JPH11305752A - Smoothing device for multilevel image and image forming device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To smooth only characters, graphics, etc., in a multilevel image without affecting a natural image. SOLUTION: A binarizing circuit 3 binarizes multilevel image data read out of an image memory 1 into binary image data consisting of bits each for a pixel. In this case, pixel values are so binarized that 255 corresponds to a value '1' and 0 to 254 correspond to a value '0'. Consequently, only characters and graphics are binarized to '1' and a natural image is binarized almost to '0'. An area of, for example, 8×9 pixels of this binary image data is put in a register 5 and image elements having the value '1' are smoothed according to the binarized data of the 8×9 pixels in the register 5. The drawing signal obtained by this smoothing processing and a drawing signal obtained by a half-toning processing are ORed to obtain a final drawing signal.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to an image forming apparatus for displaying or printing a digital image represented by a bit map, and more particularly to a smoothing technique for smoothing outlines of characters and figures.

[0002]

2. Description of the Related Art In a digital image, since a bitmap expression is used, zigzag irregularities called jaggies occur in contours of characters and figures. Smoothing is a technique that corrects jaggies to finer irregularities so that the contour lines appear to the human eye as smooth lines.

[0003] Elements constituting an image are generally classified into three main types: "characters" such as characters and symbols, "graphics" such as diagrams and drawings, and "natural images" such as photographs and paint paintings. it can. Of these three elements, those that generally require smoothing are elements having a clear outline such as “character” and “graphics”. On the other hand, since a “natural image” is an element whose image value changes substantially continuously, applying a smoothing results in an unnatural image.

[0004] Characters and graphics can typically be represented by binary image data (data in which one pixel value for one color is represented by a 1-bit word). On the other hand, a natural image needs to be represented by multivalued image data (data in which one pixel value for one color is represented by a plurality of bits (for example, 8 bits) words). Under these circumstances, the conventional smoothing technique has been developed exclusively for processing binary image data, and cannot be used for multi-valued image data.

[0005]

However, in many cases, characters and graphics are represented in the form of multi-valued image data. In the case of a color image, multivalued image data is more common. A typical example of this type of image is an image in which characters and graphics are overwritten on a natural image. Since smoothing cannot be performed on such a multi-valued image, jaggies in characters and graphics become noticeable. On the other hand, it is possible to perform the smoothing after the multi-valued image is once dropped into the binary image, but this has the side effect that the natural image that does not need to be corrected is destroyed.

SUMMARY OF THE INVENTION Accordingly, an object of the present invention is to allow a character or graphics to be selectively smoothed in a multi-valued image without affecting a natural image.

[0007]

According to a first aspect of the present invention, there is provided a multi-valued image smoothing apparatus, comprising: a binarizing section for binarizing a multi-valued image; and a binary output from the binarizing section. A smoothing unit that performs a smoothing process on the converted image to output a smoothed image signal; a halftoning unit that performs a halftoning process on the multi-valued image to output a halftone image signal; and the smoothed image signal and the halftone image. And an output unit that outputs a final image signal by combining the signal with the signal. According to this smoothing device, by appropriately performing binarization in the binarizing unit, a smoothing target such as a character or graphics can be effectively distinguished from an image element that is not a smoothing target such as a natural image. The obtained binary image can be obtained. By performing a smoothing process on the binary image, a smoothed image signal in which only the smoothing target is smoothed can be obtained. Then, by combining this smoothing image signal with a half-toning image signal obtained by separately performing a half-toning process on a multi-valued image, the natural image naturally has a continuous density, and the characters and graphics have no jaggy. A final image signal that can be reproduced with a clear outline can be obtained.

As the threshold value for binarization, a value should be used that can substantially distinguish an image element to be smoothed from an image element not to be smoothed. For example,
A value near the maximum value of the multi-valued image value can be used as the threshold. Further, the threshold value may be made variable in accordance with the state of the target image, the user's preference, and the like.

According to a second aspect of the present invention, there is provided a multi-valued image smoothing apparatus for generating a binary image in which image elements to be smoothed and non-smoothed image elements in the multi-valued image are distinguished. Unit, a smoothing unit that performs a smoothing process on the binarized image to output a smoothed image signal, a halftoning unit that performs a halftoning process on the multi-valued image and outputs a halftone image signal, and the smoothing image. An output unit that outputs a final image signal by combining the signal and the halftone image signal. According to this smoothing device, by performing smoothing on a binary image in which a smoothing target and a non-smoothing target are distinguished, a smoothed image signal in which only the smoothing target is smoothed can be obtained. By combining this smoothing image signal with a separately prepared half-toning image signal, it is possible to obtain a final image signal that can reproduce natural images naturally and characters and graphics with clear outlines. Can be.

In a preferred embodiment, a binary image as described above is generated based on attribute data specifying an image element to be smoothed in a multi-valued image. The attribute data also specifies a region where correction by the smoothing process can be performed and a region where the correction should not be performed. Based on this information, smoothing correction is performed on the region that should not be corrected. A smoothed image signal is generated so as not to be applied. Further, a smoothed image signal is generated based on the image value of the multi-valued image so that the image value of the smoothing target is reflected in the image value indicated by the smoothed image signal.

The present invention can be implemented by dedicated hardware, a computer, or a combination thereof. Also, one like a printer
It can be implemented in a single device or by multiple devices such as a host and a printer connected to it. When a computer is used, the computer program can be installed or loaded on the computer through various media such as a disk storage, a semiconductor memory, and a communication network.

[0012]

FIG. 1 shows the overall configuration of a smoothing apparatus according to an embodiment of the present invention applied to a page printer.

A page image memory 1 such as a DRAM stores bitmap image data for at least one page or one band obtained by dividing one page. This bitmap map image data is 1
This is multi-valued image data in which one pixel value for a color is represented by an 8-bit (1 byte) word. In the case of a color image, image data is usually configured as a set of planes of three or four color components, but the configuration shown in FIG. 1 is a configuration for processing a plane of one color component. Of course, if different color component planes are to be processed serially, only one device shown in FIG. 1 is required, but if processing is performed in parallel, a plurality of devices shown in FIG. 1 are required.

Since the pixel value of each color component of the image data in the page image memory 1 is a one-byte word,
A density within a range of 0 to 255 having a resolution of 256 steps is shown. The source data for this image data is
Normally, the image data is supplied from a host connected to the page printer. Therefore, the pixel value of each color component of the image data is determined by what kind of image is created by the host. However, in practice, (1) the pixel value of the natural image: 0 to 254 (2) at least one of the character and the graphics
In many cases, the pixel value of the color component is 255. The present embodiment is configured as follows so as to perform smoothing of characters and graphics paying attention to this point.

Each pixel value (1 byte word) of an image in the page image memory 1 is read out in the order of raster scan and input to an 8-raster image register 2 using, for example, an SRAM. The eight raster image register 2 includes four raster byte registers 25 to 28 and four raster bit registers 21 to 24. Each of the raster byte registers 25 to 28 can store a byte word of a pixel for one raster. Each of the raster bit registers 21 to 24 can store a bit word of a pixel of one raster which has been binarized by a binarization circuit 3 described later. Here, "raster" means a main scanning line when raster-scanning a page. For example, a resolution of 600 dpi and a size of A4 (8.
For a 27 inch × 11.69 inch page, one raster is composed of 8.27 × 600 = about 4960 pixels.

Each of the raster byte registers 25 to 28 and the raster bit registers 21 to 24 functions as a shift register or a FIFO memory, and outputs the oldest written pixel value each time a new pixel value is written. Each pixel value (1 byte word) read from the page image memory 1 is first written to the fourth raster byte register 28 shown at the bottom. Each pixel byte word output from the fourth raster byte register 28 is, on the one hand, written to the third raster byte register 27 and, on the other hand, input to the binarization circuit 3, where the fourth binarization unit is provided. It is binarized by 34 and converted to a 1-bit word. Each pixel byte word output from the third raster byte register 27 is
, And on the other hand, is binarized by a third binarization unit 33 in the binarization circuit 3 and converted into a 1-bit word. Each pixel byte word output from the second raster byte register 26 is:
On the other hand, it is written into the first raster byte register 25,
On the other hand, it is binarized by a second binarization unit 32 in the binarization circuit 3 and converted into a 1-bit word. Each pixel byte word output from the first raster byte register 25 is converted to a first binarization unit 31 in the binarization circuit 3.
And converted into a 1-bit word.

Each pixel value (1-bit word) binarized by the first binarization unit 31 is written to a fourth raster bit register 24. Each pixel bit word output from the fourth raster bit register 24 is written to the third raster bit register 23.
Each pixel bit word output from the third raster bit register 23 is stored in the second raster bit register 23.
Is written to. This second raster bit register 22
Are written to the first raster bit register 21. In this way, the eight raster image registers 2 store pixel values for eight consecutive rasters.

As described above, the binarizing circuit 3 comprises four 2
Each of the pixel byte words output from the four raster byte registers 25 to 28 is binarized and converted into a 1-bit word. This 2
The quantification uses a predetermined threshold value, and converts a pixel byte word having a value equal to or higher than the threshold value to a bit word “1” and a pixel byte word having a value lower than the threshold value to a bit word “0”.

The purpose of this binarization is to extract only image elements to be smoothed such as characters and graphics from the image, distinguishing them from non-smoothed areas such as natural images. For this purpose, the pixel value of the character and graphics to be smoothed is converted to “1” in the threshold for binarization.
A value that can be converted to “0” is selected as a pixel value of an area that is not a smoothing target. Typically, the highest value of the pixel byte word "255" (or its neighbors,
For example, “250” or “240”) is selected as the binarization threshold. As described above, in many images, the characters and graphics to be smoothed have a maximum pixel value of “255” for at least one color component, and those of natural images that are not to be smoothed are “0”.
This is because it is distributed to “254”. However, since some images do not correspond to this, an appropriate threshold may be selected according to the actual situation of the images or the user's preference. For example, when the pixel value of the natural image is “0” to
It is distributed in the range of “200”, and the pixel values of the character and graphics to be smoothed are “100” to “25”.
For an image in the range of "5", if the user desires to reliably extract all the characters and graphics and perform smoothing processing, for example, "80" to "80"
The threshold value should be selected from a range of about “150”. On the other hand, if the user attaches importance to not affecting the natural image by smoothing, for example, “150” to “2”
The threshold may be selected from a range of about 50 ". In any case, image elements to be smoothed can be substantially distinguished from other non-smoothed image elements (that is, even if they cannot be completely distinguished, they can be roughly distinguished within a practically acceptable range). A threshold is chosen.

The binarization threshold may be set manually by a user, or may be automatically set by a printer driver or a printer. As a method of automatic setting,
For example, when the user designates one or a plurality of image element regions to be smoothed and other image element regions on the application screen of the host, the two are effectively determined based on the pixel value of the designated image element. A method is conceivable in which the printer driver or the printer automatically determines a distinguishable threshold.

The pixel bit words of the fifth to eighth rasters output from the four binarization units 31 to 34 of the binarization circuit 3 shown in FIG. Out of the eight nine-stage shift registers 51 to 58 to be input to the fifth to eighth shift registers 55 to 58, respectively. Further, the pixel bit words of the first to fourth rasters output from the first to fourth raster bit registers 21 to 24 of the 8 raster image register 2 are passed through the delay circuit 4 to the 8 × 9 pixel register. 5th 1st to 4th
Are input to the nine-stage shift registers 521 to 54, respectively. The delay circuit 4 includes four delay units 41 to 4
4 and the first from the raster bit registers 21 to 24.
For the pixel bit words of the fourth to fourth rasters, the binarizing circuit 3 gives a delay to the same length received by the pixel values of the fifth to eighth rasters, and in the sub-scanning direction of the raster scan on the image. The bit words of the pixels arranged in a line are simultaneously input to the 8 × 9 pixel register 5. Therefore, the 8 × 9 pixel register 5 stores a bit word of 72 pixels in the 8 × 9 pixel area on the image. The 8 × 9 pixel area moves one pixel at a time over the image in a raster scan system. Hereinafter, the pixel at the center of the 8 × 9 pixel area, that is, the pixel corresponding to the fifth stage 100 of the fifth shift register 55 of the 8 × 9 pixel register 5 is referred to as “pixel of interest”.

The smoothing circuit 6 fetches a 72-pixel bit word in the 8 × 9 pixel area in the 8 × 9 pixel register 5 and, based on the fetched image element (bit word value “1”) in the 8 × 9 pixel area. ), The drawing laser pulse width for the pixel of interest when the contour line is smoothed is calculated, and a pulse width modulated laser drive pulse signal is generated according to the calculation result. Here, the relationship between the pulse width of the drawing laser and the size of the colorant dot formed in one pixel region by the drawing laser is as shown in FIG. 2, for example. That is, according to the laser pulse having a pulse width corresponding to one pixel full size indicated by reference numeral 201, one pixel region 2
10, full-size dots 211 are formed. Also,
With a laser pulse that is slightly shorter than the full pulse width as indicated by reference numeral 202, a slightly smaller dot 212 is formed in one pixel region 210. Shorter laser pulse 203
According to the above, the dots 213 are further formed. The dot size can be adjusted by changing the laser pulse width in this manner.

FIG. 3 shows an example in which smoothing is performed using this principle. FIG. 3A shows a binarized image of the 8 × 9 pixel area before smoothing, and FIG. 3B shows an image after smoothing. In the image before smoothing in FIG. 3A, pixels corresponding to image elements to be smoothed (oblique straight lines in the example of the figure) are indicated by hatching, but those pixels are stored in the 8 × 9 pixel register 5. Has a bit word value "1".
Other pixels are outlined, but they are 8
It has a bit word value “0” on the × 9 pixel register 5. The smoothing circuit 6 shown in FIG. 1 reads the binarized image data representing the image shown in FIG. 3A from the 8 × 9 pixel register 5, and outputs the image after smoothing shown in FIG. The drawing laser pulse width for the pixel of interest 300 is determined so that can be drawn. As a result, in the example of FIG. 3B, the drawing laser pulse for the target pixel 300 is determined to be the pulse 202 shown in FIG. The smoothing circuit 6 for performing such processing
Various specific structures are known.

The half-toning circuit 7 shown in FIG.
And performs halftoning processing, that is, processing for converting the density indicated by the pixel byte word into the presence / absence and size of a dot that makes the human eyes feel the same density. As a result of this half-toning process, the half-toning circuit 7 determines the drawing laser pulse width for the target pixel, and outputs the laser drive pulse for the target pixel pulse-width-modulated according to the determination result from the smoothing circuit 6 to the target pixel. Are output in synchronism with the output of the laser drive pulse with respect to.

The OR circuit 8 performs a logical OR operation on the laser drive pulse signal for the pixel of interest output from the smoothing circuit 6 and the laser drive pulse signal for the pixel of interest output from the half-toning circuit 7 to form a drawing laser driver ( (Not shown) to generate a drawing laser pulse.

With the above configuration, among the image elements in the page image, the image element whose bit value is set to "1" in the binarization processing, that is, mainly the character and the graphics are selectively smoothed. . When the character or graphics is a primary color (a color composed of only one color component with a density of 255) or a color composed of two or three color components with a density of 255, it is an intermediate color including a color component with a density of less than 255. In some cases, if the density of at least one color component is 255, a smoothing effect can be obtained.

FIG. 4 shows a state in which a diagonal straight line of an intermediate color is smoothed in this embodiment. FIG. 4A shows a state before smoothing, and FIG. 4B shows a state after smoothing.

The oblique line 401 before smoothing in FIG.
Is, for example, a magenta dot 402 of a full pixel size corresponding to a density of 100% (value 255) and a density of 25%
It is drawn from a ド ッ ト pixel size yellow dot 403 corresponding to (value 64). When this is passed through the above-described embodiment, the smoothing process is not applied to the yellow dot 403 having the value 64, but the smoothing process is applied to the magenta dot 402 having the value 255.
The image is printed as indicated by the oblique line 404 in FIG. Therefore, although the smoothing process is not applied to the yellow dot 403 having the value of 64, the effect of the smoothing appears as a whole.

FIG. 5 shows a second embodiment of the present invention.
The same components as those of the first embodiment shown in FIG. 1 are denoted by the same reference numerals, and duplicate description will be omitted.

The smoothing device according to the second embodiment includes, for example, D in addition to the device configuration of the first embodiment.
A page attribute memory 9 which is a RAM is provided. The page attribute memory 9 stores attribute data indicating the attribute of each pixel of the image stored in the page image memory 1. In other words, the page image is composed of one, three, or four color component planes stored in the page image memory 1 and an attribute plane stored in the page attribute memory 9. In this embodiment, the `` attribute '' of each pixel means whether the pixel is an image element to be smoothed, and whether a drawing laser pulse for smoothing correction may be output on the pixel, (Of course, other attributes may be added, and only one point of whether or not it is a smoothing target may be used). The attribute data is 2 bits per pixel. (1) 1st bit: “1” = subject to smoothing, “0” = no (2) 2nd bit: “1” = correction pulse may be output , “0” = No. For example, as shown in FIG. 6, if the attribute data of each pixel included in the region 501 indicated by fine hatching is “11”, this region 501 is a region to be smoothed and is a region where a smoothing correction pulse can be output. It means there is. In addition, regions 502 and 5 indicated by rough hatching on both sides of the smoothing target 501.
If the attribute data of each pixel included in 03, 504, and 505 is “01”, these areas 502, 503, and 50
4, 505 means that a corrected drawing laser pulse as a result of smoothing the smoothing target 501 may be output. In addition, a white area 506 outside the area,
If the attribute data of the included pixel is “00” in 507, it means that those areas 506 and 507 are not smoothing targets and no correction pulse should be output. Based on the attribute data illustrated in FIG. 6,
Smoothing as illustrated in FIG. 3 can be performed.

Such attribute data may be automatically generated by a host or a printer, or may be generated by a user on a host application. For example, in the source data of an image created by an application or in print data sent from a host to a printer, characters are character lacquer codes, graphics are vector data or function calls, and natural images are bitmap data. If different image elements are represented by different types of data, the host application, the printer driver or the printer's imaging process can automatically distinguish characters, graphics, and natural images based on the data format. Attribute data can be created. In addition, when users create and edit documents, paintings, and photographs on the application,
The user can also create attribute data on an application by specifically designating a smoothing target, or specifically designating a region where correction pulses can be or cannot be emitted. .

Referring again to FIG. 5, in synchronization with the reading of the byte word of each pixel from page image memory 1, the attribute data (2-bit word) of the corresponding pixel is read from page attribute memory 9. For example, SRAM
Is written to the 4-raster attribute register 10 using. 4
The raster attribute register 10 includes four raster attribute registers 101 to 101 each capable of storing attribute data for one raster.
104. Four raster attribute registers 101-1
04 operate as a shift register or a FIFO memory, and when attribute data of a new pixel is written,
The attribute data of the oldest written pixel is output.

The attribute data (2-bit word) of each pixel read from the page attribute memory 9 is first written to the fourth raster attribute register 104 shown at the bottom. Each pixel attribute data output from the fourth raster attribute register 104 is written to the third raster attribute register 103 on the one hand, and is input to the fourth delay unit 114 in the delay circuit 11 on the other hand. Each pixel attribute data output from the third raster attribute register 103 is written to the second raster attribute register 102 on the one hand, and the third delay unit 113 in the delay circuit 11 on the other hand.
Is input to Each pixel attribute data output from the second raster attribute register 102 is written to the first raster attribute register 101 on the one hand, and is input to the second delay unit 112 in the delay circuit 11 on the other hand. Each pixel attribute data output from the first raster attribute register 101 is input to a first delay unit 111 in the delay circuit 11.

As described above, the delay circuit 11 has four delay units 111 to 114, and each pixel attribute data output from the four raster attribute registers 101 to 104 is converted by the binarization circuit 3 Delay by the binarization processing time. Of the four raster attribute data output from the four delay units 111 to 114, the first bit is used as a control signal by a selector provided between the binarization circuit 3 and the 8 × 9 pixel register 5. Added.

The selector 12 has four selector units 1
21 to 124, and these selector units 121
To 124 receive the pixel bit words of the four rasters and the “0” value bit word output from the binarization circuit 3 as signals to be selected, and add the pixels added from the delay units 111 to 114 described above. In response to the signal of the first bit of the attribute data, the pixel bit word is selected if the signal value of the first bit is “1”, and “0” if the signal value of the first bit is “0”. A value bid word is selected and output to the 8 × 9 pixel register 5. Also, the first
The bit word output from the selector unit 121 is also written in the fourth raster bit register 24 in the eight raster image register 2.

The delay circuit 4 generates a delay corresponding to the processing time of the binarization circuit 3 in the first embodiment shown in FIG. 1, but in the second embodiment, the binarization circuit 3 And a delay corresponding to the sum of the processing time at the selector 12 and the processing time at the selector 12, so that the eight raster pixels aligned in the sub-scanning direction are simultaneously written into the 8 × 9 pixel register 5. To

With the above configuration, the 8 × 9 pixel register 5
Is written with a binarized pixel value only for a pixel to be smoothed in which the first bit of the attribute data is "1", and a "0" value is uniformly written for a pixel not to be smoothed. Become.

By the way, in the first embodiment shown in FIG. 1, the binarizing circuit 3 distinguishes between a smoothing target and a non-smoothing one. Is fulfilled, the role of the binary circuit 3 is different from that of the first embodiment. That is, the binarization circuit 3 plays a role of limiting the object to be actually smoothed from the smoothing objects specified by the attribute data by the pixel value. For example, if all characters and all graphics are specified as smoothing targets by attribute data, the binarization circuit 3
For example, if "200" is set as the threshold value, the actual smoothing process will be performed only on the pixel values of "200" or more among all the characters and all the graphics. . This smoothing target limitation function is used when it is desired to exclude a specific image element having a relatively small pixel value from among the smoothing targets specified by attribute data, or when creating attribute data, for example, a black character on a white background. This can be used, for example, when a specific area in which text is written is specified as a smoothing target, and when printing is desired, only black text in that area is to be smoothed. On the other hand, if the smoothing process can be actually performed on all of the smoothing targets specified by the attribute data, a threshold “0” is set in the binarization circuit 3,
Alternatively, the binarizing circuit 3 and the selector 12 are removed, and the first bit of the pixel attribute data from the delay circuit 11 is directly set to 8 bits.
The data may be written into the × 9 pixel register 5.

As described with reference to the first embodiment, the smoothing circuit 6 converts the smoothed laser drive pulse signal for the target pixel based on the data of 72 pixels in the 8 × 9 pixel register 5 as described above. Generate. The laser drive pulse output from the smoothing circuit 6 is input to the pulse adjustment circuit 14. The pulse adjusting circuit 14 inputs the pixel byte word output from the fifth raster register 25 of the eight raster image register 2 through the delay circuit 13 and, based on the value of the pixel byte word, outputs the laser driving pulse from the smoothing circuit 6. Is adjusted. That is, since the pulse width of the laser drive pulse output from the smoothing circuit 6 is the pulse width when the pixel value to be smoothed is “255”, the pulse width is output from the fifth raster register 25. The width is adjusted to match the pixel value of the actual pixel to be smoothed. here,
The delay circuit 13 adjusts the timing so that the pixel byte word of the target pixel is always input to the pulse adjustment circuit 14. Since the pulse width adjustment by the pulse width adjustment circuit 14 may reduce the effect of smoothing, the pulse width adjustment circuit 14 may be omitted. Also,
The pulse width adjustment circuit 14 is removed, and in another method,
Means for reflecting the actual pixel value of the smoothing target on the image after smoothing may be provided.

The laser driving pulse output from the pulse width adjusting circuit 14 is input to the gate 15 next. Gate 1
5 receives the second bit of the pixel attribute data output from the first delay unit 11 of the delay circuit 11 through the delay circuit 16, and operates the laser drive only when the value of the second bit is "1". The pulse is passed and output to the OR circuit 8. Here, the delay circuit 16 adjusts the timing so that the second bit of the attribute data of the pixel of interest always enters the gate 15. By the operation of the gate 15, the correction pulse of the target pixel from the pulse adjustment circuit 14 enters the OR circuit 8 only when the target pixel is in a region where the correction pulse can be output.

The laser drive pulse of the pixel of interest after the half-toning generated by the half-toning circuit 7 is passed through the delay circuit 17 to the OR circuit 8 at the same timing as the laser drive pulse of the pixel of interest from the gate 15. Is entered. Then, the laser drive pulse logically ORed by the OR circuit 8 is sent to the drawing laser driver.

FIG. 7 shows an example of the result of the smoothing process according to the second embodiment.

This example is an image in which an oblique straight line runs on a background of 50% density, and FIG. 6A shows a case where the image is printed without smoothing processing. FIG. 6B shows attribute data. The area indicated by fine hatching is a slant line to be smoothed, and the areas indicated by rough hatching on both sides thereof are areas where correction pulses can be output. FIG. 6C shows a print example after performing a smoothing process using the attribute data, in which oblique lines are smoothed.

Although one embodiment of the present invention has been described above, these embodiments are merely examples for describing the present invention, and are not intended to limit the present invention only to these embodiments. Therefore, the present invention can be implemented in various forms other than the above-described embodiment. The present invention can be applied not only to a laser page printer but also to any other type of ink jet printer, a display device for displaying an image on a screen, and the like. In the above embodiment, all the processing is performed by hardware, but at least a part of the processing may be performed by computer software.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a first embodiment of the present invention.

FIG. 2 is a diagram illustrating a relationship between a pulse width of a laser driving pulse and a dot size.

FIG. 3 is a diagram showing an example of a result of smoothing.

FIG. 4 is an example showing a smoothing result of an intermediate color.

FIG. 5 is a block diagram showing a configuration of a second embodiment of the present invention.

FIG. 6 is a diagram showing an example of attribute data.

FIG. 7 is a diagram showing an example of a smoothing result.

[Explanation of symbols]

 1 page image memory 2 8 raster image memory 3 binarization circuit 5 8 × 9 pixel register 6 smoothing circuit 7 half toning circuit 8 OR circuit 10 4 raster attribute register 12 selector 14 pulse adjustment circuit 15 gate

Claims (15)

[Claims] 1. A binarizing unit for binarizing a multi-valued image, a smoothing unit for performing a smoothing process on a binarized image output from the binarizing unit and outputting a smoothed image signal, A halftoning unit that outputs a halftone image signal by performing a halftoning process on the value image; and an output unit that outputs a final image signal by combining the smoothing image signal and the halftone image signal. Value image smoothing device. 2. The binarization unit according to claim 1, wherein the binarization unit performs the binarization using a threshold value that can substantially distinguish an image element to be smoothed from an image element that is not to be smoothed in the multi-valued image. The smoothing device as described. 3. The smoothing device according to claim 2, wherein the threshold value is a value near a maximum value of a multi-valued image value of the multi-valued image. 4. The smoothing device according to claim 2, wherein said threshold value is variable. 5. A binary image generating unit for generating a binary image that distinguishes an image element to be smoothed from an image element not to be smoothed in a multi-valued image, and performs a smoothing process on the binary image. A smoothing unit that outputs a smoothing image signal, a halftoning unit that performs a halftoning process on the multi-valued image and outputs a halftone image signal, and combines the smoothed image signal and the halftone image signal to obtain a final image. An output section for outputting an image signal; 6. The smoothing device according to claim 5, wherein the binary image generation unit generates the binary image based on attribute data specifying an image element to be smoothed in the multi-valued image. 7. The method according to claim 1, wherein the smoothing unit performs the correction on the unacceptable area based on attribute data that specifies an area in the multi-valued image that may be corrected by the smoothing process and an area on which the correction is not performed. 6. The smoothing device according to claim 5, wherein the smoothing image signal is generated such that the image signal is not applied. 8. The smoothing device according to claim 6, wherein the attribute data is included in the multi-valued image. 9. The smoothing unit according to claim 1, wherein the smoothing unit generates the smoothed image signal based on an image value of the multi-valued image such that an image value of the smoothing target is reflected in an image value indicated by the smoothed image signal. The smoothing device according to claim 5. 10. A binarizing unit for binarizing a multi-valued image, a smoothing unit for performing a smoothing process on a binarized image output from the binarizing unit and outputting a smoothed image signal, A halftoning unit that performs a halftoning process on the value image to output a halftone image signal; an output unit that combines the smoothing image signal and the halftone image signal to output a final image signal; And an image forming unit for reproducing the multi-valued image based on a simple image signal. 11. A binary image generating section for generating a binary image in which image elements to be smoothed and image elements not to be smoothed in a multi-valued image are distinguished, and performing a smoothing process on the binary image. A smoothing unit that outputs a smoothing image signal, a halftoning unit that performs a halftoning process on the multi-valued image and outputs a halftone image signal, and combines the smoothed image signal and the halftone image signal to obtain a final image. A multi-value image forming apparatus, comprising: an output unit that outputs an image signal; and an image forming unit that reproduces the multi-value image based on the final image signal. 12. A step of binarizing the multi-valued image, a step of performing a smoothing process on the binarized image output from the binarizing step and outputting a smoothed image signal, A multi-valued image smoothing method, comprising: performing a toning process to output a halftone image signal; and combining the smoothed image signal and the halftone image signal to output a final image signal. 13. A step of generating a binary image in which an image element to be smoothed and an image element not to be smoothed in a multi-valued image are differentiated; and performing a smoothing process on the binarized image to generate a smoothed image signal. Outputting, performing a halftoning process on the multi-valued image to output a halftone image signal, and outputting a final image signal by combining the smoothed image signal and the halftone image signal. A multi-valued image smoothing method having 14. A step of binarizing a multi-valued image, a step of performing a smoothing process on a binarized image output from the binarizing step and outputting a smoothed image signal, A step of outputting a halftone image signal by performing a toning process, and a step of outputting a final image signal by combining the smoothing image signal and the halftone image signal. A computer-readable recording medium storing a program to be executed by a computer. 15. A step of generating a binary image in which image elements to be smoothed and image elements not to be smoothed in a multi-valued image are differentiated; and performing a smoothing process on the binary image to generate a smoothed image signal. Outputting, performing a halftoning process on the multi-valued image to output a halftone image signal, and outputting a final image signal by combining the smoothed image signal and the halftone image signal. A computer-readable recording medium storing a program for causing a computer to perform the method of smoothing a multi-valued image having the method.