JPH03181277A - Image processing method - Google Patents

Abstract

PURPOSE:To prevent color blurring at black character part by discriminating and extracting a black character edge part in picture information, setting a slice level of binarization based thereon and binarizing multi-value picture information. CONSTITUTION:The machine is provided with an edge extraction means 4-1 discriminating and extracting a black character edge part in picture information, setting means 4-2, 4-3 setting a binarizing slice level based on the result from the black edge extraction means 4-1, binarizing means 4-2, 4-5 binarizing multi- value picture information according to the level set by the setting means 4-2, 4-3 and a copying means 4-6 outputting binarized picture information by the binarizing means 4-2, 4-5 and copying the information. Thus, the black character edge part of the picture information is discriminated and extracted to set the binarizing slice level. Thus, an excellent picture is obtained.

Description

[Detailed description of the invention]

[Industrial Field of Application] The present invention relates to an image copying apparatus that performs image processing on input multivalued image information and copies the binarized image information.

[Prior Art] Conventionally, this type of device reads a document using an image sensor such as a COD, performs various image processing on the A/D-converted multivalued image information, and converts it into binary image data using a laser beam. Images are copied by outputting to a beam printer or the like. Then, by dither method, error diffusion method (ED), etc., 2
It is common to convert it into a value. [Problems to be Solved by the Invention] However, the binarization process in the conventional example described above expresses density by the area density of dots, and for example, pixels at the edge of a character are not lined up and are missing. Also, since it is read using a color image sensor,
Color misalignment (normally, image sensors that read R, G, and B signals are lined up in sequence, and their positions shift) may occur in the reading system. These phenomena are particularly common in the contours of the output image. In addition, in recent years, there has been an increase in the use of color in documents in offices, but as the majority of documents are still in black,
Color copying machines have the disadvantage that black text parts may have color misalignment, making it impossible to faithfully reproduce originals. The present invention was made to solve the above problems, and
An object of the present invention is to provide an image copying apparatus capable of obtaining a good image by determining and extracting a black character edge part of image information and making it possible to set a binarization slice level. Means for Solving the Problems In order to achieve the above object, an image copying apparatus of the present invention has the following configuration. That is, in an image copying apparatus that performs image processing on input multivalued image information and copies the binarized image information, an edge extraction means that determines and extracts black character edge portions in the image information, and a setting means for setting a binarization slice level based on the result from the extraction means; a binarization means for binarizing multivalued image information according to the level set by the setting means;
A copying means is provided for outputting and copying the image information binarized by the digitization means. [Operation] In the above configuration, black character edge portions in input image information are determined and extracted, a slice level for binarization is set based on the result, and multi-valued image information is converted into binary data according to the set level. make a change. Then, it operates to output the binarized image information and copy it. [Embodiments] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. First Embodiment> FIG. 1 is an external perspective view showing an image copying apparatus according to an embodiment. In FIG. Including a line-shaped CCD as an element and various image processing units described below,
Moreover, COD in Examples is R (red) and G (green). 3 consisting of various B (blue) optical filters
It is a line color image sensor. 1-2 is a digitizer which also serves as a pressure plate for pressing the original. 1-3 is the operation unit, which instructs to start reading the original;
Select the number of copies, select various image processing, etc. Reference numeral 1-4 is a printer that outputs binary data, and in the following embodiments, a color laser beam printer is used. Next, the configuration of the document reading section and image processing section 1-1 in the embodiment will be explained below with reference to the block diagram shown in FIG. Reference numeral 2-1 denotes an original, which is placed below the original pressing plate 1-2 in FIG. 2-2 is an optical lens that forms an image of the original on the CCD. 2-3 is a CCD, R (red),
It consists of a 3-line CCD with G (green) and B (blue) optical filters. Here, the number of light-receiving parts for one line of the CCD consists of the number of light-receiving parts for one direction of the document table, for example, the entire X direction;
If the direction is 30cm and 16pe 1 reading, then 30
It is composed of x 160=4800 or more light receiving sections. When scanning a document, the imaging lens 2-
2. C0D2-3 is scanned in the y direction to read the original. The CCD 2-3 converts the document information into analog electrical signals, and outputs the converted R, G, and B various signals to the analog signal processing section 2-4. 2-4 is an analog signal processing section, which performs sample & hold, dark level correction dynamic range control, etc. for each color, performs A/D conversion (analog-digital conversion), and input image processing section 2. Outputs R, G, and B signals to -5. 2-5 is an input image processing unit, which performs shading correction, which is correction of the light-receiving sensitivity of each light-receiving part of the CCD, for each of the input R, G, and B colors.For details, refer to the image processing unit 2-6, which will be described later. Output. 2-6 is an image processing unit, which includes image area separation for separating the black character edge area, which is the center of the embodiment, and a circuit for varying the binarization slice level according to this separation information, R, G.
, a logarithmic conversion unit that converts the color information of B into C (cyan), 1M (magenta), and Y (yellow) corresponding to the color ink of the printer; a masking conversion unit that performs conversion according to the output color characteristics of the printer; It is composed of an under color removal unit (OCR) that separates gray components from M and Y and replaces them with BX (black). The above C, M, Y, BK density information is provided in the printer section 2-7.
and is copied onto paper by an output device. 2-8
is a control unit (CPU) that controls each of the above-mentioned processes, and 2
-9 is an operation unit corresponding to 1-3 shown in FIG. Next, the timing of the image signal read by the above-mentioned CCD 2-3 will be explained below with reference to FIG. In the figure, VSYNC is vertical
It is an abbreviation for "nize" and is a synchronization signal for the VIDEO signal for the original page. Also, PE is page enab
It is an abbreviation for le signal, and it is VIDEOENAB for one side of the original.
This is an LE signal. HSYNC is horizontal
It is an abbreviation for 5ynchronize, and is a synchronization signal for one line of image signal read by C0D2-3. By moving the line-shaped CCD 2-3 vertically (in the sub-scanning direction), image signals for one page are read, so HSYNC is generated for the number of lines in the sub-scanning direction. Become. Also, VE is VIDEOEN
It is an abbreviation for ABLE, and is an ENABLE signal in the horizontal direction, indicating the effective pixel area of the CCD 2-3. Further, VCLK is an abbreviation for VIDEO clock, and is a lock for image information transfer. RVIDEO, GVIDEO
, BVIDEO. The KVIDEO signals are image information of red, green, pull, and brightness, respectively, and the brightness signal includes two tails (R+G
, B). Next, each processing block of the above-mentioned image processing section 2-6 will be explained below with reference to FIG. In the figure, R, G, and B are the image signals explained earlier,
This is multivalued data. 4-1 is a black character edge detection block, which detects the edge of black characters from each image signal, and is configured so that if it is an edge, the J output becomes active 1".The detection result is then detected at the slice level. Variable ED circuit 4-2
is input. This variable ED circuit 4-2 is a binarization circuit using the ED (error diffusion) method, and performs binarization while varying the slice level of the black signal (six inputs) in response to the G input signal. It is structured like this. Further, B and D are outputs of image signals, E and C are inputs, and F is an output of a binarized black image signal. And FIFO memory 4-
3 constitutes the ED diffusion matrix. On the other hand, each signal for R, G, and B is subjected to log conversion, UCR, and masking in block 4-4, and C (cyan), 1M (magenta), Y (yellow), and B , (black) IIBK converted to each signal
The signal is binarized by the variable ED circuit 4-2 described above. Regarding Y, it is binarized at a fixed slice level by the ED circuit 4-5. The ED circuit 4-5 also includes a FIFO memory forming a diffusion matrix. And slice level variable ED circuit 4-2. ED circuit 4-
The C, M, Y, and BK signals binarized by 5 are outputted by a printer 4-6. A detailed explanation will be added below for each block 4-1 and 4-2 shown in FIG. Note that the other blocks are well known and will not be described here. First, regarding the black character edge detection block 4-1, the fifth
This will be explained with reference to the figures. In the figure, R, G, B, and J correspond to each terminal of the black character edge detection block 4-1 shown in FIG. 5-
1 is a MAX minus MIN calculation circuit, and its function is to extract the maximum signal and minimum signal from each R, G, and B signal, and output the result of subtracting the minimum signal from the maximum signal. It is configured. Here, a detailed circuit block of this circuit 5-1 is shown in FIG. In FIG. 6, R, G, B. Each terminal of E corresponds to 5-1 in Fig. 5, and 6-1
~6-3 is a comparison circuit, 6-4 to 6-9 are AND circuits, 6
-10 to 6-15 are buffer circuits with output control, and 6-16 is a subtraction circuit. The R, G, and B signals are input to the buffer circuits 6-10 to 6-12, one of which is selected, and the maximum color information is input to the six inputs of the subtraction circuit 6-16. Similarly, the minimum color information is input to the B input of the subtraction circuit 6-16. This subtraction circuit 6-16 subtracts the minimum from the maximum and sends the result to E.
Output to. Next, the result from the subtraction circuit 6-16 is input to the comparison circuit 5-3 B shown in FIG. 5, and if the difference between the maximum and minimum color information is less than a certain value, the color is achromatic. It is judged as. The register 5-2 is used to set a slice level for determining an achromatic color. Further, 5-4 is an edge component extraction block, which will be explained with reference to FIG. In the figure, K is luminance information, and B is binary information that is "1" if it is an edge component. Further, VCLK is the image transfer lock described above. 7-1.7-2 is a D-type flip-flop (DFF), 7-3 is a doubling multiplication circuit, 7
-4 is a subtraction circuit that calculates BA, and 7-5 is a subtraction circuit that calculates AB. Now, image information crab 1. Kg,
Ks is transferred to DFF7-2, but DFF7
Considering the case where 2 is latched to -1, the subtraction circuit 7-4 calculates 1 to 2·Ko, and the subtraction circuit 7-4 calculates 1 to 2.
5, (2·Ka Kl) K- is calculated. Here, assuming that the image of interest is 2, the result of the above-mentioned calculation is positive when the image of interest is larger than the images on both sides. In other words, if the calculation result is greater than or equal to a certain value, it can be determined that is an edge component. Further, in this case, edge components were extracted from three consecutive pixels, but in this case, edge components only in the horizontal direction are extracted. Generally, edge components in the horizontal and vertical directions are detected from nine pixels of 3×3, but in this embodiment, for simplicity, three pixels are used. Also, in register 7-6,
A slice level for determining whether or not it is an edge component is set and determined by the comparison circuit 7-7, and if it is an edge component, binary information of "1" is output to B. The AND circuit 5-5 in FIG. 5 which inputs this output result and the determination result from the comparison circuit 5-3 shown in FIG. Output. Next, the slice level variable ED circuit 4-2 shown in FIG.
The details are explained in Fig. 8. In Fig. 0, A, B, C, D
, E, F, and G correspond to the terminals shown in FIG. 8-1 to 8-12 are adder circuits, 8-13 to 8-22 are D-flip-flops (D-FF), 8-26 is a comparison circuit, 8-27 is a NOT circuit, 8-25 is a selector, 8-26 is a comparison circuit, 8-27 is a NOT circuit, 8-25 is a selector; −
23.8-24 is a register, and 8-28 is a memory. Three lines of black information are input to each terminal input of A, C, and H, and a diffusion matrix of 5.5.2 pixels is configured by each D-FF 8-13 to 8-22. The amount of diffusion to each pixel is determined by the table memory 8-28. The ratio of errors distributed to each pixel is 1:3 for each column.
5:3:1.3:5 near: 5:3.5 near,
The configuration is such that the error is saved in total. For example, if the width of the image is sign +8 bits, the pixel information of interest is Bi, and a dot is placed on the pixel of interest, the A<B output will be "O" and the information from the memory 3-28 to each pixel will be , -((256-Bi)÷48) = a, then a, 3
a, 5a, 3a, a, 3a, 5a, 7a. It will be spread as 5a, 3a, 5a, and 7a. Also, if no dot is hit, if Bi÷48=b, then b, 3b, 5b, 3b, b,
3b, 5b, 7b, 5b, 3b, 5b. 7b and diffused. It is the B input of the comparator circuit 8-26 that determines whether or not to place a dot. If the image information is larger than the value of the B input, the output of the comparator circuit 8-26 is "O".
is inverted by the NOT circuit 8-27, and '1'' is output to F.Here, as mentioned above, the G input is a signal indicating whether or not it is a black character edge part. Set the slice level at a certain time in register 8-23, and the slice level at no time in register 8-24.By setting the value of register 8-24 larger than register 8-23, the black character If it is an edge part, it is possible to easily place a dot.As explained above, according to the first embodiment, there is a means for extracting the edge part of black characters in a document, and a binary value is set according to this information. By varying the slice level of the conversion circuit (specifically, by lowering the slice level at the edges of black characters to make it easier to binarize), it is possible to improve the In addition to faithfully reproducing the gradation, the black text makes it easier to connect the black dots at the edges, making it possible to reproduce clearly.Therefore, when reproducing conventional color originals with a binary output color printer, , it is possible to solve the drawback that it is difficult to faithfully reproduce both photographic parts. [Other Embodiments] Next, in the above-mentioned embodiments, ED (error diffusion method) is used for the binarization process. Although the example has been described, the case where the dither method is used as the second embodiment according to the present invention will be explained below.The configuration of the second embodiment is as shown in the slice level variable ED circuit 4- 2 and the ED circuit 4-5 are replaced with a dither circuit, and since it is the same as the first embodiment, the explanation here will be omitted. The block will be explained with reference to Fig. 9. In the case of dithering, for example, there is no need to store 3 x 3 = 9 pixels, so there is no need for something equivalent to the FIFO memory 4-3. In the figure, G (image area separation information) , A (VIDEO black)
. F (Binary data) is an input/output terminal, and the alphabetic characters correspond to those in FIG. Reference numeral 9-2 is a table memory, which stores dither slice levels. Reference numeral 9-1 is an address generation section of the memory 9-2, and reference numeral 9-3 is a comparison circuit for binarizing multivalued image information in accordance with the dither slice level. Address generation unit 9-1 counts addresses AO to A3, image area separation information G is input to address A4, and memory 9-2
The bank will be switched. Referring to FIG. 10, the address timing output by the address generator 9-1 will be explained. PE stands for page enable, HSYNC stands for ho
This is an abbreviation for rizontalSyrlcronaize, and these correspond to FIG. 3 in the first embodiment. In the figure, address is an address generated by the address generator 9-1, and for each HSYNC (one line), the address is ro, 1°2.0, 1. ...”-r3,4,5,
3. ..." → "6.7, 8, 6. ..." → ro, 1
, 2, 0, 1°...'' are repeated. The relationship between this address and data will be explained with reference to FIG. 11-1 and 11-3 respectively indicate the address of the buzzer matrix corresponding to the normal document part and the character edge part. In the case of the character edge part, image area separation information is input to address A4, and It has the form of adding 10 to the address of the section. Next, data values for each address will be explained. 11-2 and 11-4 are the normal manuscript section and
The data (slice level) corresponds to the dithering of the character edge portion, and in this example, the image information width is 8bi.
The dither size of 3×3 in t and the slice level allocation pieces are not limited to this example. 11
-2゜As you can see from 11-4, the slice level of the character edge part is lowered by 10 (H) compared to the normal document part (10 → 01 (H) is an exception), and the character edge part is black. It is configured so that it is easy to dot the dots. Also, the part corresponding to the ED circuit 4-5 is a part that binarizes the color outside the black clothing, and is the same as in the case where image area separation information is not input in FIG. 9, so a detailed explanation will be omitted. . <Third Embodiment> In the first embodiment, when it is determined that it is the edge of a black character, the slice level of the binarization is lowered to make it easier to hit the black dots and to line up the black dots. However, as shown in Figure 12, the edges of characters are emphasized because they are easier to dot, but when the error diffusion method (ED method) is used for binarization, the error is always preserved. This has the negative effect of making white spots more likely to occur on the inside of the edge as a reaction to dotting the edge. In order to solve this problem, FIG. 13 shows an embodiment in which the binarization slice level is lowered at the black character edge portion and the error is not preserved. FIG. 13 is basically the same as FIG. 8, except that the black character area determination signal G is input to the address of the table memory 13-28. In this embodiment, all error distribution is realized by the table memory 13-28, so it is realized by changing the banks of the table memory 13-28 according to the black character determination signal. The other circuits in FIG. 13 are all the same as in FIG.
The adder circuit 3-1 is similar. Regarding the specific distribution, if a dot is placed in a black character area, the negative error is made smaller than the actual value, and if no dot is placed, the positive error is made larger than the actual value. As a result, the density is not preserved in the edge portion of the black text, and the density becomes darker than it actually is, resulting in a result similar to that obtained by emphasizing only the black edge of the black text. Fourth Embodiment In addition, color misalignment may occur in the copy result due to color misalignment in the reading system due to the configuration of the color-compatible CCD, or misalignment of the overlapping of each color toner or each ink in the output system. Something happened. Therefore, the edges of black characters in the copy result may not be reproduced clearly, and in areas determined to be edges of black characters, the binarization slice level of the black dots is lowered, and the binary slice level corresponding to the color outside the black A configuration in which the slice level of digitization is increased will be explained. The image processing block of the fourth embodiment will be explained with reference to FIG. The difference from the image processing block shown in Fig. 4 is that in Fig. 4, only B (black) has a variable slice level ED, but in Fig. 14, only C (cyan), 1M (magenta)
, Y (yellow). The difference is that variable ED is applied to both B and (black) colors. 14-1 shown in FIG. 14 is a black character edge detection block, and 14-2 is a log conversion and UCR. In each masking process, 14-3 is the aforementioned slice level variable ED circuit, and 14-4 is a FIFO memory for configuring an ED diffusion matrix, which corresponds to four colors. In this embodiment, the black character edge detection block 14-1
C, M depending on the detection result (E output). The binarization slice level of Y and BX is variable,
The configuration of the variable slice level ED 14-3 includes 24 blower circuits shown in FIG. The setting value of each slice level register 8-23.8-24 is R1llKili for black.
Rmxxxa, cyan is Rcxx, Rci4, magenta is R&1gs, RM□4. Yellow is Ryzs+RY24
Then, the black character area judgment result human power G is “l” in the black character area.
”, otherwise “0”, R*Kxs< RmK*a+
Rcps >Rc*4+Rvzm>R□a + R
Each register is set as v*s > Rr*a. Furthermore, since there is a possibility that the influence described in FIG. 12 will occur in this case as well, a configuration may be adopted in which the black error shown in the third embodiment is not saved. Furthermore, for the same reason, for colors other than black, the dots at the edges of black characters are controlled so that they are difficult to hit, as shown in Figure 15. ), false contours may occur in the inner part. In order to suppress this effect, the concept of the third embodiment should be used so that errors in colors other than black are not stored in the opposite direction.In other words, the error distribution should be , the negative error may actually be made larger, and the positive error may be made smaller than it actually is if no dot is hit. Further, as for the hardware configuration, the variable slice level ED shown in FIG. 13 may be used for both the four color ED circuits. Although a detailed explanation will be omitted, this is realized by switching the memory area of the allocation memory 13-28 depending on whether the area is a black character area or not.

【Effect of the invention】

As described above, according to the present invention, a good image can be obtained by determining and extracting the black character edge portion of image information and making it possible to set the binarization slice level.

[Brief explanation of drawings]

FIG. 1 is an external perspective view showing the appearance of the image copying apparatus in the embodiment, FIG. 2 is a schematic block diagram showing the configuration of the image copying device in the embodiment, and FIG. 3 is a diagram showing the timing of image information in the embodiment. , Fig. 4 is a diagram showing the image processing block in the first embodiment, Fig. 5 is a diagram showing the configuration of the edge detection block, Fig. 6 is a diagram showing the maximum-minimum calculation section, and Fig. 7 is a diagram showing the edge detection block. FIG. 8(a) is a diagram showing a slice level variable ED circuit block; FIG. 8(b) is a diagram showing an error diffusion matrix; FIG. 9 is a diagram showing the dither method in the second embodiment. 10 is a diagram illustrating the address timing of the dither memory, FIG. 11 is a diagram illustrating the dither matrix, FIG. 12 is a diagram illustrating a copy example in the embodiment, and FIG. 13 (a) 13(b) is a diagram showing a slice level variable ED circuit block in the third embodiment, FIG. 13(b) is a diagram showing an error diffusion matrix, and FIG.
FIG. 4 is a diagram showing an image processing block in the fourth embodiment, and FIG. 15 is a diagram illustrating the influence on color dots other than black. In the figure, 4-1...black character edge detection block, 4-
2...Slice level variable ED circuit, 4-3...F
IFO memory, 4-4...log, UCR, masking, 4-5...ED open circuit 4-6 printer. ] Figure

Claims (1)

[Scope of Claims] An image copying apparatus that performs image processing on input multivalued image information and copies the binarized image information, comprising an edge extraction unit that determines and extracts black character edge portions in the image information. and a setting means for setting a slice level for binarization based on the result from the edge extraction means, and a binarization for binarizing multivalued image information according to the level set by the setting means. An image copying apparatus comprising: means; and a copying means for outputting image information binarized by the binarization means and copying the same.