JP2598395B2 - Image processing device - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to an image processing apparatus, and more particularly, to an image processing apparatus applicable to a laser printer, a digital color copying apparatus, and the like.

(Prior Art) Recently, in an image processing apparatus applied to a digital color copying machine or the like, a small matrix is often used as a means for representing a halftone by combining an area gradation by a density pattern method and a method of modulating a pulse width of a laser beam. There is a method for expressing gradation. For example, JP-A-58-85671 and JP-A-59-204
In No. 060, for example, in a 4 × 4 dot matrix, 4 × 4 × 3 + 1 = 49 values are realized by further setting the value within one dot to three values and changing the pulse width in steps of 1/3. 0 level).

By the way, the reentrant line of the latent image pattern on the photoreceptor generated by the irradiation of the beam energy obtained by the above method is a rectangular wave or a triangular wave. The roughness or noise of the image generated in the digital image is caused by the harmonic component due to the rectangular wave or the triangular wave. The spatial frequency component (MTF) of a rectangular wave or a triangular wave is performed by means such as Fourier transform. However, when the above waveform is subjected to Fourier transform, a harmonic is generated at a frequency higher than the fundamental frequency of the dot.

On the other hand, in the case of an analog copying apparatus, the point image becomes a blurred image due to the aberration of the optical system or the oblique light image shift caused by the curved photoreceptor, and the blurred image is almost Gaussian distribution. It seems to be formed from the aggregate of The MTF based on the Fourier transform of a Gaussian distribution is also a Gaussian distribution and does not generate high frequencies. That is, it is determined that there is a difference between the digital image and the analog image.

The transfer function of an analog image is a Gaussian distribution, while the transfer function of a digital image is a pulse distribution including harmonics.

FIG. 9 shows an example of area gradation by the density pattern method without pulse modulation.
The case of × 2 is shown. In the figure, each pixel 1 is 2 × 2
The number of fine pixels 2 is arranged in a matrix, and the halftone is obtained by sequentially increasing the number of black fine pixels as shown in (b), (c), (d) and (e) of FIG. Expresses the brightness of the key.

FIG. 10 shows a graph of area gradation [FIG. 10 (a)] and gamma characteristics [FIG. 10 (b)] reproduced by a conventional density pattern method in a 4 × 4 matrix. As shown in the figure, in the case of only the area gradation, the linearity of the gamma characteristic is not good due to the dots to be reproduced.

This is especially true in the reproduction of highlights, in which the rise of the density becomes steep and the gradation becomes poor.

FIGS. 11 (a), (b) and (c) show the moving state of the laser beam, the width of the pulse signal, and the light quantity distribution of the obtained fine pixels. From this figure, it is understood that in the state where the pulse signal width is shorter than the moving state of the laser beam, the peak value of the obtained light quantity distribution of the fine pixels is lower than the peak value in the state where the pulse signal width is longer than the moving state of the laser beam. Is done.

FIGS. 12 (a) and 12 (b) show a conceptual diagram and a gamma characteristic, respectively, of a pattern in the case of four-level equally-spaced pulse width modulation performed on a 4 × 4 matrix. According to an experiment using this method, even when multivalued pulse width modulation is performed, the density rises at a level corresponding to the first to third values of the highlight rises quickly, as shown in FIG. 12 (b). It has been confirmed that the gamma characteristic is not a straight line, and the energy distribution is rectangular as described above, and as shown in the pattern example of FIG. It has a rough feeling.

(Purpose) The present invention has been made to eliminate the above-mentioned drawbacks of the related art, and it is an object of the present invention to provide an image processing apparatus capable of reducing image roughness, which is a drawback of digital images. It is in.

(Constitution) In order to achieve the above object, the present invention processes image data, modulates the pulse width of a pulse signal of a laser beam together with image processing to represent halftone density by area gradation by a density pattern method. In an image processing apparatus for performing image processing representing a multi-value density, a laser beam density pattern is of a concentrated type or a spiral type, and a pulse width distribution of each level of a multi-value is such that a reference distribution approximates a Gaussian function. It was made.

Hereinafter, a specific description will be given based on an embodiment of the present invention.

FIG. 1 is a schematic view showing the periphery of a photosensitive drum of a digital color copying apparatus to which the present invention is applied.
Is a photosensitive drum, 4 is a charger, 5 is a yellow developing device, 6
Is a magenta developing device, 7 is a cyan developing device, 8 is a black developing device, 9 is a pulse sensor, 10 is a pre-transfer lamp, 11 is a pulse counter, 12 is a transfer drum, 13 is a transfer charger, 14
Is a static elimination before cleaning, 15 is a separation charger, and 16 is a cleaning device. 17 is a yellow supply signal line, 18 is a magenta supply signal line, 19 is a cyan supply signal line, 20 is a black supply signal line, 21 is a yellow developing bias signal, 22 is a magenta developing bias signal, and 23 is a cyan developing bias signal. , 24 are a black developing bias signal, 25 is a drum rotation pulse signal line, and 26 is a pulse counter output line.

As described above, this digital color copying apparatus includes developing units 5, 6, 7, and 8 for each of the four colors, the photosensitive drum 3, the transfer drum 12, and a writing laser optical element which is omitted in FIG. I'm more connected.

The toner supply control process in the above configuration will be described. First, a standard detection pattern is exposed on the photoreceptor drum 3 to the rear end of the image signal by a laser beam from the direction of the arrow to the photoreceptor drum 3 to form a detection pattern latent image. This latent image is developed with the same developing bias value as that of the image signal portion. When the detection pattern image on the photosensitive drum 3 passes through the pulse sensor 9 while rotating, the amount of reflected light of the pattern is measured by the pulse sensor 9 at the timing measured by the pulse counter 11. This process is repeated four times for each color.

FIG. 2 is a schematic sectional view showing an image scanner for photoelectrically converting and reading an original as an example of a writing laser optical system. In the figure, 27 is an image scanner, 28 is a video processing circuit, 29 is a contact glass on which an original is placed, 30 is a light source, 31 is a first scanning mirror, 32 is a second scanning mirror, 33 is an imaging lens, 34 Is a photoelectric conversion unit for electrically reading the reflected light of the document, 35 is a lighting circuit for driving the light source 30, and 36 is a DC motor for performing mechanical scanning (sub-scanning).

Since the image scanner 27 is of a fixed document scanning type so as to be able to cope with the variety of documents, the optical path length of the document reflected light 37 is always constant in the sub-scanning.
The first carriage Ca1 for mounting the light source 30 and the first scanning mirror 31 and the second carriage Ca2 for mounting the second scanning mirror 32 are sub-scan driven by the DC motor 36 at a speed ratio of 2: 1.

The photoelectric conversion unit 34 uses two sets of CCDs (solid-state imaging device arrays) to reduce the load on the microlenses of the reduced image forming optical system, so that 16 × 16 (dots / mm ² )
The original can be read at a high resolution. Therefore, the main scanning is a divisional imaging scan by these two sets of CCDs and two sets of imaging lenses 33.

FIG. 3 is a block diagram for explaining the schematic operation of the image scanner 27. In the figure, 30 is a light source, 34 is a photoelectric conversion unit, 34a and 34b are CCDs, 34c and 34d are amplifiers, 35 is a lighting circuit, 36 is a DC motor, 38 is a timing circuit, 38a is a timing generator, 38b is an oscillator, 39 is a scanner control circuit, 40 is a position sensor, 41 is a sub-scanning speed control circuit, 42 is a speed signal generation circuit, 43 is an encoder, and a is a start signal.

2 and 3, the lighting circuit 35 controls the light source.
30 is driven. The original reflected light of the reading line (sub-scanning line) irradiated by the light source 30 is reflected by the first scanning mirror 31.
And two sets of imaging lenses reflected by the second scanning mirror 32
The light is guided to a photoelectric conversion unit 34 by 33. Here, the main scanning is performed by dividing and forming an image on the light receiving surface of each of the two sets of CCDs 34a and 34b. The main scanning by the CCDs 34a and 34b is performed by the shift pulses from the timing generator 38a.
The reading by b is sequentially shifted.

FIG. 4 is a block diagram showing a schematic configuration for performing color image processing. In the figure, the image processing unit A
The scanner unit B and the printer unit C are shown centered on FIG. In FIG. 4, 44 is a shading correction circuit, 45 is a gamma (γ) correction circuit, 46 is a masking processing circuit, 47 is a UCR processing circuit, 48 is a density pattern processing, 49 is a multi-value processing circuit, and 50 is a laser driver Reference numeral 51 denotes a laser unit, 52 denotes a main body control unit, and 53 denotes a synchronization control circuit.

In FIG. 4, an image signal read by a CCD light receiving unit (not shown) of the scanner unit B is subjected to a shading correction circuit 44 for correcting optical illuminance unevenness for each of yellow (Y), magenta (M), and cyan (C). Correction is performed. Next, the gamma correction circuit 45 corrects the gradation of each color signal.

The masking processing circuit 46 calculates optimal amounts of Y, M, and C at the time of printing. In the next UCR processing circuit 47, an appropriate black (Bk) amount for producing black is calculated from Y, M, and C.

Further, a density distribution processing 48 outputs a pattern distribution corresponding to each of the 64 gradations, and a multi-value processing circuit 49 performs pulse width modulation in accordance with 1 to 4 values in the pattern.

The pulse width of this pulse width modulation will be described with reference to FIG. FIG. 5 shows a portion according to the present invention, in which four pulse widths are provided. The value of this pulse width is determined by a Gaussian function. Fig. 5 (a)
Indicates the Gaussian distribution and the setting state of each level by a dotted line. This Gaussian function is It is represented by Here, M represents a pulse width (unit: time nsec) allowed for one dot, and in this case, 125 nsec.
The values of each level in FIGS. 5 (a), (b) and (c) are determined as follows. First, Level 1 is the minimum pulse width that can be reproduced by an actual printer and depends on the beam diameter, but is about 10 nsec in an experimental machine. Therefore Level 1
Is determined to be 10 nsec. Next, since M is determined to be 125 nsec, substituting 125 for M in Equation (1), 10 for y, and 3 for x (the distance of level 1 is 3 from the center of Gauss), and the shape parameter Determine a.

For the other levels 2 and 3, x and 2, respectively, are substituted for x to obtain the value of y. Thus, four values of the pulse width are obtained.

FIG. 5 (b) shows the pulse width of each level within one dot. FIG. 5 (c) shows the pulse width values for each of the four values calculated according to the above equation. FIG. 5D shows an example of the gamma characteristic in this case. This gamma characteristic is equal to the gamma characteristic of halftone dots in printing.
The reason is considered to be that the shape of the dot approximates to the Gaussian distribution and shows the same tendency as in the case of the halftone dot.

6 (a), 6 (b) and 6 (c) show the case where 7-level pulse width modulation is performed in the same manner as above, and FIG. 6 (a) shows the Gaussian distribution and the setting state of each level. Is shown by a dotted line, FIG. 6 (b) is a diagram showing one dot and seven values, and FIG. 6 (c) is the same expression as in FIG. 5 (c). Shows the value of the pulse width of each of the seven values calculated according to the following.
In this case, the Gaussian function that determines the pulse width is Therefore, the pulse width of each level is determined as shown in the table.

FIG. 7A shows a 4 × 4 matrix concentrated type density pattern. This indicates the order of priority for filling patterns within 4 × 4. In addition, FIG.
This shows the pulse level distribution in the matrix determined when 37 values out of × 4 values = 64 gradations. FIG. 7 (c) shows a processing flow of a level distribution determination program. In this processing flow, a gamma curve is set by inputting a gamma parameter (x ² + y ² = a), and a 7-level level pulse width input P (1)-
The energy calculation E (x) of 0 to 63 gradations is performed by P (7), and if the loop K = 0 to 63, K> 63, each P (x) is arranged in the order of the density pattern, and × 4 matrix 0 to 63 gradations are output. On the other hand, the loop K = 0 to 63, K>
If it is not 63, from the calculation of the number of 7-valued P (7),
The number calculation of (1) is performed.

8 (a) shows the determined horizontal and oblique level values of the 4 × 4 matrix, FIG. 8 (b) shows its exposure energy distribution, and FIG. 8 (c) shows the latent image electric field. The distribution curve is shown. The electric field distribution of the latent image in FIG. 8 (c) has a substantially Gaussian distribution. FIG. 8D conceptually and three-dimensionally shows a toner image distribution of a 4 × 4 matrix.

As described above, according to the present embodiment, the toner adhered to the photosensitive drum 3 in FIG. 1 has a shape approximate to a Gaussian distribution for each density pattern, so that the density distribution of dots of the formed image becomes analog. Therefore, no harmonics are generated, and thus the reproduced image is an image with little roughness. In this embodiment, the laser beam having a concentrated laser beam pattern is described.

(Effects) As described above, according to the present invention, the concentration pattern of the laser beam is of a concentrated type or a spiral type, and the pulse width distribution of each level of the multi-value is such that the reference distribution approximates a Gaussian function. As a result, since the dot density distribution of the formed image becomes analog, no harmonics are generated, and thus the reproduced image can be an image with little roughness.

[Brief description of the drawings]

FIG. 1 is a schematic view showing the periphery of a photoreceptor drum of a digital color copying apparatus to which the present invention is applied, FIG. 2 is a schematic sectional view showing an image scanner, and FIG. 3 is a block diagram for explaining the schematic operation of the image scanner. FIG. 4 is a block diagram showing a schematic configuration for performing color image processing, FIG. 5 (a) is a graph showing a Gaussian distribution, and FIG. 5 (b) is a graph showing the pulse width of each level within one dot. FIG. 5 (c) is a diagram showing the relationship between each level and pulse width, FIG. 5 (d) is a graph showing gamma characteristics, and FIG. 6 (a) is a case of 7-level pulse width modulation FIG. 6 (b) is a diagram showing one dot and seven values, FIG. 6 (c) is a diagram showing the relationship between each level and pulse width, and FIG. 7 (a) is a 4 × 4 graph. FIG. 7B shows a density pattern of a matrix, and FIG. 7B shows a pulse level distribution in a 4 × 4 matrix. FIG. 7 (c) is a processing flowchart of a level distribution determination program, and FIG. 8 (a) is an explanatory diagram showing horizontal and diagonal level values of the determined 4 × 4 matrix. 8 (b) is a graph showing an exposure energy distribution, FIG. 8 (c) is a graph showing a latent image electric field distribution, and FIG. 8 (d) is a conceptual diagram showing a toner image distribution of a 4 × 4 matrix in a three-dimensional manner. Figure 9 (a),
(B), (c), (d), and (e) show the brightness of five gradations in the density pattern method without pulse modulation, and FIG. 10 (a) shows the area by the density pattern method of a 4 × 4 matrix. FIG. 10 (b) is a graph showing the gamma characteristic reproduced, and FIGS. 11 (a), (b) and (c) are the moving state of the laser beam, the width of the pulse signal, and FIGS. 12 (a) and 12 (b) are diagrams showing a light quantity distribution of the obtained fine pixels, and are a pattern conceptual diagram and a gamma characteristic diagram in a case where three-value equal-interval pulse width modulation is performed on a 4 × 4 matrix. A: Image processing unit, B: Image scanner unit, C: Printer unit, 48: Density pattern processing, 49: Multi-value processing circuit.

Claims (1)

(57) [Claims] 1. An image processing device which processes image data and expresses a halftone density by an area gradation by a density pattern method.
In an image processing apparatus that modulates the pulse width of a pulse signal of a laser beam to perform image processing representing a multi-level density, the density pattern of the laser beam is a concentrated type or a spiral type,
An image processing apparatus wherein a pulse width distribution of each level of a multi-value is such that a reference distribution approximates a Gaussian function.