JPH07184051A - Image processor - Google Patents

Abstract

PURPOSE:To apply the processor to a high speed system with comparatively simple configuration by providing a means for generating two specific patterns and a means for generating a binary signal to the processor, thereby eliminating broken or interrupted character or line and improving the resolution and reproducibility while keeping the gradation reproducibility of an intermediate tone image area. CONSTITUTION:Two triangle wave having a period twice the picture element period with high gradation reproducibility generated by pattern generating circuits 103, 104 are used for pattern signals to generate a binarizing signal with pulse width modulation. An edge direction detection circuit 107 detects the presence and direction of an edge based on the absolute value of a difference between two photoelectric conversion element with an attentional picture element inbetween and on the relation of quantity. If the signal is an edge, a proper triangle wave in the two triangle waves is selected depending on the direction, and the selected triangle wave and analog signal obtained by convering a digital picture signal are compared to select a generated binary signal with a time width. Thus, a picture having high quality is obtained.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention enables an image processing apparatus for obtaining a binary image signal which is pulse width modulated by comparing a multi-valued image signal with a pattern signal having a predetermined period, and enables pixel attraction. The present invention relates to an image processing device that performs halftone generation.

[0002]

2. Description of the Related Art Conventionally, as a method of binarizing a digital image signal and forming an image by an electrophotographic digital copying machine, a laser beam printer, etc., the digital image signal is once converted into an analog image signal, and this signal is converted into a triangular wave. It has been proposed to generate a binarized signal subjected to pulse width modulation by comparing it with a periodic pattern signal such as (JP-A-62-39972). In these methods, a high-frequency triangular wave pattern is used when there are many properties with high spatial frequencies such as characters over the entire area of the original image, and a high-frequency triangular wave pattern is used when there are many properties with low spatial frequencies such as photographs. When applying a low frequency triangular wave pattern that emphasizes tonality reproducibility, or for a document image in which both text and photo areas are mixed, high frequency is applied to each pixel according to the image area separation and edge detection results. A typical example is to switch and apply a low frequency triangular wave pattern.

FIG. 15 shows the latter conventional example.
For example, in a 400 dpi digital copying machine system, 200 lines (lines / inch) and 2 lines of 400 lines are used.
The kind of triangular wave pattern is switched according to the result of image area separation processing or edge detection processing. Figure 15
, The digital image signal is converted into an analog signal by the D / A converter 101, and then the two comparators 10
2 are input respectively. On the other hand, the pattern generation circuit 10
In 3, the 200-line triangular wave signal is output by the pattern generation circuit 10
In 4, the triangular wave signals of 400 lines are respectively generated and input as the reference comparison signals of the two comparators 102. The two comparators 102 compare the magnitudes of the two triangular waves and the image signal, and generate binary output signals having time widths corresponding to the magnitudes. The image area separation or edge detection processing circuit 105 detects a character or a photograph area, an edge and the like. The image area separation or edge detection processing circuit 105 collects pixels in a 4 × 4 block, for example, as shown in FIG. 16A, finds the difference between the maximum density and the minimum density in the block, and sets it as a threshold value. If the difference between the maximum density and the minimum density is greater than or equal to the threshold value, it is determined to be a character area, otherwise it is determined to be a photograph area, or as shown in FIGS. 16 (b) and 16 (c). As described above, the first-order differential filter composed of the 3 × 3 matrix detects the edge at the position where the differential value is maximum. According to the detection result of the image area separation or edge detection processing circuit 105, the selector 106 selects one of the two binary output signals generated by the two triangular waves, whichever is more suitable for edge reproduction. Select as final output signal. With such a configuration, a high-resolution screen of 400 lines is selectively applied to the area of characters and the like, and a screen of 200 lines is selectively applied to the area of photographs and the like, and both can be reproduced at a certain level.

[0004]

However, when considering the application to a system that requires a higher quality reproduced image, especially in the case of characters / lines, a screen equivalent to the resolution of the system (400 lines here) is used. Even if it is used, it is not at a satisfactory level. FIG. 17 explains the reason for this. As shown in FIG.
128,128,0 (density level 0-255) and the following 2
Reproduction of dot lines will be described as an example. The 400-line triangular wave and the density level 128 are compared, and the density level 1
A binarized signal that is "1" when 28 is higher than the triangular wave signal level and "0" when the density level 128 is lower than the triangular wave signal level is obtained as shown in FIG. The line width reproducibility is almost good. However, as can be seen from FIG. 17B, the line is divided. If the reproducibility of the output device is not so high, the influence of this line division can hardly be detected, but as the reproducibility of the output device improves, line breakage will be detected. As a method for improving this, a 400 line having a period of a pixel unit descends to the right,
It is also proposed that two sawtooth waves falling to the left are selected according to the direction of the edge to prevent line breakage as much as possible, and to improve the line width reproducibility of lines and the like (JP-A-62-62).
-233981). However, since a 400-line screen is used in this method, resolution reproducibility is good for characters with high spatial frequency, but gradation reproducibility is reduced for photos with low spatial frequency. In order to improve the gradation reproducibility even in the same area, it is necessary to have a separate 200-line screen.

As described above, when a screen having the same resolution as the system (400 lines in this case) is used, the reproducibility of resolution is good, but the reproducibility of gradation is not so good that the reproducibility of halftone images such as photographs is improved. There is a drawback to decrease.

Pixel pulling processing is known for edge correction and smoothing, and an image processing apparatus capable of generating this halftone controls a pixel generation position using a sawtooth wave. The method was common. This is disclosed in JP-A-62-233980 and JP-A-62-2.
There are those which use only a sawtooth wave such as 33981 and Japanese Patent Laid-Open No. 2-217175, and those which combine a sawtooth wave and a triangular wave such as Japanese Patent Laid-Open No. 2-47973. However, as the image processing speed has increased in recent years, the halftone generation operating frequency has increased, and when a sawtooth wave is generated, the waveform tends to be distorted or accentuated, which tends to adversely affect the gradation characteristics. Further, when the sawtooth wave is combined, a timing shift of the sawtooth wave may cause a crack or a step in the output signal, which adversely affects the gradation characteristics. In order to prevent this, the timing of the sawtooth wave generation signal is adjusted, but it takes a stable adjustment because adjustment costs and parts costs etc. are required to fully respond to changes due to changes in each device and external elements. This is a cause of sexual deterioration and increased costs.

As one of the methods for solving the problem when the sawtooth wave is used, a method using a triangular wave is disclosed in JP-A-5-244526 and JP-A-5-32810.
Proposed in 8. By using the triangular wave, the gradation characteristics deteriorate due to the distortion and accent of the waveform, which is a problem in the case of the sawtooth wave, and the deterioration of the gradation due to the deterioration of the stability of the output signal due to the timing deviation of the sawtooth wave and the step difference. It can be prevented. However, the triangular wave configuration requires three types of triangular waves to be used when the pixel pulling position is in three directions of right / left / middle, which leads to a complicated circuit configuration and an increase in cost.

The present invention has been made in order to solve the problems found in the conventional method, and detects the presence or absence of an edge and the direction based on the relative level of the pixel adjacent to the pixel of interest to obtain a halftone image. Characters while maintaining the tone reproduction of the range
It is an object of the present invention to provide an image processing apparatus that eliminates cracks and breaks in lines and the like, improves reproducibility of resolution, and obtains a high-quality output image.

It is another object of the present invention to have a relatively simple structure and reduce the operating speed by half as compared with the conventional system, and to enable application to a higher speed system.

It is another object of the present invention to provide an image processing device for generating a halftone which is capable of attracting pixels with a minimum configuration using a triangular wave.

[0011]

An image processing apparatus according to the present invention comprises means for detecting the presence or absence of an edge and a direction based on the absolute value and the magnitude relation of the difference between at least two pixels sandwiching the pixel of interest, and the pixel cycle A means for generating two patterns having a phase that is two times the phase of each other and one or both of them and a digital image signal after being converted into an analog signal are compared and compared with each other. It is characterized in that it comprises means for generating a binary signal.

The present invention also uses two types of reference waves,
In the image pulling mode, a centered pulse output is obtained by comparing one reference wave with the image signal, and another one of the reference waves and the image signal is compared with the image pulling mode depending on whether the image pulling mode is right aligning or left aligning. By inverting or non-inverting the signal, or by inverting or non-inverting the binary output depending on the comparison result, it is possible to pull the pixels centered, right-justified and left-justified as in the case of using three types of reference waves. It is characterized by having done.

[0013]

According to the image processing apparatus of the present invention, as a periodic pattern signal for generating a binarized signal by performing pulse width modulation, phase signals having a cycle twice as long as a pixel cycle having high gradation reproducibility are provided. Using two triangular waves in which is reversed, the presence or absence of an edge and the direction are detected from the absolute value of the difference between the two pixels sandwiching the pixel of interest and the magnitude relationship, and if there is an edge part, depending on the direction, By selecting an appropriate one of the two triangular waves and comparing the selected triangular wave with the analog signal obtained by converting the digital image signal, the binary signal of the time width generated is selected to increase the Achieved high quality image output. Further, it is possible to perform edge correction (smoothing) of the output image by pulling the output pulse in the right direction, the left direction, and the middle direction (intermediate) of the pixel by using pulse width modulation using two types of triangular waves.

[0014]

FIG. 1 is a diagram showing an embodiment of the present invention. In FIG. 1, a digital image signal is converted into an analog signal by a D / A converter 101, and then two comparators 1
It is input to each of 02. On the other hand, the pattern generation circuits 103 and 104 generate two triangular wave signals whose phases are inverted by 180 degrees, and these are input as comparison reference signals to the comparator 102. The period of the triangular wave generated here is twice the period of the pixel. Two
The two comparators 102 compare the magnitudes of the two triangular waves with the image signal, and generate binary output signals having time widths corresponding to the magnitudes. Edge / direction detection circuit 1
In 07, the edge direction (downward rightward or downward leftward) in the edge portion is detected by the method described later, and this detection result signal is output to the selector 106. The selector 106 selects, as a final output signal, one of the two binary output signals generated by the two triangular waves whose phases are inverted, whichever is more suitable for edge reproduction, as described later. To do.

Next, an edge portion and an edge direction (downward rightward, downward leftward) detection method will be described. FIG. 2 is a diagram showing a flowchart of edge part and edge direction detection. Considering three consecutive pixels ABC, a pixel B sandwiched between two pixels A and C is set as a pixel of interest. Target pixel B
The absolute value E (= the density difference level of two pixels A and C sandwiching
| AC |) is detected (step), and this is compared with a preset threshold Th (step),
If it is larger than Th, it is determined as an edge region (step). If it is smaller than Th, it is determined to be a non-edge region. Then, the selector 106 of FIG. 1 is controlled so that the triangular wave is selected according to the magnitude relationship between E and Th. When E is larger than Th and is determined to be an edge area, the magnitude relationship of the densities of pixels A and C is compared (step), and when the density of pixel A is large, it is determined to be a right-down edge (step). When the density of the pixel C is higher, it is determined that the edge is a left-down edge (step). FIG. 3 shows a specific configuration of the edge and edge direction detector. Two latch circuits 201 delay one pixel by one pixel and fetch three consecutive pixels ABC. Subtractor 202
Then, the absolute value E of the density difference levels of the pixels A and C sandwiching the pixel of interest B is calculated, and the calculated absolute value E and the threshold value Th are compared by the comparator 203. Further, the comparator 204 compares the densities of the pixels A and C, and compares the densities of the pixels 203 and 204 with 0 and 1 for each pixel.
Are input to the pattern determination logic circuit 205 to derive the final select signal of the triangular wave pattern. The selection logic of the triangular wave select pattern in the pattern determination logic circuit 205 is set as shown in Table 1.

[0016] In Table 1, the pixel flag is a flag indicating the position of the pixel, that is, whether it is an odd number or an even number from the start pixel. If the right tilt and the left tilt at the start pixel of the triangular wave pattern are known, each pixel flag indicates The waveform of the pattern corresponding to the pixel can be specified.

When the edge reproduction changes in a triangular wave pattern (cycle, phase) to achieve high-quality reproduction,
It is necessary to select the optimum one, and the selection of the triangular wave pattern will be described with reference to Table 1 and FIGS. 4 and 5.
FIG. 4 is an explanatory diagram when the pattern 2 is selected, and FIG. 5 is an explanatory diagram when the pattern 1 is selected. As in the case of FIG. 17,
Reproduction of 0,128,128,0 (density level 0-255) and subsequent 2-dot lines is considered as image data.
In addition, in FIG. 4 and FIG.
It is assumed that the 00-line triangular wave pattern 1 and pattern 2 are 180 degrees out of phase with each other and the pixel flags are as shown in the figure. In FIG. 4, when comparing the densities of the pixels A and C sandwiching the pixel of interest B, since A <C, the edge direction is downward to the left, and the pixel flag corresponding to the pixel of interest is “0”. Pattern 2 is selected from 1. Also, when the pixel of interest is shifted to the right by one pixel, the edge direction goes down to the right and the pixel flag becomes "1".
Pattern 2 is selected. When pattern 2 is selected, as shown in FIG. 4 (c), line cracks do not occur,
It enables sharp and high-quality line reproduction. In the pattern 1, line cracks are generated as shown in FIG.

In this way, the triangular wave pattern 1 and it and 1
In the selection of the triangular wave pattern 2 with a phase shift of 80 degrees, the pattern of falling to the right and the pattern of falling to the left appear repeatedly for each pixel, so the selection logic must be set to be inverted for each pixel, as shown in Table 1. As described above, if the pixel flag is 0 and the pixel is falling to the right, the pattern 1 is set.

Further, as shown in FIG. 5, when the line position of the image data is shifted to the right by one pixel with respect to FIG. 4, when the densities of the pixels A and C sandwiching the pixel of interest B are compared, A <C Therefore, the direction of the edge is downward to the left, and the pixel flag is "1" at this time, so pattern 1 is selected from Table 1. When the pixel of interest is shifted to the right by one pixel,
Since the edge direction is falling to the right and the pixel flag is "0", pattern 1 is selected from Table 1, and as shown in Fig. 5 (b), line cracks do not occur, and sharp and high-quality line reproduction is possible. become. In addition, in the pattern 2, FIG.
Line cracks are generated as shown in.

As described above, by matching the dot growth direction from the center of the line with the reproduction direction of the waveform pattern of the 200-line triangular wave, it is possible to reproduce a high-quality line without line cracks, and in principle, all There are only two line / dot growth directions at the edge part of the right-down or left-down direction, and if two 200-line triangular waves with phases inverted by 180 degrees are held, you must select either one. Can be matched.

FIG. 6 is a diagram showing a second embodiment of the present invention. In this embodiment, a selector 106 is inserted immediately after the pattern generation circuits 103 and 104 to select a pattern before the binarization process, and only one comparator is used. This is the same as the first embodiment.

In the above embodiment, the two pattern generating circuits are provided to generate the 200-line triangular wave that is 180 degrees out of phase with each other. However, as shown in FIG. 7, only one pattern generating circuit is used. Two triangular wave patterns may be generated by selecting the selected pattern and the pattern that has passed through the phase inversion circuit with the selector.

Next, the pixel attraction using a triangular wave, which is the premise of the pixel attraction of the present invention, will be described. FIG. 8 is an overall block diagram of a pixel attracting circuit using a triangular wave, and FIG. 9 is a block diagram showing a pixel attracting circuit portion. In FIG. 8, the image processing apparatus 301 has a function of performing image processing such as dark correction, shading correction, color correction, and reduction / enlargement on the read image signal, and its 8-bit video data is the interface 302. , F
Temporarily stored in the IFO or line memory 303, D / A
The converted data is converted into analog data by the converter 304 and added to one input of the comparators 321, 322, 323 through the operational amplifier 305 functioning as a buffer.
The other inputs of the comparators 321, 322, 323 are 400-line triangular wave, 200-line A-phase triangular wave, and 2 line, respectively.
00 line B phase triangular wave is added. The 400-line triangular wave directly buffers the video clock signal from the oscillator 310.
In addition to the 400-line triangular wave generator 315, the 200-line A-phase triangular wave and the 200-line B-phase triangular wave generate the video clock signal from the oscillator 310 through DF / F.
The frequency is divided into 1/2 by the frequency divider 311 composed of F, and the 200-line A-phase triangular wave generator 3 is passed through the buffers 313 and 314.
In addition to the 16,200-line B-phase triangular wave generator 317, they are generated by shifting their phases by 180 °. Each triangular wave has a coupling capacitor to cut the DC component,
The DC component from the C component generators 318 to 320 is added, the level is adjusted, and the DC component is added to each comparator. On the other hand, a 2-bit edge direction signal (Tag) obtained by the difference calculation in the image processing apparatus as described in FIG.
Is temporarily stored in the interface 306, the FIFO or the line memory 307, delayed by the timing adjustment circuit 308, adjusted in timing, and added as a select signal to the selector.
1, the output of one of the comparators 322 and 323 is selected, and the laser diode 332 is passed through the buffer 331.
Is output to.

The pixel attraction will be described with reference to FIG. 9. As shown in FIG. 9A, the video data from the image processing device 301 in the preceding stage before the pixel attraction circuit is the DAC (D / A converter) as described above. At 304, it is converted into an analog video signal. The reference wave (triangular wave) to be compared with the analog video signal is, as shown in FIG. 9B, a triangular wave 400 line, a triangular wave 200 line with a phase difference of 180 °, and an A phase and a B phase by three types of pattern generators. Is generated,
Three types of output pulses are obtained by the comparator in comparison with the analog video signal, and the pixel pulling is performed by selecting three types of triangular wave comparison output pulses according to the edge direction signal from the identification circuit 301a. The identification circuit 30
1a is included in the image processing apparatus 301 in FIG. Pixel pulling in the present invention means centering (middle), right aligning, and left aligning as shown in FIG. 9C, but center aligning means that a pulse grows from an intermediate position in one pixel. The right alignment means that the pulse grows from the right position in one pixel, and the left alignment means that the pulse grows from the left position in one pixel.

FIG. 10 shows an operation phenomenon of drawing a pixel using a triangular wave. In FIG. 10, odd / even of the video data output in synchronization with the video clock corresponds to the pixel flag in Table 1 and indicates whether it is an odd number or an even number from the start pixel at which the main scanning synchronization signal rises. . The video data is converted into analog data as shown by the DAC. The reference wave (triangular wave) is 400-line triangular wave, 2
It becomes a 00-line triangular wave A phase and a 200-line triangular wave B phase. 20
The phase of 0-line triangular wave A phase and 200-line triangular wave B phase is 180
Shows triangular waves that are offset. Needless to say, three types of triangular waves having a frequency ratio of 2: 1, A phase and B phase (phases differ by 180 °), such as 600-line triangular wave, 300-line triangular wave A-phase and 300-line triangular wave B-phase, can be used. Nor. Pixel drawing can be centered, right-justified, or left-justified within one pixel as a basic direction, but 400 lines of a conventional character or map or 200 lines of a photograph or the like can be obtained by using each triangular wave. It goes without saying that a combination of 600 lines and 300 lines may be used for letters, maps and photographs.

First, the pixel pulling position will be described. In the case of centering, the 400-line triangular wave shown in FIG. 10 is used, and by comparing with analog video data, the output pulse becomes a pulse that grows from the middle position (intermediate) of the pixel as shown in FIG. . Here, it is assumed that the output pulse is obtained when the triangular wave is larger than the video data. In the case of right alignment, the 200-line B phase is selected when the video data is the odd pixel and the 200-line B phase is selected after the main scanning synchronization signal shown in FIG. 10 is valid, and when the video data is the even pixel. , 200 line A phase is selected, the output pulse shown in FIG.
You can get a pulse that grows from the right. In the case of left-justification, if the video data is an odd-numbered pixel after the main scanning synchronization signal shown in FIG.
By selecting the 0-line A phase and selecting the 200-line B phase when the video data is an even pixel, the B-phase of FIG.
A pulse that grows from the left can be obtained, such as the output pulse shown inside. Also, although it is not a pixel drawing, 400 lines are used for conventional characters and maps, and 2 for photos.
00 line A phase is selected.

As described above, the pixels can be attracted by using the triangular wave, but when three types of attraction such as centering, right justification, and left justification are performed, three kinds of triangle waves are used, so that the conventional characters, maps, etc. Compared to the case of 400 lines or 200 lines such as photographs, the number of circuits is increased by one circuit, resulting in cost increase, circuit complexity, and board area increase. In the present invention,
Using a 400-wire / 200-wire circuit, 3 with 2 types of triangular waves
The same effect as the pixel attraction using the types of triangular waves can be obtained.

Examples of the present invention will be described below. Figure 11
Is an overall block diagram of a pixel attracting circuit using a triangular wave of the present invention, and FIG. 12 is a block diagram showing a pixel attracting circuit portion. In FIGS. 11 and 12, the video data is D
The reference wave (triangular wave) converted into an analog video signal by AC 304 is generated by two types of pattern generators.
As shown in (b), 400 triangular wave lines and 200 triangular wave lines A
A phase is generated, the comparator compares the analog video signal with the reference wave (triangular wave) to obtain three types of output pulses, and the edge direction signal from the identification circuit 301a selects three types of triangular wave comparison output pulses to select pixels. Attract. The selector 340 of FIG.
1. The non-inverting amplifier 342 selects the inverted and non-inverted 8-bit video data by the edge direction signal and extracts it. The comparator 324 extracts the outputs having 180 ° different phases. The triangular wave 200 line B phase can be omitted as shown below.

Next, the operation phenomenon of pixel attraction using two types of triangular waves will be described with reference to FIGS. 13 and 14. Two types of triangular waves are used, 400 lines and 200 lines A phase, and 200 lines B phase is realized by 200 lines A phase when three types of triangular waves are used. Make sure you get the same effect as if you used it. Figure 1
When analog video data as shown in 3 is obtained,
By comparing the analog video data with the reference 200-line B-phase triangular wave, a reference 200-line B-phase output pulse as shown in the figure is obtained. Therefore, when the 200-line B-phase is selected in the pixel drawing mode, the selector 340 inverts the analog video data (in the case of 8-bit digital data, 0 → 255, 255 → 0 is inverted).
Data is extracted and compared with the 200-line A-phase triangular wave, and in the conventional comparator, when the triangular wave and the analog video data have a larger relationship (higher level) in the triangular wave, a Hi level pulse (laser diode lighting at Hi,
If the analog video data is larger (the level is higher), the Hi level pulse is output instead of the (image formation). The same virtual B-phase output pulse as the phase output pulse can be obtained. This is the comparator 324
Similarly, by selecting the opposite phase output of, the virtual B phase output pulse created by using the 200 line A phase can be obtained. In this way, three types of pixel pulling can be performed with only two types of triangular waves, centering, right aligning, and left aligning.

The centering operation, the right-justifying operation, and the left-justifying operation using only two types of triangular waves according to the present invention will be described individually. In the case of the centering operation, by using the 400-line triangular wave shown in FIG. As shown in the centering of the triangular wave operation shown in 14, it is possible to obtain a pulse growing from the middle position (intermediate position) of the pixel. In case of right alignment,
When the video data is an odd-numbered pixel after the main scanning synchronization signal shown in FIG. 14 is enabled, the virtual 200-line B-phase using the 200-line A-phase is selected (the inverted analog data in FIG. 14 is used. ), When the video data is an even-numbered pixel, by selecting the 200-line A phase (using the analog data in FIG. 14), a pulse that grows from the right like the output pulse shown in FIG. 14 is generated. Obtainable. In the case of left alignment, after the main scanning synchronization signal shown in FIG. 14 becomes effective, when the video data is an odd pixel, the 200-line A phase is selected (using the analog data in FIG. 14), When the video data is an even-numbered pixel, by selecting the virtual 200-line B-phase using the 200-line A phase (using the inverted analog data in FIG. 14), the output pulse shown in FIG. , You can get the pulse growing from the left. In this way, with two types of triangular waves and two circuit configurations similar to the conventional one, pixels can be attracted, and problems such as cost increase, circuit complexity, and board area increase can be solved. Also, although it is not a pixel drawing, 400 lines are selected for conventional characters and maps, and 200 A lines are selected for photographs and the like.

By the way, the output pulse width adjustment (density (area) by image formation) has been performed by AC coupling the triangular wave input to the comparator and applying the DC component, as shown in FIGS. However, in this embodiment, the 200-line triangular wave A phase can be adjusted by adjusting the application of the DC component as in the conventional case. However, in the virtual B-phase operation of the 200-line triangular wave A phase, the analog data to be compared is inverted. Therefore, the DC component of the 200-line triangular wave A phase cannot be adjusted, and the output pulse width can be adjusted by adjusting the triangular wave amplitude value or the analog video data amplitude value.
However, pixel pulling is performed at the time of edge correction and smoothing in the first place, and tends to work on data in the dark direction in terms of density, and when the gradation of a photograph or the like is required, the 200-line A phase is used. The B phase tends to be used only when the pixels are attracted. In such a case, it is possible to fix the triangular wave amplitude value or the analog video data amplitude value in a certain range and leave the B phase unadjusted.
This is not the case when pixels are drawn even when the density is low. ), It is possible to attract pixels by the conventional circuit configuration, control method, and adjustment method.

[0032]

As described above, according to the present invention, only two phase-inverted triangular wave patterns having a period twice the pixel period having high gradation reproducibility are used, and each pixel of the image data is used. By selectively switching according to the detected edge direction, characters, lines, etc. can be reproduced without interruption and sharpness, and the reproducibility of halftone images, etc. is not deteriorated. The operating frequency is 1 / the processing speed.
With the triangular wave that can be realized with 2 clocks, high-quality reproduction of characters and lines / photos is possible, which is advantageous for future high-speed copying systems and printers, and has excellent stability.

Further, by using two types of triangular waves, it is possible to attract pixels in the same manner as when using three types of triangular waves, and it is possible to attract pixels at a low cost with ease and stability.

[Brief description of drawings]

FIG. 1 is a diagram illustrating an embodiment of the present invention.

FIG. 2 is a diagram showing a flowchart of edge and edge direction detection.

FIG. 3 is a diagram showing a configuration of an edge and edge direction detector.

FIG. 4 is a diagram illustrating selection of a triangular wave pattern.

FIG. 5 is a diagram illustrating selection of a triangular wave pattern.

FIG. 6 is a diagram for explaining another embodiment of the present invention in which one comparator is used.

FIG. 7 is a diagram for explaining another embodiment of the present invention in which one triangular wave pattern generation circuit is used.

FIG. 8 is an overall block diagram showing a configuration for explaining pixel attraction using a triangular wave.

FIG. 9 is a diagram illustrating a pixel drawing circuit using a triangular wave.

FIG. 10 is a diagram illustrating a triangular wave operation phenomenon when a pixel is drawn using a triangular wave.

FIG. 11 is an overall block diagram showing a configuration of pixel attraction using a triangular wave according to the present invention.

FIG. 12 is a diagram illustrating a pixel attracting circuit using a triangular wave according to the present invention.

FIG. 13 is a diagram illustrating a triangular wave operation phenomenon of the present invention.

FIG. 14 is a diagram illustrating a triangular wave operation phenomenon when a pixel is drawn using the triangular wave according to the present invention.

FIG. 15 is a diagram showing a conventional example in which a high frequency / low frequency triangular wave is switched depending on a result of image area separation or edge detection.

FIG. 16 is a diagram illustrating an image area separation or edge detection method.

FIG. 17 is a diagram illustrating image reproduction when a 400-line triangular wave is used.

[Explanation of symbols]

101 ... D / A converter, 102 ... Comparator, 103
... pattern generating circuit, 104 ... pattern generating circuit, 10
5 ... Image area separation or edge detection circuit, 106 ... Selector,
107 ... Edge and edge direction detection circuit, 201 ... Latch circuit, 202 ... Subtractor, 203, 204 ... Comparator, 2
05 ... Pattern determination logic circuit, 301 ... Image processing device,
304 ... DC converter, 315 ... 400-line triangular wave generator, 316 ... 200-line A-phase triangular wave generator, 317 ... 20
0-line B-phase triangular wave generator, 318-320 ... DC component generator, 321-323 ... Comparator, 330 ... Selector, 340 ... Selector.

─────────────────────────────────────────────────── ─── Continuation of front page (51) Int.Cl. ⁶ Identification number Reference number within the agency FI Technical display location G06F 15/68 320 A (72) Inventor Yoshihiro Terada 2274 Hongo, Ebina City, Kanagawa Prefecture Fuji Xerox Co., Ltd. (72) Inventor Hiroshi Sekine 2274 Hongo, Ebina City, Kanagawa Prefecture Fuji Xerox Co., Ltd. (72) Inventor Waho Kiya 2274, Hongo, Ebina City, Kanagawa Prefecture Fuji Xerox Co., Ltd.

Claims (2)

[Claims] 1. An image processing apparatus for binarizing a multi-valued image signal to obtain a binary image signal, and means for generating a pair of phase-inverted pattern signals having a period twice the pixel period. , A binarizing means for comparing the pattern signal with the image signal and outputting a binary signal having a time width corresponding to the magnitude, and a pattern signal discrimination for selecting the pattern signal based on the density level determination of pixels around the target pixel And an image processing apparatus configured to obtain a binary signal by selecting a pattern signal for each pixel. 2. An image processing apparatus for binarizing a multi-valued image signal to obtain a binary image signal, the first reference wave and a second reference wave having a period twice as long as that of the first reference wave. A reference wave generating means for generating the reference wave, and a binarizing means for comparing the first and second reference waves with the image signal and outputting a binary signal having a time width corresponding to the magnitude, and a pixel of interest. And a selecting unit for selecting a binary signal corresponding to the first or second reference wave based on the density level judgment of the surrounding pixels, and the pixel pulling mode is the second.
When a binary signal corresponding to the reference wave is selected, a means for inverting / non-inverting the image signal level to be compared according to the pixel pulling mode, or a means for selecting an inverted / non-inverted signal output from the binarizing means is provided. An image processing device characterized by the above.