JPH0583576A - Color reader - Google Patents

Abstract

PURPOSE:To correct color slurring at black and white edge parts. CONSTITUTION:This device is equipped with an image reading means 112 to read black and white images in a linear area on an original 101 while decomposing the colors into plural color components, extracting means 113 to successively extract image data for the prescribed number of picture elements having respective picture elements at the heads from the image data corresponding to the linear area obtained by the image reading means 112, holding means 114 to hold the image data of the picture elements at the starting point and ending points of the extracted range, a judging means 115 to judge whether the value of one component at least in the image data is fixed at both terminal parts of the extracted range or not, and a replacing means 116 to replace the image data of the respective picture elements contained in the starting point side area and the ending point side area with the image data of the picture elements at the starting point and ending point held in the holding means 114 respectively corresponding to the judged result showing that the value is fixed at both terminal parts of the extracted range, and to transmit the replaced image data.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a color reading device, and more particularly to a color reading device for correcting a read image.

A color reading device is used as an image input device for a facsimile machine or a computer. For example, an image sensor using a charge-coupled device (CCD) or the like is used to convert a color image into R (red) and G (green). , B
What is measured and read as the intensity distribution of each component of (blue) is known.

In such a color reading apparatus, there is a case where a color image and a document expressed in black and white or a line drawing are mixed, such as a catalog or a pamphlet, and the document is read together with the color image. There is a demand for highly accurate reading of the black and white binary image.

[0004]

2. Description of the Related Art FIG. 14 shows the structure of a conventional filter switching type color reading device. In FIG. 14, a linear portion (hereinafter, referred to as a reading line) 613 on the original 611 is illuminated by a lighting device 612 such as a fluorescent lamp. The reflected light from the reading line 613 is sent to a sensor (for example, CCD line image sensor) 61 by a lens 614.
5 is imaged.

Further, a filter 62 for transmitting a red (R) component, a green (G) component and a blue (B) component, respectively.
1r, 621g, 621b, and the filter 621 described above in the optical path from the document 611 to the sensor 615.
Any of r, 621g, and 621b is selectively inserted (FIG. 14 shows a state in which the filter 621g is inserted in the optical path). That is, by switching the filter to be inserted in the optical path, the sensor 6
The color components of the light reaching 15 are switched, and the reading line 6
The intensity distribution of each color component of the reflected light from 13 is measured in a time division manner, and the color image on the reading line 613 is read.

After that, the original 611 is read on the reading line 6
A two-dimensional color image is read by moving a predetermined length in the sub-scanning direction perpendicular to 13 (indicated by arrow A in FIG. 14) and repeating the same processing.

As an example of another color reading device, FIG. 15 shows the configuration of a filter-adhesion type color reading device. Similar to the filter switching type color reading device described above, the document 61
The reading line 613 on the upper part 1 is the lighting device 61 such as a fluorescent lamp.
Illuminated by 2. A rod lens array 714 configured by arranging a plurality of lenses having a diameter of about 1 mm in a line is arranged in close contact with the reading line 613,
The rod lens array 714 is configured to form an image of the reflected light from the reading line 613 described above on a sensor 715 such as a CCD line image sensor.

Here, the length of each element of the above-mentioned sensor 715 in the arrangement direction is equal to the width of the original 611. In addition, as shown in FIG. 16, each pixel on the read line 713 corresponds to a set of three elements on the sensor 715, and each of these three elements corresponds to R, G, and B.
Each component is covered with a filter that has a separated color.
Therefore, these three elements are color sensors that measure the intensity of the reflected light from the corresponding pixels on the read line 613, and the color image of the corresponding pixels is R, G by these three elements. , B components are color-separated and read.

In this case, the intensity distribution of the R, G, B components of the reflected light from the color image of each pixel on the reading line 613 can be read at the same time, so that the reading of the color image can be speeded up. ..

After the reading operation for one line is completed, the original 611 is read in the sub-scanning direction perpendicular to the reading line 613 (see FIG. 15).
(Indicated by arrow B in FIG. 2) by a predetermined length, and the same process is repeated to read a two-dimensional color image.

[0011]

By the way, in the above-mentioned conventional method, a color image is once decomposed into R, G, and B components, the intensity distribution of each component is measured, and then these measurements are performed. A color image is obtained by combining the results. Here, the lens 614 is included in the measurement result of each separated color.
And errors due to aberrations of the imaging optical system such as the rod lens array 714 are included.
It is different for each separation color.

For example, the filter switching type color reading apparatus shown in FIG. 14 is a reduction optical system for connecting an image obtained by reducing the reading line 613 on the sensor 615. Therefore, due to the chromatic aberration of the lens 614, the image formation state and the image formation position on the sensor 615 change for each separated color, and a reading error occurs. Further, since the reading for each separated color is performed with a time difference, when the original 611 is read while moving in the sub-scanning direction, the reading position in the sub-scanning direction differs for each separated color.

In the filter contact type color reading apparatus shown in FIG. 15, the reading positions of the three elements of the sensor 715 corresponding to each pixel on the reading line 713 are 1/3 pixel in the main scanning direction. Because each is different,
In addition to the aberration of the rod lens array 714, the difference in the reading position of each color component is the reading error.

As described above, since the measurement results of each separated color include various measurement errors, a color shift is inevitably generated in the color image obtained by combining these measurement results. This color shift is not so noticeable when reading a color image such as a landscape photograph or a person, but in a binary image such as a document or a line drawing, it is perceived by the human eye as a remarkable deterioration of the read image. ..

In particular, when an image having a black and white edge such as a document or a line drawing is read by using the contact type color reading mechanism described above, the reading positions of the elements corresponding to the R, G and B components are different. Due to the generated color shift, the position at which the white level corresponding to white changes to the black level corresponding to black differs for each separated color. For example, in FIG.
The read image of the step image as shown in FIG.
As shown in (b), the reading result of each color component blunted by the aberration of the rod lens array 714 is shifted and synthesized according to the shift of the reading position. Therefore, in the read image, a color that should not be present (false color) occurs in the color shift area where at least one color component is at an intermediate level (indicated by an arrow in FIG. 7 (b)). Resulting in.

As shown by hatching in FIG. 18, this false color appears so as to border the edge portion of the pattern forming the character or line drawing, and is easily recognized by the human visual sense. , The quality of the read image is greatly reduced.

As a method for preventing the color shift as described above, the factors of the error (chromatic aberration, etc.) for each separated color are investigated, and the error for each separated color is analytically evaluated based on these factors. Then, there is a method of correcting the measurement result. However, since the color reading device is composed of a large number of constituent elements, there are a large number of factors, and it is difficult to analytically determine the correction amount in consideration of all the factors.

It is an object of the present invention to provide a color reading device that corrects a color shift of an edge portion caused by a reading error of each separated color.

[0019]

FIG. 1 is a view showing the arrangement of a color reading apparatus according to the first aspect of the present invention. The invention of claim 1 is
A line image sensor 111 in which a plurality of light receiving elements are arranged in one row is provided, and a black and white image of a linear region on the original 101 is converted into a plurality of color components for each pixel corresponding to each light receiving element of the line image sensor 111. An image reading unit 112 that reads out by color separation and outputs image data composed of a plurality of components indicating the intensity of each separated color, and a predetermined number of pixels starting from each pixel from the image data corresponding to the linear area. Extracting means 113 for sequentially extracting minute image data, and extracting means 11
The holding unit 114 for holding the image data of the pixels at the start and end points of the extraction range by 3 and the image data extracted by the extracting unit 113 are input, and at least one of the image data is input.
Extraction is performed according to the determination means 115 that determines whether the values of the two components are constant at both ends of the extraction range and the determination result that the values of the corresponding components at both ends of the extraction range are constant. A replacing unit that divides the range into a region on the starting point side and a region on the ending point side, and replaces the image data of each pixel included in each region with the image data of the starting point and the ending point pixel held in the holding unit 114 and sends the replacement data. 116 and are provided.

FIG. 2 is a view showing the arrangement of the color reading apparatus according to the second aspect. According to a second aspect of the present invention, in the color reading apparatus according to the first aspect, the image data obtained by the image reading unit 111 is converted into color difference data representing a corresponding color in the color difference space, and the linear data on the original 101 is converted. A color conversion unit 121 for sending to the extraction unit 113 as image data corresponding to the area, and a color reverse for sending the image data from the replacement unit 116 after converting it into image data composed of components corresponding to the separated colors in the image reading unit 112. And a conversion means 122.

FIG. 3 is a view showing the arrangement of the color reading device according to the third and fourth aspects. According to a third aspect of the present invention, in the color reading apparatus according to the first aspect, the reference pattern 10 has at least one pattern corresponding to a point light source.
2 and the reference pattern 102 by the image reading means 112.
The range of the line image sensor 111 to which the light from the pattern corresponding to the point light source arrives is determined according to the input of the image data obtained by reading, and the extraction unit 113 uses the number of pixels corresponding to this range as the extraction width. An extraction width determining means 131 used for extraction processing is provided.

According to a fourth aspect of the invention, in the color reading apparatus according to the first aspect, a reference pattern 103 having at least one edge pattern represented by white and black,
The image data obtained by reading the reference pattern 103 by the image reading unit 112 is input, and the line image sensor 111 that obtains image data having a level between a white level corresponding to white and a black level corresponding to black.
Is obtained, and the number of pixels corresponding to this range is used as the extraction width, and the extraction width determination means 132 to be used for the extraction processing by the extraction means 113 is provided.

[0023]

According to the first aspect of the present invention, the replacement unit 116 stores the image data of each pixel extracted by the extraction unit 113 in the holding unit 114 according to the determination result of the determination unit 115. The pixel data is used for replacement. Here, the portion where the value of at least one component of the image data is constant for a plurality of pixels is the portion where no color misregistration occurs,
By the extraction unit 113 and the determination unit 115 described above, it is possible to determine the region including the portions where the color misregistration does not occur at both ends. This area is divided into an area on the starting point side and an area on the ending point side, and it is assumed that there is an edge at the boundary of these areas, and the image data included in each area is replaced, whereby the image reading unit 112 It is possible to reproduce the sharp edge image by removing the color shift of the edge portion caused by the reading error of each separated color.

According to the second aspect of the present invention, the color difference data obtained by the color conversion means 121 is sent to the extraction means 113 as image data. Therefore, the determination means 115 determines each pixel indicated by each component of the color difference data. The determination process described above can be performed based on color or brightness. In general,
In color difference data such as the L ^* a ^* b ^* color system, the value of each component tends to be a constant value, so it is necessary to accurately determine whether both ends are at the white level or the black level. You can

According to the third aspect of the present invention, the range of the image sensor 111 to which the light from the pattern corresponding to the point light source of the reference pattern 102 reaches reaches the spread of the image due to the reading error of each separated color by the image reading means 112. The extraction width determining means 131 can determine an appropriate extraction width by utilizing the corresponding fact.

According to the invention of claim 4, the range of the line image sensor 111 in which the image data having an intermediate level between the white level and the black level is obtained is the image reading means 11.
It is possible to obtain an appropriate extraction width by the extraction width determining unit 132 by utilizing the fact that the image width of the edge pattern of the reference pattern 103 according to 2 corresponds.

[0027]

Embodiments of the present invention will now be described in detail with reference to the drawings. FIG. 4 shows the configuration of an embodiment of the color reading device of the present invention.

In FIG. 4, the image reading means 112 reads the original 101 by the filter contact type color reading mechanism 211 shown in FIG.
The analog output of the sensor 715 corresponding to 11 is converted into a digital value by the analog-digital (A / D) converter 213. Further, the conversion result obtained by the analog-digital conversion unit 213 is input to the shading correction unit 220, and is normalized by the shading correction unit 220 using the standard image data stored in the memory 221 in advance. , R, G, and B components are obtained to obtain image data.

Here, as the standard image data described above, a white plate or the like is read by the color reading mechanism 211 prior to reading the original 101, and the image data for one line obtained at this time is stored in the memory 221. You can keep it in.

As a result, the influence of the unevenness of the light amount of the illumination device 612 provided in the color reading mechanism 211 and the variation of the sensitivity of each element of the sensor 715 is removed, and the shading of the image appearing due to such characteristics. Can be corrected to obtain image data that faithfully reflects the color and brightness of the image on the reading line 613 on the original 101.

The image data of each pixel thus normalized is sequentially input to the color misregistration correction unit 230. In FIG. 4, the color misregistration correction unit 230 includes a color conversion unit 121.
The color conversion unit 231 corresponding to converts the color system of the image data of each pixel from the RGB color system to the L ^* a ^* b ^* color system,
The obtained color difference data is held in the buffer 232, and after the color difference data is subjected to the correction processing described later, the color conversion unit 233 corresponding to the color reverse conversion means 122 reverses the RGB color system image data. It is configured to be converted and transmitted.

The above-mentioned buffer 232 is configured to hold the input color difference data for m pixels at the addresses in the input order and output the oldest color difference data in response to the input of new color difference data. And functions as the extraction unit 113. Below, this buffer 23
Each address of No. 2 is shown in the order of input with numbers "1" to "m". The writing process and the reading process for the buffer 232 are performed by the reading processing unit 234 and the writing processing unit 235, respectively.

Further, in FIG. 4, four registers 23 are provided.
6 _{1-236 4} and two comparators 237a, and 237b and the AND gate 238 forms a judgment unit 115, registers 236 _1, 236 ₂ and the register 236
_3, 236 compare ₄ of the inner container 237a, 237b are respectively compared, AND gate 238 has a configuration for outputting a logical product of the comparison result as the determination result.

[0034] register 236 _{1-236 4} described above,
Addresses "1", "2", "m- of the buffer 232"
1 ”and“ m ”respectively, and the read processing unit 23
4, the components of the color difference data read from these addresses are set respectively. Further, each of the comparators 237a and 237b described above may be configured to output a logical "1" as a comparison result when the contents of the corresponding registers 236 are equal and output a logical "0" when the contents are different.

The color misregistration correction process will be described below with reference to the flowchart showing the color misregistration correction operation shown in FIG. 5 and FIG. When starting the color misregistration correction processing for the image data of each line, first, each address of the buffer 232 is initialized, and then the following processing is started.

First, the color conversion unit 231 sequentially converts the image data for one pixel from the shading correction unit 220 into color difference data and inputs it to the buffer 232, and the buffer 232 holds the color difference data for m pixels. Until step 302, step 301 to step 302 are repeated as a negative determination in step 302.

When m pixels of color difference data are held in the buffer 232, an affirmative decision is made in step 302,
In response to this, the read processing unit 234 determines that the color difference data of the m pixel width stored in the buffer 232 is 2 at both ends.
Reads the lightness of pixels L ^*, is set in the register 236 _{1 to 236 4} described above (step 303). At this time, the read processing unit 234 sets the lightness index L ^* read from the addresses “1” and “2” of the buffer 232 in the registers 236 ₁ and 236 ₂ , respectively, and the address “m” is set.
The brightness index L ^* read from "-1" and "m" may be set in the registers 236 ₃ and 236 ₄ , respectively.

As a result, the comparators 237a and 237b compare the lightness index L ^* set in the corresponding registers 236 (step 304), and the respective comparison results are sent to the AND gate 238.

Therefore, the two comparators 237a and 237b are provided.
Output of both become logic "1", and AND gate 237
Thus, when the logical "1" is output as the determination result, the level of the lightness index L ^{* at} both ends of the extraction range described above is kept constant, and an affirmative determination in step 305 may be made.

In this case, the writing processor 235 copies the color difference data of the address "1" to the address "2" to the address "k" of the buffer 232, and copies it to the address "k + 1" to the address "m-1". The color difference data of the address "m" is copied (step 306). At this time, the write processing unit 235
First, the read processing unit 234 may be requested to read the color difference data at the address “1” and the address “m”, and the obtained color difference data may be written in each of the addresses described above. Further, the above-mentioned address "k" is, for example, m
The address may correspond to the pixel at the center of the pixels.

As a result, the color difference data of the m pixels held in the buffer 232 is divided by the above-mentioned kth pixel.
The first half is replaced with the color difference data of the address "1", and the second half is replaced with the color difference data of the address "m". As described above, the writing processing unit 235 performs the above-described step 30.
The function of the replacement unit 116 is realized by performing the process of 6 and rewriting the color difference data of each pixel.
In this case, the buffer 232 serves the function of the holding unit 114.

Here, in a read image obtained by reading a black-and-white image such as a document or a line drawing, a portion where at least one component of the color difference data is constant over a plurality of pixels is a portion in which no color misregistration occurs. By the above-mentioned step 305, it is possible to discriminate the area including the portions where the color misregistration does not occur at both ends.

In a read image obtained by ideally reading a black-and-white image, the color difference data changes discontinuously between the white level and the black level only at the edge portion, and at the portion other than the edge portion, a plurality of continuous color difference data is obtained. The pixel color difference data has a constant value indicating white or black. That is, in step 306, the pixel corresponding to the address “k” is set as a virtual edge position, the color difference data of each pixel is replaced with this pixel as a boundary, and the color difference data of each pixel that is continuous with the pixels at both ends is replaced. The color shift can be removed by keeping the value of a constant value.

FIG. 6 shows an explanatory diagram of the color misregistration correction operation of the present invention. FIG. 6A shows color difference data obtained by color-converting the image data of the edge portion shown in FIG. 17B. In this way, when the color difference data including the color misregistration area in the area where the color misregistration does not occur is extracted on both sides of the color misregistration area, the affirmative determination is made in step 305 described above, and the replacement process of step 306 is performed. Done. As a result, the color difference data of each pixel included in this extraction range is
As shown in FIG. 6B, one of the pixels is replaced with black-level color difference data and the other is replaced with white-level color difference data with the central pixel as a boundary.

This color difference data corresponds to an ideal reading result of an image having an edge at the center of the extraction range, and this m pixel worth of color difference data is sent to the color conversion section 233 (step 307). By inversely converting into the RGB color system, as shown in FIG. 6 (c), image data in which each RGB component changes stepwise from the black level to the white level with the central pixel as a boundary is obtained.

That is, by the above-described processing, it is possible to correct the color shift caused by the reading error of each of the RGB components and reproduce the step image. After that, it is determined whether or not the processing for one line is completed (step 309), and in the case of a negative determination, the processing returns to step 301 and the above-described processing is repeated. In this case, the above step 301,
302 is repeated m times, and when the color difference data for m pixels is held in the buffer 232, step 3 is executed again.
The process starting from 03 is started, and the levels of the brightness index L ^{* at} both ends of the newly extracted color difference data for m pixels are determined.

On the other hand, in the case of a negative determination in step 303, the color difference data held in the address "1" of the buffer 232 is inversely converted into image data of the RGB color system by the color conversion section 233 (step 308). After the output, the process proceeds to step 309, and the processes of step 301 to step 309 are repeated until the positive determination of step 309 is obtained. With this, it is possible to find all the portions where the lightness index L ^* is constant at both ends while shifting the extraction range by one pixel, and replace the color difference data at this portion with the color difference data at both ends.

When the affirmative determination is made in step 309, it is determined that the processing for one line is completed, and the color difference data held in the buffer 232 is sequentially reverse-converted into image data and output, and the buffer 232 is output. After initializing, the processing of the image data for the next one line may be started.

In this way, by correcting the image data for all lines, the color misregistration in the main scanning direction is corrected, and the false color associated with the edge portion in the sub-scanning direction is removed to obtain a high quality image. The read image can be output.

Further, by performing the same processing in the sub-scanning direction to correct the color misregistration in the sub-scanning direction, it is possible to remove the false color associated with the edge portion in the main scanning direction, so that a higher quality reading can be performed. Images can be output.

Here, the lightness index L ^* of the color difference data is R
Since each of the GB components is reflected, by determining the position of the edge based on the variation of the lightness index L ^* and performing the replacement process, the color misregistration can be corrected while considering the reading error of each of the RGB components. ..

Further, in the color difference space represented by the L ^* a ^* b ^* color system or the like, the value of each component of the color difference data corresponding to the image data obtained by reading a black and white image tends to be a constant value. Since it is particularly strong, the extraction unit 113 and the determination unit 115 described above can accurately determine a portion having a region where color misregistration does not occur at both ends.

The above-mentioned processing may be performed using the perceptual chromaticities a ^* and b ^* instead of the lightness index L ^* . Further, the image data represented by the RGB color system may be directly corrected.

For example, as shown in FIG. 7, the color shift correction unit 230 shown in FIG.
And the image data from the shading correction unit 220 is retained in the buffer 232 as it is, and the read processing unit 234 reads the G component of the image data from the buffer 232 and sends it to the determination unit 115. Good.

In this case, since the color conversion section is unnecessary,
It is possible to simplify the configuration of the device and speed up the correction process. However, the position of the edge is determined based on the variation of the reading result of the G component regardless of the variation of the reading result of the R component and the B component, and the image data replacement process is performed with the edge position as a boundary. When the reading error of the R component or the B component is extremely large, there is a possibility that the color misregistration cannot be completely corrected.

[0056] In correspondence to the RGB components, registers 236 _{1 to 236 4} and the comparator 237a, 3 pairs comprising by determination means a combination of 237b and AND gate 238 115
And the logical product of the determination results for each of the RGB components is sent to the writing processing unit 235 as the overall determination result, the edge correction is performed in consideration of the variation in the reading result of each of the RGB components. It can be carried out.

By the way, if the extraction width of the color difference data (or image data) by the above-mentioned extraction means 113 is too narrow, as shown in FIG. 8A, when the color difference data of the edge portion is extracted, the determination means 115 As a result, the level of at least one end is not constant, so that the edge portion cannot be corrected. On the other hand, if the extraction width is expanded too much, the replacement process described above will be performed when the color difference data including a plurality of edge portions is extracted, as shown in FIG. Pattern may disappear.

Therefore, in order to correct all edges, the extraction means 113 can be used as shown in FIG. 8 (c).
It is necessary to extract the color difference data (or the image data) with an extraction width that is not excessive or insufficient by adding a predetermined margin to the width of the color shift area.

Here, each color component of the read image is represented by the convolution integral of each color component of the original image and the point spread function representing the reading error of each color component of the color reading mechanism 211.
By examining the spread of the point spread function of each color component,
The spread of the blurred image of the edge, that is, the width of the color shift area can be obtained.

FIG. 9 shows a block diagram of an embodiment of the color reading apparatus of the third aspect. In the color reading apparatus shown in FIG. 9, the image reading unit 112 reads the reference pattern 102 prior to reading the original 101, and the components of the obtained image data are line buffers 311r and 311 respectively.
g, 311b.

As shown by hatching in FIG. 10, the reference pattern 102 is a pattern corresponding to a point light source, and a black pattern having a size corresponding to one pixel of the sensor 715 is mainly provided at predetermined intervals. It is formed so as to be arranged in the scanning direction. Conversely, a white pattern may be arranged on a black background to form the reference pattern 102.

In the extraction width determining means 131, the spread width detecting sections 312r, 312g, 312b are supplied with the white level input thereto and the above-mentioned line buffers 311r, 311g,
The image data stored in 311b are sequentially compared,
The range in which the image data other than the white level is continuously distributed is detected as the spread of the image of the pattern corresponding to the point light source described above, and the determination processing unit 313 determines the extraction range based on the detection result. It is configured to do. The above-mentioned spread width detection units 312r, 312g, 312b
Is an address of the line buffers 311r, 311g, 311b indicating the start position and the end position of the image spread corresponding to each of the patterns corresponding to the point light sources, and this address is used as the detection result of the image spread range. The configuration may be such that it is sent to the decision processing unit 313.

These spread width detecting sections 312r, 312
The image spread of the pattern corresponding to the point light source obtained by g, 312b is the spread of the point spread function of each color component,
In addition, the range of the light receiving element of the sensor 715 having the reading sensitivity for the pixel at the position of the above-described pattern is shown.

As described above, in the filter contact type color reading mechanism, the reading position of the light receiving element corresponding to each color component is deviated, so that the reading sensitivity distribution of each color component is as shown in FIG. In addition, they are spread around different positions.

Therefore, the determination processing unit 313 may set the extraction width as a value obtained by adding a predetermined margin to both sides of the range given by the intersection of the image expansion ranges of the respective color components. For example, by the spread width detection units 312r, 312g, and 312b,
The minimum value of the address is subtracted from the maximum value of the address obtained for each of the patterns corresponding to the point light source, and the intersection of the spread range of each color component of the corresponding image is obtained, and the maximum value of these intersections is calculated. The value obtained by adding a margin for two pixels may be set as the extraction width, and the capacity of the buffer 232 included in the color misregistration correction unit 230 may be set.

As a result, it is possible to accurately evaluate the width of the color misregistration area due to the reading error of the color reading mechanism 211 and set the extraction width that is just enough, so that the color misregistration of the edge portion included in the read image is not lost. Can be corrected.

Further, the image data of one line may be divided into a plurality of areas corresponding to each of the patterns corresponding to the point light sources, and the determination processing unit 313 may determine the extraction width for each area. ..

As shown in FIG. 12, a reference pattern 103 having an edge pattern is provided in place of the reference pattern 102, and the extraction width determining means 132 determines the extraction width based on the reading result of the reference pattern 103. It may be configured to determine.

As the above-mentioned reference pattern 103, as shown in FIG. 13, a pattern in which black strip-shaped patterns and white strip-shaped patterns which are sufficiently long in the sub-scanning direction are alternately arranged may be used. However, in FIG. 13, a black pattern is shown by hatching. For this strip pattern,
What has a width sufficiently wider than the reading sensitivity distribution of each color component may be used.

The extraction width determining means 132 is provided with spread width detecting portions 314r, 314g, 314b instead of the spread width detecting portions 312r, 312g, 312b shown in FIG. This spread width detection unit 314r,
314g and 314b are line buffers 311r and 31
The image data of each pixel held in 1g and 311b is compared with the white level and the black level, and the range in which the image data having an intermediate level between the white level and the black level is continuously distributed is defined as the spread width of the edge pattern. And is sent to the decision processing unit 313.

Thus, the spread width of the edge pattern due to the reading sensitivity distribution of the color reading mechanism 211, that is, the width of the color misregistration area is directly evaluated, and an extraction width that is just enough and corresponds to the width of the color misregistration area is obtained. You can

Further, in this case, since the same reading result can be obtained within a predetermined range in the sub-scanning direction, the alignment in the sub-scanning direction is easy. Therefore, the line width 311 of the extraction width determining unit 132 surely holds the image data corresponding to the read intensity distribution of each color component, and the spread width detection unit 31.
Since it can be subjected to the spread width evaluation process of 4, the width of the read intensity distribution of each color component can be accurately evaluated. With this, the determination processing unit 313 can accurately obtain the extraction widths that are just enough, so that the color misregistration of the edge portion can be completely corrected, and a sharp edge can be reproduced.

The image reading means 112 may be constructed by using the filter switching type color reading mechanism shown in FIG. 14, and the obtained reading result may be corrected by the color shift correcting section 230.

In the filter switching type color reading mechanism, the reading positions of the respective color components in the main scanning direction are the same, but since the reduction optical system is included, the periphery of the original 101 having a large image height is included. The chromatic aberration of the part is large.

The color shift due to such chromatic aberration can be similarly corrected by the above-described correction processing, so that the original 10 is corrected.
It is possible to correct the color misregistration of the edge portion of the peripheral portion of No. 1 and reproduce the sharp edge as in the central portion. As a result, a good read image can be obtained over the entire screen.

[0076]

As described above, according to the present invention, an area including a part where no color shift occurs at both ends is discriminated, and image data of each pixel included in the corresponding area is separated by a virtual edge. As a result, by replacing the image data at both ends with the color misregistration of the edge portion caused by the reading error of each separated color in the image reading means, a sharp edge image without false color is reproduced and a high quality read image is obtained. Obtainable.

Further, since the judging means makes the judgment based on the color difference data, it is possible to accurately judge the region including the portions where the color misregistration does not occur at both ends. In addition, based on the reading result of the reference pattern, the extraction width is determined in consideration of the spread of the reading result of the pattern corresponding to the point light source or the edge pattern, and the edge portion is corrected without omission, and high It is possible to obtain a quality read image.

[Brief description of drawings]

FIG. 1 is a diagram showing a configuration of a color reading apparatus according to claim 1.

FIG. 2 is a diagram showing a configuration of a color reading device according to a second aspect of the invention.

FIG. 3 is a diagram showing a configuration of a color reading device according to claims 3 and 4;

FIG. 4 is a diagram showing a configuration of an embodiment of a color reading device according to claim 1;

FIG. 5 is a flowchart showing a color misregistration correction operation.

FIG. 6 is an explanatory diagram of a color misregistration correction operation.

FIG. 7 is a diagram showing the configuration of another embodiment of the color reading apparatus according to claim 1;

FIG. 8 is a diagram showing a relationship between a width of a color shift area and an extraction width.

FIG. 9 is a diagram showing a configuration of an embodiment of a color reading device of claim 3;

FIG. 10 is a diagram showing an example of a reference pattern.

FIG. 11 is an explanatory diagram of a reading sensitivity distribution.

FIG. 12 is a diagram showing a configuration of an embodiment of a color reading device according to claim 4;

FIG. 13 is a diagram showing an example of a reference pattern.

FIG. 14 is a configuration diagram of a filter switching type color reading device.

FIG. 15 is a configuration diagram of a filter contact type color reading device.

FIG. 16 is an explanatory diagram of a sensor.

FIG. 17 is an explanatory diagram of color misregistration.

FIG. 18 is a diagram showing an example of false color generation.

[Explanation of symbols]

 101 Originals 102, 103 Reference Pattern 111 Line Image Sensor 112 Image Reading Means 113 Extracting Means 114 Holding Means 115 Judging Means 121 Color Converting Means 122 Color Inverse Converting Means 131, 132 Extraction Width Determining Means 211 Color Reading Means 213 Analog-Digital Converter (A / D) 220 Shading correction unit 221 Memory 230 Color shift correction unit 231,233 Color conversion unit 232 Buffer 234 Read processing unit 235 Write processing unit 236 Register 237 Comparator 238 AND gate 311 Line buffer 312, 314 Spread width detection Part 313 Determination Processing Part 611 Original Document 612 Illumination Device 613 Reading Line 621 Filter 614 Lens 615, 715 Sensor 714 Rod Lens Array

Claims (4)

[Claims] 1. A line image sensor (111) in which a plurality of light receiving elements are arranged in one line, and each pixel corresponding to each light receiving element of the line image sensor (111),
An image reading unit (11) that reads a black and white image of a linear region on a document (101) by separating it into a plurality of color components, and outputs image data consisting of a plurality of components indicating the intensity of each separated color.
2), extraction means (113) for sequentially extracting a predetermined number of pixels of image data starting from each pixel from the image data corresponding to the linear area, and extraction by the extraction means (113). Holding means (114) for holding image data of pixels at the start and end points of the range
And the image data extracted by the extraction means (113) is input, and a determination means (115) for determining whether or not the values of at least one component of the image data are constant at both ends of the extraction range. ), And the extraction range is divided into an area on the start side and an area on the end side according to the determination result that the values of the corresponding components at both ends of the extraction range are constant and included in each area. A color reading device comprising: a replacement unit (116) for replacing the image data of each pixel with the image data of the starting and ending pixels held in the holding unit (114) and transmitting the image data. 2. The color reading device according to claim 1, wherein the image data obtained by the image reading means (111) is converted into color difference data representing a corresponding color in the color difference space, and the line on the original (101) is converted. Conversion means (121) for sending to the extraction means (113) as image data corresponding to the striped area
And a color reverse conversion means (122) for converting the image data from the replacement means (116) into image data composed of components corresponding to the separated color in the image reading means (112) and sending the image data. And a color reading device. 3. The color reading device according to claim 1, wherein a reference pattern (102) having at least one pattern corresponding to a point light source, and the reference pattern (1) by an image reading means (112).
02), the range of the line image sensor (111) where the light from the pattern corresponding to the point light source reaches is determined according to the input of the image data, and the number of pixels corresponding to this range is set as the extraction width. , Extraction means (11
Extraction width determination means (131) used for the extraction process according to 3)
And a color reading device. 4. The color reading device according to claim 1, wherein at least one edge pattern represented by white and black is provided.
The reference pattern (103) having the two, and the reference pattern (1) by the image reading means (112).
03) is input, and the range of the line image sensor (111) from which image data having a level between a white level corresponding to white and a black level corresponding to black is obtained, A color reading device, comprising: an extraction width determining unit (132) that is used for extraction processing by the extraction unit (113), using the number of pixels corresponding to the range as an extraction width.