JP2002247292A - Image reader - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image reader which appropriately corrects deviation of reading position in accordance with a situation and effectively prevents a problem that resolution is largely damaged, a thin line is blurred, a black line and a black character are colored, a black/white original is erroneously recognized as a color original and coloring largely changes due to magnification. SOLUTION: The image reader is provided with a scanning means which optically scans the original at prescribed speed in an auxiliary scanning direction, the line sensors of a plurality of stages, which at arranged in parallel with prescribed intervals, to which light from the scanned original is inputted and which output image signals, and a correcting means correcting deviation of reading position in the auxiliary scanning direction, which is included in the outputted image signals with a prescribed position as a reference. The correcting means sets the prescribed position being the reference at an arbitrary position in the auxiliary scanning direction.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image reading apparatus for optically reading an image, and more particularly to processing for outputting image signals from a plurality of line sensors arranged in parallel.

[0002]

2. Description of the Related Art Heretofore, a color image reading apparatus using three line sensors respectively corresponding to red, green, and blue, which are arranged in parallel at a fixed interval, has been widely known. In such a color image reading apparatus, the reading position of each line sensor is slightly shifted due to the distance between each line sensor, the reading magnification (scanning speed), and the like. Gap (gap)
Needs to be corrected.

[0003] From such a viewpoint, Japanese Patent Laid-Open Publication No.
No. 66, Japanese Patent Application Laid-Open No. 9-270988, etc.
A line sensor corresponding to the rear end in the sub-scanning direction (for example,
With the reading position of the line sensor corresponding to blue (fixed) as a reference, the image signal from the line sensor (for example, the line sensor corresponding to red and green) preceding the line sensor in the sub-scanning direction is delayed. A technique has been proposed in which the reading position of each line sensor is made to match by interpolation.

FIG. 21 schematically illustrates such an image reading apparatus. This image reading apparatus converts an original D placed on a plante glass through a reduction optical system,
R: Red, G: Green, B: Read by line sensors corresponding to each color of blue. here,
The distance between the line sensors is equal to two lines (pixels).

FIG. 1 explains the reading position of each line sensor on the platen glass. FIG. 21 (a),
As shown in FIG. 22A, when the reading magnification is 100%, the line sensor corresponding to green actually advances in the sub-scanning direction (−) by two lines more than the line sensor corresponding to blue. I'm reading the position. Further, the line sensor corresponding to red reads a position advanced by four lines in the sub-scanning direction (-) from the line sensor corresponding to blue. Further, as shown in FIGS. 21 (b) and 22 (b), when the reading magnification is 25% (25% reduction), the line sensor corresponding to green is actually higher than the line sensor corresponding to blue. Also read a position advanced by 0.5 line in the sub-scanning direction (-). Further, the line sensor corresponding to red reads a position advanced by one line in the sub-scanning direction (-) from the line sensor corresponding to blue.

FIG. 23 is a diagram for explaining the structure of a conventional concrete means for correcting a deviation of a reading position. Of the image signals RGB simultaneously output from each line sensor, R
For the image signals G and G, the reading position is corrected by the correction means. As a specific configuration of this correction means,
A gap memory that corrects the displacement of an integer number of pixels by delaying the output of the image signal, and a line memory and an interpolation circuit that corrects the displacement of a few pixels of the displacement by interpolating the preceding and following image signals, respectively. It is provided independently for each image signal (R and G). Further, a CPU for transmitting a control command to the gap memory, the interpolation circuit, and the like at a predetermined timing is provided.

[0007] For example, FIG.
In the case of 100% reading magnification as shown in FIG.
The PU controls each gap memory so as to output after holding (temporarily) the image signal R for four lines and the CPU outputs the image signal G after holding (temporarily) two lines. . Also, FIG. 21 (b), FIG. 22 (b)
In the case of a reading magnification of 25% as shown in FIG. 7, the CPU holds (temporarily) the image signal R for one line and then outputs the image signal R. The CPU interpolates the image signal G with the interpolation coefficient 0.5. The gap memory and the interpolation circuit are controlled so as to output the result. By performing such displacement correction, the displacement can be corrected as shown by arrows in FIGS. 22 (a) and 22 (b).

[0008]

However, in some cases, when interpolating positional deviations of sub-pixels, the resolution is greatly impaired, thin lines are blurred, black lines and black characters are colored, and black and white originals are erroneously recognized as color originals. And the color tone changes greatly depending on the magnification.

FIGS. 24 and 25 illustrate such a problem. Here, in the conventional image reading apparatus, a document D horizontal to the main scanning direction (perpendicular to the sub-scanning direction) is used.
A case where a black line horizontal in the main scanning direction (perpendicular to the sub-scanning direction) is read at a reduced scale of 25% will be described. FIG. 24 (a)
Indicates a black line portion in the document D, and the black line portion means that the reflectance is low (black). This is read by a line sensor RGB (see FIG. 21), and each image signal R
FIG. 24 shows the result of correcting the displacement of the integer pixels of GB.
This is shown in FIG. The image signal R is delayed by the gap memory and coincides with the image signal B. on the other hand,
Here, the displacement of the sub-pixel of the image signal G has not been corrected yet.

FIG. 25A shows a state in which the displacement of the sub-pixels of the image signal G is interpolated by the line memory and the interpolation circuit, and the displacement of the reading position of the image signal RGB is completely corrected. However, the interpolation coefficient of the image signal G is 0.5, and the distribution of the image signal G component has spread. Therefore, when the obtained image signal RGB is subjected to chroma conversion, it becomes as shown in FIG. 25B, and relatively large chroma components appear at both ends of the black line in the sub-scanning direction. Further, when the obtained image signal RGB is subjected to luminance conversion, FIG.
As shown in (c), the graph becomes a mountain-like graph that is duller than the original luminance.

The appearance of a saturation component in a place where it should not exist causes problems such as black lines and black characters being colored, and a monochrome document being erroneously recognized as a color document.
Further, since most of the luminance information depends on the image signal G, if the image signal G becomes dull, there arises a problem that the resolution is greatly impaired or a thin line is blurred when binarizing the image. Further, the magnitude of the interpolation coefficient, that is, the value of a sub-pixel of the positional deviation depends on the magnification of image reading (scanning speed). Therefore, if the magnification of image reading is different, the RGB balance is lost and the color tone is changed. .

The present invention has been made in view of these technical problems, and an object of the present invention is to perform appropriate reading position deviation correction according to the situation, and to greatly reduce the resolution and blur fine lines. It is an object of the present invention to provide an image reading apparatus capable of effectively preventing problems such as coloring of black characters, erroneous recognition of a black-and-white document as a color document, and a large change in color depending on magnification.

[0013]

That is, the present invention relates to a scanning means for optically scanning a document at a predetermined speed in a sub-scanning direction, and a scanning means which is disposed parallel to each other at a predetermined distance and is scanned. An image reading apparatus comprising: a plurality of stages of line sensors that receive light and output image signals; and a correction unit that corrects a reading position shift in the sub-scanning direction included in each output image signal with reference to a predetermined position. The correction means can set the predetermined reference position to an arbitrary position in the sub-scanning direction.

Since the reference position is set arbitrarily as described above, the degree of freedom in correcting the reading position deviation is increased, and the reading position deviation can be appropriately corrected according to the situation. Can be effectively prevented from being greatly damaged, thin lines are blurred, black lines and black characters are colored, a black-and-white document is erroneously recognized as a color document, and a color tone is largely changed by magnification.

Here, the intervals between the line sensors may be different or the same. In general, it is not necessary to perform an interpolation process when reading an image with a reading magnification of 100%, which is the most frequently used image. Therefore, it is preferable that the interval between each line sensor is N pixels (N is an integer). When light is input, each line sensor outputs an image signal for one pixel in the sub-scanning direction, and the line sensors in a plurality of stages output image signals (substantially) simultaneously.

The reading position shift in the sub-scanning direction occurs due to the scanning speed of the document, the reading magnification of the document, the interval between the line sensors, and the like. Therefore, it is preferable that the correction unit sets (selects again) the predetermined position to be the reference based on the scanning speed of the document, the reading magnification of the document, and whether the scanning is going forward or backward.

This read position shift is expressed in units of the number of read units (lines, pixels) in the sub-scanning direction, and the read position shift X
(Line), I for the integer line is I = [X]
(Line) and its fractional line D can be expressed as D = X- [X] (line). [A] is A
Represents the integer part of. For example, when X = 2.7 (lines), D = 2 (lines) and I = 0.7 (lines).

The correction means may be capable of arbitrarily setting the predetermined reference position other than the reading position corresponding to each of the plurality of stages of line sensors. Further, the correction means may be such that it can set a predetermined reference position behind the last reading position in the sub-scanning direction among a plurality of reading positions corresponding to each line sensor. Further, the correction means may include, among a plurality of reading positions corresponding to each line sensor, a predetermined position as a reference in a region from the last reading position in the sub-scanning direction to one reading unit (line, pixel) later. May be settable.

A delay unit for correcting an integer line I of a reading position shift X in the sub-scanning direction, which can be included in each image signal from the predetermined reference position, by delaying the output of the image signal; An interpolating unit that corrects a fractional line D of a reading position shift X in the sub-scanning direction that can be included in each image signal from the predetermined position as a reference by interpolation with the preceding and succeeding image signals. it can.

The correction means can set the predetermined reference position to an arbitrary position in the sub-scanning direction. The correction means can be roughly divided into two modes based on the "arbitrary position". . One is to set a predetermined position as a reference with emphasis on resolution, and the other is to set a predetermined position as a reference with emphasis on saturation.

First, a description will be given of the former mode in which a predetermined position is set as a reference with emphasis on resolution. In this aspect, the correction means sets the reference value such that the deviation between the reading position corresponding to any one of the line sensors and the predetermined reference position does not always include the D element for the decimal line (D = 0). Is set. Therefore, for example, when the plurality of line sensors are three-stage line sensors corresponding to the three colors of red, green, and blue, the correction unit may determine the reading position of each of the red, green, and blue line sensors. The reference position can be set so that the deviation from one of the reading positions to the reference position does not include the D element for the decimal line.

Here, when the reference position is set so that the difference between the reading position of the red line sensor and the reference position does not include the D element for a fractional number of lines, the red resolution priority mode in which the resolution of red is hardly deteriorated. Becomes Similarly, when the reference position is set such that the difference between the reading position of the green line sensor and the reference position does not include the D element for a fractional number of lines, the green resolution is less likely to degrade, and the green resolution priority mode is used. When the reference position is set such that the difference between the reading position of the line sensor and the reference position does not include the D element for the decimal line, the blue resolution priority mode is set in which the blue resolution is hardly deteriorated. Furthermore, since most of the luminance information is carried by green light, the above-described green resolution priority mode can be said to be simultaneously a luminance priority mode. It is preferable that the image is read in the brightness priority mode when the document size is detected or when a monochrome document is read.

Next, a description will be given of the latter mode in which the predetermined position is set as the reference with emphasis on saturation. In this aspect, when the number of sub-lines is D ≠ 0, the reference value is set so that all of the image signals output from the line sensors in a plurality of stages can include the sub-line D of the reading position shift in the sub-scanning direction. Is set. As a result, the correction means can correct all of the image signals output from the plurality of stages of line sensors.

[0024] The correcting means may adjust the reference position and the reference position so that the sub-scanning position deviation D in the sub-scanning direction included in each image signal output from (all of) the plurality of line sensors is equivalent. The reference position may be set such that the sub-scanning reading position shift fractional line number D included in each image signal output from a part of the plurality of stages of line sensors is equivalent. May be set.

Here, "equivalent" means that, for example, when there are a plurality of line positions D (1) to D (n) corresponding to a plurality of line sensors (1) to (n) and the read position is shifted.
D (1) ≒ D (2) ≒... ≒ D (n) or D
.. = D (n), or | D
(1) | ≒ | D (2) | ≒... | D (n) | or | D (1) | = | D (2) | =... = | D (n) | D (1) -0.5 | ≒ | D (2) -0.5 | ≒ ...
≒ | D (n) −0.5 | or | D (1) −0.
5 | = | D (2) -0.5 | = ... = | D (n) -0.5
| Is satisfied.

More specifically, when the plurality of line sensors are three-stage line sensors corresponding to three colors of red, green and blue, the correction means outputs the signals from these three-stage line sensors. The predetermined reference position may be set so that the sub-scanning direction read position shift sub-fractional lines D included in each image signal are equivalent. When the predetermined position is set as described above, the color balance of the image to be read becomes good, so that the saturation priority mode can be set. At the time of reading a color original, at the time of automatic color original identification, at the time of automatic density correction level detection, etc., it is preferable to perform image reading in the saturation priority mode.

Any two of these three-stage line sensors (red and green, green and blue,
The reference position may be set such that the sub-scanning read position deviation fraction line D included in each image signal output from any of blue and red) is equivalent.

[0028]

Embodiments of the present invention will be described below based on embodiments.

Embodiment FIG. 1 is a schematic sectional view of a color image reading apparatus according to the present embodiment. This image reading apparatus is used as a scanner of a copying machine. The image reading apparatus is provided with a casing 10 and a full-length scanning direction within the casing 10 along a sub-scanning direction (thick arrow in FIG. A full-rate carriage (scanning means) 11 that moves by a stroke, a half-rate carriage (scanning means) 12 that also moves by half a stroke in the scanning direction, a platen glass 13 on which a document D is placed, and a placed document is pressed. And a platen cover 14.

In the full-rate carriage 11, a lamp (scanning means) La as a light source and a mirror (scanning means) are provided.
The half-rate carriage 12 houses mirrors (scanning means) M2 and (scanning means) M3. Further, an imaging lens (scanning means) L is arranged on the optical axis constituted by the mirrors M1, M2, M3, and a CCD (Charge Coupled D) is provided at an imaging position of the imaging lens L.
evice) 1 is arranged. An image processing unit 2 formed on a substrate is disposed below the housing 10, and a CCD 1 is provided.
And are electrically connected via a harness. A white reference plate WR is arranged adjacent to an end of the platen glass 13. The white reference plate WR is used when capturing shading correction data and white fluctuation correction data.

To briefly explain the image reading by this image reading apparatus, first, an original is placed on the platen glass 13 and the platen cover 14 is closed. Thereafter, light is emitted from the lamp La toward the document along with the reading instruction. For example, the full-rate carriage 11 moves all the strokes in the forward scan direction (outbound), and the half-rate carriage 12 moves half a stroke in the scan direction. I do. By these movements, the reflected light from the original is transferred to mirror
1, M2, M3, reflected by the imaging lens L and CCD
It is taken into 1. The CCD 1 converts the optical image of the document into an electric signal in line units in synchronization with the scanning, and sequentially sends the electric signal to the image processing unit 2.

FIG. 2 illustrates the structure of the CCD 1 shown in FIG. 1 in more detail. In this embodiment, the CCD 1
Are optical images L (1) (2) (2) corresponding to red (R), green (G), and blue (B) input through reduction optical systems (mirrors M1, M2, M3, and imaging lens L). 3) take in each 3)
The line sensors LS (1), (2), and (3) of the stage are arranged in parallel in the main scanning direction (hereinafter, sometimes abbreviated as FS direction) at equal intervals of two lines. Output signals O (1) of channels Ch1 and Ch2 are output from the red (R) line sensor, output signals O (2) of channels Ch3 and Ch4 are output from the line sensor of green (G), and a blue (B) line. The output signals O (3) of the channels Ch5 and Ch6 are output from the sensors to the corresponding analog LSs of the image processing unit 2.
Sent to I200.

Each of the line sensors LS (1) to LS (3)
Can be arbitrarily selected and designed to correspond to any one of RGB. In this embodiment, the reading position on the platen glass 13 (L (3)) that precedes in the SS direction (in the forward path) is indicated by red. (R) and the last stage in the SS direction (L (1)) corresponded to blue (B). It is assumed that the CCD 1 used in this embodiment includes three line sensors each including 5000 pixels.

FIG. 3 is a block diagram for explaining the structure of the image processing section 2 shown in FIG. 1 in more detail. Hereinafter, a description will be given from the upstream side to the downstream side according to the flow of the output signal (image signal) O (1) to (3).

The analog LSI 200 outputs an output signal O (1) from each of the line sensors LS (1) to LS (3) of the CCD 1.
To (3), AGC (Auto Gain Control) processing and AOC (Auto Offset Control) processing are performed. A / D
(Analog / digital conversion) circuit 201
Obtain output signals O (1) to O (3) from SI 200 and convert them to digital signals. Light intensity distribution correction unit 202
Performs shading correction, W-ref correction (correction of the white level of the white reference plate), and white variation correction corresponding to each of the line sensors LS (1) to LS (3).

The gap correction section (correction means) 3 includes an upstream selection section 203, an integer line correction section 204, a decimal line correction section 205, and a downstream selection section 206. These are further described below.

The color conversion unit 207 converts L * a *
It performs color conversion to the b-system, and is divided into a detection system color conversion unit 207a and an image data output system color conversion unit 207b. The color conversion units 207a and 207b perform appropriate color conversion based on different LUTs (look-up tables) and operation coefficients according to their purposes. The output from the detection system color conversion unit 207a is input to the detection unit 208,
Therefore, document detection, ACS (automatic color document identification), AE
(Automatic density correction level detection) and the like. On the other hand, the output from the image data output system 207b is input to the text / image determination unit 209 and the color conversion unit 210.

The color conversion unit 210 performs color conversion (L *) optimized for the characteristics of the image forming apparatus that finally outputs an image based on the detection result from the detection unit 208 and the attribute from the text / image determination unit 209. a * b system to YMCK system)
Is performed. The scaling unit 211 scales both the attribute from the determination unit 209 and the image signal. In the filter unit 212, filtering is performed based on the attribute from the determination unit 209 and the detection result from the detection unit 208. Gamma correction unit 2
In step 13, the gamma correction and base processing optimized for the configuration of the image forming apparatus and the copy mode are performed, and the image forming apparatus IO
An image signal is output to T.

Each element 200 constituting the image processing section 2
213 are provided with a CPU 20 for performing various arithmetic controls, a ROM 21 for storing arithmetic methods and control methods as programs, and a RAM 22 as a work area for the CPU 20.
1, a RAM 22, and each of the image processing elements 200 to 213 are connected by a bus 24, and various control commands are exchanged.

Next, the structure of the gap correction unit 3 will be described with reference to FIGS.
This will be described in detail with reference to FIG.

FIG. 4 illustrates the structure and operation of the upstream selector 203 and the downstream selector 206. The upstream selector 203 determines which of the three output signals O (1) to (3) should be delayed. The downstream selector 206 prevents the upstream selector 203 from affecting the downstream image processors 207 to 213.

In the present embodiment, the forward scan direction (outgoing path)
That is, when scanning is performed on the minus side in the SS direction (see FIG. 2A), the selection units 203 and 204 output the output signals O (1) to O (1) to ( 3) Select the route. As a result, the output signals O (1) and O (2) are
Is corrected by On the other hand, when scanning is performed in the reverse scan direction (return), that is, in the plus direction in the SS direction (FIG. 2).
(See FIG. 4A). In response to an instruction from the CPU 20, each of the selectors 203 and 204 outputs an output signal O as shown in FIG.
The routes (1) to (3) are selected. As a result, the output signals O (3) and O (2) are corrected by the integer line correction unit 204. Also in the case of this reverse scan, the image processing unit 20 on the further downstream side
Processes 7 to 213 are performed in the same manner as in the case of the forward scan.

In the following description, it is assumed that scanning is performed in a forward scan unless otherwise specified. As described later, the output signal O in the case of the forward scan is output.
(3) Delay correction may be performed by the decimal line correction unit 205 on the output signal O (1) in the case of reverse scanning.

FIG. 5 illustrates the configuration of the integer line correction unit 204. This integer line correction unit 204
Delay memory 30 provided for output signal O (1)
(1) and a delay memory 30 (2) provided for the output signal O (2). Each of these delay memories 30 (1) and (2) is a FIFO (First In First
Out) memory, the capacity of which is determined by the interval between the line sensors LS (1) to LS (3) and the maximum reading magnification. In this embodiment, the line sensor LS
The interval between (1) and the line sensor LS (3) is four lines (see FIG. 2B), and the maximum magnification of this image reading device is 4
Assuming that it is 00%, 4 × (400/100) = 16, so that the capacity of the delay memory 30 (1) needs to have a capacity to store output signals for 16 lines. On the other hand, the distance between the line sensor LS (2) and the line sensor LS (3) is two lines (see FIG. 2B), and if the maximum magnification of this image reading apparatus is also 400%, 2 × (400/400/400)
Since 100) = 8, the capacity of the delay memory 30 (2) needs to have a capacity to store output signals for eight lines.

Then, when scanning is actually performed,
The number of lines to be stored in each delay memory 30 (1) (2) is determined based on the delay coefficient I (1) (2) specified by the CPU 20. For example, if the delay coefficient I (1) = 2 and the delay coefficient I (2) = 1, the delay memory 30 (1) stores the output signals O (1) for two lines and sequentially outputs them by FIFO, The delay memory 30 (2) outputs the output signal O for one line.
(2) is stored and sequentially output by FIFO. If the delay coefficient I (1) = 1 and the delay coefficient I (2) = 0, the delay memory 30 (1) stores the output signal O (1) for one line and sequentially outputs the signal by FIFO. Delay memory 30 (2)
Outputs the output signal O (2) as it is (without delay).

FIG. 6 illustrates the structure of the decimal line correction unit 205. This decimal line correction unit 205
Line memories 31 (1) to (3) and an interpolation circuit 32 provided for each of the output signals O (1) to (3)
It is composed of (1) to (3). Each of the line memories 31 (1) to 31 (3) has a capacity capable of storing one line of the output signals O (1) to O (3).

When scanning is actually performed,
More than the interpolation coefficients D (1) to (3) specified by the CPU 20, the input output signals O (1: i) to (3: i), and the output signals O (1: i) to (3: i). Decimal line correction is performed based on output signals O (1: i-1) to (3: i-1) input one line before and stored in the line memories 31 (1) to 31 (3). . For example, when the interpolation coefficient D specified by the CPU 20 is 0.5, the output signal O ′ (i) output from the decimal line correction unit 205
= O (i) × 0.5 + O (i−1) × 0.5 is calculated by the interpolation circuit 32. When the interpolation coefficient D specified by the CPU 20 is 0.3, the output signal O ′ (i) = O (i) output from the decimal line correction unit 205
× 0.3 + O (i−1) × 0.7 is calculated.

FIG. 7 is a flowchart showing the procedure of the scanning operation in the image reading apparatus. Hereinafter, the scanning operation will be described in detail with reference to this flowchart. The order of each operation and processing shown in this flowchart is not unique, and a part of the order can be changed.

First, a delay color is selected (step S1).
As described with reference to FIG. 4, when the scan is a forward scan (outbound path), the output signal O (1) shown in FIG.
(3) is selected, and when the scan is a reverse scan (return path), the output signals O (1) to O (1) shown in FIG.
The route of (3) is selected.

Next, the relative reading position shift amount X is calculated (step S2). This is a displacement X1 indicating how much the reading position of the line sensor LS (2) is ahead of the reading position of the line sensor LS (3) corresponding to the last stage in the scanning direction. Line sensor LS
The line sensor LS relatively to the reading position of (3)
The position shift amount X2 indicating how far the reading position of (1) precedes is calculated. The positional deviation amounts X1 and X2 are defined as follows: the interval between the line sensors LS (3) and (2) is N1 (line); the interval between the line sensors LS (3) and (1) is N2 (line); If P (%), X1 = N1 × (P / 100), X2 = N2 ×
It can be calculated as (P / 100).

In the image reading apparatus according to this embodiment, N1 =
Since 2 (lines) and N2 = 4 (lines), for example, when the scan magnification is 75%, X1 = 2 × 0.
It can be calculated as 75 = 1.5 (line) and X2 = 4 × 0.75 = 3.0.

Next, a reference position S is set (step S
3). As the reference position S, an appropriate position is selected according to the purpose of scanning as described later. Next, the absolute reading position deviation amount Y is calculated (step S4). The absolute reading position deviation amount Y is different from the reference position S with respect to each line sensor LS.
This indicates how far the reading positions (1) to (3) precede.

FIG. 8 illustrates the relationship between the reference position S, the relative reading position Y, and the absolute reading position X. First, the reference position S is a blue line sensor LS (3), which is the last stage (in the minus direction in the SS direction) at the time of forward scanning.
One line is selected from the area on the plus side in the SS direction by one line (the shaded area in FIG. 8). The distance from the reference position S to the reading position B of the line sensor LS (3) is determined by the absolute reading position deviation amount Y (3), and the distance from the reference position S to the line sensor L
The distance from the reading position G of S (2) to the reading position G of the line sensor LS (1) is defined as the distance from the reference position S to the reading position R of the line sensor LS (1). I do. The relationship between the absolute displacement Y and the relative displacement X is as follows: Y (2) = X1 + Y (3), Y (1) = X2 + Y
(3).

Next, the delay coefficient I is converted to an integer line correction unit 204
(Step S5). Here, the delay memory 30
The delay coefficient I (1) set to (1) is I (1) =
[Y (1)], and the delay coefficient I (2) set in the delay memory 30 (2) is I (2) = [Y (2)] (see FIG. 5). It should be noted that I (1) ≧ 0 and I (2) ≧ 0, and [Y] is an operation symbol for operating the integer part of Y.

Next, the interpolation coefficient D is converted to the decimal line correction unit 20.
5 (step S6). Each interpolation circuit 32 (1)
The interpolation coefficients D (1) to D (3) set to D (1) to Y (1) and D (2) =
Y (2)-[Y (2)], D (3) = Y (3)-[Y
(3)] (see FIG. 6). Note that 0 ≦ D (1) <
1, 0 ≦ D (2) <1, 0 ≦ D (3) <1.

Then, scanning is actually performed (step S7), and one scanning operation ends.

The image reading apparatus according to the present embodiment sets an appropriate position as the reference position S according to the purpose of scanning. For example, when detecting the document size,
When reading a black-and-white document, it is preferable to obtain output signals O (1) to O (3) having a high resolution (even if the saturation is slightly inferior), and such a reference position S is set. On the other hand, it is preferable to obtain high chroma output signals O (1) to (3) at the time of reading a color document, at the time of automatic color document identification, at the time of automatic density correction level detection (although the resolution is slightly inferior). , Such a reference position S is set. Hereinafter, a scan magnification = 75% (X1 = 1.2) is divided into a resolution priority mode in which resolution is prioritized and a saturation priority mode in which saturation is prioritized.
5 (line) and X2 = 3.0 (line)), a method of setting the reference position S will be described in detail.

In the resolution priority mode, the reference position S is set so that D (2) = 0 (see step S3 in FIG. 7). Such a reference position S
As shown in FIG. 9A, the line sensor LS
It is conceivable that the reading position G of (2) is set as the reference position S. The interpolation coefficients D (1) to D (3) at that time are as follows:
It can be determined based on the graph shown in FIG. However, considering the scan direction and the gap correction method (see FIGS. 5 and 6), 0 ≦ D (1), 0 ≦ D
(2), it is necessary that 0 ≦ D (3).

Therefore, as shown in FIG. 9B, the reference position S is set on the plus side in the SS direction from the reading position B of the last line sensor LS (3) in the forward scanning (step in FIG. 7). S3). At this time, Y (1) = 3.
5, Y (2) = 2.0 and Y (3) = 0.5 (FIG. 7)
, D (1) = 0.5, D (2) =
0.0, D (3) = 0.5 (Step S5 in FIG. 7)
), I (1) = 3, and I (2) = 2 (see step S6 in FIG. 7). Also, the scan magnification is 50% to 1
FIG. 10B is a graph showing a change in each of the interpolation coefficients D (1) to D (3) up to 00%. It can be seen that D (2) = 0 always irrespective of the change in scan magnification. By storing the parameters of the graph in the ROM 21 (see FIG. 3) or the like in advance, the CPU 20 quickly sets the predetermined reading reference position S, the interpolation coefficients D (1) to (3), and the delay coefficient I (1) ( 2) can be calculated.

In the saturation priority mode, the reference position S is set so that D (1), D (2), and D (3) are equivalent (see step S3 in FIG. 7). More specifically, | D (1) −0.5 | = | D (2) −
The reference position S is set so that 0.5 | = | D (3) −0.5 |. As such a reference position S, FIG.
As shown in (a), it can be considered that D (2) = − 0.25. In addition, each interpolation coefficient D (1) to D
(3) can be determined based on the graph shown in FIG. However, considering the scan direction and the gap correction method (see FIGS. 5 and 6), 0 ≦ D
(1), 0 ≦ D (2), 0 ≦ D (3).

Therefore, as shown in FIG.
Consider D (2) = 0.75. At this time, Y (1) = 3.
25, Y (2) = 1.75, Y (3) = 0.25 (see step S4 in FIG. 7), D (1) = 0.25, D
(2) = 0.75, D (3) = 0.25 (see step S6 in FIG. 7), and | D (1) −0.5 | = | D
(2) −0.5 | = | D (2) −0.5 | = 0.25. Also, I (1) = 3 and I (2) = 1 (FIG. 7)
Step S5). FIG. 10B is a graph showing a change in each of the interpolation coefficients D (1) to D (3) when the scan magnification is 50% to 100%. Regardless of the change in scan magnification, | D (1) −0.5 | ≒ | D (2) −0.
It can be seen that 5 | ≒ | D (3) −0.5 |. By storing in advance an arithmetic expression indicating this graph in the ROM 21 (see FIG. 3) or the like, the CPU 20 quickly sets the predetermined reading reference position S, the interpolation coefficients D (1) to (3), and the delay coefficient I (1). ) (2) can be calculated.

Comparative Example Hereinafter, a gap correction method of a conventional image reading apparatus will be described for comparison. As shown in FIG. 13, the reading reference position S is always the reading position B corresponding to the last line sensor LS (3). Each interpolation coefficient D (1) to 50% to 100% of scan magnification
FIG. 14 is a graph showing a change in D (3). And
Scan magnification is 100%, 75%, 50%, 2
The relative displacement X of the conventional image reading apparatus at 5%
1, X2, the absolute displacement amounts Y (1), Y (2), Y in the resolution priority mode of the image reading apparatus according to the present embodiment.
(3) Absolute displacement Y (1) in the saturation priority mode,
Table 1 shows Y (2) and Y (3).

[0063]

[Table 1] Here, comparing FIG. 10B with FIG. 14, the interpolation coefficient D (2) changes in the conventional image reading apparatus, and in some cases (when the scan magnification is 75%), the maximum value D
(2) = 0.5, whereas in the resolution priority mode, the interpolation coefficient is always D (2) = 0 irrespective of the scan magnification. However, it can be seen that a higher resolution is maintained as compared with the conventional one.

FIG. 15 shows the effect of the resolution priority mode.
This is explained in comparison with No. 5. FIG. 15A shows a state in which the displacement of the sub-pixels of the image signal G is interpolated by the line memory and the interpolation circuit, and the displacement of the reading position of the image signal RGB is completely corrected. here,
The interpolation coefficient D (2) of the image signal G is 0, and the image signal G
The distribution of the components is not wide. On the other hand, the distribution of the image signals R and B is widened. Therefore, the obtained image signal R
FIG. 15B shows the result of the saturation conversion of GB, and large saturation components appear at both ends of the black line in the sub-scanning direction. However, when the obtained image signal RGB is subjected to luminance conversion, the result is as shown in FIG. 15C, and it can be seen that a high-resolution image can be obtained with almost no deterioration in original luminance.

On the other hand, comparing FIG. 12B with FIG. 14, the conventional image reading apparatus uses interpolation coefficients D (1) and D (1).
(2), the values of D (2) and D (3), and the values of D (3) and D (1) are different.
At 5%) the difference can be up to 0.5,
In the saturation priority mode, each interpolation coefficient is equivalent (| D (1) −
0.5 | ≒ | D (2) -0.5 | ≒ | D (3) -0.5
|), Indicating that there is little change in hue even when the magnification changes.

FIG. 16 shows the effect of the saturation priority mode in FIG.
This will be described in comparison with the above. FIG. 16A shows a state in which the displacement of the sub-pixels of the image signal G is interpolated by the line memory and the interpolation circuit, and the displacement of the reading position of the image signal RGB is completely corrected. Here, the interpolation coefficient of the image signal is | D (1) −0.5 | = | D
(2) −0.5 | = | D (2) −0.5 | = 0.25 (equivalent), and the distribution of each component of the image signal is slightly spread uniformly. However, when the obtained image signal RGB is subjected to chroma conversion, the result is as shown in FIG. 16B, and a large chroma component does not appear at both ends of the black line in the sub-scanning direction. When the obtained image signal RGB is subjected to luminance conversion, FIG.
As shown in (c), a mountain-shaped graph slightly lower than the original luminance is obtained.

In this embodiment, the interpolation coefficients D (1) to D (1)
(3) is calculated based on the graphs shown in FIGS. 10 (b) and 12 (b) (more precisely, the arithmetic expressions stored in the ROM 21 in advance) and the given scan magnification. The shape of the graph (more precisely, ROM
The arithmetic expression stored in 21 is not uniquely determined, and may take various forms.

Modification 1 In the embodiment, | D (1) −0.5 | = | D (2) is set between the resolution priority mode in which the interpolation coefficient D (2) = 0 and each interpolation coefficient. −0.5 | =
| D (3) -0.5 | is established, but the method of determining the interpolation coefficient (reference position S
Is not limited to these. For example, between each interpolation coefficient, | D (1) −0.5 | = | D (3) −0.5 |
(Modification 1 mode) in which the relationship (1) is satisfied can also be introduced.

That is, | D (1) −
The reference position S is set such that the relationship of 0.5 | = | D (3) −0.5 | is satisfied (see step S3 in FIG. 7).
As such a reference position S, the value of D (3) can be considered to be negative as shown in FIG. The interpolation coefficients D (1) to D (3) at that time are as follows:
It can be determined based on the graph shown in FIG. However, considering the scan direction and the gap correction method (see FIGS. 5 and 6), 0 ≦ D (1), 0
≦ D (2) and 0 ≦ D (3).

Therefore, as shown in FIG.
For a 75% scan magnification, Y (1) = 0.0,
Y (2) = 1.5, Y (3) = 3.0 (see step S4 in FIG. 7), D (1) = 0, D (2) = 0.5,
D (3) = 0 (see step S6 in FIG. 7), and | D
(1) −0.5 | = | D (2) −0.5 | = | D (2)
-0.5 | = 0.25. Also, I (1) = 3, I
(2) = 1 (see step S5 in FIG. 7). Also,
Each interpolation coefficient D for scan magnifications from 50% to 100%
FIG. 17B is a graph showing changes in (1) to D (3). Regardless of the change in scan magnification, | D (1) -0.
It can be seen that 5 | ≒ | D (3) −0.5 |. By storing in advance an arithmetic expression representing this graph in the ROM 21 (see FIG. 3) or the like, the CPU 20 can quickly execute the predetermined reading reference position S, the respective interpolation coefficients D (1) to (3), and the delay coefficient I (1). ) (2) can be calculated.

In the first modification, the output signal O
Since there is no relative color shift between the colors corresponding to (1) and O (2), red and blue, reproducibility of magenta color is particularly excellent. Therefore, the first modification mode can be said to be a magenta saturation priority mode.

Modification 2 Decimal line correction unit 2 of the embodiment
Reference numeral 05 includes line memories 31 (1) to (3) and interpolation circuits 32 (1) to (3) provided for the output signals O (1) to (3), respectively. However, in the case of the brightness priority mode, at least one of the values of the complementary relation number becomes 0, that is, it is not necessary to perform decimal line correction on any one of the output signals. For example,
In the case of the resolution priority mode (G brightness priority mode),
No decimal line correction is performed on the output signal corresponding to G. Therefore, in an image reading apparatus having no saturation priority mode, as shown in FIG. 18, a correction color selection unit is provided upstream and downstream of an output signal, and a part of the output signal ( In this case, a line memory and an interpolation circuit can be provided for (two in this case).

Modification 3 Integer line correction unit 204
As shown in FIG. 19, it may be configured by a write delay circuit and an image memory provided for each of the output signals O (1) to O (3). That is, the output signal O stored in each image memory is output at a predetermined timing by the write delay circuit, so that it can function as an integer line correction unit.

Modification 4 The order in which the integer line correction unit 204 and the decimal line correction unit 205 are provided is as follows from the upstream side to the downstream side of the output signal from the integer line correction unit 204 to the decimal line correction unit. Although it may be provided in the order of the unit 205, it may be provided in the order of the decimal line correction unit 205 and the integer line correction unit 204 as shown in FIG.

[0075]

As described above in detail, according to the present invention, a proper reading position deviation correction is performed according to the situation, the resolution is greatly impaired, thin lines are blurred, black lines and black characters are colored, and black and white originals are colored. It is possible to provide an image reading apparatus that can effectively prevent problems such as false recognition as a color original and a large change in color depending on the magnification.

[Brief description of the drawings]

FIG. 1 is a schematic diagram illustrating a configuration of a color image reading apparatus according to an embodiment;

FIG. 2 is a diagram illustrating a C of the color image reading apparatus illustrated in FIG. 1;
It explains a CD in detail.

FIG. 3 illustrates a configuration of an image processing unit of the color image reading device illustrated in FIG. 1;

FIG. 4 is a diagram for explaining the configuration and operation of an upstream and downstream selection unit of the image processing unit shown in FIG. 3;

FIG. 5 illustrates a configuration of an integer line correction unit of the image processing unit illustrated in FIG. 3;

FIG. 6 illustrates a configuration of a decimal line correction unit of the image processing unit illustrated in FIG. 3;

FIG. 7 is a flowchart illustrating a basic operation of the color image reading apparatus illustrated in FIG. 1;

FIG. 8 illustrates a basic method of selecting a reference position.

FIG. 9 is a view for explaining selection of a reference position in the case of a resolution priority mode and a 75% scan magnification.

FIG. 10 shows a resolution priority mode and 50 to 10
9 is a graph for explaining selection of an interpolation coefficient when the scanning magnification is 0%.

FIG. 11 illustrates selection of a reference position in the case of a saturation priority mode and a 75% scan magnification.

FIG. 12 shows a saturation priority mode and 50 to 100.
9 is a graph illustrating selection of an interpolation coefficient in the case of% scan magnification.

FIG. 13 illustrates selection of a reference position in the case of a 75% scan magnification for comparison.

FIG. 14 is a graph illustrating selection of an interpolation coefficient in the case of a 50 to 100% scan magnification for comparison.

FIG. 15 illustrates an effect of the resolution priority mode.

FIG. 16 illustrates the effect of the saturation priority mode.

FIG. 17 shows a magenta saturation priority mode and 50
10 is a graph illustrating selection of an interpolation coefficient when the scan magnification is 100%.

FIG. 18 illustrates a modification of the decimal line correction unit.

FIG. 19 illustrates a modification of the integer line correction unit.

FIG. 20 illustrates a modification of the order of the decimal line correction unit and the integer line correction unit.

FIG. 21 is a diagram for explaining a reading position shift occurring in a line sensor.

FIG. 22 illustrates a conventional gap correction method.

FIG. 23 illustrates a configuration of a conventional gap correction unit.

FIG. 24 is a diagram for more specifically explaining a conventional gap correction method.

FIG. 25 illustrates a problem of a conventional gap correction.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... CCD, LS ... Line sensor, 11 ... Full rate carriage (scanning means), 12 ... Half rate carriage (scanning means), 13 ... Platen glass, 2 ... Image processing part, 20 ... CPU (correction means), 21 ... ROM , 22 ...
RAM, 3 ... gap correction unit (correction means), 203 ... upstream selection unit, 204 ... integer line correction unit (delay unit), 205 ... decimal line correction unit (interpolation unit), 206 ...
Downstream select section

──────────────────────────────────────────────────の Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) H04N 1/393 H04N 1/04 D 5C079 1/48 1/46 A F term (Reference) 5B047 AA01 AB04 BB03 BC05 BC09 BC11 BC23 CB09 CB10 CB17 DB01 DC13 EA07 EB03 5B057 AA11 BA02 BA13 CA01 CA08 CA12 CA16 CB01 CB08 CB12 CB16 CC01 CE16 CH01 CH11 5C051 AA01 BA03 DA03 DA06 DB01 DB07 DB22 DB24 DB28 DC02 DE12 FA01 DE01 EA01 DA02 FB04 QA10 QA17 UA06 UA13 UA18 XA01 5C076 AA21 AA22 AA26 BA01 BA06 BA08 BB04 BB13 BB31 5C079 HB01 HB06 JA03 JA23 JA25 LA03 LA05 LA12 LA17 LA28 LA37 MA03 MA11 NA03 NA04

Claims (10)

[Claims] A scanning means for optically scanning a document in a sub-scanning direction at a predetermined speed; and a scanning means disposed in parallel with each other at a predetermined distance and receiving light from the scanned document and outputting an image signal. An image reading apparatus comprising: a plurality of stages of line sensors; and a correction unit that corrects a reading position shift in the sub-scanning direction included in each output image signal with reference to a predetermined position. An image reading apparatus, wherein the predetermined position to be set can be set to an arbitrary position in the sub-scanning direction. 2. The method according to claim 1, wherein the correction unit is configured to set a predetermined reference position behind the last one in the sub-scanning direction among a plurality of reading positions corresponding to each line sensor. 2. The image reading device according to 1. 3. A delay unit for correcting an integer line I of a reading position shift in the sub-scanning direction, which may be included in each image signal from the predetermined reference position, by a delay of an output of the image signal. An interpolating unit that corrects a fractional line D of a reading position shift in the sub-scanning direction that can be included in each image signal from the reference predetermined position by interpolation with the preceding and succeeding image signals. The image reading device according to claim 1. 4. The correction unit according to claim 1, wherein the correction unit is configured to set the reference position such that a deviation between the reading position corresponding to any one of the line sensors and the reference position does not always include the D element for the decimal line. The image reading device according to claim 3, wherein the image reading device can set. 5. The multi-stage line sensor is a three-stage line sensor corresponding to three colors of red, green, and blue, and the correction unit is configured to determine a reading position corresponding to the green line sensor and a predetermined reference. Decimal line D for deviation from position
The image reading apparatus according to claim 4, further comprising a resolution priority mode for setting a predetermined reference position so that elements are not always included. 6. The image reading apparatus according to claim 5, wherein the image reading is performed in a resolution priority mode when detecting a document size and / or when reading a black and white document. 7. The reference position is set so that all of the image signals output from the line sensors in a plurality of stages include a fractional line D of the reading position shift in the sub-scanning direction. The image reading device according to claim 3, wherein the image reading device can set the image reading device. 8. The multi-stage line sensor is a three-stage line sensor corresponding to three colors of red, green, and blue, and the correction means applies each image signal output from the three-stage line sensor to the image sensor. The image reading apparatus according to claim 7, further comprising a saturation priority mode for setting the predetermined reference position such that the sub-scanning read position deviation D included in the sub-scanning direction is equivalent. 9. The image reading apparatus according to claim 8, wherein at least one of reading of a color document, automatic color document identification, and detection of an automatic density correction level performs image reading in a saturation priority mode. 10. The image according to claim 1, wherein the correction unit sets or selects a predetermined position to be the reference based on the reading magnification of the document and / or whether the scanning is going forward or backward. Reader.