JPH0834541B2 - Image processing device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION <Industrial Application Field> The present invention relates to an image processing apparatus.

<Prior Art> Conventionally, there has been known a technology for discriminating a character region in a document image from a halftone region such as a photograph and performing a binarization process suitable for each region.

<Problems to be Solved by the Invention> However, when data such as characters and numbers given as code data separately from the above-mentioned document image is inserted into the document image, the code is placed on the photographic area in the image. When the characters and numbers converted from the data are superimposed, the above-mentioned discrimination cannot be performed well, and the problem that the photographic area cannot be reproduced well occurs.

That is, when it is determined that the area where the photo area in the document overlaps with the character is the character area and the halftone processing for the character is performed, the area up to the area where the original halftone processing for the photo should be performed is the character area. There is a problem that the image quality is deteriorated because the processing is performed. For example, in the photograph of a person, a portion such as eyelashes around the pupil is mistakenly determined to be a character, and when halftone processing for the character is performed, an image having a remarkable discomfort is reproduced only in such a portion of the photograph. .

On the other hand, as described above, when a character or a number converted from code data is superimposed on the photographic area of the original in addition to the image, the character or the number may be erroneously discriminated and cannot be reproduced well. Also happened.

Such a problem has been a big problem depending on the type of the algorithm for distinguishing the character area from the halftone area such as a photograph.

Furthermore, when characters, numbers, etc. converted from code data are to be superimposed on a color image, there is a large amount of color image data, so an efficient configuration for superimposition is desired.

In view of the above point, the present invention makes it possible to excellently reproduce an image such as characters and numbers generated separately from a color original and a character area in the color original and a halftone area such as a photograph, and to reproduce the color original and the character. It is an object of the present invention to provide an image processing device in which the configuration of the device when synthesizing images such as numbers is simplified.

<Means and Actions for Solving the Problem> In order to solve the above problems, the image processing apparatus of the present invention is an input unit for inputting a plurality of color component signals (corresponding to the solid-state image sensor 14, 16, 18 in the embodiment). ), Processing means for processing a plurality of color component signals input by the input means, and outputting reproduced color component signals of a plurality of colors in a frame-sequential manner (also UCR processing unit 119); Discriminating means for discriminating a halftone region from a plurality of color component signals (also a halftone discriminating circuit 127), and the reproduced color component signals outputted by the processing device are discriminated to be in the halftone region. Halftone processing for halftone area reproduction is performed for the areas that have been marked, and halftone processing for non-halftone area reproduction is performed and output for areas that are determined not to be the halftone area. Halftone processing means (also dither processing unit
124), generating means (also CG in FIG. 6) for generating an image signal corresponding to a given character code signal, and the input means for generating an image signal from the generating means. And a synthesizing unit (also 210 in FIG. 3) for synthesizing the predetermined character with a predetermined color on an image represented by a plurality of input color component signals. Of the reproduced color component signals of a plurality of colors output to, the reproduced color component signal corresponding to the predetermined color is combined with the image signal generated by the generating means.

This embodiment is shown in FIG. Original 1 is a transparent plate 2 on the original table
And is fixed by the original mat 3. The photosensitive drum 24 and the transfer drum 53 rotate in the arrow direction to execute the color process. Reference numeral 12 is a spectroscopic dichroic mirror, and reference numerals 14, 16 and 18 are CCDs that sense the spectrum and generate color signals B, G, and R, respectively. The lamp 8 and the mirrors 9 and 10 reciprocate to scan the original 1 and simultaneously output color signals B, G, and R from each CCD to generate a Y signal for reproduction, and then reciprocate again to output an M signal. The above scanning is repeated four times to form Y, M, C, and BK signals, and the laser is controlled to sequentially form each color latent image on the drum 24. Each color is sequentially developed, the transfer drum 53 is rotated four times, and repeatedly transferred to the paper on it to obtain a full-color copy.

The optical system emits light from the illumination lamps 5 and 6, and the reflecting mirrors 7 and 8
The original is irradiated with light from the moving reflection mirrors 9 and 10 and passes through the lens 11 and 12
Pass through the dichro filter. Here, the light is separated into light having a blue wavelength, light having a green wavelength, and light having a red wavelength. The decomposed light of the blue wavelength of each decomposed light passes through the blue filter 13 and is received by the solid-state imaging device 14. Similarly, light of a green wavelength passes through the green filter 15 and is received by the solid-state image sensor 16. Light having a red wavelength passes through the red filter 17 and is received by the solid-state image sensor 18. That is, the original 3 is moved by the moving reflecting mirror 9 that moves integrally with the illumination lamps 5 and 6 and the moving reflecting mirror 10 that moves at the half speed of the moving reflecting mirror 9 and moves in the same direction. The solid-state image sensor of each color through the lens 11 and the dichro filter 12 while maintaining
It is imaged on 14, 18 and 16. The output of each of the solid-state image pickup devices 14, 16 and 18 is applied as a light output from the semiconductor laser 21 to the polygon mirror 22 via an image processing unit 27 described later. Since the polygon mirror 22 is rotated by the scanner motor 23, laser light is scanned perpendicularly to the rotation direction of the photosensitive drum 24. Further, there is a photo sensor 64 at a position 11 mm before the laser beam starts scanning the drum, and when the laser beam hits this, B,
Generate a D signal.

The photosensitive drum 24 is negatively charged by the negative charger 25, which is supplied with a negative high voltage current from the high voltage power supply 25. Then, when it reaches the exposure unit 26, the original 1 on the transparent plate 2 of the original table is illuminated by the illumination lamps 5 and 6, and the movable reflecting mirrors 9 and 10, the lens 11, the dichro filter 12 and the blue filter.
The image output from the CCD imaged on the solid-state image pickup device 14, 16, 18 by the green filter 15, green filter 15 and red filter 17 passes through the shading unit 104 for each color by the image processing circuit of FIG. , Γ correction unit 105, masking processing unit 109, UCR processing unit 119,
The laser processing unit 124, the multi-value processing unit 125, and the laser driver unit 126 output the laser light to the laser 21 to form an image on the photosensitive drum 24. There, an electrostatic latent image is formed and enters the four-color developing devices 36, 37, 38, 39. Here, the three colors are separated by one exposure scan, and the above respective processes are performed.
The UCR output is sequentially selected for each of B, G, R, and black scan. When the one-color separation light signal in the image processing unit 27 is selected by the signal of the main body control (E signal to UCR), the developing device corresponding to it is selected. The developing device selected there is performed by powder development using a magnetic blade method, and the electrostatic latent image is visualized. After that, the ghost miniature lamp 40 for erasing the electrostatic latent image and the negative post electrode 41 supplied from the negative voltage power source 25.
Is erased by the electrostatic latent image.

Next, the upper and lower cassettes 43 and 44 are selected from the cassette selected from the operation unit 45, and the copy paper 48 sent by rotating the paper feed rollers 46 and 47 passes through the upper and lower 49, 50 of the first registration roller. The transfer roller 53 passes through the second registration roller 52 and is wound around the transfer drum 53. Then, the toner on the photosensitive drum 24 is transferred to the copy paper 48 by the transfer electrode 54. After the transfer is completed, the photosensitive drum 24 is discharged from the copy paper 48 by the discharging electrode 55 supplied with a high voltage from the high voltage generator 25.

In the case of a normal color original, the above operation is repeated four times for four colors and the transfer drum is rotated four times to superimpose each color. If an original with only one black color is detected when the original has only one black color at the time when one optical movement is completed, as described later, the G, R scan, development, and transfer processes are skipped. Then, the copy operation of the black image is started. That is, in the case of a color original, the operating time for four colors is required, but in the case of an original of one black color, the operating time for two colors or one color can be shortened. The copy paper, which has been transferred twice or four times, is peeled off from the gripper 57 and adsorbed on the belt 59 by the transport fan 58 and fixed on the fixing section 60.
It is sent to the outside of the machine after being guided by and fixed.

2 and 3 show image processing circuit diagrams. The light of the original separated into three colors from the dichro filter is CCD14,1
When 6 and 18 are irradiated, the output is amplified by the CCD substrates 101, 102 and 103 and A / D converted to the next shielding unit 104.
It is sent in bit-parallel. When the incident light quantity of the CCD is the same (when it is white), the shading unit 104 corrects so that the output data for each bit of the CCD becomes equal and the variations of the CCDs 14, 16 and 18 for the three colors are eliminated. RAM
This is a configuration, and the data is used as an address and is properly output.

Next, a switch 106, 107, 108 is used to find the optimum γ curve for linearizing the gradation between input and output by the γ correction unit 105.
Select by switching the value of. Incidentally, the reason for changing the processing from 8 bits to the upper 6 bits is to perform the processing in a significant level area.

Next, the masking unit 109 performs arithmetic processing on each B, G, R signal at the same time to change the mixing ratio of each color component to perform color correction. This calculation is performed by the coefficient multiplication ROM and the addition / subtraction ROM. The mixing ratio of each color is performed by switching the values (coefficients) of the switches 110 to 118. Incidentally, the reason that the calculated value is set to the upper 4 bits is that it is limited to a significant area level.

Next, in the UCR processing unit 119, each comparator COMP
And the minimum value signal of B, G, R is determined by the logic of each gate MiN. A value obtained by multiplying the minimum signal by an arbitrary coefficient according to the value of the switch 120 is used as a black level signal. That value each
Subtract from each color by UCR. The signal is branched into two systems, and one of them is sent to the dither processing unit 124 by a select signal 121, 122, 123 from the main body control unit. In the dither processing unit 124, each color signal is a deep signal and, for example, a 6-bit signal is used as a table-referenced dither.
A ROM is used for comparison so that a 1-bit digital signal of 1 or 0 is converted to facilitate laser modulation. Next, this signal passes through the RW line memory and is multi-valued by the multi-value processing unit 125, and the laser 21 is driven by the laser driver unit 126. The dither processing unit has ROM ₁ in which low threshold levels are arranged, ROM ₃ in high arrangement, and ROM ₂ in the middle.The input signal is simultaneously compared with these ROM outputs, and the output from each comparator is compared. Put in memory, latch, and divide 1 pixel into 3 equal parts. In other words, 1 pixel is divided by pulses with different widths of φ ₁ to φ ₃ and output corresponding to ROMs ₁ to ₃ , respectively, and the four-valued output is used to perform pulse width modulation in four ways for 1 pixel Can be represented.

It should be noted that the processing X, Y and Z shown in FIGS. 2 and 3 are performed in real time (real time) substantially at the same time.

On the other hand, the other branched signal is sent to the halftone judging circuit 127. Memories 128-1, 128-2, 128-3 are from θθ address to θF
Θ is stored in the address, 1 is stored in the addresses 1θ to 2E, and θ is stored in the addresses 2F to 3F. The 6BiT signal input to the memory B128-1 varies from θθ when there is a lot of light (low density of the original) to 3F when there is little light (the high density of the original) due to the intensity of the light input to the CCD. Change.

As an example, when the signal of θθ to θF is low density, the signal of θ2E is intermediate density, and when the signal of 2F to 3F is high density signal. When the intermediate density signal is input to the memory B, 1 is output,
Otherwise, θ is output. The latch circuit 129 inputs this to the hold circuit 130 in synchronism with the clock.

This operation will be described with reference to the flow chart of FIG. From FIG. 1 69, the RESET signal 200 is input just before the optical scan 201 and the outputs Q _{1 to} Q ₃ are reset. If there is an intermediate density signal even once in the first optical scanning, the hold circuit outputs 1 and, as a result, the output 140 of the OR gate becomes H.
The control circuit 69 (Fig. 1) sends this signal to the optical scanning completion 20
Check immediately after 2, and if 203 is an H signal, control circuit 69
Switches the selectors 141 to 143 to the dither side by the switching signal 144 (step 204), and if the signal is an L signal, switches the selectors 141 to 143 to the fixed data side by the switching signal 144 (step 205) to perform the dither processing. Omit.

It is also possible to perform dither selection control in advance by performing a forward scan (not forming a reproduced image) on the image at a high speed, instead of the main scan that is directly involved in the reproduced image formation as the first optical scan.

Further, the range for determining the intermediate density can be freely determined by switching the memories in which the memory patterns of 1 and θ are changed.

A circuit as shown in Fig. 5 is added to or replaced with x, y, z in Fig. 2, and the BD signal from Fig. 1-64 (signal by the detection of the end of the beam scanning of one line) is input to the coun 145. If an appropriate count value (for example, 4 × 4 dither matrix) is applied to the RESET signal at 4 and the signal 140 is input to the switching signal 144 of FIG. Sequential dither processing can be performed. Therefore, the dither can be selectively controlled for each area.

With such a simple circuit, it is possible to obtain a high-resolution print by subjecting an original having an intermediate density to dither processing and printing a high gradation without performing a dither processing.

Note that the present invention is effective even if the input signals B, G, and R in FIG. 2 are from the host computer, and the host and the CCD reader are connected at the connection points of X, Y, and Z as necessary. Can be switched. In this case, when the command signal without halftone is added to the head of the transmission signal from the host, it is possible to judge this and perform selector control so as to omit the dither processing. In addition, 1 pixel 4 dot type printer,
The present invention can be applied to thermal printers and ink jet printers.

In this example, when the dither processing is omitted, the pulse width modulation of the laser drive signal is performed according to the above-mentioned four values, so that a slight halftone can be reproduced and a low-level halftone (there may be a dither-omit at the end). Can be reproduced. In addition, the determination can be performed every several pixels, and the above-mentioned selection control can be partially performed by making the synchronization accurate.

FIG. 6 is a circuit diagram for inserting characters, numbers and the like into the above color image.

200 is a code generating means for generating characters and numerical values as code data, M ₁ is a buffer memory for storing the code data at the same time as the code is generated, ADC ₁ is an address counter for controlling the address for writing and reading the memory, CG Is a well-known character generator that outputs characters and numerical values as dot pattern image data according to the code data read from the memory M ₁ , and M ₂ is a buffer memory that stores the dot data from the CG at the same time as the output, and one pixel of the image data. It is stored so that 1 dot from CG is corresponded to,
That is, a dot pattern for a plurality of characters and numbers is stored with each character and each numerical value having a predetermined interval similar to that of the reproduced image. AD
C ₂ is an address counter that controls the address for writing and reading the memory M ₂ , and this determines the timing of the reading start in synchronization with the processing progress of the color image data. The character composition position of can be determined. 201 is a signal input source for presetting the timing. When the image processing shown in FIG. 2 reaches the preset coordinates X, Y by 201, the reading from the memory M ₂ is started and the color reproduction corresponding to the above position after dither processing is performed. The character output is synthesized from the output.

205 is a CG when the output of the comparator 206 is L (white, halftone)
Is a gate that outputs the dot output of H (black), and 204 is an inverter that outputs L (white) when the output of the comparator is H (black). The comparator 206 outputs H when the black component output A in FIG. 2 is above a certain level L ₁ and outputs L when it is below the level L ₁ . Accordingly, when the image is a dark background color, the signal of the above level is output from B in order to clear the character image from the CG from the background color and to make the character image black when the tone is light. The signal from B is input to 210 in FIG. 3 and is combined with the image data after dither processing. In this case, since the inserted character is not dithered, the resolution is not impaired. Note that 21
0 has a function of replacing the image data after dithering with the signal from B for the pixel to which the dot data from CG should be output.

The R / W signal is a write / read signal, and the read signal of the memory M ₂ is a black scan signal synchronized with the black processing step.
It is applied to the black process and it is output to form black as characters. If you want a color image is to insert the character-intensity when the single-color images of blue, as to output as well as further executes a-intensity step B signals only the process in the memory M ₂ in synchronization with the read signal to-intensity processing Given to.

The address counter ADC ₂ counts the dot (CLK) of the image data processing and the line to determine the read start timing of the memory M ₁ , and the count value is 20
When the preset X, Y by 1 is reached, the reading of the memory M ₂ is started in synchronization with the clock CLK. The counting of dots starts counting the bit for each one line end signal, and the line counting starts from the beam detection signal in the laser scanner indicating the end of one line scan or the bit count end signal for one line from the start of color data processing. The processing of FIGS. 5 and 6 is also performed in real time.

As shown in FIG. 7, when the code "1984" is stored in the memory M ₁ by the key or transmission line attached to the copying machine of FIG. ₁ , the CG converts it into a dot pattern and stores it in the memory M ₂ . To do. After that, the above-mentioned color data processing becomes possible (in this example, scanning of the original document by the optical system is possible. Until the command input indicating that there is no inserted data, it is prohibited until then.) Color data processing Start and
201 preset coordinates in the Brasksky process
When it reaches X and Y, the reading of the memory M ₂ is started, the B signal of the character is sequentially output in synchronization with the CLK, and the latent image of the black character is formed on the drum and combined as shown in FIG. Is transferred and synthesized into a color image of to obtain a color print with characters. If necessary, other color characters such as red and blue can be inserted as described above.

However, when a character is added to a position where a part of a dark color (black) is used as a background, as described above, the background is automatically determined from the signal A and a white character signal is formed to generate a signal B. Output as. This procedure is for color image processing
Make a circuit configuration with accurate synchronization so that it can be achieved in real time. The preset data 201 can be the key of the copying machine shown in FIG. 1 or the transmitted code data. It may be a memory ROM that stores the format 200.

By the way, when the background color is repeated in a short range such as a stripe pattern, it may be unsightly to treat the white color accordingly, and in order to eliminate the inconvenience, the W in FIG.
It is possible to provide a delay circuit at each point so that whitening can be performed only when the background is dark in a predetermined range.

Incidentally When adding all characters in the white, the memory M ₂ to allow the treatment with the color image processing step at that time
The read timing of can be determined.

<Effects of the Invention> As described above, according to the present invention, it is possible to satisfactorily reproduce both an image such as characters and numbers generated separately from a color original and a character area in a color original, and a halftone area such as a photograph, and color. It is possible to simplify the configuration of the device when synthesizing a document and an image such as characters and numbers.

That is, according to the present invention, the image signal corresponding to the character code signal is combined with the reproduced color component signal which is output in the frame order from the processing means, so that the halftone region is discriminated from the plurality of color components. No means is identified,
The character can be reproduced well, and the configuration of the device can be simplified as compared with the case of combining with a plurality of input color component signals.

[Brief description of drawings]

FIG. 1 is a sectional view of a color copying machine to which the present invention can be applied, and FIGS. 2, 3, 5, and 6 are circuit diagrams of an image processing unit.
FIG. 4 is a flow chart for halftone judgment, and FIG. 7 is an explanatory view of a reproduced image. In the drawing, 27 ... Image processing unit 127 ... Black judgment circuit.

Claims (1)

[Claims] 1. Input means for inputting a plurality of color component signals, processing means for processing a plurality of color component signals input by the input means, and outputting reproduced color component signals of a plurality of colors in a frame sequential manner. A discriminating unit discriminating a halftone region from a plurality of color component signals input by the input unit, and a discriminating unit discriminating the reproduced color component signal output by the processing unit as a halftone region. The halftone processing means performs halftone processing for halftone area reproduction for the area, and performs the halftone processing for non-halftone area reproduction for the area determined not to be the halftone area, and outputs the halftone processing means. Generating means for generating an image signal corresponding to a character code signal representing a predetermined character, and the input device according to the image signal generated by the generating means. And a synthesizing unit for synthesizing the predetermined character with a predetermined color in an image represented by a plurality of color component signals input by the synthesizing unit. Of the reproduced color component signals of 1., the image signal generated from the generating means is combined with the reproduced color component signal corresponding to the predetermined color.