JP3143480B2 - 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 for digitally processing image information, such as a copying machine, a facsimile, a word processor, and a printer.

[0002]

2. Description of the Related Art Conventionally, in this type of apparatus, as a method for improving the recording reproducibility of a black character portion, an operation mode dedicated to a character original is provided.

In this operation mode, the amount of edge enhancement is increased. Increase the level of the black signal. The recording information and the sharpness of the black character portion are corrected by appropriately selecting or combining the above processes to convert the image information. That is, in Japanese Patent Application Laid-Open No. H11-27566, by increasing the amount of edge enhancement, the resolution at the edge portion of the character portion is improved, and the black character portion is intended to be reproduced more clearly. Further, in Japanese Patent Application Laid-Open No. H11-157, the recording density (black component) is reproduced densely, so that the contrast difference between the character portion and the peripheral portion is increased, so that the black character portion is reproduced more clearly.

[0004]

However, in the above-mentioned conventional example, when the image information to be input for switching the operation mode itself and performing processing is an image in which a black character portion and a gradation image portion are mixed. Then, the following adverse effects occur. However, there is a problem that the entire image is reproduced with a rough feeling because the edge enhancement amount in the gradation image portion becomes excessively large. In particular, when the gradation image portion is a halftone image, However, there is a problem that moiré is generated and remarkable image quality is deteriorated. In the above, the recording density (black component) is reproduced densely, so that the black component contained in the gradation image portion is also emphasized and reproduced at high density.
There is a problem that the whole image is reproduced in a cloudy feeling,
In particular, when the gradation image portion is a halftone image, there is a problem that image collapse occurs in a high density portion and remarkable image quality degradation occurs.

There is also known a technique for identifying whether an image is a gradation image or a character image and performing different processing on each of them, but there is still room for improvement in terms of control of spatial frequency characteristics. was there.

[0006] The present invention has been made in view of such a conventional technique, and has as its object to provide an image processing apparatus capable of reproducing a high-quality image.

[0007]

According to the present invention, there is provided an input means for inputting image information corresponding to an image, and detecting a black character portion of the image based on the input image information. First detecting means for outputting a signal indicating the black character portion, detecting a portion of the image adjacent to the black character portion based on the input image information, and generating a signal indicating a portion adjacent to the black character portion. A second detection unit for outputting, and a conversion unit for converting a spatial frequency characteristic to the image information, wherein the conversion unit includes a signal indicating the black character;
The spatial frequency is converted according to a signal indicating a portion adjacent to the black character portion.

[0008]

According to the following embodiments of the present invention, there are provided means for inputting image information, means for detecting a black character portion in the image information input from the means, and outputting a first black character detection signal. An image recording apparatus comprising: means for detecting a portion adjacent to a black character portion and outputting a second black character detection signal; and sharpness conversion means for converting sharpness of the image information. Means for area delay of the input image information, means for calculating the edge amount of the predetermined area,
The input image information is corrected by the edge amount, and the control unit controls the level of the edge amount based on a black character detection signal obtained from the input image information. Even in the case of an image in which a gradation image portion is mixed, it is possible to record a reproduced image in which image quality is not deteriorated and characters are clear.

FIG. 1 shows the configuration of the present invention.
In FIG. 1, reference numeral 1 denotes an in-line type image reading sensor which reads an original and separates the original into R, G, B color signals, and is constituted by a CCD.

2 is an analog amplifier, 3 is an A / D converter for converting an analog signal into a digital signal, and 4 is
A shading correction circuit for correcting characteristics of a reading system, for example, characteristics of a document illumination lamp and a sensor, 5
Is a color shift correction circuit for correcting the influence of the arrangement of R, G, B of the sensor 1, and 6 is for extracting black level characters from the original and adding black character information to the image signal. A black character detection unit 7 for changing the size of the original by inflating and thinning out the image data;
Displays image data on a monitor display or FA
A multi-level data interface (I / F) section 9 for sending to other external devices such as a transmission unit for X transmission, a storage unit for storage, and receiving image data from the external device;
Are control signal generators for separating an image processing (control) signal included in the image signal from the image signal;
Is a logarithmic converter for converting an image signal from R, G, B to C, M, Y. 11 is a minimum value extractor for extracting the smallest level among the C, M, Y signals. , 12 are
A masking processing unit 13 for correcting an image signal in consideration of a color separation characteristic of a reading system and a color development characteristic of ink;
Is an edge emphasizing / smoothing unit that enhances or smoothes a change point of the image signal; 14 is a head shading correction unit that corrects the image signal in accordance with the characteristics of the ink ejection amount of the head; Is a gamma conversion unit for adjusting the image to the desired density of the user, 16 is a binarization unit for converting a multi-valued image signal into a binary image signal, 1
Reference numeral 7 denotes a black character processing unit for converting binary image data based on black character information, and reference numeral 18 denotes a head-to-head adjustment processing unit for delaying the binary image data as necessary according to the width of the print head. , 19 are circuits for converting binary image data into a head drive signal, and 20 is a color recording head using a bubble jet method.

In this recording head, a state change such as film boiling occurs in the ink by thermal energy to generate air bubbles (bubbles), and the inks are discharged from the ejection ports (nozzles) to the recording material using the bubbles. The recording head is a bubble jet type recording head that discharges toward and records characters, images, and the like.
In this recording head, the size of a heating resistor (heater) provided in each nozzle is much smaller than that of a piezoelectric element used in conventional ink jet recording, and high-density multiple nozzles are possible. High quality recorded images can be obtained, and it has characteristics such as high speed and low noise.

In the above configuration, the document information is converted into R, G, B image signals by scanning the document with the sensor 1. The image signal converted into R, G, and B is amplified to a target level by the amplifier 2, and the A / D converter 3
Thus, the analog signal is converted into a digital signal (8-bit multi-value data in the present invention). The image signal converted into a digital signal is subjected to shading correction 4 to perform correction relating to characteristics such as reading and conversion, such as 0FFH for the lightest image and 0O for the darkest image.
H. The image signal that has been subjected to shading correction is subjected to color shift correction 5, and RGB, RGB,...
For example, correction is performed between adjacent signals of the same color system so that a color shift caused by a physical shift due to the arrangement of the colors is eliminated.

The image signal subjected to the color misregistration correction performs black character detection 6, extracts a black level character portion from the image, and adds black character information to the image signal.

FIG. 2A shows an example in which a black character (number) 1 shown in FIG. 2A is read by the sensor 1 (assuming that the sensor 1 is located at two points or a line).

FIG. 2B is an enlarged view of the black characters in FIG. FIG. 2C shows image data when the black character signal of the document is input to the black character detection unit 6. An X symbol is present in the image data, but the X symbol indicates that it is secured as a space for writing data generated in other processes (from the black character detection unit 6). FIG. 2D shows signals KB * and KW * obtained by a black character detection unit 6 described later. KB * is active (Low because the KB * is negative logic) at the pixel (a set of R, G, B) determined to be a black character. KW * is active (KW) in surrounding pixels (a set of R, G, B) determined to be a black character.
* Is Low) because of negative logic. FIG. 2 (E)
Indicates image data at the output of the black character detection unit 6. KB * and KW * are inserted in the position of the X signal of the image data, and thereafter, data of one pixel is set by R, G, B and X.

The image which has been detected as a black character is subjected to scaling 7 and the image data is padded and thinned out for each pixel so that the image has the desired size. At this time, the black character signals KB * and KW * are also scaled similarly to the R, G and B data. Further, the image signal that has been processed from the sensor 1 to the magnification 7 can be sent to another external device using the multi-valued I / F 8.

Further, necessary processing is performed on an image signal that has been processed by the external device from the sensor 1 to the magnification unit 7 and a signal output from the multi-valued I / F 8, and the obtained image signal is multiplied. It is also possible to input from the value I / F8.

An image signal that has undergone scaling or a multi-valued I /
The image signal input from F8 is stored in a D-FF (flip-flop) 30 as shown in FIG. X (PO of control signal)
HTO *, MASK *, KW *, and KB *) signals are separated, and a necessary time delay is output. For example, the PHOTO * signal, which is a signal indicating which of the edge enhancement and the smoothing is to be performed, is the time (the number of pixels) that the image signal is delayed until the edge enhancement and smoothing unit 13 after the X signal is separated from the image signal. ), Just like P
By delaying in advance by the HOTO * delay circuit 30, the image signal and the corresponding PHOTO * at the position of the edge enhancement / smoothing unit 13 are matched. FIG.
(B) shows the content of the X signal in the control signal generator 9. KB * / KW * is as described above. here,
PHOTO * is a signal for determining whether to enhance the edge of the image signal or perform smoothing.
Is a signal for determining whether to print an image or not to print (mask).

The image signal from which the control signal is separated is LOG
The conversion unit 10 converts the signals into C, M, and Y signals. L
The image signal that has been subjected to the OG conversion is a minimum value min among the C, M, and Y pixel signals in the minimum value extracting unit 11.
(C, M, Y) is selected and the selected minimum value min
(C, M, Y) is inserted into the timing position of the X signal as a lower color signal, and C, M, Y, the minimum value min (C, M,
The pixel configuration is changed to Y) and sent to the next processing.

The image signal with the minimum value signal is subjected to a masking process in a masking section 12 so that C,
The image signals are converted into four color image signals of M, Y, and K (black level signals, which are located at the position of X). The signal subjected to the masking process is subjected to a process of emphasizing or smoothing a change point of the image signal in the edge emphasizing / smoothing unit 13.

The signal subjected to the edge emphasis and smoothing processing is corrected in accordance with the recording characteristics of each discharge nozzle of the recording head in head shading correction 4, and the same printing is performed for the same image data even if the discharge characteristics of the head are different. The image data is corrected so as to have the density.

The signal subjected to the head shading correction processing is subjected to γ conversion in the γ conversion section 15 so that the density of the density set by the user in advance and the color balance of C, M, Y and K are obtained. do.

The γ-converted signal is converted by the binarizing section 16 from an 8-bit multi-level image signal to a 1-bit binary image signal.

The binarized image signal is supplied to a black character processing unit 1
7, the KB *, KW * obtained by the black character detection unit 6
The signal undergoes a conversion process based on the signal. For example, KB *
Is active (when it is 0), convert 0, K of C, M, Y signals to 1 and print with black ink only, or
If * is active and one of the C, M, and Y signals is 1, the 0, K of the C, M, and Y signals are converted to 1 and the image is printed using only black ink. Transform the signal. When KW * is active (0), C, M,
By converting the Y and K signals to 0 and whitening the periphery of the black character, the sharpness of the black character can be added.

The image signal that has been subjected to the black character processing is delayed in the head-to-head adjustment processing unit 18 for a time corresponding to the length between the heads 20.

The signal subjected to the head-to-head adjustment processing is converted into a signal necessary for driving the head 20 in the head driving unit 19, and the head 20 is driven to discharge ink and form an image.

Next, the details of the black character detecting section 6 will be described below.

<Black Character Detecting Unit> The black character detecting unit 6 shown in FIG.
As shown in the figure, a black signal generation unit 61 and a binarization unit 6 for the signal
A signal K indicating that it is a part of a character from the binarized data
B and a character determination unit 63 that generates a 1-bit KW both-bit determination signal indicating a pixel adjacent thereto.

FIG. 13A shows the black signal generator 31.
In the figure, a maximum value detection unit 61 and a minimum value detection unit 62 compare the levels of the RGB signals represented by 8 bits for each pixel, and determine the maximum value and the minimum value as max (RGB) min (RGB), respectively. Get as. The adder 63 calculates the difference max (RGB) -min (RG
B) is calculated, and the result is multiplied by a predetermined constant α in a multiplier 66 and added to max (RBG) in an adder 64. When the addition result exceeds the 8-bit width by the limiter (limiter) 65, the addition result is limited to 255 to obtain the black signal d.

FIG. 13B is an explanatory diagram of this processing. In this embodiment, it is assumed that the larger the value of each of the color signals R, G, and B, the whiter, that is, the blacker when R = G = B = 0. Therefore, in FIG. 13B, the portion A is a thin line having a color, and the portion B is a black line.

D = max (RGB) + α [max (RG
(B) -min (RGB)] (max (RGB) is a gray component signal, max (RGB) -min (RGB) is a signal indicating tint, α is a color suppression constant) The physical meaning of this processing is max ( If RGB) -min (RGB) is regarded as a color and max (RGB) is a gray component (brightness), this value is obtained when max (RGB) -min (RGB) is large, that is, when the color has a color. Is multiplied by a constant α and max (R
GB) can be converted to a brighter (white) direction. Therefore, as the value of the constant α increases, α is referred to as a color suppression constant as a value representing the degree to which the tint is suppressed in order to make a point having a tint a whiter point. That is, the degree of black component detection indicated by d can be changed by changing the value of α set in the multiplier 66 by a CPU (not shown). In FIG. 13 (b), α = 1, but max
(RGB) is large and max (RGB) -min (R
When GB) is large, the limiter 65 shown in FIG. 13A reaches a maximum value 255 representing a completely white point. Therefore, it is considered that the change in the generated signal d substantially means a change in the black component.

The black signal d is input to the binarizing section 32 shown in FIG.

<Binarization section> In FIG. 15, the data for one line of the black signal d is held and delayed using the memory 71, and the data of five pixels sequentially line-delayed are added by the adder 72. Further, the added value is set to 1 by F / F73.
The five additional values held delayed in the pixel are further added by the adder 75. The output of the adder is delayed by two pixels from the input data, that is, the output of the memory 71-2 is output to the F / F 73.
If the output position of the F / F 73-6 delayed by -5 and 73-6 is set as the target pixel position, it is the integrated value of the surrounding 25 pixels. Is obtained.

The comparator 79 compares the data d at the pixel position of interest with the average value m as a threshold value to obtain a finer binary signal B.

That is, when d <m, B = 1 (black) When d ≧ m, B = 0 (white).

Further, a difference between the average values m and d is obtained by an adder 77, converted into an absolute value by an absolute value circuit 78, and then compared with a constant δ by a comparator 74 to obtain a binary signal C.

That is, C = 1 when | dm |> δ, and C = 0 when | dm | ≦ δ.

The physical meaning of both signals is that the former B is a signal obtained by binarizing a black signal with high definition, and the latter C is a signal obtained by binarizing the level change amount at a target pixel. That is, when B = 1 and C = 1, the density change at the target pixel is larger than the constant δ,
In addition, it can be determined that it has changed in the black direction. That is, it can be said that there is a high probability that this point is a part of the character thin line.

However, since the pixel of interest may be a halftone image portion represented by a halftone dot, the same two-bit signal is input to the black character discrimination unit 33 to identify the character portion in order to remove the halftone dot. .

<Black Character Recognition Unit> FIG. 16 shows a black character recognition unit. First, the input binary signal B is stored in the line memory 80-
1, 80-2, 80-3, and 80-4 to sequentially hold the delay one line at a time, and to use flip-flops 81-0 to 81-0.
At 81-9, the delay is held for each pixel. Therefore, if the target pixel position is the F / F 81-4 output position, the binary data B of the eight pixels adjacent to the target pixel is F / F 81-
Input positions of 2, 81-4, 81-6 and F / F 81-
Nine adjacent pixel data including the target pixel are input to the gate circuit 83-2, located at the output terminals 2, 81-6, 81-3, 81-5, and 81-7.

In the same manner, the gate circuit 83-1 stores pixel data one line after the target pixel, that is, F / F8.
The data of nine points adjacent to the 1-2 output position, and the data of nine points adjacent to the F / F 81-6 output position one line before the target pixel are input to the gate circuit 83-4. In the gate circuit 83, whether or not the center point is inverted from the eight binary levels (0 or 1) of the eight adjacent points, that is, the center point is isolated from the periphery and is "0", as will be described in detail later. Alternatively, 0 is set for each pixel depending on whether or not the pixel has the level of “1”.
Assign a value S of ~ 4. It can be said that the larger the value S is, the greater the possibility of a halftone pixel is, and conversely, if the value S is 0, it is more likely to be a part of a character or the like. This is because characters and lines are a set of one-dimensionally continuous dots. However, since it is not always possible to judge whether a character is a single point or not, the value S indicating the isolation assigned to each pixel is two-dimensionally integrated and determined. That is, multi-value data indicating the degree of isolation is used for the determination. First, the adder 85-
1 adds three pixels in the line direction, and outputs the result as F / F
The adder 85 holds the delay of 6 pixels at 84-1 to 84-6 and holds it.
If the addition is performed at −2, if a pixel delayed by two lines and 4 pixels from the input image data B is set as the target pixel position, an added value Pf of data S of 3 × 7 pixels centered on the target pixel is obtained. The feature amount Pf means a two-dimensional spatial frequency. In other words, as the Pf value increases, the value of the binary data B is more frequently inverted from “0” to “1” in the vicinity of the target pixel, that is, the spatial frequency is high, and therefore, a point where dots are isolated two-dimensionally. Means a lot.

E = 1 when Pf> K, C = 1 and B = 1, otherwise E = 0

Therefore, the Pf value is compared with a predetermined constant K (approximately 4 to 5) by the comparator 86, and the position of the pixel of interest, that is, two lines in the memories 80-5 and 80-6 and the F / F 82-1
87 output value E = 1 obtained by ANDing binary signal C and binary signal B at the same position further delayed by 4 pixels at .about.82-4.
Are some points of black characters.

One line and F from the signal E and the target pixel position
The signal B delayed by two pixels in / F89-1 and / F89-2 is input to the black character signal generator 88 to generate black character signals KB and KW as final outputs.

Now, the inside of the gate circuit 83 will be described in detail. FIG. 17 is a diagram showing the gate circuit 83, in which a, b, c, d, e, f,
g and h are the input / output signals B of the F / F 81 described above.
EX-OR gates 831-1 and 831-2 are a, i, i
It is detected whether or not the target pixel i is inverted in the direction. That is, if both EX-OR gate outputs are "1", the output of the AND gate 832-1 becomes 1 and the target pixel i is aih
Isolated in the direction. Similarly, an EX-OR gate 831-
3, 831-4, AND gate 832-2 are in the cif direction, EX-OR gates 831-5, 831-6, AND
The gate 832-3 is in the big direction, and the EX-OR gate 83
1-7, 831-8, AND gate 832-4 is die
Detects directional isolation. Here, the AND gate 833-
1, 833-2 and the OR gate 834 indicate that the target pixel i is bi
The case where the same level “0” or “1” is detected continuously in the g direction or the die direction is detected.
4 outputs become “0”. This signal is applied to an AND gate 835-
1 to 835-4, the AND gates 832-1 to 83 described above.
If the output is ANDed to 2-4 and the output result is added by an adder 836, S indicating a value of 0 to 4 is obtained.

The physical meaning of the condition by the AND gate 833 and the OR gate 834 is likely to be a part of a character because its continuity is a line segment orthogonal to the paper surface (document or recording paper surface). Accordingly, in this case, the characteristic amount Pf is uniquely reduced as S = 0. The area to be integrated as the Pf value is desirably a square of about 7 × 7, and a wider area than in the present embodiment can identify characters with higher definition. However, the block shape of the block size is not limited to the above example, and can be appropriately set according to the detection accuracy and the like.

Next, the inside of the black character signal generator 88 is shown in FIG.
This will be described in detail with reference to FIG. The processing part is K, which indicates a part of black characters.
An object is to generate a B = 1 (1 bit) signal and a pixel position signal KW = 1 (1 bit) adjacent to the signal and adjacent to a black character with B = 0. The above-mentioned 1-bit signals E and B are input, and the E signal is line-delayed to F / F880-1 to F / F880-1.
If the target pixel position is set to the F / F 880-3 output position by further delaying one pixel at 880-5, the OR gate 88
1 indicates that one of eight pixels adjacent to the target pixel is E = 1
At the time "1" is obtained at the output. Therefore, the B signal is changed to F / F
If the signal is delayed at 880-7, adjusted to the target pixel position, inverted, and input to the AND gate 882, a KW signal is obtained.

<Another Embodiment (1) of Black Signal Generating Unit> FIG.
FIG. 4A shows another embodiment of black signal generation. In the previous embodiment, the tint suppression constant is a constant value α irrespective of the gray component max (RGB), but in general, the gray component is large, that is, the lighter the gray component is, the more the tint is suppressed. Make it bigger. By doing so, a black signal d having higher contrast can be generated.

[0049]

[Outside 1]

Max (RGB) obtained by the adder 63
The value of min (RGB) is multiplied by the multiplier 661 to the maximum value detection unit 6.
Multiplies one output max (RGB) and divides the result by a divider 66
The value of β divided by the constant β by 2 is preferably, for example, about 128. Further, the result is added to max (RGB) by the adder 64 and clamped to 255 by the limiter 65.

<Another Embodiment (2) of Black Signal Generating Unit> In the previous embodiment, the tint is defined by the difference between max (RGB) and min (RGB), and the suppression is performed by an addition operation. Is ma
The suppression is defined by the ratio of x (RGB) and min (RGB) and the suppression is m
ax (RGB) may be defined. I mean

[0052]

[Outside 2] Here, the constant γ is preferably about 63. According to the above expression, the larger the gray component, that is, the brighter the gray component, the more effective the suppression of the color tint.
Since both ax (RGB) and min (RGB) are alleviated when they are small, a black signal d having a high contrast can be generated without increasing the value of d even when the color is slightly colored from the black component. . In the embodiment of FIG. 14B, first, a constant γ is added to the output of the minimum value detector 62 by the adder 663, and the output of the maximum value detector 61 is divided by the divider 664. Thereafter, the multiplier 665 multiplies the value by max (RGB).

As described above, the algorithm for generating multi-valued data in which the tint is suppressed from each of the R, G, and B data is not limited to the present invention, and is particularly applicable to recording colors limited to the primary colors RGB. It goes without saying that the same effect can be obtained by using another color component signal such as YMC.

Next, one process of a characteristic portion in the embodiment of the present invention will be described in detail with reference to the drawings.

In this embodiment, the edge emphasizing / smoothing block 13 includes a target pixel E and eight pixels (A, B, C, D, and D) adjacent to the target pixel E as shown in FIG.
An edge change amount is calculated from the pixel data of (F, G, H, I), and the sharpness of the input image is converted by correcting the data of the target pixel E with a correction value corresponding to the change amount.

FIG. 4 is a block diagram showing a circuit configuration of the edge enhancement / smoothing block 13.

In the figure, image data output from the masking block 12 is input to an area delay circuit 1310 to delay data for a predetermined period. FIG. 8 is a diagram showing the configuration of the area delay circuit 1310. The input pixel data is pixel-delayed by a pixel synchronization clock (not shown) using D flip-flops 1317 and 1318.
Are output as pixel data corresponding to the positions G, H, and I shown in FIG. On the other hand, the input pixel data is supplied to a line delay unit 13.
12, the output of the delay unit, that is, the pixel data input one line before, is output to the D flip-flop 131.
5, 1316, and is output as pixel data corresponding to the positions D, E, and F shown in FIG.

Similarly, the output of the line delay unit 1311, that is, the pixel data input two lines before, is pixel-delayed by a pixel synchronization clock (not shown) using D flip-flops 1313 and 1314, and the position shown in FIG. A,
Output as pixel data corresponding to B and C. These 9
The data of the pixel position is output at the same timing and output to the edge amount calculation circuit 1320.

The edge amount calculation circuit 1320 calculates the edge amount based on preset parameters (m, n, s, w). FIG. 7 is a diagram showing a configuration of the edge amount calculation circuit 1320. The data of the target pixel E is
At 321, the signal is multiplied by an integer by a predetermined multiplication coefficient m set in a register, and then becomes the input A of the subtractor 1322. On the other hand, the data of the eight pixels adjacent to the pixel of interest is added to only the pixel data at a predetermined position in the addition circuit 1323 based on the addition position selection signal s, and the subtracter 132
2 is input B. The subtractor 1322 calculates and outputs the difference (A−B) between the two inputs A and B. Note that the set value of the multiplication coefficient m is set to be the same as the number of pixels to be added based on the addition position selection signal s. If all the pixels have the same value, the subtractor 1322
Becomes zero. The 1 / n circuit 1324 is a subtractor 1
322 is multiplied by 1 / n with a predetermined division coefficient n set in a register, and then the offset circuit 132
5 is output. When the absolute value of the output data of the 1 / n circuit 1324 is equal to or larger than the predetermined offset data w set in the register, the offset circuit 1325 outputs the output data of the 1 / n circuit 1324 as it is to the edge amount calculation circuit 1
It is output as the output Q of 320. On the other hand, if it is smaller than the offset data w, 0 is set to the edge amount calculation circuit 132.
Output as an output Q of 0. The output is connected to the input A of the multiplier 1301. As an example of selection using the addition position selection signal s, FIG. 6 shows actual operation coefficient values assigned to each pixel position when the number of pixels to be added is eight.
Shown in

On the other hand, a predetermined multiplier value selected by the selector 1300 is input to the input B of the multiplier 1301, and the multiplier 1301 multiplies the output Q by a predetermined value to generate edge amount data. The edge amount data is input to the input B of the adder / subtractor 1302.

To the selector 1300, two different multiplier values (multiplier 1 and multiplier 2) are input.
A multiplier value to be output to 01 is selected based on a black character detection signal.

The adder / subtractor 1302 adds or subtracts the edge amount data output from the multiplier 1301 to the data of the target pixel E input to the input A, and performs a sharpness conversion. Output image data. In addition, in the adder / subtractor 1302, when the conversion switching signal (PHOTO *) is set to “1”, the addition processing is performed.
When the value is set to “0”, the subtraction process is performed.

Here, in order to clarify the basic operation of the edge emphasizing / smoothing block 13, the state of the signals of the respective parts in FIG. 9 will be described.

FIG. 9A shows the image data input to the edge emphasis / smoothing block in chronological order. FIG. 9A shows a black thin line portion, and FIG. 9B shows a white thin line portion. Is shown.

FIG. 9 (B) shows how the data of the edge amount changes in a time-series manner.

FIG. 9C shows a conversion switching signal (PHOTO).
This shows the output of the edge emphasis / smoothing block when *) is set to "1", and shows how sharpness is emphasized (edge emphasis) at the thin line edge portion.

FIG. 9D shows a conversion switching signal (PHOTO).
This shows the output of the edge emphasizing / smoothing block when *) is set to "0", and shows how sharpness is reduced (smoothed) at the thin line edge portion.

Next, referring to FIG. 10, a black character detection signal (K
B *, KW *) to control an edge emphasis / smoothing block.

FIG. 10A shows the image data input to the edge emphasizing / smoothing block in chronological order, and FIG. 10C shows a black thin line portion determined as a black character by the black character detection block. , (D) show a thin black line portion that is not determined as a black character by the black character detection block.

FIG. 10B shows a black character detection signal (BK
*, KW *) in chronological order. (BK *, KW * are negative logic) FIG. 10C shows the selector 130 using only KB * of the two black character detection signals.
This is a time-series diagram showing how the edge amount data changes when 0 is controlled. A multiplier 1 is selected when KB * is High, and a multiplier 2 is selected when KB * is Low. (Multiplier 1 <Multiplier 2). Therefore, the change amount of the edge amount data increases only in the plus (density) direction at the position corresponding to the black thin line portion (c).

FIG. 10D shows a conversion switching signal (PHOT).
This shows the output of the edge emphasis / smoothing block when O *) is set to "1", and particularly shows how sharpness (edge emphasis) is emphasized in the thin black line portion (c).

FIG. 10E shows two black character detection signals K
The time series shows how the edge amount data changes when the selector 1300 is controlled by both B * and KW *. When both KB * and KW * are “H”, the multiplier is 1 , KB *, and KW *, the multiplier 2 is selected in the case of “L” (multiplier 1 <multiplier 2). Therefore, the amount of change in the edge amount data increases in both the plus and minus directions only at the position corresponding to the black thin line portion (c).

FIG. 10F shows a conversion switching signal (PHOT).
This shows the output of the edge emphasis / smoothing block when O *) is set to "1", and particularly shows how sharpness (edge emphasis) is emphasized in the thin black line portion (c).

As is clear from the above description, by configuring the circuit of the edge emphasis / smoothing block 13 as shown in FIG. 4, the black character detection signals (KB *, KW
*) By controlling edge amount data according to *), black characters can be reproduced clearly while reducing the adverse effects (moire generation, increase in rough feeling, etc.) due to the change in sharpness. is there.

FIG. 11 is a block diagram showing another circuit configuration of the edge emphasis / smoothing block 13.

The configuration example shown in FIG. 4 differs from the configuration example shown in FIG. 4 in that the configuration example shown in FIG. By switching the multiplier value for multiplying the output value of the circuit 1320 by the black character detection signal (KB *, KW *),
While the size of the edge amount data is controlled, in the configuration example shown in FIG. 11, edge amount calculation circuits 1330 and 1340 having two different outputs set in advance.
Is controlled by the black character detection signals (KB *, KW *) to control the size of the edge amount data.

Here, the edge amount calculation circuit 13 in FIG.
Numerals 30 and 1340 are equivalent to the circuit configuration shown in FIG. 7, and obtain different output values with respect to the same input due to differences in preset parameters (m, n, s, w). is there. As a result, the black character portion can be clearly reproduced while reducing the adverse effects as in the above configuration example.

The settable ranges of the parameters (m, n, s, w) in the edge amount calculation circuit used in the description of each of the above embodiments are shown below. (When the calculation area of the edge change amount is 9 pixels) Multiplication coefficient m: 0 to 8 Position selection signal s: 8 bits (each bit corresponds to a pixel position) Division coefficient n: 0 to 255 Offset data w: 0 to 255

In the configuration examples shown in FIGS. 4 and 11,
The case where the calculation area of the edge change amount is 9 pixels centered on the target pixel shown in FIG. 5 has been described, but the calculation area of the edge change amount is not limited to this. May be constituted by an area composed of an arbitrary number of pixels including.

Next, in the embodiment of the present invention, in particular, details of the γ-conversion unit 15 which is one of the main elements are shown in FIG. 19, and the operation thereof will be described below.

In FIG. 19, 150 is a fixed value for inputting data corresponding to an image signal.
M, Y, and K are set independently.

Reference numeral 151 denotes a selector for selecting an 8-bit signal, and 152 denotes a ROM in which a table for γ conversion is written.

A multiplier 153 multiplies the input signals A and B and outputs the result. When the output exceeds 8 bits (at the time of overflow), a multiplier having a function of fixing the output to 0FFH is provided. A multiplier circuit that subtracts B from signal A and outputs the result. When the output is underflow (when B is larger than A), the multiplier has a function of fixing the output to 0H. In 155, data (fixed value of data “1”) that makes the multiplier 152 pass through is set.

156 is a coefficient of the multiplier 154 (that is, 1H
０0FFH, and optimal values are set independently for C, M, Y, and K. ) Are set by a CPU (not shown), 157 is a register in which data (fixed value of data “0”) that makes the subtractor 154 through is set, and 158 is a coefficient of the subtractor 154 (that is, a fixed value of “0”). 0H-0
This is FFH, and optimal values are set independently for C, M, Y, and K. ) Is set in the register 159,
The selector for selecting an 8-bit signal, 160
A selector 170 for selecting a signal t is a ROM in which a table for γ conversion is written.

In the above configuration, first, the selector 151
Either the image signal or the predetermined fixed value 150 is selected. The selection of the image signal or the fixed value is performed by the KB * signal. For example,
When the image is a black character, KB * = 0 and the fixed value 150 is selected, and the image signal is set to a fixed value (for example, C = 0H, M =
0H, Y = 0H, K = 0FFH, etc.). When the image is not a black character, KB * = 1 and an image signal is selected. Further, a function of always setting the KB * signal to KB * = 1 manually from the operation unit may be provided, and the selector 151 may be fixed so as to select only the image signal.

The image signal selected by the selector 151 is applied to addresses A7 to A0 of the ROM 152. At the same time, addresses A11 to A8 of the ROM 152 indicate signals corresponding to the density or color balance selected by the user, and addresses A13 to A12 indicate which color of the C, M, Y and K signals is being processed. A color switching signal is added, and γ conversion is performed. A table for converting an image signal based on an image signal, a density selection signal, and a color switching signal is previously stored in the ROM 152 for γ conversion, and data output terminals D7 to D7 of the ROM 152 are stored in advance.
Output from D0.

The gamma-converted image signal is supplied to the multiplication block 1
53 to the input terminal A. The multiplication block 153
To the input terminal B, the coefficient selected by the selector 159 is added, and the image signal is multiplied by the coefficient and output.
The multiplication factor is selected by KB *. For example, when the image is a black character, KB * = 0 and the magnification block 156
In the data previously stored (2 to 2
55) operates to increase the image signal data (make the image darker). If the image is not black text, KB *
= 1, and the selector 159 selects the data (1) previously stored in the magnification block 155, and acts so as to directly output the data of the image signal. Although not described in the present invention, the selector 15 has a function of always setting the KB * signal to KB * = 1.
9 may be fixed to the magnification block 155 so that the image signal data is output as it is.

The image signal output from the multiplication block 153 is applied to the input terminal A of the subtractor 154. Further, the offset value selected by the selector 160 is added to the input terminal B of the subtractor 154, and the offset value is subtracted from the image signal and output. The offset value is selected by KW *. For example, when the image is a peripheral pixel of a black character, KE * = 0, and the data (1 to 255) stored in advance in the offset block 158 is selected to reduce the image signal data (lighten the image). Work like that. When the image is other than pixels surrounding black characters, K
W * = 1, and the selector 160 selects the data (value 0) previously stored in the offset block 1157, and functions so as to output the image signal data as it is. Although not described in the present invention, a function of always setting a KW * signal to KW * = 1 is provided,
The selector 160 can be fixed to the offset block 157.

That is, in the γ conversion processing, KB * =
When 0 (pixels of black characters), the image signal is processed in the dark direction, and when KW * = 0 (pixels around the black characters), the image signal is processed in the light direction, thereby improving the print quality of the image. It is a measure.

Another embodiment of the present invention will be described below with reference to FIG.

First, the selector 151 determines whether an image signal
Alternatively, one of the predetermined fixed values 150 is selected. The selection of the image signal or the fixed value is performed by the KB * signal. For example, if the image is black, KB *
= 0, a fixed value 150 is selected, and the image signal is set to a fixed value (for example, C = 0H, M = 0H, Y = 0H, K = 0FF).
H). If the image is not black text, KB *
= 1 and an image signal is selected. Further, similarly to the above, a function of always setting the KB * signal to KB * = 1 can be provided, and the selector 151 can be fixed to only the image signal selection.

The image signal selected by the selector 151 is applied to addresses A7 to A0 of the ROM 170. At the same time, signals corresponding to the density or color balance selected by the user are stored in the addresses A11 to A8 of the ROM 170.
Addresses A13 to A12 indicate color switching signals indicating which color of the C, M, Y, and K signals are being processed. Address A14 indicates a KB * signal, and address A15 indicates a color switching signal.
The KW * signal is added, and γ conversion is performed. R for γ conversion
A table for converting the image signal based on the image signal, the density selection signal, and the color switching signal is stored in advance in the OM 170, and is output from the data output terminals D7 to D0 of the ROM 170. Below, KB *, K
The combination of W * indicates what kind of data is stored.

When KB * = 1 and KW * = 1, data is stored in the γ-conversion ROM 170 so that the above-described conversion similar to that of the γ-conversion ROM 152 is performed.

When KB * = 0 and KW * = 1, data is stored inside the γ-conversion ROM 170 so that data obtained by multiplying the data of the γ-conversion ROM 152 by a certain coefficient is output.

When KB * = 1 and KW * = 0, data is stored in the γ-conversion ROM 170 so that data obtained by subtracting a certain offset value from the data of the γ-conversion ROM 152 is output. .

When KB * = 0 and KW * = 0, the γ-conversion ROM 170 is stored in the γ-conversion ROM 170 so that data obtained by subtracting a certain offset value from data obtained by multiplying the data of the γ-conversion ROM 152 by a certain coefficient is output. Data is stored.

As described above, the common black character signal K
Both spatial frequency characteristic conversion such as edge emphasis and smoothing and γ conversion can be controlled based on B and KW, and efficient and high-quality image processing can be realized.

Further, since not only the pixel to be processed but also the detection signals of the peripheral pixels are used, it is possible to reproduce a higher quality image.

The form of the output data is not limited to binary, but may be multi-valued data such as ternary or quaternary data.

The method for detecting a black character is not limited to the method described above.

A matrix for converting the spatial frequency characteristics can be set arbitrarily.

As described above, since the level of the edge amount generated in the black character portion can be selectively switched based on the black character detection signal, the image information to be input can be switched between the black character portion and the gradation image portion. There is no problem that the whole image is reproduced with a rough feeling or a cloudy feeling even when the image is a mixture of Even in the case of, there is no problem that moiré is generated or the image is crushed in a high-density portion to cause a remarkable deterioration in image quality. In addition, there is an effect that a reproduced image with clear characters can be recorded.

[0103]

As described above, according to the present invention, an input means for inputting image information corresponding to an image, and a black character portion of the image is detected based on the input image information.
First detecting means for outputting a signal indicating the black character portion, detecting a portion of the image adjacent to the black character portion based on the input image information, and outputting a signal indicating a portion adjacent to the black character portion And a conversion unit for converting a spatial frequency characteristic to the image information, the conversion unit according to a signal indicating the black character and a signal indicating a portion adjacent to the black character portion. Since the spatial frequency is converted, it is possible to take into account not only the detection result of the black character portion to be processed but also the detection result of the portion adjacent to the black character portion, and a problem associated with converting the spatial frequency, For example, black characters can be reproduced clearly while reducing problems such as occurrence of moiré and increase in roughness.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating a configuration of an image recording apparatus embodying the present invention.

FIG. 2 is a diagram illustrating a relationship between a black character portion and a black character detection signal.

FIG. 3 is a diagram showing a configuration of a control signal.

FIG. 4 is a diagram showing a circuit configuration of an edge enhancement / smoothing block.

FIG. 5 is a diagram showing an example of an edge change amount calculation area.

FIG. 6 is a diagram illustrating an example of an operation coefficient value assigned to each pixel position in an edge change amount operation area.

FIG. 7 is a diagram showing a configuration of an edge amount calculation circuit.

FIG. 8 is a diagram showing a configuration of an area delay circuit.

FIG. 9 is a diagram illustrating a state of a signal of each unit of an edge enhancement / smoothing block.

FIG. 10 is a diagram illustrating a state of a signal of each part of an edge emphasis / smoothing block when controlled by a black character detection signal (KB *, KW *).

FIG. 11 is a diagram illustrating another circuit configuration of the edge enhancement / smoothing block.

FIG. 12 is a diagram showing a black character detection unit.

FIG. 13 is a diagram showing black signal generation.

FIG. 14 is a diagram showing another circuit configuration for generating a black signal.

FIG. 15 is a diagram showing a binarization unit.

FIG. 16 is a diagram showing a black character identification unit.

FIG. 17 is a diagram showing a circuit for halftone dot determination.

FIG. 18 is a diagram showing a black character generation unit.

FIG. 19 is a diagram showing a circuit configuration of a γ conversion unit.

FIG. 20 is a diagram showing another circuit configuration of the γ conversion unit.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 In-line type image reading sensor 2 Analog amplifier 3 A / D converter 4 Shading correction block 5 Color shift correction block 6 Black character detection block 7 Zooming block 8 Multi-level interface (I / F) block 9 Control signal generation block 10 Logarithmic conversion Block 11 Minimum value extraction block 12 Masking block 13 Edge emphasis / smoothing block 14 Head shading correction block 15 γ conversion block 16 Binarization block 17 Black character processing block 18 Head adjustment processing block 19 Head conversion signal conversion block 20 Bubble jet Color print head 30 using a system 30 Delay device 150 Blocks 151, 159, 160 storing data for replacing image signals when black characters are used Block block 157 and 158 offset data of ROM 153 multipliers 154 subtracter 155 and 156 magnification data for conversion 2 gamma is stored is stored

──────────────────────────────────────────────────続 き Continuation of the front page (56) References JP-A-2-186876 (JP, A) JP-A-2-253380 (JP, A) JP-A-2-84879 (JP, A) JP-A-63-1988 205238 (JP, A) JP-A-63-180273 (JP, A) JP-A-3-72778 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) H04N 1 / 40-1 / 409 H04N 1/46 H04N 1/60

Claims (3)

(57) [Claims] An input unit for inputting image information corresponding to an image; a first detection unit for detecting a black character portion of the image based on the input image information and outputting a signal indicating the black character portion A second detection unit that detects a portion adjacent to a black character portion of the image based on the input image information, and outputs a signal indicating the portion adjacent to the black character portion; An image processing apparatus comprising: a conversion unit configured to convert a spatial frequency characteristic, wherein the conversion unit converts a spatial frequency according to a signal indicating the black character and a signal indicating a portion adjacent to the black character portion. . 2. The image processing apparatus according to claim 1, further comprising image forming means for forming a visible image of an image corresponding to the image information whose spatial frequency characteristics have been converted by said converting means. 3. The image processing apparatus according to claim 2, wherein the image forming unit includes a head of a type that causes droplets to be ejected by causing film boiling due to thermal energy.