JPH041872A - Graphic processor - Google Patents

Abstract

PURPOSE:To obtain an area rate at a high speed without performing the division of subpixels nor counting the number of area-filled subpixels by providing an edge deciding means and a 2nd arithmetic means which calculates the area rate of each edge part picture element based on the coordinate values of two intersection, the area of a triangle, and the type of the edge respectively. CONSTITUTION:A 1st arithmetic means calculates the area S of a triangle from the difference between the coordinate values X0 and X1 of two intersections when the vector data cross the scan lines Y0 and Y1. An edge deciding means decides the type of an edge, i.e., decides whether the vector data show the right or left edge. Then a 2nd arithmetic means calculates the area rate of each unit part picture element based on the values X0 and X1, the area S, and the type of the edge. Thus the area S is obtained at a high speed. In particular, the area rate is accurately obtained in the form of an ideal numerical value at a normal vector edge part having no edge point. As a result, the area rate can be obtained at a high speed without performing the division of subpixels nor counting the number of area-filled subpixels.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a graphic processing device that performs anti-aliasing processing to remove jagged edges of an output image.
More specifically, the present invention relates to a graphic processing device that can perform anti-aliasing processing at high speed.

[Conventional technology]

In the field of computer graphics, when displaying an image on a CRT, which is an output medium, a technique called anti-aliasing processing is used to make the displayed image more beautiful. This process is performed on the jagged part (called alias) on the stairs as shown in Fig. 25(a).
The visually displayed image is shown in Figure 25 ('b
) as shown in the figure.

In conventional graphic processing apparatuses, (1) uniform averaging method, (2) weighted averaging method, (2) convolution integral method, etc. are generally applied as anti-aliasing processing methods.

■The uniform averaging method calculates each pixel (picture element) to N*M (N,
After decomposing the image into sub-pixels (M is a natural number) and performing rask calculation at high resolution, the brightness of each pixel is determined by taking the average of N*M sub-pixels. Figure 26(a)
, (b), anti-aliasing processing using the uniform averaging method will be specifically explained. If the edge of the image overlaps a certain pixel (here, it is assumed that the image is connected to the bottom right of the diagonal line), and when anti-aliasing processing is not performed, as shown in Figure (a),
The brightness kid of this pixel is assigned the highest brightness of the displayable gradations (for example, kid·255 for 256 gradations). When performing anti-aliasing processing on this pixel using the uniform averaging method with N=M=7, the same figure (
As shown in b), the pixel is decomposed into 7*7 sub-vixels and the number of sub-vixels covered by the image is counted. The count number (2 days) is divided by the total number of sub-vixels in one pixel (49 in this case), normalized (averaged), and multiplied by the maximum brightness (255) to calculate the brightness of that pixel. In this way, in the uniform averaging method, the brightness of each pixel is determined by taking into consideration how the image covers each pixel.

■ Weighting is an averaging method The weighting and averaging method is a partial modification of the uniform averaging method. In contrast to the weighted averaging method, which gives each sub-vixel a weight so that the influence on the brightness kid of that sub-vixel differs depending on which sub-vixel the image falls on. ing. Note that the weight at this time is given using a filter.

Referring to FIGS. 27(a) and (b), an example is shown in which the same image data as in FIG. 26(a) is subjected to the same dividing method (N=M=7> and weighting is the averaging method). .

FIG. 27(a) shows a filter (here, conef
ilter), and the corresponding sub-vixel is given the same weight as this characteristic. For example, the weight of the sub-vixel in the upper right corner is 2. When an image is applied to each sub-vixel, the weight value given by the filter characteristics becomes the count value of that sub-vixel. Same figure (c)
In the figure, the display pattern of the image is changed depending on the weight of the sub-vixels. In this case, if we count the weighted sub-vixels of the image, we get
It becomes 199. This value is divided by the sum of the filter values (336 in this case) corresponding to uniform averaging, averaged, and multiplied by the maximum brightness to calculate the brightness of this pixel. In addition, as a filter, Fig. 28 (a),
, (C), and (d) are known.

■Convolution integral method The convolution integral method is a method that refers to the appearance of surrounding pixels when determining the brightness of one pixel. That is, the area around one pixel whose brightness is to be determined is N
' The target pixel is indicated by 2901.

The image continues to the bottom right of the diagonal line, and the sub-vixels painted in black are the sub-vixels that are counted. Each pixel is divided into 4*4. Therefore, in this case, a $12 filter will be used.

This method has the effect of removing high frequency components contained in a vector image.

On the other hand, a publishing system using a personal computer,
With the spread of so-called DTP (desk top publishing), systems for printing vector images, such as those used in computer graphics, have become widely used. A typical example is a system using Adobe's Post Script. PostScript is a page description language.
criptionLanguage: Hereafter, PDL
It belongs to a language genre called ``Japanese language'', and describes the form of the contents of a single document, including the text (letter part), graphics, and their arrangement and appearance. It is a programming language for systems such as this, and vector fonts are used as character fonts in such systems. Therefore, even if the characters are scaled, the print quality can be significantly improved compared to systems using bitmap fonts (for example, conventional word processors).
Another advantage is that character fonts, graphics, and images can be mixed and printed.

However, the resolution of the laser printers used in these systems is at most 240dpi to 400dp.
There are many i, computer graphics C
Similar to RT display, there is a problem in that aliasing occurs due to the low resolution. For this reason, even in printing using a laser printer, there is a need to perform anti-aliasing processing to improve the quality of the printed image.

[Problem to be solved by the invention]

However, according to a graphic processing device that applies a conventional anti-aliasing processing method, one pixel is divided into a plurality of sub-vixels (for example, 49 sub-vixels), and the number of filled-in sub-vixels is counted to calculate the area. Since the ratio (brightness) is calculated, it takes time to calculate the area ratio, which poses a problem in that it becomes an impediment to improving display speed or printing speed. In particular, the convolution method has the problem that it requires a large amount of calculation and affects multiple pixels, making it difficult to improve the processing speed.

The present invention has been made in view of the above, and it is an object of the present invention to obtain the area ratio at high speed without performing sub-vixel division or counting the number of filled pixels.

[Means to solve the problem]

In order to achieve the above object, the present invention adjusts the output of pixels at the edges of vector data based on the area ratio to be filled, and performs anti-aliasing processing to smoothly express jaggedness (alias) at the edges of the output image. A first calculation means calculates the area of a triangle based on the difference in coordinate values of two intersection points when the vector data crosses a scan line, and a first calculation means calculates the area of the triangle, and determines whether the edge of the vector data is the left or right edge. an edge determination means for determining the type of edge; coordinate values of two intersection points; area of a triangle;
and a second calculation means for calculating the area ratio of each edge pixel based on the type of edge.

In addition to the above-described configuration, the sub-pixel division means calculates the area ratio of the edge pixel by sub-pixel division, and when the vector data has an end point on the scan line, the sub-pixel division means divides the edge pixel. The present invention provides a graphic processing device that calculates area ratio.

[Effect]

In the graphic processing device of the present invention, the first calculation means calculates the area of a vector-specific triangular part that is common to the edge portion of each scan line, and the second calculation means calculates the area of the vector-specific triangular part that is common to the edge portion of each scan line. coordinate values of two intersection points,
Based on the area of the triangle and the determination result (type of edge) of the edge determining means, the area ratio of each edge pixel on the scan line is determined. Further, when the vector data has an end point on the scan line, the area ratio of the edge portion pixel is determined by the subpixel dividing means.

〔Example〕

Hereinafter, an image forming system incorporating the graphic processing device of the present invention as a PDL controller will be described as an example. and operation.

A detailed explanation will be given in the following order: 1) the configuration of the image processing device, 2) the configuration and operation of the multi-valued color laser printer, and 2) the multi-valued drive of the driver.

■Overview of anti-aliasing processing The graphic processing device of the present invention (hereinafter referred to as PDL controller) is capable of processing when a straight line with a certain slope (straight line vector) intersects multiple parallel lines (scanning lines). , the area of the edge pixel can be easily determined by utilizing the fact that each parallel line has a common triangular part. Hereinafter, with reference to FIG. 1(a) to 1, the area ratio of the edge area pixels is calculated using the area of the triangular part unique to the vector that is common to the edge area of each scan line, which is the main part of the present invention. The principle of anti-aliasing processing will be explained in detail.

As shown in Figure 1(a), the vector data is Yo and Y.
When crossing one scan line, two intersections are made on the Y side and the Y1 side, as shown in the figure. This is because the scan line has a thickness of one pixel (that is, the thickness of YlYo). Hereinafter, among the two intersections described above, the intersection near the image area will be referred to as X0, and the intersection farther from the image area will be referred to as Xl.

Here, taking as an example vector data (straight line vector) with an inclination of 30 degrees at the left edge, a common triangular portion unique to the vector formed on each scan line will be described.

As described above, since each scan line is a straight line running in parallel and having a thickness of one pixel, triangles of the same size are formed by intersecting the vector data, as shown in the figure. Since each pixel is a 1×1 square, the area S of the triangle of these common parts is calculated using the two intersection points when the vector data crosses the scan line.
It can be easily calculated using the following formula.

S= 1/2

As shown in FIGS. 1(b) and 1(c), when a common triangle extends over a plurality of pixels on the scan line, the area S is divided into the plurality of pixels using the similarity relationship of the triangles. At this time, the area is distributed from the edge area pixels away from the image area (in the same figures), edge area pixels including Xl in (C), to the pixels adjacent to the image area (in the figure (5),
In (C), the processing is performed sequentially toward edge pixels (including χ0).

First, the case of the left edge will be explained with reference to FIG. 1(G). If the area of the edge pixel (1) with Xl is Sl, then the following relationship holds true from the triangular similarity relationship. Here, cl indicates the decimal part of the intersection X1, and 1-c corresponds to the base of the triangle of the edge pixel (1) as shown.

Therefore, the area S1 of the edge pixel (1) is the area S,2
intersection point (Xo.

It can be easily calculated from the coordinate values of Xl and the decimal part c+).

Next, edge pixel (2) adjacent to edge pixel (1)
) can be determined by the following equation using the area S1 of the edge pixel (1).

This calculates the area of a triangle whose base is 2-01 based on the similarity relationship from the triangular area S of the common part, and calculates the area of the edge pixel (2) by excluding the triangular area S1. .

Finally, since the edge pixel (3) adjacent to the image area (edge pixel including XO) includes a rectangular part, the area S3 of the common part is triangular area S The area of the triangle whose base is 2-01 is removed from , and the area of the rectangle of 1-00 is added to the value. Here, co indicates the decimal part of the intersection X0.

The areas S, , S, and S3 obtained in this way indicate the actual areas of the image portions of edge pixels (1), (2), and (3), respectively, and the ideal value can be obtained as the area ratio. Moreover, since the area is determined without dividing the pixel into sub-vixels, high-speed processing is possible.

FIG. 1(b) shows the case where the common triangle spans three pixels, but if the common triangle spans three or more pixels, it can be determined by the following general formula.

The area S3 of the edge pixel (3) adjacent to the SI image portion, which is the area of the pixel farthest from the i1 image portion, can be determined by the area Sf of the edge pixel adjacent to the i11 image portion.

Furthermore, its general formula is as follows.

i8 Area of the pixel farthest from the image portion S1 Next, with reference to FIG. 1(C), the determination of the area for the right edge will be explained. If the area of a certain edge pixel (1) of XI is 81, the area Sf of the edge pixel adjacent to the ii0 image part holds true from the triangular similarity relationship and can be easily calculated.

Furthermore, the edge pixel (2) adjacent to the edge pixel (1)
), as explained above, calculate the area S of the triangle of the common part from the coordinate values of the two intersection points X, , X in the above general formula
By performing calculations by substituting o,

Note that if there is an end point (start point or end point of a vector) on the scan line as shown in FIG. 1(d), the above equation cannot be applied because there is no common triangular part. In such a case, for example, an approximate area ratio is calculated using a conventional sub-pixel division method. However, since the probability that an end point of a vector exists on a scan line in boat fishing is smaller than the probability that a common triangular part is formed, most of the edge pixels of the scan line can be can be obtained, and speeding up can be achieved.

■Block diagram of the image forming system The image forming system of this embodiment uses the Page Description Language output from DTP (Desk Top Publishing).
ge (hereinafter referred to as PDL language)) and image information read by an image reading device. The configuration of the image forming system of this embodiment will be described below with reference to FIG.

The image forming system includes a host computer 100 that creates a document written in PDL language (Postscript language is used in this embodiment);
A PDL controller (graphic processing device of the present invention) that applies anti-aliasing processing to the PDL language sent page by page from ) 200, an image reading device 300 that reads image information via an optical system unit, and a PDL controller 200. Or.

An image processing device 400 that inputs the image output from the image reading device 300 and performs image processing (details will be described later), and a multi-value color laser printer that prints the multi-value image data output from the image processing device 400. printer 50
0, PDL controller 2002 image reading device 30
09 Image processing device! 400. and a system control section 600 that controls the multivalued color laser printer 500.

■Configuration and operation of PDL controller FIG. 3 shows the configuration of the PDL controller 200, which includes a receiving device 201 that receives the PDL language sent from the host computer 100, and storage control and control of the PDL language received by the receiving device 201. A CPU 202 that executes anti-aliasing processing, an internal system path 203, a RAM 204 that stores PDL language to be transferred from the receiving device 201 via the internal system bus 203, and a ROM 205 that stores anti-aliasing programs and the like.
, a page memory 206 that stores multivalued RGB image data subjected to anti-aliasing processing, a transmitting device 207 that transfers the RGB image data stored in the page memory 206 to the image processing device 400 , and a system control unit 600 . It consists of an I10 device W2O3 that performs transmission and reception.

Here, the CPU 202 receives the P received by the receiving device 201.
The DL language is transferred to the RAM 204 through the internal system control unit 203 according to the program stored in the ROM 205.
Store in. After that, when the PDL language for one page is received and stored in the RAM 204, anti-aliasing processing is applied to the graphic elements in the RAM 204 based on the flowchart described later, and the multilevel RGB image data is transferred to the plane of the page memory 206. (The page memory 206 includes a plain memory section for R, G, and B,
(consisting of a feature information memory section). The data in the page memory 206 is then sent to the image processing device 400 via the sending device 207.

The operation of the PDL controller 200 will be described below with reference to FIGS. 4(a) and 4(b).

FIG. 4(a) shows a flowchart of processing performed by the CPU 202. As described above, the PDL controller 200 performs anti-aliasing processing on the PDL language sent page by page from the host computer 100, and converts it to red (R).

Develop into three-color images: a (G) and blue (B).

In the PDL language, graphics and characters are all described using vector data, and as the name "page description language" indicates, image information is processed in units of pages. Furthermore, one page is made up of at least one path, with each path being made up of one or more elements (graphic elements and text elements).

First, when PDL language is input, it is determined whether the element is a curved vector, and if it is a curved vector, it is approximated to a straight line vector and registered in the work area as a straight line element (line). This is performed for all graphic and character elements within one path, and linear elements are registered in the work area for each path (processing 1).

Then, the straight line elements of the work area registered in this path unit are sorted by the starting X coordinate of the straight line (processing 2)
.

Next, in process 3, while updating the X coordinate one by one,
Performs filling processing using scanning lines. For example, in Figure 4 (
When performing the path filling process shown in b), the scan line yc to be processed (in this example, a line with a thickness of one pixel, such as the Y, -Y, and scan lines mentioned above, is described as a scan line. However, when a scan line is a straight line with no thickness, it is described as a scan line), and the real value of the X coordinate across the scan line yc (
Xl X2 X3 Xa) shown in Cb) of the same figure and A
It is registered in an ET (Active Edge Table: a table that records the X coordinates of edge portions appearing on a scanning line). Here, the order of the elements registered in the work area is the order in which they were registered in process 1, so they are not necessarily registered in order of decreasing X coordinate across the scanning line yc. For example, in process 1, FIG.
), if a straight line element passing through scanning lines yc and X is processed first, then
The coordinate X is first registered in the AET. Therefore,
After registering the AET, the elements on each side within the AET are sorted in descending order of X coordinate. Then, the first two elements of AET are made into a pair, and the space between them is filled in (specifically, for example, a filling process is performed using a scan line formed by scanning line yc and scanning line yc±1).

Anti-aliasing processing is achieved by adjusting the density and brightness of pixels in the edge portion in accordance with the near-value area ratio in this filling processing. Then remove the processed edge from the AET, update the scanline (update the y coordinate), and so on until all edges in the AET have been processed, in other words, all elements in one pass have been processed. Repeat the process.

Above processing 1. Processing 2. The work in process 3 is executed for each bus and repeated until all passes for one page are completed.

Next, the anti-aliasing process executed during the scan line filling process of process 3 described above will be described in detail with reference to the flowchart of FIG. 4(C).

Here, for example, in process 1 of FIG. 4(a),
), this figure has the following elements.

(a) Five line vectors AB, BC, CD, DE, and EA (represented by real numbers) (b) Color and brightness values inside the figure This figure is created by the above-mentioned operation as shown in Figure 5 (b). , into seven straight line vectors (expressed as real numbers) extending in the main scanning direction. At this time, in this embodiment, the following information is added to the starting points and ending points of the seven straight line vectors. That is, (c) Starting point coordinate values (real number expression) of the vector elements ((a) above) forming the starting point and ending point of the straight line vector (d) Slope information (near ) Characteristic information of the start point and end point of a straight line vector (right edge, left edge, apex of a figure, line of 1 dot or less, intersection of straight lines, etc.) When an edge pixel is detected in the scan line filling process, the Antialiasing processing shown in the flowchart of FIG. 4(C) is executed. Figure 1 (a
), (b), and (C), in the present invention, the input/output coordinates X of vector data with respect to the scan line
11 Based on +χ1, find the area S of the triangle of the common part, and based on the area S and the type of edge, perform the calculation by substituting Xo, By doing this, the area ratio of each edge pixel on the scan line is determined.

First, it is determined whether the end points of the vector are included on the scan line (S401).
). If there is no endpoint of the vector, find the intersection XO+X between the vector data and the scan line (S402
), Next, it is determined whether or not scan line processing of the corresponding vector data is being performed for the first time (5403). If it is the first time, the area S of the triangle of the common part is determined (S404).
, and registers the area S in the AET (5405). If it is not the first time, the area S is already registered in the AET, so the area S is referenced from the AET (5406). Next, the decimal parts CO+CI are extracted from the intersections XO+Xl (S407). After that, determine the left and right sides of the edge (3
408) In the case of a left edge, the area ratio of each edge pixel is determined according to Table 1 (5409), and in the case of a right edge, the area ratio of each edge pixel is determined according to Table 2 (S410). ).

Table 1 Table 2 On the other hand, if there is an end point of the vector on the scan line, the area ratio of the edge pixel is determined by subpixel division (5411).

The process shown in the flowchart of FIG. 4(C) is called as a subroutine when an edge pixel (or edge pixel group) is detected in the scan line filling process of process 3.

The CPU 202 repeats the above processing up to the last pixel of the scanning line (y coordinate), and at the same time updates the contents of (c) above with the information of (d) above. The approximate area ratio of the figure in FIG. 5(a) obtained by the anti-aliasing process in this way has a value as shown in FIG. 6.

Here, if the figure in FIG. 5(a) has a white background color (
The figure color is red (maximum brightness: 25) on top of (maximum brightness = 255)
5), then from the approximate area ratio k (see Figure 6), the brightness value Kr (red) for each color of the figure,
(green) and Kb (blue) are obtained based on the following formula.

Kr= K□Xk + KczX<x h)K*
= KGIxlc + KczX(1k)Kb =
Km + Xk + Kmzx (1k) However, K...
KGI, , respectively indicate the brightness values of the colors (red, green, and blue, respectively) of the figure given in (II) above, and K, ,
,KGZ,indicates the luminance value of each previously painted color. In addition, KHz, KGZ, KII2 is page memory 2
The data in each plane memory section corresponding to RGB of 06 is referred to.

The luminance value of 9+Km to the luminance value Kr+ obtained in this way is stored as RGB image data in the corresponding plane memory section of the page memory 206, as shown in FIGS. 7(a), (b), and (C). Stored. Here, for comparison, RGB image data without antialiasing processing is shown in Figures 8(a), (b), and (C).
).

(2) Configuration of Image Processing Apparatus The configuration of the image processing apparatus 400 will be explained with reference to FIG. 9.

The image processing device 400 is a CC in the image reading device W300.
Black (BK) necessary for recording the three-color image signals read by D7r, 7g, and 7b.

Yellow (Y), magenta (M), and cyan (C)
Convert to each recording signal. In addition, the RGB image data given from the PDL controller 200 described above is similarly converted into black (BK), yellow (Y), magenta (
M) and cyan (C) recording signals. Here, the mode for inputting image signals from the image reading device 300 is set to copy machine mode, and the mode for inputting image signals from the PDL controller 200 is set to copy machine mode.
The mode in which GB image data is input is called a graphics mode.

The image processing device 400 includes CCDs 7r, 7g, and 7b.
The color gradation data obtained by A/D converting the output signal of
a shading correction circuit 401 that performs correction for sensitivity variations, etc. of internal terminal elements of 7r, 7g, and 7b;
a multiplexer 402 that selectively outputs either the color gradation data output from the shading correction circuit 401 or the color gradation data (CRGB image data) output from the PDL controller 200 according to the aforementioned mode;
A γ correction circuit 40 receives the 8-bit data (color gradation data) output from the multiplexer 402, changes the gradation according to the characteristics of the photoreceptor, and outputs it as 6-bit data.
3, (R) and green (G) output from the γ correction circuit 403.
), 6-bit gradation data indicating the gradation of blue (B) and their respective complementary colors, cyan (C) and magenta (M).

A complementary color generation circuit 405 converts into yellow (Y) gradation data (6 bits), and a masking processing circuit 406 performs predetermined masking processing on each Y, M, and C gradation data output from the complementary color generation circuit 405. , a UCR processing/black generation circuit 40 that inputs Y, M, and C gradation data after masking processing and executes UCR processing and black generation processing.
7 and Y output from the UCR processing/black generation circuit 407
,M.

Each 6-bit gradation data of C2 and BK is converted into 3-bit gradation data Yl, Ml, C1, and BKI, and outputted to the laser drive processing section 502 inside the multivalued color laser printer 500. A gradation processing circuit 408;
It is composed of a synchronization control circuit 409 for synchronizing each circuit of the image processing device 400.

Although details will be omitted, the γ correction circuit 403 has a configuration in which the gradation can be arbitrarily changed using an operation button on the console 700.

Further, as an algorithm used in the gradation processing circuit 408, a multi-value dither method, a multi-value error diffusion method, etc. can be applied. For example, a dither matrix of the multi-value dither method is
x3, multilevel color laser printer 500
The number of gradations is the product of 3×3 area gradations and 3-bit (that is, 8 steps) multivalue level, and is 3×3×8=72 (gradations).

Next, the processing of the masking processing circuit 406 and the UCR processing/black generation circuit 407 will be explained.

Generally, the arithmetic expression for the masking process of the masking process circuit 406 is as follows: YA, M, , C,: Data before masking process YO, M,
, C0: Data after masking processing Also, the arithmetic expression for UCR processing of the UCR processing/black generation circuit 407 is generally expressed as follows.

Therefore, in this embodiment, new coefficients are obtained from these equations using the product of both coefficients.

In this embodiment, a new coefficient (such as "all") that performs this masking processing and UCR processing at the same time is calculated in advance, and
Furthermore, using the new coefficients, the masking processing circuit 40
6 scheduled input values Y,,M.

Output value (Y0', etc.) corresponding to Ci (6 bits each):
A value that is the calculation result of the OCR processing/black generation circuit 407 is obtained and stored in a predetermined memory in advance.

Therefore, in this embodiment, the masking processing circuit 406 and U
The CR processing/black generation circuit 407 is composed of a set of ROMs, and the data at the address specified by the inputs Y, M, and C of the masking processing circuit 406 is given as the output of the UCR processing/black generation circuit 407.

In addition, for boat fishing, the masking processing circuit 406 performs Y and M according to the characteristics of the spectral reflection wavelength of the toner for forming a recorded image.
, C signals, and the UCR processing/black generation circuit 407 performs correction for color balance in overlapping toners of each color. UCR processing/black generation circuit 4
07, black component data BK is generated by combining the input three color data of Y, M, and C, and the output Y,
The M and C color component data are corrected to values obtained by subtracting the black component data BK.

In the above configuration, the T correction circuit 403 executes the process based on the γ correction conversion graph shown in FIG. 10, and the complementary color generation circuit 405 executes the process as shown in FIG. 11(a).

(b) Assuming that processing is executed based on the conversion graph for complementary color generation shown in (C), and then the masking processing circuit 406 and the UCR processing/black generation circuit 407 execute processing based on the following equation. Figure 7(a).

The RGB image data shown in (b) and (C) is T
Correction circuit 403. Complementary color generation circuit 405. Masking processing circuit 406. Then, through the UCR processing/black generation circuit 407, it is converted as shown in FIGS. 12(a), (b), (C), and (d).

Furthermore, if the gradation processing circuit 408 uses a Bayer type 3×3 multivalued dither matrix shown in FIG. 13, Y in FIGS. ,M
.

The data for C and BK are shown in Figure 14 (a) and (b), respectively.
, (C).

It is converted into the data shown in (d).

For comparison, when data without anti-aliasing processing (data in FIGS. 8(a), (b), and (C)) is processed by the image processing device 400, the first
The conversion is performed as shown in Figures 5 (a), (c), (C), and (d).

■Configuration of multilevel color laser printer First, the 16th
The schematic configuration of the multivalued color laser printer 500 will be described with reference to the control block diagram shown in the figure.

A photoreceptor development processing unit 501 uniformly charges the surface of a photoreceptor drum (described later), exposes the charged surface to a laser beam to form a latent image, develops the latent image with toner, and transfers it to recording paper. The details will be explained later, but the development and development of BK data is
A black developing/transfer section 501bk that performs transfer, a cyan developing/transfer section 501C that develops and transfers C data,
Cyan development/transfer section 501 that develops/transfers M data
m, and a cyan developing/transfer section 501y that develops and transfers Y data.

The laser drive processing unit 502 includes the image processing device 4 described above.
The laser beam is output by inputting the 3-bit data of Y, M, C, and BK output from 00 (in this case, image density data). Input buffer memory 503y', 503
m, 503c, and a laser diode 504y that outputs laser beams corresponding to Y, M, C, and BK, respectively.
, 504m, 504c, and 504bk, and drivers 505y, 505m, and 505c that drive the laser diodes 504y, 504m, 504c, and 504bk, respectively:
505b.

In addition, the black developing/transfer section 5 of the photoreceptor development processing section 501
01bk, the laser drive processing unit 502, the laser diode 504bk, and the driver 505bk are called a black recording unit BKU (see FIG. 17). Similarly, the combination of cyan developing/transfer section 501c, laser diode 504c, driver 505c, and buffer memory 503C is combined into cyan recording unit CU (see FIG. 17) and magenta developing/transfer section 501.
m, laser diode 504 m, driver 505 m, and buffer memory 503 m are combined into a magenta recording unit) MU (see Figure 17), yellow developing/transfer section 501 y, and laser diode 504
y, driver 505 y, and.

The combination of buffer memories 503y is called a yellow recording unit (YU) (see FIG. 17). As shown in the figure, each of these recording units is connected to a conveyor belt 50 that conveys the recording paper.
6 (around the black recording unit from the recording paper transport direction)
BKU, cyan recording unit CU, magenta recording unit MU, and yellow recording unit (YU) are arranged in this order.

With this arrangement of each recording unit, exposure starts first, so the laser diode 5 for black exposure
04bk, and the laser diode 504y for yellow exposure starts exposure last. Therefore,
There is a time difference in the order of exposure start between each laser diode,
Data recorded during the time difference (output of the image processing device 400)
In order to hold the data, the laser drive processing unit 502 includes the three sets of buffer memories 503y, 503m, and 503C described above.
is provided.

Next, the configuration of the multivalued color laser printer 500 will be specifically explained with reference to FIG.

The multilevel color laser printer 500 includes a conveyor belt 506 that conveys recording paper, and recording units YU, MU, and YU arranged around the conveyor belt 506 as described above.
Paper cassette 507 containing CU, BKU, and recording paper
a, 507b, and paper feed rollers 508a, 508 that feed the recording paper from cassettes 507a, 507b, respectively.
b, and a registration roller 509 that aligns the recording paper sent out from the paper feed cassettes 507a and 507b.
Recording unit) BKU, CU, by conveyor bell) 506
It is composed of a fixing roller 510 that sequentially transports MU and YU and fixes the transferred image on the recording paper, and a paper discharge roller 511 that discharges the recording paper to a predetermined discharge section (not shown). Here, each recording unit YLJ, MLI, CU, B
KU is photosensitive drum 5123F, 512m, 512c
, 512bk, and photoreceptor drums 512y, 51, respectively.
Charger 51 that uniformly charges 2m512c and 512bk
3y, 513m, 513c, 513bk, and polygon mirrors 514y, 514 for guiding the laser beam to the photosensitive drums 512y, 512m, 512c, 512bk.
m, 514c, 514bk and motor 515y, 515
m, 515c, 515bk, and photosensitive drum 512y,
Toner developing devices 516y, 516m, 516c, and 516bk that develop the electrostatic latent images formed on 512m, 512c, and 512bk using toners of corresponding colors, respectively.
and a transfer charger 5 that transfers the developed toner image onto recording paper.
17y, 517m, 517c, 517bk, and photosensitive drums 512y, 512m, 512c, 512b after transfer.
A cleaning device 51 that removes toner remaining on k
8y, 518m.

It consists of 518c and 518bk. In addition, 519y
, 519m, 519c, and 519bk indicate COD line sensors for reading predetermined patterns provided on the photoreceptor drums 512y, 512m, 512c, and 512bk, respectively. Although details are omitted, this allows multi-value color laser - Detects the process status of the printer 500.

In the above configuration, the operations of the yellow recording unit YU will be explained using exposure, development, and transfer as examples.

Figures 18(a) and 18(b) show the yellow recording unit YU.
The configuration of the exposure system is shown. In the figure, a laser beam emitted from a laser diode 504y is reflected by a polygon mirror 514y, passes through an f-theta lens 502y, and then passes through a mirror 521y.

522y and is irradiated onto the photosensitive drum 512y through the dustproof glass 523y. At this time, since the polygon mirror 514y is rotated at a constant speed by the motor 515y, the laser beam moves in a direction along the axis of the photosensitive drum 512y (main scanning direction). Furthermore, in this embodiment, a photosensor 524y is provided to detect the base point for tracking the scanning position of the main scan, so that the laser beam at the non-exposed position is detected. The laser diode 504y outputs recording data (
Since the photoreceptor drum 504y is activated to emit light based on the 3-bit data from the image processing device 400, multivalue exposure corresponding to the recording data is performed on the surface of the photoreceptor drum 504y. The surface of the photosensitive drum 504y is uniformly charged in advance by the charger 513y as described above, and an electrostatic latent image corresponding to the original image is formed by the exposure. The electrostatic latent image is developed by a yellow developing device f516y to become a yellow toner image. This toner image is transferred from the cassette 507a (or 507b) to the paper feed roller 50, as shown in FIG.
8a (or 508b), and is transferred by a registration roller 509 onto a recording sheet conveyed by a conveyor belt 506 in synchronization with the toner image formation in the black recording unit BKtJ.

The other recording units BKU, CU, and MU have similar configurations and perform similar operations, but the black recording unit BKU is equipped with a black toner developing device 516bk and forms and transfers a black toner image, and the cyan recording unit C
U is equipped with a cyan toner developing device 516c, which forms and transfers a cyan toner image, and is connected to a magenta recording unit L-
The MU includes a magenta toner developing device 516m, which forms and transfers a magenta toner image.

■The multi-value drive drivers 505y, 505m, 505c, and 505b are as follows:
Y sent from the image processing device 400.

Based on the 3-bit data of M, C, and BK, the corresponding laser diodes 504y, 504m, 504c, and 5
It performs control for multi-value driving of 04bk,
As the driving method, power modulation, pulse width modulation, etc. are generally used.

Hereinafter, the multivalue drive by power modulation applied in this embodiment will be explained in detail with reference to FIGS. 19(a), (c), (c), and (d). Furthermore, the drivers 505y and 505m.

505c, 505b, and laser diode 504
y, 504m, 504c, and 504bk each have the same configuration, so here, the driver 505y and laser diode 504y will be explained as an example.

As shown in FIG. 19(a), the driver 505y turns the laser diode 504y on and off based on a predetermined LD drive clock.
n10ff circuit 550 and 3-bit image density data (
Here, the D/A converts Y data) into an analog signal.
From a converter 551 and a constant current circuit 552 that inputs an analog signal based on the image density value from the D/A converter 551 and supplies a current (LD drive current) Id for driving the laser diode 504y to the laser diode on10ff circuit 550. configured.

Here, the LD drive clock is "'1" on "0"
As shown in FIG. 19('b), the laser diode on10ff circuit 550 turns off the laser diode 504y accordingly. Furthermore, since the LD drive current Id and the laser beam power are in a proportional relationship, by generating the LD drive current Id based on the image density data value, the laser beam power output corresponding to the image density data value can be obtained. become.

For example, as shown in FIG. 19(b), when the image density data value is "4" (data N-1 in the same figure), the corresponding LD drive current Id is supplied by the constant current circuit 552, The laser beam power of the laser diode 504y is level 4. Also, the image density data value is “7”.
” (data N in the figure), the constant current circuit 552
The corresponding LD drive current 1d is supplied, and the laser beam power of the laser diode 504y becomes level 7.

Next, referring to FIG. 19(C), laser diode on10ff circuit 550. A specific circuit configuration of the D/A converter 551 and the constant current circuit 552 is shown. The laser diode on10ff circuit 550 includes TTL inverters 553 and 554, differential switching circuits 555 and 556 that perform an onloff toggle operation, and a VGI
> When VC2, the differential switching circuit 555 is on,
The differential switching circuit 556 is off, VGl<
When VC2, the differential switching circuit 555 is off.

Resistors Rz and R form a voltage dividing circuit that generates VC2 that satisfies the conditions for the differential switching circuit 556 to be turned on.
It consists of s. Therefore, when the LD drive clock is "'1", the output of the inverter 554 generates VC,1, the above condition (VGI>VC2) is satisfied, the differential switching circuit 555 is turned on, and the differential switching circuit 556 is turned off and the laser diode 504y is turned off.
Do n.

Conversely, when the LD drive clock is "'0",
Since there is no output from the inverter 554, the condition (VC;
1<VC, 2), and the differential switching circuit 55
5 is turned off, the differential switching circuit 556 is turned on, and the laser diode 504y is turned off.

The D/A converter 551 includes a latch 557 that latches the input image density data while the LD drive clock is "1", a V ref generator 558 that provides the maximum output value V rmf, and a V ref generator 558 that provides the image density data and the maximum output value. ,. ,
3-bit D/ that outputs analog data Vd based on
A converter 559. Furthermore, here Vd
The relationship between Vrmf, image density data, and maximum output value Vrmf is expressed by the following equation.

The constant current circuit 552 includes the laser diode 504 in the laser diode ON10ff circuit 550 as described above.
y current, and the transistor 560
, and resistors Rn and Rs. The output Vd from D/A converter 551 is applied to the base of transistor 560 to determine the voltage applied to resistor R1.

In other words, the current flowing through the resistor R4 is the current flowing through the transistor 5.
60, the current Id flowing through the laser diode 504y is controlled by Vd.

FIG. 19(d) shows the output of the latch 557 mentioned above.

5 is a timing chart showing the relationship between vGl, Vd, and Id. Here, Vd is based on the image density data (3 bit data: 8 gradation data from 0 to 7), ■r, f×
Values are taken in 8 steps from 0/7 to 7/7, and Id is this Vd.
Based on the O value, ■. It shows eight levels from ~■.

The laser diode 504y is placed on the photosensitive drum 512y at the 20th level according to the 8 levels of this Id (Io = level 0°I, = levelt..., ■, = level 7).
A latent image as shown in the figure is formed.

In the image forming system to which the anti-aliasing processing and the apparatus of the present invention are applied, the toner image shown in FIG. 21 is finally created for the pentagon ABCDE shown in FIG. 5(a) by the above-described configuration and operation. Formed on recording paper. -Resolution of laser printer for boat fishing is 240~
Considering that the resolution is 400 dpi, the density of the edges of the figure becomes visually thinner due to the anti-aliasing process. FIG. 22 shows a toner image of a pentagon ABCDH without anti-aliasing processing, and as is clear from comparing FIG. 21 (toner image of the present invention) and FIG. The jagged parts (aliases) on the stairs that appear as diagonal lines become visually smoother.

Further, in this embodiment, multi-value driving using power modulation is applied, but it goes without saying that similar effects can be obtained by using multi-value driving using pulse width modulation.

For reference, FIG. 23 shows the change in latent image form depending on the level of pulse width modulation, and also shows the toner image when pulse width modulation is applied to the pentagonal ABCDH shown in FIG. 5(a). It is shown in FIG.

〔Effect of the invention〕

As explained above, the graphic processing device of the present invention adjusts the output of pixels at the edges of vector data based on the area ratio to be filled, and smoothly expresses the jaggedness (alias) at the edges of the output image. In a graphic processing device that performs anti-aliasing processing, a first calculation means calculates the area of a triangle based on the difference in coordinate values of two intersection points when vector data crosses a scan line; edge determination means for determining the type of edge; and second calculation means for calculating the area ratio of each edge pixel based on the coordinate values of the two intersection points, the area of the triangle, and the type of edge. Because of this, the area ratio can be determined at high speed, and in particular, the area ratio can be determined with an ideal value (accurately) in a normal vector edge portion that does not have an end point.

In addition to the above-described configuration, the sub-pixel division means calculates the area ratio of the edge pixel by sub-pixel division, and when the vector data has an end point on the scan line, the sub-pixel division means divides the edge pixel. Since the area ratio is determined, the overall accuracy can be improved and the speed can be increased compared to the case where the approximate area ratio is determined only by sub-vixel division.

[Brief explanation of drawings]

FIGS. 1(a) to (d) are explanatory diagrams showing the principle of anti-aliasing processing in the graphic processing apparatus of the present invention, FIG. 2 is an explanatory diagram showing the configuration of the image forming system of this embodiment, and FIG. FIG. 4(a) is an explanatory diagram showing the configuration of a PDL controller (graphic processing device of the present invention), FIG. 4(a) is a flowchart showing the operation of the PDL controller, FIG. C) is a flowchart showing anti-aliasing processing, FIG. 5(a),
(b) is an explanatory diagram showing linear vector division of a figure, Fig. 6 is an explanatory diagram showing the approximate area ratio after performing anti-aliasing processing, and Figs. 7 (a), (b), and (C) are page memory An explanatory diagram showing RGB image data stored in the plane memory section of the page memory, FIGS. An explanatory diagram showing RGB image data, FIG. 9 is an explanatory diagram showing the configuration of the image processing device, FIG. 10 is an explanatory diagram showing the T correction conversion graph of the T correction circuit, and FIGS. 11 (a) and (b). , (C) is an explanatory diagram showing a conversion graph for complementary color generation used in the complementary color generation circuit, FIGS. 12(a), (b), (C), (d)
are shown in Figures 7(a) and (b). An explanatory diagram showing the state in which the RGB image data shown in (C) is output from the UCR processing/black generation circuit, FIG. 13 is an explanatory diagram showing a Bayer type 3×3 multilevel dither matrix, and FIG. 14 (a), (b), (c), (d) are Y in Figure 12 (a), (b), (C), (d),
Explanatory diagram showing the state in which M, C, BK data is converted by the gradation processing circuit, Fig. 15 (a), (b), (c
), (d) are Fig. 8(a), (b). An explanatory diagram showing the state in which the data in (C) is processed by the image processing device, Fig. 16 is a control block diagram showing a multi-value color laser printer, and Fig. 17 shows the configuration of the multi-value color laser printer. Explanatory diagram, Fig. 18(a),
19(b) is an explanatory diagram showing the configuration of the exposure system of the yellow recording unit, and FIGS. 19(a), 19(b), and 19(C). (b) is an explanatory diagram showing multi-value drive by power modulation,
Figure 0 is an explanatory diagram showing the state of the latent image depending on the level of power modulation, and Figure 21 is the pentagon ABCDE shown in Figure 5 (a).
FIG. 22 is an explanatory diagram showing the final toner image of pentagon ABCDE without anti-aliasing processing. FIG. 23 is an explanatory diagram showing the state of the latent image depending on the level of pulse width modulation. Figure 24 is Figure 5 (
Explanatory diagram showing a toner image when pulse width modulation is applied to the pentagonal ABCDH shown in a), FIG. 25(a), (
b) is an explanatory diagram showing conventional anti-aliasing processing;
FIGS. 26(a) and 26(b) are explanatory diagrams showing antialiasing processing using the uniform averaging method, and FIGS. 27(a) and
(b) is an explanatory diagram showing anti-aliasing processing using weighted averaging method; Fig. 28 (a), (b), (C
), (d) is an explanatory diagram showing an example of a filter used in the weighted averaging method, and FIG. 29 is an explanatory diagram showing a convolution method with reference to 3×3 pixels. Explanation of symbols 100...-Host combiator 200...-
PDL controller 201-------One receiving device 202--------
CPU U2O5------Internal system bus 204
−・−・・RAM 205−・・ROM2O6・
・--Page memory 207 ・-- Transmitting device 208
-------I10 device 300--Image reading device 400-----Image processing device 500--Multi-valued color laser printer 600
−・−System control unit

Claims (2)

[Claims] (1) In a graphic processing device that adjusts the output of pixels at the edge of vector data based on the area ratio to be filled, and executes anti-aliasing processing to smoothly express jaggedness (alias) at the edge of the output image, A first step that calculates the area of a triangle based on the difference in coordinate values of two intersection points when the vector data crosses the scan line.
an edge determining means for determining whether the vector data is a left or right edge or the type of edge; A graphic processing device comprising: second calculation means for calculating an area ratio of a partial pixel. (2) In claim 1, further comprising subpixel division means for determining the area ratio of the edge pixel by subpixel division, and when the vector data has an end point on a scan line, the subpixel division means A graphic processing device characterized by determining an area ratio of edge pixels.