JPH1032722A - Picture reader - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable precise color/black judgment by means of foreseeing color smear by discriminating whether picture data is a color or a monochrome based on distribution on a chromaticity drawing in the uniform color space of picture data. SOLUTION: An original loaded on original stand glass P is irradiated with light beams from a light source K and the reflected light is reflected on mirrors M1-M3 so as to converge them in a lens L. They are received by a picture read part 1 provided with a three line color image sensor. For processing a signal which is read by the picture read part 1, an input gradation correcting part 2 converts RGB output signals into signals proportional to lightness and a first color signal conversion part 3 converts the signals into L*a*b. A third color signal conversion part 10 calculates α'β' based on a a*b* in L*a*b* outputted from the first color signal conversion part 3. A color/black identification part 11 judges whether the original is the color picture or the monochromatic pictured by comparing α'β' with a threshold.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to an image reading apparatus having a color linear image sensor for reading a plurality of pieces of color information.

[0002]

2. Description of the Related Art An image reading apparatus for reading image information of a document includes three (R), G (green) and B (blue) on one chip.
A three-line color linear image sensor having a color pixel row is widely used.

In a three-line color linear image sensor, original information is reduced and projected on a sensor surface by an image forming lens constituting a reduction optical system, and one line of main scanning image information is read. Reading in the sub-scanning direction is sequentially performed by mechanically changing the positional relationship between the formed image and the sensor by the system moving unit.

FIG. 14 is a schematic diagram for explaining the pixel column interval of a three-line color linear image sensor. The layout (element arrangement) in the vicinity of each of the RGB three-color pixel columns includes a photosensitive pixel column corresponding to each color, and a transfer electrode for transferring signals from odd-numbered pixels and a transfer electrode for transferring signals from even-numbered pixels. And two upper and lower horizontal transfer electrodes. That is, the pixel row 20R corresponding to R is configured with the R photosensitive pixel row 21R and the two horizontal transfer electrodes 22R, and the G pixel row 20G is configured as the G photosensitive pixel row 21G and the two horizontal transfer electrodes 22G. And a pixel row 20B corresponding to B is a photosensitive pixel row for B 21B.
And two horizontal transfer electrodes 22B.

[0005] Further, the interval between the rows of photosensitive pixels of the three-line color linear image sensor along the sub-scanning direction (vertical direction in the drawing) is usually an integral multiple of one line corresponding to the pixel pitch in the main scanning direction. , For example, are opened by 12 lines.

For this reason, the image signals output simultaneously from each of the RGB read pixel columns are read lines at different positions along the sub-scanning direction on the document. Therefore,
The document surface is sequentially read out in chronological order by mechanical movement in the sub-scanning direction, and readout signals output in chronological order are read out in chronological order using an image buffer memory for alignment in line units corresponding to the intervals between readout lines of each color. By adding a delay operation, the displacement of the reading lines of the three colors in the sub-scanning direction is corrected so that the image data of the same line on the document surface.

Further, in an image reading apparatus comprising a digital color copying machine, it is determined whether the read image on the document is a color image or a monochrome image, and the processing in the image output unit is switched. .

FIG. 15 is a block diagram for explaining an image reading apparatus comprising a conventional digital color copying machine. In this image reading apparatus, an RGB signal proportional to the reflectance output from the image reading unit 1 ′ is converted into an L signal by an input tone correction unit 2 that performs reflectance → brightness conversion and a first color signal conversion unit 3 that performs matrix calculation. Convert to ^* a ^* b ^* signal.

The flow of the L ^* a ^* b ^* signal is divided into a main route video signal for editing processing such as hue rotation and an a ^* b ^* signal for color / black discrimination. The video signal of the main route is passed from the image editing unit 4 to the second color signal conversion unit 5 and converted into a YMC signal. The black generation processing unit 6, the image switching unit 7, the output gradation correction unit 8, the image output unit 9 is output as a YMCK full-color image.

The image switching section 7 selects a full-color YMCK based on a signal from the color / black identification section 11 ',
Determine whether to select only the L ^* signal. That is, in the case of a color original, the color / black discriminating section 11 'to the image switching section 7
Signal that the document is in color is passed to
Select the MCK signal. On the other hand, in the case of a monochrome document, a signal indicating that the document is monochrome is passed from the color / black discriminating section 11 'to the image switching section 7, and the L ^* signal is selected.

The color / black discrimination by the color / black discriminating section 11 'is performed by first converting an RGB signal output from a color linear image sensor into an L ^* a ^* b ^* signal at the time of reading a document, and indicating chromaticity information among these signals. Calculating the value of the saturation C ^* from the a ^* b ^* data using the following equation, comparing the value of the saturation C ^* with a preset threshold value, and determining whether or not the value is greater than the threshold value. Thus, whether the image on the document is a color image or a monochrome image is identified.

C ^* = {(a ^* ) ² + (b ^* ) ² } ^1/2

Here, the boundary line based on the threshold value for determining whether the image is a color image or a monochrome image has a radius C _TH centered on the origin on the a ^* b ^* chromaticity plan view as shown in FIG. When the saturation C ^* is inside the circle, the image is determined to be a monochrome image, and when the saturation C ^* is outside the circle, the image is determined to be a color image.

[0014]

However, the conventional image reading apparatus has the following problems. That is, if scanning speed unevenness due to vibration or the like exists in the mechanical sub-scanning, the positional accuracy of the read line in the sub-scanning direction is deteriorated, and the positional deviation correction by the line-by-line delay operation becomes incomplete. Color shift (hereinafter referred to as chromaticity error) occurs in the image data to be obtained.

The scanning speed unevenness is periodic, and its frequency is caused by the number of teeth of a belt and a gear for transmitting the rotation of the motor to the scanning system, and the frequency can be adjusted by changing the number of teeth. When the frequency is reduced, the amplitude tends to increase. On the other hand, when the frequency is increased, there is a danger that resonance will occur due to overlap with the natural frequency of the scanning mirror, and adjustment is very difficult.

Further, when determining whether a document is a color image or a monochrome image, a color shift occurs in the read image data due to vibration and uneven speed during sub-scanning, and the read data has a saturation C ^* There is a problem that the component becomes large and this value exceeds a preset threshold, and a document of a monochrome image is erroneously determined to be a color image.

In the three-line color linear image sensor, the chromaticity error caused by vibration and uneven speed at the time of sub-scanning can be improved by reducing the interval between the rows of photosensitive pixels along the sub-scanning direction. However, it is not possible to improve the erroneous determination when determining whether the image is a color image or a monochrome image based on the value of the saturation C ^* obtained by Equation 1 described above.

[0018]

SUMMARY OF THE INVENTION The present invention is an image reading apparatus made to solve such a problem. That is, the image reading apparatus of the present invention includes a color linear image sensor that reads image information of a document, and a chromaticity diagram in a uniform color space of image data obtained when monochrome image information is read by the color linear image sensor. , A conversion unit that converts a non-uniform color space of a color signal output from a color linear image sensor into image data of a uniform color space, and the image data converted by the conversion unit is color. Determining means for determining whether the image data is monochrome or monochrome based on the distribution of the image data stored in the storage means on a chromaticity diagram in a uniform color space.

Further, a photoelectric conversion element array in which a plurality of photoelectric conversion elements are arranged for one line in the main scanning direction with respect to the original is
A plurality of color linear image sensors arranged at intervals of a predetermined number of lines to read different color information along a sub-scanning direction perpendicular to the main scanning direction, and collects reflected light from a document in the sub-scanning direction. In an image reading apparatus having an optical imaging scanning mechanism for scanning, an interval corresponding to the number of lines of each photoelectric conversion pixel array of the color linear image sensor is
When scanning the optical system in the sub-scanning direction by the optical system image scanning mechanism, 1 / one of the number of lines read in one cycle of the natural vibration of the optical system.
In addition to the number of photoelectric conversion elements, the number of photoelectric conversion elements is set to be smaller than 3, and a plurality of photoelectric conversion element rows are arranged around a photoelectric conversion element row for reading green color information as color information.

In the image reading apparatus of the present invention, the distribution of the image data obtained when the monochrome image information is read by the color linear image sensor on the chromaticity diagram in the uniform color space is stored in the storage means. Based on this distribution, it is determined whether the image data is color or monochrome.
That is, since the distribution stored in the storage unit is based on the color shift of the read image data unique to the image reading apparatus, it is possible to perform accurate color / black determination in anticipation of the color shift. .

A photoelectric conversion element for reading green color information by setting the interval between the photoelectric conversion pixel rows in the sub-scanning direction smaller than 1/3 of the image length read in one period of the natural oscillation of the optical system. By arranging the columns in the center, the amount of color shift of the read image data due to vibration of the scanning mechanism of the optical system or the like can be reduced, and the original information can be faithfully read.

[0022]

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram showing an embodiment of an image reading apparatus according to the present invention. 1A and 1B are diagrams for explaining the present embodiment. FIG. 1A is a block diagram of a signal processing system, and FIG. 1B is a schematic diagram of a scanning system.

As shown in FIG. 1B, the image reading apparatus according to the present embodiment irradiates a document (not shown) placed on a document table glass P with light from a light source K, and reflects the reflected light. Is reflected by the mirrors M1 to M3, collected by the lens L, and received by the image reading unit 1 provided with the three-line color image sensor.

As shown in FIG. 1A, in order to process a signal read by the image reading unit 1, an input gradation correction unit 2 converts an RGB output signal into a signal proportional to brightness.
A first color signal conversion unit 3 composed of a multiply-adder for converting this signal into an L ^* a ^* b ^* signal, an image editing unit 4, L ^* a
a second color signal conversion unit 5 for generating a YMC signal from the ^* b ^* signal, a black generation processing unit 6 for generating a K signal, an image switching unit 7, an output gradation correction unit 8, and an image output unit 9; It has.

Further, the image reading apparatus of the present embodiment
L output from the one-color signal converter 3 ^*a^*b^*Of
a^*b^*A third color signal for calculating α′β ′ described later based on
Signal conversion unit 10 and a comparison between α′β ′ and a predetermined threshold value.
Therefore, whether the original is a color image or a monochrome image
Color / black discriminating unit 11 for determining
And

Here, the result of the conversion to a ^* b ^* data using the conversion coefficient, the conversion to the saturation C ^* data, and the determination of the color black is obtained by the image reading using the three-line color image sensor. A description will be given of how the reading scanning speed of the unit changes due to a change in the arrangement order of the pixel rows of the three colors of the three-line color image sensor.

FIG. 2 is a diagram showing a change in read signal level due to scanning speed unevenness, and shows a case where achromatic modulation image data is read and scanned. The displacement between the three colors in the read pixel position can be corrected by replacing the positional displacement with a time displacement and giving a delay to the output signal from the sensor if the reading scan speed is constant. If there is, the reading position shifts.

FIG. 2A shows the difference between the levels of the three colors when the modulation information such as the ladder chart is read due to the displacement of the three colors. When determining whether the original is a color image or a monochrome image, the determination is performed based on the result of converting the reading level (reflectance) of each color of RGB into an L ^* a ^* b ^* signal. Among the values of the L ^* a ^* b ^* signal, L ^* indicates brightness information, and a ^* b ^* indicates hue / saturation information.

The origin of the a ^* b ^* coordinates represents an achromatic color (white to gray to black), and the closer to the origin, the closer the color becomes. The fact that the result of reading the achromatic color modulation data deviates from the origin of the a ^* b ^* coordinates indicates an information reading error. This results in a chromaticity error.

The reading level of each color of RGB when reading the image information of the original changes depending on the phase of the reading scanning vibration and the phase of the modulation of the reading signal, and the change causes the color determination data a ^* b ^*. Also changes.
In this manner, when the phase of the reading scanning vibration and the phase of the modulation of the reading signal are respectively varied, and the line spacing of the three-line color image sensor and the arrangement of the reading colors of each pixel column change, the chromaticity error is FIGS. 3 to 8 show plots of the change. The condition at this time is
The case where the vibration frequency of the scanning system is converted into the image length is 36 lines, which is based on the actual apparatus.

FIG. 3 shows a case where each pixel column is arranged at an interval of 8 lines around the G pixel column, and FIG. 4 shows a case where each pixel column is arranged at an interval of 8 lines around the B pixel line. 5 shows a case where each pixel column is arranged at an interval of 8 lines around the R pixel column, and FIG.
FIG. 7 shows a case where each pixel column is arranged at a line interval, FIG. 7 shows a case where each pixel column is arranged at a line interval of 6 lines, and FIG. This is a case where pixel columns are arranged.

According to these figures, when the pixel rows of R and B among the three colors are arranged at the center, the shift of a ^* b ^* from the achromatic color due to the phase of the vibration modulation becomes large, and the worst value becomes large. Become. On the other hand, it can be seen that when the G pixel columns are arranged at the center, the distortion of the distribution is reduced and the worst value can be reduced. Further, the smaller the pixel column interval, the greater this tendency.

FIG. 9 is a diagram showing a case where the relationship between the reading position of the three-line color image sensor and the phase of the scanning vibration is the same. Thus, the cycle at which the scanning unevenness occurs is
When the interval is three times the pixel column interval of the three-line color image sensor, the arrangement becomes unrelated to the arrangement of the RGB three-color pixel columns.

FIG. 10 shows the relationship between the pixel row interval and the maximum chromaticity error as a parameter of an array of three pixel rows of a three-line color image sensor (which color RGB row is centered). FIG.

As described above, when the pixel row interval is 12 lines, the vibration frequency of the scanning system is 1/3, and it is understood that the arrangement order of the RGB three-color pixel rows and the maximum chromaticity error are unrelated. However, if the pixel column interval is reduced to reduce the chromaticity error, the arrangement order of the RGB three-color pixel columns appears as a characteristic difference of the chromaticity error.

As described above, the pixel row interval is set to 12 lines (1/1 of the number of lines read in a cycle in which scanning unevenness occurs).
3) To reduce the chromaticity error by making it smaller,
In addition to simply reducing the pixel column interval, the G pixel column is arranged at the center of the RGB three-color pixel columns.

On the other hand, the pixel column interval is reduced as described above.
, The chromaticity depends on the arrangement order of the RGB three-color pixel columns.
Error tendency changes. Is this an RGB signal
L ^*a^*b^*Matrix used when converting to signals
Is caused by the coefficient.

The first color signal converter 3 shown in FIG.
Matrix coefficient to be used is input gradation correction to RGB data
Processed value L_R, L_G, L_BFor a^*= −0.24 × L_B+ 1.11 × L_G-0.87x
L_R b^*= 1.08 × L_B-0.81 × L_G-0.27 × L
_R Multiplied by a^*b^*(Strictly speaking, a^*b^*The value
Scale conversion to allocate to 8-bit data
)). This coefficient is the gray balance of three colors
Is considered so as not to break_R= L_G=
L_BIn the case of ^*= B^*= 0.

[0039] Among the coefficients, coefficients of L _B for a ^*, the magnitude of L _R factor for b ^* is smaller than the other coefficients. Due to this property, the variation in the amount of a ^* b ^* due to the variation of the three color data from the gray is as shown in FIG. 11, and the variation varies depending on the center of the color.

This means that the original definitions of the values of a ^* b ^* correspond to the spectral sensitivity characteristics x (x bar) and y (y bar) shown in FIG.
A = 500 ^× (X ^1/3 −Y ^1/3 ) b ^* = 200 ^× (X ^1/3 −Z ^{1 / 3)} obtained by converting as, the Y ^1/3 approximately L _G, Z ^1/3
Is almost equivalent to L _B , and X ^{1/3 is} almost equivalent to L _R.

[0041] For this reason, the coefficient of L _B for a ^*, b
The magnitude of the _LR coefficient for ^* is smaller than the other coefficients. Since the L ^* a ^* b ^* color system was originally created to approximate the visual characteristics of the human eye,
The meaning of the matrix coefficients a ^* b ^* is derived from the characteristics of the human eye.

Since a matrix coefficient due to the nature of the human eye is used, the color signal output read using the three-line color linear image sensor is calculated as L ^*
is converted to the color system of the a ^* b ^*, etc., to perform the color black determination using two a ^* b ^* values with these color information, there are the following properties. When the G pixel column is located at the center of the three lines as the arrangement of the three color pixel columns of the three-line color linear image sensor, the amount of chromaticity error due to the reading vibration of the scanning system is reduced. The occurrence of a chromaticity error due to scanning vibration has directionality on the a ^* b ^* coordinates.

As described above, this property appears when the pixel column interval is smaller than 1/3 of the number of lines to be read in a cycle in which scanning unevenness occurs.

The above characteristic also appears when “reading frequency × (number of lines between pixel columns)” is sufficiently smaller than the vibration frequency of the scanning system. That is, 12 shown in FIG.
In the case of the line interval (when “reading frequency × (the number of lines at the pixel column interval)” is １ ／ of the vibration frequency), the distribution of the chromaticity error appears in a region relatively close to a circle, but FIG. 5 to 8 shown in FIG. 5, the chromaticity error has a specific direction due to the arrangement order of the colors.

From the above, when a three-line color image sensor is used, among the pixels having a small pixel column interval, the one having the G pixel column at the center is used, and the area for the color / black determination is used. Is an elliptical area enlarged in a direction in which a chromaticity error is likely to occur.

This makes it possible to improve the accuracy of the color / black determination. Here, in order to make the region of color / black determination an ellipse as shown in FIG. 13, variable conversion may be performed as in the following Expression 2.

[0047]

(Equation 1)

In order to carry out the equation (2), a matrix conversion as shown in the equation (3) may be performed.

[0049]

(Equation 2)

FIG. 1A shows such a matrix conversion.
Is performed by the third color signal converter 10 shown in FIG. The third color signal conversion unit 10 converts a ^* b ^* into α′β ′, and the color / black discrimination unit 11 determines whether or not α′β ′ is within the inner circle of the radius m. It is determined whether the image is a monochrome image or a monochrome image.

This makes it possible to accurately determine the color or blackness, particularly when the pixel line interval of the three-line color image sensor is reduced.

[0052]

As described above, the image reading apparatus according to the present invention has the following effects. That is, it is possible to improve the determination accuracy by determining whether the image data is color or monochrome in consideration of the color shift of the read image data unique to the image reading apparatus.

Further, since the amount of color misregistration of the read image data caused by the vibration of the scanning mechanism of the optical system can be reduced, the read image information more faithful to the original information can be obtained.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating an embodiment.

FIG. 2 is a diagram illustrating a change in a read signal level due to uneven scanning speed.

FIG. 3 is a diagram (part 1) illustrating a change in a chromaticity error amount;

FIG. 4 is a diagram illustrating a change in a chromaticity error amount (part 2);

FIG. 5 is a diagram (part 3) illustrating a change in a chromaticity error amount.

FIG. 6 is a diagram (part 4) illustrating a change in the chromaticity error amount.

FIG. 7 is a diagram illustrating a change in the chromaticity error amount (No. 5).

FIG. 8 is a diagram (part 6) illustrating a change in the chromaticity error amount.

FIG. 9 is a diagram illustrating a relationship between a change in a reading position and scanning vibration.

FIG. 10 is a diagram illustrating a relationship between a pixel column interval and a chromaticity error maximum value.

11 is a graph showing changes in C ^* values for variations in the value of L _R L _G L _B.

FIG. 12 is a diagram illustrating spectral sensitivity characteristics of a colorimeter.

FIG. 13 is a diagram illustrating a color / black determination area according to the present embodiment.

FIG. 14 is a schematic diagram illustrating a pixel column interval.

FIG. 15 is a configuration diagram illustrating a conventional example.

FIG. 16 is a diagram illustrating a conventional color / black determination area.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 Image reading part 2 Input gradation correction part 3 First color signal conversion part 4 Image editing part 5 Second color signal conversion part 6 Black generation processing part 7 Image switching part 8 Output gradation correction part 9 Image output part 10 Third Color signal converter 11 Color / black discriminator 12 System CPU

Claims (2)

[Claims] 1. A color linear image sensor for reading image information of a document, and a distribution on a chromaticity diagram in a uniform color space of image data obtained when monochrome image information is read by the color linear image sensor. A conversion unit that converts a non-uniform color space of a color signal output from the color linear image sensor into image data of a uniform color space; and that the image data converted by the conversion unit is color or monochrome. A determination unit that determines whether the image data is stored in the uniform color space on the chromaticity diagram in the uniform color space. 2. A method for reading a row of photoelectric conversion elements in each of which a plurality of photoelectric conversion elements are arranged for one line along a main scanning direction with respect to a document to read different color information along a sub-scanning direction perpendicular to the main scanning direction. A color linear image sensor arranged at a plurality of intervals of a predetermined number of lines, and an image reading apparatus including an optical imaging scanning mechanism that condenses reflected light from the original and scans in the sub-scanning direction. An interval corresponding to the number of lines of each photoelectric conversion pixel row of the color linear image sensor is read in one cycle of natural vibration of the optical system when the optical system scans the optical system in the sub-scanning direction by the optical system imaging scanning mechanism. The number of lines is smaller than 1/3, and among the plurality of photoelectric conversion element arrays, the photoelectric conversion element arrays for reading green color information as the color information are arranged around the center. Image reading device.