JP3379462B2 - Image processing device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus mounted on a digital color copying machine or the like, and more specifically, to a positional deviation between R, G and B in a sub scanning direction of a reduction type color CCD sensor. The present invention relates to a line-to-line correction process for correcting in accordance with a scaling ratio.

[0002]

2. Description of the Related Art Image reading units such as color copying machines
For example, as described in Japanese Patent Application Laid-Open No. 9-261491, a reduction type color CCD sensor generally reads information obtained by reducing and projecting an original image through an optical system from the viewpoint of cost. Such a reduced color CC
As shown in FIG. 10, in the D sensor, the R (red), G (green), and B (blue) element rows in which pixels are arranged in the main scanning direction are separated by a predetermined distance d in the sub scanning direction. The structures are arranged in parallel with each other.

In the case of image reading using the reduction type color CCD sensor as described above, R in the sub-scanning direction, which is the direction in which the original and the CCD sensor move mechanically relative to each other,
Due to the positional deviation of G and B (distance d), a temporal deviation, that is, a phase deviation occurs between the image signals of R, G, and B colors obtained from the CCD sensor.

In order to correct the phase shift between R, G and B (hereinafter also referred to as position shift), the first R output image data is delayed by a time corresponding to an interval 2d (for example, 8 lines), The G output image data generated next is separated by the interval d.
By delaying the time corresponding to (for example, four lines), the correction processing for matching the phase with the R output image data that occurs last is performed.

Further, when the scaling factor for reducing and projecting an original image changes, such as when the scanning speed in the sub-scanning direction changes in a color copying machine having a reduction / enlargement function, the phase between R, G and B changes. The shift may not be an integral multiple of one line, and a fraction (fractional part) may occur. In such a case, it is necessary to correct the phase shift between R, G and B by interpolation processing as accurately as possible. That is, when the corrected position is a position between the lines, the density value of each color at that position is obtained by the weighted average of the values on the lines on both sides.

[0006]

However, as a result of the subsequent studies, in the correction of the phase shift between R, G, and B when the above-mentioned reduction type color CCD sensor is used, the decimal part is corrected by interpolation processing. It was found that the reproducibility of the black fine line was deteriorated by carrying out. It is considered that this is because the balance of the reading characteristics of R, G, and B colors is greatly disturbed when the above-described interpolation process is performed when the reduced projection of the black thin line becomes close to, for example, one dot width.

The present invention has been made in view of the above problems, and corrects the phase shift between R, G, and B in the case of using a reduction type color CCD sensor as accurately as possible. An object of the present invention is to provide an image processing device capable of enhancing the reproducibility of a black fine line.

[0008]

An image processing apparatus according to the present invention is an image obtained from an image sensor having a structure in which a plurality of element rows long in the main scanning direction are arranged in parallel with each other at a predetermined pitch in the sub scanning direction. An apparatus for performing data correction processing, comprising an inter-line correction unit that corrects a positional deviation between element rows in the sub-scanning direction of an image sensor, and the inter-line correction unit performs integer-line correction that performs correction on a line-by-line basis. Units and a fractional part correction unit that performs correction of less than one line by interpolation calculation, and it is configured so that activation / non-operation of the fractional part correction unit can be selected. Correction for less than one line is usually performed by interpolation processing.

[0009]

[0010] Then, the main scanning direction included in the image data
A black thin line detection unit that detects a black thin line that extends to the black thin line detection unit is provided.If the width of the black thin line is larger than a predetermined value based on the output signal of the black thin line detection unit, the decimal part correction unit is activated to set the width In the following cases, the operation of the fractional part correction unit is prohibited. That is, when the black thin line detection unit line width after reduction projection detects black thin line of approximately one dot, you prohibit the correction by interpolation of the fraction correction unit based on the detection output.

Further, preferably, an area discriminating section for discriminating an area to be subjected to image processing such as edge enhancement is further provided, and a method of creating lightness data in the area discriminating section is switched according to an output signal of the black fine line detecting section. . For example, in the normal processing, the minimum value of the R, G, B color image data is output as the lightness data, but the maximum of the R, G, B color image data is output only when the black thin line detection unit detects the black thin line. The value is output as brightness data. This improves the contrast. Further, it switches the threshold for determining the black edge in the realm discriminator in accordance with the output signal of the black thin line detection unit. For example, when the black fine line detection unit detects a black fine line, the threshold value for determining the black edge portion is set to be higher than that in the case where it is not. This prevents erroneous discrimination due to noise.

According to the fifth aspect of the present invention, the threshold value for determining the black edge portion in the area discriminating unit is switched according to the output signal of the black fine line detecting unit. For example, when the black fine line detection unit detects a black fine line, the threshold value for determining the black edge portion is set to be higher than that in the case where it is not. This prevents erroneous discrimination due to noise.

[0013]

1 is a block diagram showing the overall configuration of an image processing apparatus M1 according to the present invention.

In FIG. 1, information obtained by reducing and projecting an original image through an optical system is read by a reduction type color CCD sensor 12. The obtained R, G, and B color image signals are input to the A / D converter 13. A / D converter 1
3 converts an R, G, B color image signal which is an analog signal into R, G, B color image data which is 8-bit digital data (density data of 256 gradations). The obtained R, G, B color image data is subjected to shading correction for correcting light amount unevenness in the main scanning direction by the shading correction unit 14, and then the interline correction unit 15
Entered in.

The line-to-line correction unit 15 is a reduction type color CC.
This is a circuit for correcting the phase shift of the R, G, B color image signals (data) due to the positional shift between the R, G, B lines of the D sensor 12. Correction is performed by delaying R and G color image data using a field memory.
The specific circuit configuration will be described later. The R, G, B color image data output from the interline correction unit 15 is corrected by the chromatic aberration correction unit 16 for color misregistration due to chromatic aberration of the lens system. In addition, scaling including line memory for scaling
The movement processing unit 17 performs enlargement / reduction processing in the main scanning direction according to the scaling ratio.

R output from the scaling / movement processing unit 17,
The G and B color image data are input to the color conversion unit 18,
After the adjustment between R, G, B is performed, the color correction unit 19 performs R
The color image data of GB (addition color system) is converted into CMY (color reduction system) color image data C (cyan), M (magenta), Y (yellow), and Bk (black). The C, M, Y, and Bk color image data are subjected to processing such as edge enhancement and smoothing by the MTF correction unit 20 and then given to the printer unit via the printer interface 21.

Further, R, which is output from the color conversion unit 18,
The G and B color image data is also given to the area discriminating unit 22, and the area discriminating unit 22 discriminates whether the read image is a halftone dot image, a character image or a photographic image.
By providing the determination result to the MTF correction unit 20, the MTF correction unit 20 switches whether to perform correction processing such as edge enhancement and smoothing according to the type of image in the area.

In FIG. 1, a 1-dot-width black fine line detection unit 2
Reference numeral 3 is a circuit for detecting whether or not the image currently being read is a black thin line having a width of approximately 1 dot. The 1-dot width black thin line detection unit 23 outputs the output signal BI of the shading correction unit 14.
Based on N and the R image delay data RMD and the G image delay data GMD output from the interline correction unit 15, the image portion currently being processed in the image information projected on the reduction type color CCD sensor 12 has a 1-dot width. Whether or not the line is a black thin line is determined as described below. If it is determined that the line is a 1-dot-width black thin line, its output signal SEL is set to L (low) level, and otherwise, it is set to H (high) level. This output signal SEL is used for the interline correction unit 15 and the area determination unit 22.
Given to.

FIG. 2 is a block diagram of the interline correction unit 15.

In FIG. 2, signals RIN, BIN, GI
N is color image data input of R (red), B (blue), and G (green), respectively. Input these color image data R
IN, BIN, and GIN are subjected to delay correction for integer lines by the first correction unit 30, and are fractional (decimal) by the second correction unit 31.
Minute image interpolation processing is performed, and color image data output ROU
T, BOUT, GOUT.

The first correction section 30 includes a field memory 33.
~ 35 for color image data input RIN and GI
Delay N with respect to color image data input BIN.
That is, as described in the description of the prior art, the color image data input RIN is set to R in the sub scanning direction of the CCD sensor 12.
The color image data input GIN is delayed by a time corresponding to a distance (element row interval) 2d between the element row and the B element row, and the color image data input GIN is divided between the G element row and the B element row in the sub scanning direction of the CCD sensor 12. It is delayed by a time corresponding to the element row interval d between them. The element row interval d is represented by the number of lines shifted between both element rows. In the present embodiment, the element row interval d between the G element row and the B element row is “4”, and the element row interval 2d between the R element row and the B element row is “8”.

The field memories 33 to 35 are used to delay the image data in units of a plurality of lines. For example, each field memory 33-35 has 256K
If the memory capacity is Bytes and the image data capacity of each color per line is 5 KBytes (for 5,000 pixels), the image data for 51 lines per field memory can be delayed. As shown in FIG. 2, the color image data input RIN can be delayed up to 102 lines by two field memories 33 and 34 connected in series, and the color image data input GIN can be delayed by 51 lines by one field memory 35. Can be delayed until.

The actual amount of delay is determined by each field memory 33.
This is performed by controlling the timing of the reset signals given to the read reset terminal RRES and the write reset terminal WRES of ~ 35. Note that, in FIG. 2, "-" added to the head of the sign of each signal means a negative logic signal, and "-" is omitted in this description. The same applies to the other figures and their description.

Each of the field memories 33 to 35 starts writing input data when a reset signal is applied to the write reset terminal WRES, and starts output of accumulated data when a reset signal is applied to the read reset terminal RRES. Therefore, after the reset signal is applied to the write reset terminal WRES, the read reset terminal RRES
The period until the reset signal is applied to is the delay amount.

In FIG. 2, the reset signal R is applied to the write reset terminals WRES of the field memories 33 and 35.
ES0 is applied, and the reset signal RES1 is applied to the read reset terminal RRES. Further, the reset signal RES1 is applied to the write reset terminal WRES of the field memory 34, and the reset signal RES2 is applied to the read reset terminal RRES. Therefore, the color image data input RINR is reset by the reset signal R by the two field memories 33 and 34 connected in series.
The color image data input GINR is delayed by the period from ES0 to the reset signal RES2, and the color image data input GINR is reset by the field memory 35 from the reset signal RES0 to the reset signal R.
It is delayed by the period until ES1. The color image data input BINR is passed to the second correction unit 31 without delay.

FIG. 3 shows reset signals RES0 and RES.
1 shows a timing chart of RES2 and R, G, B output image data.

FIG. 3 shows that the G image delay data GMD is obtained with an n line delay with respect to the B image data, and the R image delay data RMD is obtained with an n line delay. The value of n, that is, the number of lines corresponding to the delay time from the reset signal RES0 to the reset signal RES1, is the integer part (in
t). In the case of the same size, the element row interval itself is, for example, 4 lines, but, for example, in the case of a scaling factor of 0.6, it is 4
The integral part 2 is × 0.6 = 2.4. Actually, the value obtained by adding 1 to this value is the final delay amount (number of lines). This is for facilitating the interpolation process in the second correction unit 31 described later. Similarly, regarding the number of lines corresponding to the delay time from the reset signal RES0 to the reset signal RES2, the value obtained by adding 1 to twice the integer part (int) of the value obtained by multiplying the element column interval by the scaling factor is finally determined. The delay amount (number of lines) is used.

As described above, the R delayed image data and the G delayed image data obtained are given to the second correction section 31 together with the B image data which is not delayed. Second correction unit 31
Outputs the G delayed image data as it is as the color image data output GOUT without performing the correction processing,
The delayed image data and the B image data are interpolated based on the G delayed image data. This is R (red)
This is for avoiding a decrease in density due to the interpolation processing for the image of G and the image of G (green) which is more noticeable than the image of B (blue).

The second corrector 31 is provided with R delay image data and
For each of the B image data, one image data
And FIFO memory 36 and 37 for delaying
It has arithmetic units 38 and 39. A input terminal of the interpolation calculation unit 38
R delayed image data R for the child_nIs input and F is input to the B terminal.
R delay image further delayed by one line in the IFO memory 36
Data R_n-1Is entered. Also, the K terminal has an interpolation coefficient
α is input. The interpolation calculation unit 38 uses the interpolation coefficient α.
Data R according to the formula described below_nAnd data R _n-1With
Interpolation data R_n’Is calculated.

On the other hand, F is connected to the A input terminal of the interpolation calculation section 39.
B image data B delayed by one line in the IFO memory 37
_n-1 is input, and B image data B _n before delay is input to the B terminal.
Is entered. Further, the interpolation coefficient α is input to the K terminal. The interpolation calculation unit 39 uses the interpolation coefficient α to interpolate the data B _n and the data B _n-1 according to the equation described later.
Calculates _n '.

The interpolation coefficient α is a fraction (fractional part) when the element row interval is multiplied by the scaling factor. In the above example, the element row interval "4" is multiplied by the scaling factor of 0.6. It is 0.4 of 0.4. Therefore, the interpolation coefficient α is obtained from the following equation.

[0032]

## EQU00001 ## .alpha. = Element row interval.times.magnification-int (element row interval.times.magnification) where int () is an operator for extracting the integer part of the numerical value in ().

Using this interpolation coefficient α, R after interpolation calculation
And B image data, that is, interpolation data R _n 'and B
_n 'is calculated from the following formula.

[0034]

[Equation 2]

[0035]

[Equation 3]

In the above equations (Equations 2 and 3), the coefficient α or (α-1) by which the data of the (n-1) th line and the data of the nth line are multiplied is the R image data and the B image data. And are reversed. In this regard, as shown in FIG. 2, the relationship between the image data input to the A and B terminals of the interpolation calculation units 38 and 39 and the delay data for one line thereof is the interpolation calculation unit for R image data. 38 and the interpolation calculation unit 39 for B image data are reversed. This is because, as described above, the reduction type color CCD sensor 12 has a structure in which the R element row and the B element row are arranged in front of and behind the G element row. That is, the relationship between the equations (Equation 2 and Equation 3) is schematically shown in FIG.

In FIG. 4, R_n, R_n-1, B_n, B
_n-1If the position of is fixed, the value of the interpolation coefficient α is 0 to 1
Change to G_nIs in position B_n-1(R_n) Side to B
_n(R _n-1) Side changes. As mentioned above, the first correction
In the part 30, the R image data is compared with the B image data.
Although it is delayed by one extra line, this second correction unit 3
By the interpolation calculation in 1, the one line is compensated,
The phases of the R, G, and B image data must be aligned.
It

As described above, in addition to the positional deviation correction for the integer line in the first correction unit 30, the decimal correction processing in the second correction unit 31 is performed, so that more detailed color misregistration correction can be performed. It can be carried out. However, since the decimal part interpolation process takes a weighted average of the densities of the preceding and following lines, in the case of a 1-dot wide thin line, the density is greatly reduced by this interpolation process. For this reason,
It was found that in the case of a black thin line, the balance of the densities of R, G, and B colors is greatly disturbed, and reproducibility is deteriorated.

A description will be added with reference to FIG. FIG. 5 shows positions (phases) and densities of R, G, and B image data in the case of a 1-dot-width black thin line, and (a) shows a state before correction by the interline correction unit 15. . (B) shows the positions (phases) and densities of the R, G, and B image data after correction when the scaling factor is equal, that is, when the correction coefficient α is 0. In this case, since the fractional part interpolation process in the second correction unit 31 is not actually performed, the density of each image data does not decrease, and the first correction unit 30 only performs the positional deviation correction for an integer line. become. In addition, in these explanations
As is apparent, the black thin line detected is the CCD sensor 1
2 also extends in the direction in which each element row extends, that is, in the main scanning direction.
Therefore, for each element row of the 1-dot-width black thin line,
The image data obtained from the sub-image data is, as shown in FIG.
There is only one pixel in the scanning direction.

In FIG. 5C, when the scaling factor is not equal and the correction coefficient α does not become 0, correction by both the first correction unit 30 and the second correction unit 31 of the interline correction unit 15 is performed. The positions (phases) and densities of the R, G, and B image data are shown when they are performed. In this case, the density of the G image data that is the reference (no interpolation processing is performed) does not change, whereas the densities of the R and B image data are greatly reduced by the interpolation processing. As a result, as described above, the density balance of each color is largely lost, and the reproducibility of the black fine line is deteriorated.

Therefore, as shown in FIG. 2, the selector 32
And a second dot based on an output signal (hereinafter referred to as an interpolation prohibition signal) SEL of the 1-dot-width black thin line detection unit 23 in FIG.
The correction unit 31 is configured to automatically select whether or not to perform the interpolation processing of the decimal part. In FIG. 5C, when only the position correction for the integer line is performed by the first correction unit 30 and the decimal part interpolation processing is not performed by the second correction unit 31, the process is as shown in FIG. 5D. In this case, R, G,
Although the positional deviation between B slightly remains, the density deterioration due to the interpolation processing does not occur, so that the balance of the density characteristics of each color is not lost, and the reproducibility of the black fine line is improved as compared with (c).

In FIG. 2, the selector 32 selects the interpolation coefficient α input to the B terminal when the interpolation prohibition signal SEL input to the S terminal is at the H level (inactive level), and the interpolation calculation unit 38, 39 K input terminal. However, when the 1-dot-width black thin line detection unit 23 detects the 1-dot-width black thin line and the interpolation prohibition signal SEL becomes L level (active level), "00" input to the A terminal is selected and interpolation calculation is performed. It is applied to the K input terminals of the parts 38 and 39. That is, the equation (Equations 2 and 3) and the interpolation coefficient α in FIG.
Is forced to 0.

In this way, in the case of a 1-dot-width black fine line, by not performing the interpolation processing of the decimal part, the reproducibility of the black fine line is prioritized over the avoidance of the density decrease over the detailed correction of the positional shift (color shift). Has improved.

The 1-dot-width black thin line detecting section 23
B For each color image data, the absolute value of the difference between the current image data and the preceding and following data is obtained, and if the six values are greater than a predetermined threshold value, it is determined to be a 1-dot-width black thin line and interpolation is performed. The prohibition signal SEL is set to L level (active level). When at least one of the values is smaller than the threshold value, the interpolation prohibiting signal SEL is set to H level (inactive level). That is, the interpolation prohibiting signal SEL becomes L level only when the following formula is satisfied.

[0045]

[Equation 4]

The respective threshold values S1 to S6 may be changed or may be the same depending on the characteristics of each color.

Next, detailed circuits of the area discriminating unit 22 in FIG. 1 are shown in FIGS. First, in FIG. 6, R,
The minimum value of the G and B color image data is obtained by the minimum value circuit 41 and output. Further, the maximum value of the R, G, B color image data is obtained by the maximum value circuit 42,
Is output. Further, the saturation calculation unit 43 causes R, G, B
The difference between the maximum and minimum values of color image data is calculated,
The saturation data W7-0 is output. In addition, R, G,
The greater the difference between the maximum value and the minimum value of the B color image data, the greater the saturation.

On the other hand, the maximum value and the minimum value of the R, G, B color image data are input to the A terminal or the B terminal of the selector 44. The selection control terminal S has an output signal SE (hereinafter referred to as a black thin line detection signal) SE of the 1-dot-width black thin line detection unit 23.
L is input. The selector 44 outputs the black thin line detection signal SE
If L is at H level, the minimum value (B terminal input) of R, G, B color image data is selected and output as lightness data V7-0. On the other hand, if the black fine line detection signal SEL is at the L level, the maximum value of the R, G, B color image data (A terminal input) is selected and output as the brightness data V7-0. This facilitates the edge detection in the area discrimination unit 22 in the case of a 1-dot-width black thin line, and improves the contrast.

The brightness data V7-0 is given to the feature quantity extraction filter 51, as shown in FIG. The feature amount extraction filter 51 calculates a primary differential and a secondary differential in each of the main scanning direction and the sub scanning direction of the brightness data V7-0. The larger one of the primary differential values in the main scanning direction and the sub scanning direction is selected by the maximum value circuit 52, and the value is input to the comparators 54 and 55.

The comparator 54 inputs 1 to the P terminal.
If the second derivative value is larger than the threshold value for determining the edge portion 1 input to the Q terminal, the edge portion 1 determination output EDG1 is set to H level, and if it is smaller, it is set to L level. There are two thresholds REF1A and RE as the threshold for determining the edge portion 1.
F1B is prepared, and the selector 50a selects either one. If the black thin line detection signal SEL is at H level (inactive), the threshold value REF1A is selected, and if it is at L level (active), the threshold value REF1A is selected. However, REF1A is larger than REF1B, that is, severe.

The comparator 55 receives 1 input to the Q terminal.
The non-edge portion 1 determination output NEDG1 is set to H level if the next differential value is smaller than the non-edge portion 1 determination threshold value input to the P terminal, and is set to L level if larger. There are two thresholds REF2A as the thresholds for determining the non-edge portion 1.
And REF2B are prepared, and the selector 50b selects either one. If the black thin line detection signal SEL is at H level (inactive), the threshold value REF2B is selected,
If it is at L level (active), the threshold value REF1A is selected. However, REF1A is smaller than REF1B, that is, severe.

Similarly, the larger one of the secondary differential values in the main scanning direction and the sub scanning direction is selected by the maximum value circuit 53,
The value is input to the comparators 56 and 57.

The comparator 56 inputs 2 to the P terminal.
If the second derivative value is larger than the threshold value for determining the edge portion 2 input to the Q terminal, the edge portion 2 determination output EDG2 is set to H level, and if it is smaller, it is set to L level. There are two thresholds REF3A and RE as the threshold for determining the edge portion 2.
F3B is prepared, and the selector 50c selects either one. If the black thin line detection signal SEL is at H level (inactive), the threshold value REF3B is selected, and if it is at L level (active), the threshold value REF3A is selected. However, REF3A is larger than REF3B, that is, severe.

The comparator 57 inputs 2 to the Q terminal.
The non-edge portion 2 determination output NEDG2 is set to H level if the next differential value is smaller than the non-edge portion 2 determination threshold value input to the P terminal, and is set to L level if it is larger. There are two thresholds REF4A as thresholds for determining the non-edge portion 2.
And REF4B are prepared, and the selector 50d selects either one. If the black thin line detection signal SEL is at H level (inactive), the threshold value REF4B is selected,
If it is at L level (active), the threshold value REF4A is selected. However, REF4A is smaller than REF4B, that is, severe.

As described above, when determining the edge portion,
When the black fine line detection signal SEL is at the L level, that is, when a 1-dot-width black fine line is detected, the threshold value is set to be more severe than when it is not. The determination output of the non-edge portion is used as a signal for preventing a black character erroneous determination as described later, but when determining the non-edge portion, when a 1-dot-width black thin line is detected, it is more likely than not. The threshold is tight.

Edge portion 1 determination output EDG1 and edge portion 2
The determination output EDG2 is input to the NOR circuit 58, and if either input is at H level, the output becomes L level. This output signal is output to another circuit as an edge portion signal EDG indicating an edge portion at the L level, and a signal B indicating a black edge portion is further processed as described later.
KEDGE is generated.

The non-edge portion 1 determination output NEDG1 and the non-edge portion 2 determination output NEDG2 are input to the AND gate 59 together with the signal COLOR indicating the color region. Both NEDG1 and NEDG2 are at H level, and
In the case of the color area, the black character erroneous discrimination signal CAN output from the AND gate 59 becomes H level. This signal C
The AN is used to prevent erroneous discrimination of black characters, as described later.

A signal COLO indicating the color area
R is obtained by ANDing the comparison result of the brightness data V7-0 and the threshold value REF5 and the comparison result of the saturation data W7-0 and the threshold value REF6 by the AND circuit 63.
In the comparator 60, the brightness data V7-0 is the threshold value R.
If it is smaller than EF5, the output is activated. In the comparator 61, the saturation data W7-0 has a threshold value REF6.
If larger, activate output.

Edge signal EDG, that is, NOR circuit 5
The output of 8 is input to the closing processing circuit 64 and 3
Enlargement processing of the edge portion is performed by a × 3 matrix calculation. The output of the closing processing circuit 64 is input to the A terminal of the selector 65. B which is another input of the selector 65
The output of the NOR circuit 58, that is, the edge portion signal EDG before the enlargement processing is input to the terminal. In the selector 65, the inner edge signal INEDG input to the selection control terminal S is L
When the level is high, the edge portion signal after the enlargement processing input to the A terminal is selected, and when the level is H, the edge portion signal before the enlargement processing input to the terminal B is selected.

After all, when the inner edge signal INEDG is at the L level, enlargement processing is performed, and when the inner edge signal INEDG is at the H level, the edge portion signal that is not subjected to the expansion processing is output from the Y terminal of the selector 65. The inner edge signal INEDG is a signal obtained by comparing the output signal of the inner edge detection filter in the feature quantity extraction filter 51 with “00” by the comparator 67, and is on the inner side of the edge region, that is, on the black line. Is a signal indicating L level. Therefore, on the outer side of the edge region, that is, on the background side of the black line, the inner edge signal IN
EDG becomes H level. This edge signal INEDG
Is also used in other circuits.

The output of the selector 65 becomes one input of the negative logic AND gate 66, and the negative logic black area signal BLK indicating the black area is input to the other input. The negative logic AND gate 66 takes the logical product (in negative logic) of the two signals and outputs it as the negative logic black edge signal BKEDG. The black area signal BLK is a signal obtained by comparing the saturation signal W7-0 with the data obtained by converting the brightness data V7-0 by the black determination threshold value table (WREF) 68. That is, the comparator 69 uses the black determination threshold value table (WREF) 68 input to the terminal P.
If the output data of is larger than the saturation signal W7-0 input to the terminal Q, the output BLK is set to L level.

The black judgment threshold value table 68 has two kinds of conversion characteristics as shown by a solid line and a broken line in FIG. 9, and either one of them is converted by the black thin line detection signal SEL input to the switching terminal A8. The characteristic is selected. When the black fine line detection signal SEL is at the H level (inactive) normally, the conversion characteristic indicated by the broken line is selected, but the black fine line detection signal SEL is L.
At the level (active), that is, when the 1-dot-width black thin line is detected, the solid line conversion characteristic is selected. That is, when the 1-dot-width thin black line is detected, the reference data input to the P terminal of the comparator 69 becomes higher (loose) than otherwise, and it is easy to determine the black area.

The black character erroneous discrimination signal CAN output from the AND gate 59 is applied to the black character erroneous discrimination number counting section 71, as shown in FIG. 8, where the number of black character erroneous discriminations in the 9 × 9 matrix is counted. It The result is given to the comparator 72 and compared with the threshold value REF7. If the number of black character misidentifications is smaller than the threshold value REF7, the output of the comparator 72 becomes L level, and if larger, it becomes H level. This output is a negative logic AND gate 7
3 and is logically ANDed with the black edge signal BKEDG (output of the negative logic AND gate 66 in FIG. 7).

The output of the negative logic AND gate 73 finally becomes the negative logic signal PAPA indicating the black character edge portion. That is, when the black edge signal BKEDG is at the L level (active) and the number of black character misidentifications counted by the black character misidentification number counting unit 71 is smaller than the threshold value REF7, the black character edge signal PAPA becomes L level. Become. These area discrimination signals are given to the MTF correction circuit 20 as the output of the area discrimination unit 22 in FIG. 1, and the MTF correction circuit 20 performs correction processing such as edge enhancement and smoothing according to the area discrimination signal.

In the above-described embodiment, the element row interval d is "4" and the element row interval 2d is "8". However, other numerical values or non-integer numerical values may be used.

[0066]

As described above, according to the present invention, R, G, and
While the phase shift between B is corrected as accurately as possible, the interpolating process can be prohibited in the case of a black thin line to improve its reproducibility.

[Brief description of drawings]

FIG. 1 is a block diagram showing an overall configuration of an image processing apparatus according to an embodiment of the present invention.

FIG. 2 is a block diagram illustrating a configuration example of an interline correction circuit that constitutes the image processing apparatus.

FIG. 3 is a timing chart showing the integer part correction by the first correction unit of the line correction circuit.

FIG. 4 is a schematic diagram for explaining a decimal part interpolation process by a second correction unit of the interline correction circuit.

FIG. 5 is a graph schematically showing the phase and density of R, G, and B image data in the case of a 1-dot-width black thin line.

FIG. 6 is a block diagram illustrating a configuration example of an area determination unit included in the image processing apparatus.

FIG. 7 is a block diagram illustrating a configuration example of an area determination unit included in the image processing apparatus.

FIG. 8 is a block diagram illustrating a configuration example of an area determination unit included in the image processing apparatus.

FIG. 9 is a graph showing conversion characteristics of a black determination threshold value table that constitutes an area determination unit.

FIG. 10 is a schematic view showing the structure of a reduction type color CCD sensor.

[Explanation of symbols]

M1 image processing device 12 Reduction type color CCD sensor 13 A / D converter 14 Shading correction circuit 15 line correction circuit 16 chromatic aberration correction circuit 17 Magnification / movement processing circuit 18 color conversion circuit 19 color correction circuit 20 MTF correction circuit 21 Printer Interface 22 Area discriminator 23 1-dot width black fine line detector

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁷ identification code FI H04N 1/48 H04N 1/04 D (58) Fields investigated (Int.Cl. ⁷ , DB name) H04N 1/40-1 / 409 H04N 1/46 H04N 1/60

Claims (4)

(57) [Claims] 1. An image processing apparatus for correcting image data obtained from an image sensor having a structure in which a plurality of element rows long in the main scanning direction are arranged in parallel with each other at a predetermined pitch in the sub scanning direction, An inter-line correction unit that corrects a positional deviation between the element rows in the sub-scanning direction of the image sensor is provided, and the inter-line correction unit includes an integer part correction unit that performs correction on a line-by-line basis.
A fractional part correction unit that performs correction of less than one line by interpolation calculation, and it is configured such that activation or non-operation of the fractional part correction unit can be selected, and a black thin line extending in the main scanning direction included in the image data. A thin black line detecting unit for detecting the black thin line detecting unit, and the black thin line is a black thin line having a one-dot width based on the output signal of the black thin line detecting unit.
In this case, the fractional part correction unit is configured to be inoperative, and the image processing device is characterized by the above. 2. A reduction type color CCD sensor reads information obtained by reducing and projecting an original image through an optical system, and A / D-converts R, G, and B color image signals from the reduction type color CCD sensor. An image processing apparatus for performing correction processing of obtained image data, comprising: R in a sub-scanning direction of a reduction type color CCD sensor;
An inter-line correction unit that corrects the positional deviation between G and B according to the scaling factor of the reduced projection is provided.
An integer part correcting unit corrects the per-line, the correction less than a line and a fractional part correction unit that performs the interpolation operation, operation or non-operation of the fractional part correction unit is configured to be selectable, A black fine line detection unit for detecting a black fine line extending in the main scanning direction included in the image data is provided, and the black fine line is a black fine line having a one-dot width based on an output signal of the black fine line detection unit.
In this case, the fractional part correction unit is configured to be inoperative, and the image processing device is characterized by the above. 3. A region discriminating unit for discriminating a region to be subjected to image processing such as edge enhancement, and a method of creating lightness data in the region discriminating unit is switched according to an output signal of the black fine line detecting unit. The image processing device according to claim 1. 4. A region discriminating unit for discriminating a region to be subjected to image processing such as edge enhancement, and a threshold for discriminating a black edge portion in the region discriminating unit depends on an output signal of the black fine line detecting unit. The image processing apparatus according to claim 1, wherein the image processing apparatus is switched over.