction process, each pixel of the B input image is set as the target pixel B (X, Y), and the same process as in steps S101 to S109 described above is performed, and Bout (X, Y). Is output.

FIG. 11A schematically shows the result of the color misregistration correction processing when Rmax ≧ G (X, Y) ≧ Rmin and Bmax ≧ G (X, Y) ≧ Bmin. In FIG. 11A, the vertical axis indicates the gradation value. The left side of the graph shows the gradation values r11, r12, r13, r21, r22 (R (X, Y)), r23, r31, r32, r33 of the virtual pixel group for the target pixel R (X, Y). . The center of the graph shows Gout (X, Y), Rout (X, Y), and Bout (X, Y). The right side of the graph shows the gradation values b11, b12, b13, b21, b22 (B (X, Y)), b23, b31, b32, b33 of the virtual pixel group for the target pixel B (X, Y). .
As shown in FIG. 11A, R (X, Y) .noteq.G (X, Y) .noteq.B (X, Y) before color misregistration correction is in a state of being colored. As a result, R (X, Y) = G (X, Y) = B (X, Y), and color misregistration is completely corrected.

Here, Japanese Patent Application Laid-Open No. 2001-203900 also discloses a color misregistration correction method. However, in this technique, only RH1 to RH15 (R interpolation pixels), Comparison is made between BH1 to BH15 (B interpolation pixels), R and B interpolation points (of which one is an input pixel) and Gn (G input value of the pixel of interest). RH10 and BH5 are selected as the closest values to Gn, but these do not always coincide with Gn, and there is usually a slight difference from Gn. Therefore, in order to improve the color misregistration correction performance, it is necessary to increase the number of interpolation points. If the sub-scanning direction is also performed, it is necessary to calculate more interpolation points.
In contrast, in the color misregistration correction process A described above, for each of R and B, it is only necessary to perform an interpolation calculation of the input pixel and 8 points, and a comparison operation between the maximum value and the minimum value G (X, Y). The circuit configuration is simple, and a better color misregistration correction result can be obtained. In addition, since color misregistration correction is performed without performing edge detection, not only coloring due to color misregistration at the edge of achromatic parts such as black characters is suppressed, but also excellent correction effects are obtained for black reproduction inside black characters. be able to.

FIG. 11B schematically shows the result of color misregistration correction processing when Rmax <G (X, Y). In FIG. 11B, the vertical axis indicates the gradation value. The left side of the graph shows the gradation values r11, r12, r13, r21, r22 (R (X, Y)), r23, r31, r32, r33 of the virtual pixel group for the target pixel R (X, Y). . The right side of the graph shows gradation values R (X, Y) and G (X, Y) before correction, and gradation values Rout (X, Y) and Gout (X, Y) after correction.
As shown in FIG. 11B, in the case of Rmax <G (X, Y), as a result of the above-described color misregistration correction processing, the value of R (X, Y) becomes Rmax closest to G (X, Y) in the virtual pixel. It is corrected. The same applies to B.

FIG. 11C schematically shows the result of color misregistration correction processing when Rmin> G (X, Y). In FIG. 11C, the vertical axis indicates the gradation value. The left side of the graph shows the gradation values r11, r12, r13, r21, r22 (R (X, Y)), r23, r31, r32, r33 of the virtual pixel group for the target pixel R (X, Y). . The right side of the graph shows gradation values R (X, Y) and G (X, Y) before correction, and gradation values Rout (X, Y) and Gout (X, Y) after correction.
As shown in FIG. 11C, when Rmin> G (X, Y), the value of R (X, Y) is the closest to R (X, Y) in the virtual pixel group as a result of the color misregistration correction process described above. It is corrected to. The same applies to B.

  In the color misregistration correction process A, the case where the color misregistration correction is performed in both directions of the main scanning direction and the sub scanning direction has been described as an example. However, only the color misregistration in any one direction can be corrected. For example, as shown in FIG. 9B, the adjacent pixel window created in step S101 in FIG. 8 is the target pixel and three pixels adjacent thereto in the main scanning direction, and the virtual pixel set in step S102 is changed from the target pixel to the main scanning direction. If it is provided at positions (two places) separated by Kh, color misregistration correction can be performed only in the main scanning direction. Further, as shown in FIG. 9C, the adjacent pixel window created in step S101 in FIG. 8 is set as the target pixel and three pixels adjacent to the target pixel in the sub scanning direction, and the virtual pixel set in step S102 is changed from the target pixel to the sub scanning direction. If it is provided at positions (two locations) separated by Kv, color misregistration correction only in the sub-scanning direction can be performed.

[Second Embodiment]
Next, a second embodiment of the present invention will be described.
The color shift of the achromatic portion of the original such as black characters has a feature that appears in the peripheral portion of the achromatic portion. On the other hand, when color misregistration correction is performed on an area other than the achromatic portion of the original, such as color halftone dots such as C, M, and Y of a color original, a phenomenon may occur in which the saturation of the edge portion of the chromatic color is lowered. . Therefore, in the second embodiment, by performing color misregistration correction only in a region where black pixels (achromatic pixels) are in the vicinity, good black character reproducibility with suppressed achromatic coloring and color Color misregistration correction that suppresses peripheral saturation reduction is realized.

  In the present embodiment, as in the first embodiment, G is the reference color, R and B are the correction target colors, that is, the G input image input from the interline correction unit 33 is the reference. The case where the R input image and the B input image are corrected will be described. However, the G input image and the B input image may be corrected using the R input image as a reference, and the B input image may be used as a reference. As an alternative, the R input image and the G input image may be corrected. Further, the configuration of each part of the image reading apparatus 1 is the same as that described in the first embodiment, and thus the description is incorporated.

  FIG. 12 shows a flowchart of color misregistration correction processing B executed by the color misregistration correction unit 34 in the second embodiment. In the color misregistration correction unit 34, the color misregistration correction processing B shown in FIG. 12 is performed for each pixel of the input image of each color while sequentially moving the target pixel (X, Y).

  In the color misregistration correction process B, first, a black area determination process is performed (step S11). The black area determination process is a process of determining the presence state of black pixels (achromatic pixels) around the target pixel.

FIG. 13 shows a flowchart of the black area determination process executed in step S11.
First, a black area determination window shown in FIG. 14A is set around the target pixel (step S201).
As shown in FIG. 14A, the black region determination window is set for the pixel of interest (X, Y) and surrounding pixels (existing from the pixel of interest up and down wy and left and right wx). For each pixel of the black area determination window, JdgBk (X + j, Y + i) (−wx ≦ j ≦ wx, −wy ≦ i ≦ wy) for storing the result of the black pixel / color pixel determination processing described later; If the pixel at the position is determined to be a black pixel, TRUE, and if it is determined to be a color pixel, FALSE) is provided. As for the values of wx and wy, good results can be obtained with one pixel or about two pixels in consideration of the degradation of the MTF of the optical system, but different values can be set according to the characteristics of the apparatus and the document. Also good.

  Next, an initial value 0 is set to the counter CountBk (step S202), -wy is set to the variable i, and -wx is set to the variable j (step S203, step S204).

  Next, a black pixel / color pixel discrimination process is executed for the pixel (X + j, Y + i) (step S205).

FIG. 15 shows a flowchart of the black pixel / color pixel discrimination process.
First, the gradation value of R (X + j, Y + i) is set in the variable R, the gradation value of G (X + j, Y + i) is set in the variable G, and the gradation value of B (X + j, Y + i) is set in the variable B. It is set (step S2051).

Next, based on the values of the variables R, G, and B, an index W relating to the saturation of the pixel (X + j, Y + i) is calculated by (Equation 1) (step S2052).
(Formula 1) W = MAX {R, G, B} -MIN {R, G, B}

Next, based on the values of the variables R, G, and B, an index V related to the brightness of the pixel (X + j, Y + i) is calculated by (Equation 2) (step S2053).
(Formula 2) V = MIN {R, G, B}

As a method for calculating W and V, a known method different from the above may be used.
For example, V may be obtained by a product-sum operation with the following constants Kr, Kg, Kb (Kr + Kg + Kb = 1).
V = Kr x R + Kg x G + Kb x B

Next, the saturation reference Wref (V) stored in the memory in the color misregistration correction unit 34 is referred to, and it is determined whether or not Wref (V)> W. It is determined whether the pixel is an achromatic pixel or a color pixel (chromatic pixel) (step S2054). As shown in FIG. 16, the saturation reference Wref (V) stores W threshold values Wref (V) (threshold values for determining whether a chromatic color or an achromatic color) for each lightness value V. It is a table.
When Wref (V)> W, the pixel (X + j, Y + i) is determined as a black pixel (achromatic pixel). When Wref (V) ≦ W, (X + j, Y + i) is a color pixel (chromatic pixel) ).

  If it is determined in step S2054 that Wref (V)> W (step S2054; Yes), TRUE is set in JdgBk (X + j, Y + i) (step S2055), and the process proceeds to step S206 in FIG. If it is determined in step S2054 that Wref (V) ≦ W (step S2054; No), FALSE is set in JdgBk (X + j, Y + i) (step S2056), and the process proceeds to step S206 in FIG.

  In step S206 of FIG. 13, it is determined whether JdgBk (X + j, Y + i) is TRUE or FALSE. When it is determined that JdgBk (X + j, Y + i) is TRUE (step S206; TRUE), CountBk is incremented by 1 (step S207), and the process proceeds to step S208. If it is determined that JdgBk (X + j, Y + i) is FALSE (step S206; FALSE), the process proceeds to step S208.

  In step S208, the variable j is incremented by 1 (step S208), and it is determined whether j> wx is satisfied (step S209). If it is determined that j> wx is not satisfied (step S209; No), the process returns to step S205. If it is determined that j> wx is satisfied (step S209; Yes), the variable i is incremented by 1 (step S210), and it is determined whether i> wy is satisfied (step S211). If it is determined that i> wy is not satisfied (step S211; No), the process returns to step S204. If it is determined that i> wy (step S211; Yes), the process proceeds to step S212.

  In step S212, it is determined whether or not the value of CountBk is greater than or equal to a predetermined threshold Th. The threshold value Th is a value determined by comparing and considering the deterioration in image quality due to weakening of correction and the deterioration in image quality due to correction other than black pixels, for example, 1. If it is determined that CountBk ≧ Th (step S212; Yes), TRUE is set in the black region determination result AreaBk (X, Y) (step S213). If it is determined that CountBk <Th (step S212; No), FALSE is set in the black area determination result AreaBk (X, Y) (step S214). And a process transfers to step S12 of FIG.

  In step S12 of FIG. 12, it is determined whether AreaBk (X, Y) = TRUE. If it is determined that AreaBk (X, Y) = TRUE (step S12; TRUE), R-color misregistration correction processing (step S13) and B-color misregistration correction processing (step S14) are executed. The process proceeds to S17. Since the R-color misregistration correction process and the B-color misregistration correction process are the same as those described in the first embodiment, the description will be cited. On the other hand, if it is determined that AreaBk (X, Y) = FALSE (step S12; FALSE), the value of R (X, Y) is set in Rout (X, Y) (step S15), and Bout (X , Y) is set to B (X, Y) (step S16), and the process proceeds to step S17.

  In step S17, G (X, Y) is set in Gout (X, Y) (step S17), and the color misregistration correction process B ends.

  In FIG. 12, the R-color misregistration correction process is followed by the B-color misregistration correction process, and then G (X, Y) is set to Gout (X, Y). The order does not matter. These processes may be performed in parallel.

The black region determination window is not limited to that shown in FIG. 14A, and the black region determination window shown in FIGS. 14B and 14C can also be used.
The black area determination window shown in FIG. 14B is an example suitable for a case where TF deterioration, MTF deterioration of the optical system, color misregistration amount, and color misregistration of the document itself are small.
The black area determination window shown in FIG. 14C is an example suitable for the case where the MTF degradation of the optical system, the color misregistration amount, and the color misregistration of the document itself are larger in the sub scanning direction than in the main scanning direction.

  In the second embodiment, in the black area determination process, the number of black pixels in the black area determination window is simply counted to determine whether or not the surrounding area of the target pixel is a black area. As a modification, each pixel in the black pixel determination window may be weighted according to a relative position from the target pixel.

  FIG. 17 shows an example of a weighting filter for the black area determination window shown in FIG. 14A (a filter storing the weighting coefficient of each pixel in the black area determination window). The left side of FIG. 17 shows the coordinates of the black area determination window, and the right side shows the weighting coefficient. In the weighting filter of FIG. 17, the weighting coefficient is larger as the pixel is closer to the target pixel. Further, the weighting coefficient is larger in the sub-scanning direction between the main scanning direction and the sub-scanning direction.

  When the weighting filter as shown in FIG. 17 is used, in step 11 of FIG. 12, processing is performed in the flow of black area determination processing B shown in FIG. In the following description, the case where the black region determination window shown in FIG. 14A is used will be described as an example. However, the black region determination may be performed using the black region determination windows shown in FIGS.

  In the black area determination process B, first, a black area determination window is set (step S301).

  Next, an initial value 0 is set in the counter Sum (step S302), -wy is set in the variable i, and -wx is set in the variable j (step S303, step S304).

  Next, the black pixel / color pixel discrimination process shown in FIG. 15 is executed for the pixel (X + j, Y + i) (step S305).

  Next, it is determined whether JdgBk (X + j, Y + i) is TRUE or FALSE. When it is determined that JdgBk (X + j, Y + i) is TRUE (step S306; TRUE), the weighting filter shown in FIG. 14C is referred to, and the weighting coefficient of WtBk (X + j, Y + i) is added to the value of the counter Sum ( In step S307), the process proceeds to step S308. If it is determined that JdgBk (X + j, Y + i) is FALSE (step S306; FALSE), the process proceeds to step S308.

  In step S308, the variable j is incremented by 1 (step S308), and it is determined whether j> wx is satisfied (step S309). If it is determined that j> wx is not satisfied (step S309; No), the process returns to step S305. If it is determined that j> wx (step S309; Yes), the variable i is incremented by 1 (step S310), and it is determined whether i> wy is satisfied (step S311). If it is determined that i> wy is not satisfied (step S311; No), the process returns to step S304. If it is determined that i> wy (step S311; Yes), the process proceeds to step S312.

  In step S312, it is determined whether or not the value of Sum is greater than or equal to a predetermined threshold Th. The threshold value Th is a value determined by comparing and considering the deterioration in image quality due to weakening correction and the deterioration in image quality due to correction other than black pixels, for example, 4. If it is determined that Sum ≧ Th (step S312; Yes), TRUE is set in the black region determination result AreaBk (X, Y) (step S313). If it is determined that Sum <Th (step S312; No), FALSE is set in the black region determination result AreaBk (X, Y) (step S314). And a process transfers to step S12 of FIG. 12, and the process after step S12 is performed.

In the black area determination process B, for example, when Th = 4, color shift correction is performed when even one of the pixels whose weighting coefficient is set to 4 in FIG. In the case of other pixels, color misregistration correction is not performed unless it is determined that at least two pixels are black pixels.
As described above, by increasing the value of the weighting coefficient at a position close to the target pixel, there is an effect that a more stable black region determination result can be obtained even at the boundary between the determination of the black pixel and the color pixel.

  The processing of the color misregistration correction unit 34 described in the second embodiment and the modification may be realized by dedicated hardware, or the color misregistration correction processing program P4 shown in FIG. A program for causing the CPU to execute the process B may be included, and the function of the color misregistration correction unit 34 may be realized by software processing in cooperation with the CPU and the color misregistration correction processing program P4.

  Further, in the present embodiment, it has been described that the pixel of interest is included in the black region determination window, but the black region determination window is set only to the peripheral pixels, and the periphery of the pixel of interest is set according to the number of black pixels of the peripheral pixels It may be configured to determine whether or not is a black region.

[Third Embodiment]
Next, a third embodiment of the present invention will be described.
As described above, the image reading apparatus 1 can switch between the two document reading methods, the ADF document reading method and the FB document reading method. In the ADF document reading method, since the document is moved in the sub-scanning direction with respect to the sensor module using a plurality of rollers as described above, color misregistration is compared to the FB document reading method in which the document is fixed on the platen glass 201. The amount is large. Therefore, in the present embodiment, the document reading method of the reading unit 20b is determined before performing the color misregistration correction processing described in the first embodiment or the second embodiment, and the document reading method is used. Thus, the values of Kv and Kh used when setting the virtual pixel are automatically switched.

In the third embodiment, the ROM of the control unit 10 stores KvFB, which is a Kv value corresponding to the FB document reading method, KhFB, which is a Kh value corresponding to the FB document reading method, and Kv, corresponding to the ADF document reading method. The value KvADF and the value KhADF which is the Kh value corresponding to the ADF document reading method are stored. KvFB <KvADF and KhFB <KhADF.
The configuration of each part of the other image reading apparatus 1 and the operation of the color misregistration correction unit 34 are the same as those described in the first embodiment or the second embodiment.

FIG. 19 is a flowchart illustrating a scan control process executed by the control unit 10 when a scan start instruction is input from the operation unit 50.
First, it is determined whether or not the current scan start instruction is a first scan start instruction (step T1). Here, scanning refers to reading a document. For example, the scan that is performed for the first time after a document is placed on the ADF unit 20a (a document detection sensor is output from the document detection sensor) is determined to be the first scan. In addition, a scan performed for the first time after the opening detection signal is output from the open / close sensor and a scan performed after a predetermined time has elapsed are determined to be the first scan.

  If it is determined in step T1 that this is the first scan (step T1; Yes), the process proceeds to step T3. If it is determined that it is not the first scan (step T1; No), it is determined whether or not the document reading method is the same as the previous scan (step T2). Here, the RAM of the control unit 10 stores Kv and Kh values corresponding to the previous document reading method. In the control unit 10, for example, when scanning is instructed in a state where a document presence detection signal is output from the document detection sensor, it is determined that the ADF document reading method is used. When scanning is instructed in a state where an open detection signal is output from the open / close detection sensor, it is determined that the FB document reading method is used. If the determination result matches the document reading method corresponding to the Kv and Kh values stored in the RAM, it is determined that the document reading method is the same as the previous one. Therefore, it is determined that the document reading method is different from the previous one.

  When it is determined in step T2 that the document reading method is the same as the previous time (step T2; Yes), scanning by the same document reading method as the previous time is instructed to the reading device 20, and scanning processing is executed in the reading device 20. (Step T7).

  If it is determined that the document reading method is different from the previous one (step T2; No), it is determined whether the current document reading method is the ADF document reading method or the FB document reading method (step T3). When it is determined that the current document reading method is the FB document reading method (step T3; FB), the Kv value stored in the RAM is replaced with KvFB, and the Kh value is replaced with KhFB (step T4). The process proceeds to step T6. When it is determined that the current document reading method is the ADF document reading method (step T3; ADF), the Kv value stored in the RAM is replaced with KvADF, and the Kh value is replaced with KhADF (step T5). The process proceeds to step T6.

  In step T6, the values of Kv and Kh stored in the RAM are set in the register of the color misregistration correction unit 34 (step T6). Then, scanning by the document reading method determined in step T3 is instructed to the reading device 20, and scanning processing is executed by the reading device 20 (step T7).

  According to the third embodiment, in the ADF document reading method having a large color shift amount compared to the FB document reading method, the Kv and Kh used when setting virtual pixels compared to the FB document reading method. The value can be increased automatically. Therefore, a good correction result according to the document reading method can be obtained.

  As described above, the color misregistration correction unit 34 of the image processing device 30 sets a virtual pixel between a target pixel and an adjacent pixel adjacent to the target pixel (R, B) in the correction target color (R, B) image. Based on the gradation value of the pixel and the gradation value of the adjacent pixel, the gradation value of the virtual pixel is calculated, and the gradation value of the reference color (G) of the target pixel is a virtual composed of the target pixel and the virtual pixel of the correction target color. When the gradation value of the pixel group is not less than the minimum value and not more than the maximum value, the gradation value of the target pixel of the correction target color is converted to the gradation value of the reference color.

  Therefore, if the tone value of the reference color (G) of the target pixel is not less than the minimum value and not more than the maximum value of the tone value of the virtual pixel group consisting of the target pixel and the virtual pixel of the correction target color (R, G), Since the gradation values of the reference color and the color to be corrected in the pixel match, color misregistration around the black characters in the main scanning direction and sub-scanning direction caused by chromatic aberration, mechanical vibration, etc. can be achieved without increasing the circuit scale. It can be completely eliminated with simple processing. Further, not only the coloring due to the color shift of the edge portion of the achromatic color portion such as a black character can be suppressed, but also an excellent correction effect can be obtained for the black reproduction inside the black character. Furthermore, it is difficult to depend on the strength of the edge, and a stable color shift suppressing effect can be obtained.

  Further, when the gradation value of the reference color of the pixel of interest exceeds the maximum value of the gradation value of the virtual pixel group, the color misregistration correction unit 34 determines the gradation value of the pixel of interest of the correction target color as the level of the virtual pixel group. When the tone value of the reference color of the target pixel is less than the minimum value of the virtual pixel group, the tone value of the target pixel of the correction target color is converted to the level of the virtual pixel group. Convert to the minimum key value. Therefore, appropriate color misregistration correction can be performed according to the relationship between the gradation value of the reference color and the gradation value of the pixel of interest and each pixel adjacent thereto. For example, when the periphery of the target pixel is a single chromatic color, the color misregistration correction is not necessary, so the color misregistration correction is not substantially applied.

  In addition, since the interval between the target pixel and the virtual pixel can be set to a different value in the main scanning direction Kh and the sub scanning direction kv, color misregistration corresponding to each color misregistration amount in the main scanning direction and the sub scanning direction. Correction can be performed.

  Further, by configuring the values of Kv and Kh so that they can be set by the operation unit 50, it is possible to change the correction amount of the color misregistration correction according to the reading resolution and the conveyance speed.

  Further, the control unit 10 switches the values of Kv and Kh depending on whether the reading device 20 reads the document in the stationary state and reads it by the ADF document reading method or moves the document and reads it by the FB document reading method. As a result, the optimum color misregistration correction can be performed for each document reading method. In particular, in the ADF original reading system that reads and conveys an original, since the original passes between a plurality of rollers and the friction is not constant depending on the image quality state of the original, an unevenness in the conveying speed is likely to occur. By setting the values of Kv and Kh to greater than those of the FB document reading method, it is possible to obtain a good correction result.

Further, the color misregistration correction unit 34 determines whether the pixel of interest is a black pixel or a color pixel with respect to peripheral pixels (each pixel in the black area determination window) existing within a predetermined range from the pixel of interest. Since the gradation value of the target pixel of the correction target color is converted only when the number of pixels determined to be black pixels exceeds a predetermined threshold, color halftone dots, color thin line colors, etc. It is possible to perform the color misregistration correction of the achromatic color portion while suppressing a decrease in saturation. In addition, a weighting factor is associated with each pixel in the black region determination window according to the positional relationship with the target pixel, and a value obtained by multiplying the pixel determined to be a black pixel by the weighting factor is set as a predetermined threshold value. By converting the gradation value of the target pixel of the correction target color only when the value exceeds the threshold value, it is possible to obtain a more stable correction result for the boundary between the determination of the black pixel and the color pixel.
It is possible to easily determine whether each of the target pixel and the peripheral pixels is a black pixel or a color pixel based on the saturation and brightness of the target pixel and the peripheral pixels.

  The descriptions in the first to third embodiments show a preferred example of the image processing apparatus 30 according to the present invention, and the present invention is not limited to this.

  For example, in the first to third embodiments, the case where the image reading apparatus 1 includes the image processing apparatus 30 according to the present invention has been described as an example. However, the present invention is not limited to this. For example, an image output device such as a printer or display device connected to the image reading device 1 is provided with the image processing device 30, and the image data output from the image reading device includes the above-described color misregistration correction processing. It is good also as a structure which performs a process. Further, the image processing apparatus 30 may be provided in a copier or MFP (Multi Function Printer).

  In the above description, an example in which a ROM is used as a computer-readable medium for the program according to the present invention is disclosed, but the present invention is not limited to this example. As other computer-readable media, portable recording media such as HDDs and CD-ROMs can be applied. A carrier wave is also applied as a medium for providing program data according to the present invention via a communication line.

  In addition, the detailed configuration and detailed operation of each device can be changed as appropriate without departing from the spirit of the present invention.

DESCRIPTION OF SYMBOLS 1 Image reading apparatus 10 Control part 20 Reading apparatus 20a ADF part 20b Reading part 21 Document placing tray 22 Paper feed roller 23 Contact roller 24 Guide roller 25 Switching claw 26 Reverse roller 27 Paper discharge tray 28 Document cover 201 Platen glass 202 Image sensor 202R R sensor 202G G sensor 202B B sensor 203 Analog processing unit 204 A / D converter 30 Image processing device 31 Shading correction unit 32 γ correction unit 33 Inter-line correction unit 34 Color shift correction unit 35 Enlargement / reduction processing unit 36 Color conversion Processing unit 37 Image correction unit 38 Image area identification processing unit 40 Communication I / F
50 Operation unit 60 Display unit 70 Bus

Claims (10)

An image processing apparatus that performs image processing on an image obtained by reading a document using a plurality of image sensors arranged in parallel to each other and separated into a plurality of colors,
One of the color-separated colors is set as a reference color, the other at least one color is set as a correction target color, a virtual pixel is set between a target pixel of the correction target color and an adjacent pixel adjacent thereto, The gradation value of the virtual pixel is calculated based on the gradation value of the target pixel and the gradation value of the adjacent pixel, and the reference pixel gradation value of the target pixel is the target pixel of the correction target color and the virtual pixel. A color misregistration correction unit that converts the gradation value of the pixel of interest of the correction target color into the gradation value of the reference color when the gradation value of the virtual pixel group consisting of Image processing device.   When the gradation value of the reference color of the pixel of interest exceeds the maximum value of the gradation value of the virtual pixel group, the color misregistration correction unit determines the gradation value of the pixel of interest of the correction target color as the virtual pixel group. When the gradation value of the reference color of the target pixel is less than the minimum value of the virtual pixel group, the gradation value of the target pixel of the correction target color is The image processing apparatus according to claim 1, wherein the image processing apparatus converts the gradation value of the virtual pixel group to a minimum value.   The image processing apparatus according to claim 1, wherein an interval between the target pixel and the virtual pixel can be set to a different value in a main scanning direction and a sub scanning direction.   The image processing apparatus according to claim 1, further comprising an operation unit configured to set an interval between the target pixel and the virtual pixel in each of a main scanning direction and a sub-scanning direction.   The interval between the pixel of interest and the virtual pixel is set to a different value depending on whether the reading device reads the document in a stationary state or moves and moves the document. An image processing apparatus according to claim 1.   The color misregistration correction unit determines whether the pixel of interest and the surrounding pixels existing within a predetermined range from the pixel of interest are black pixels or color pixels, and the pixels determined to be black pixels 6. The image processing apparatus according to claim 1, wherein the gradation value of the target pixel of the correction target color is converted only when the number exceeds a predetermined threshold.   A weighting factor is associated with the target pixel and the peripheral pixel according to a positional relationship with the target pixel, and the color misregistration correction unit multiplies the pixel determined to be the black pixel with the weighting factor. The image processing apparatus according to claim 1, wherein the gradation value of the target pixel of the color to be corrected is converted only when the integrated value exceeds a predetermined threshold value.   The color misregistration correction unit determines whether each of the target pixel and the peripheral pixel is a black pixel or a color pixel based on saturation and brightness of the target pixel and the peripheral pixel. Or the image processing apparatus according to 7. An image processing method for performing image processing on an image obtained by reading a document using a plurality of image sensors arranged in parallel to each other and separated into a plurality of colors,
A virtual pixel is set between a target pixel of the correction target color and an adjacent pixel adjacent to the target pixel of the correction target color, with one of the color-separated colors as a reference color and at least one other color as a correction target color, The gradation value of the virtual pixel is calculated based on the gradation value of the target pixel and the gradation value of the adjacent pixel, and the reference pixel gradation value of the target pixel is the target pixel of the correction target color and the virtual pixel. An image processing method including a step of converting a gradation value of a target pixel of the color to be corrected into a gradation value of the reference color when the gradation value of the virtual pixel group consisting of is not less than a minimum value and not more than a maximum value. A computer used in an image processing apparatus that performs image processing on an image obtained by reading a document using a plurality of image sensors arranged in parallel to each other and separated into a plurality of colors,
One of the color-separated colors is set as a reference color, the other at least one color is set as a correction target color, a virtual pixel is set between a target pixel of the correction target color and an adjacent pixel adjacent thereto, The gradation value of the virtual pixel is calculated based on the gradation value of the target pixel and the gradation value of the adjacent pixel, and the reference pixel gradation value of the target pixel is the target pixel of the correction target color and the virtual pixel. A color misregistration correction unit that converts a gradation value of a pixel of interest of the correction target color into a gradation value of the reference color when the gradation value of the virtual pixel group consisting of
Program to function as.

Images