JPH0696222A - Picture processor - Google Patents

Abstract

PURPOSE:To decide the density of picture elements for processing time similar to conventional RMW(read/modify/write) processing without lowering picture quality. CONSTITUTION:An LUT circuit 42 calculates successively the picture element density from density data, area rate K and density Dm by using RMW processing, buffers 43a-43c output the density of respective left, objective, and right picture elements as outputs A, B and C, and an arithmetic part 44 compares the output B with the other picture element, decides density Dn of the objective picture element and outputs it to a frame memory. Namely, the density is compared with the right and left picture elements or five picture elements including three picture elements in read adjacent lines and when the density is lower than the other density, the density of any one of adjacent picture elements or the average value of outputs A and C is defined as Dn. When the density is out of the range of outputs A and C, the intermediate value is defined as Dn. In the other case, the output B is defined as Dn. Thus, the picture quality is not lowered by running out of the density of an edge part from the area density on both of sides.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to read-modify the density which is anti-aliased in order to form areas surrounded by edges of an image formed by edges displayed in vectors at specified density. The present invention relates to an image processing device that performs write processing and writes in a frame memory.

[0002]

2. Description of the Related Art An image input in from a host machine such as a computer and described in a PDL language (frame description language) such as post script is printed by a printer such as a laser printer or a display such as a CRT. There is an image processing device that performs image processing in the middle of the image display for displaying.

In particular, when processing an image consisting of multiple gradations, areas surrounded by edges of an image formed by edges displayed as vectors (hereinafter referred to as "vector image") are respectively assigned with specified density. The filling process is frequently performed.

For example, when a character is printed, if a vector font is used, a beautiful character can be formed regardless of the size. However, a character cell formed by an edge displayed in a vector (enclosed by the edge Area)
It is indispensable to paint the inside of the black solid.

An image having continuous gradation such as a photograph is also converted into an image having multiple gradations in order to perform digital processing, and image processing is performed as a vector image surrounding areas of the same gradation.

When an image processed in this way is output by a laser printer or a CRT, discontinuity at the next line is inevitable, no matter how high the pixel density or the resolution is. Very fine jaggedness (called jaggies or aliasing) appears in the shaded area near the horizontal. Therefore, visual smoothing is performed by adjusting the edge density by anti-aliasing processing.

The anti-aliasing process is roughly classified into a method for increasing the sample rate and one pixel.
With an averaging method that divides a pixel into sub-pixels and averages
There are types. Further, the averaging method includes a uniform averaging method in which each sub-pixel is treated as an equivalent value, and a weighted averaging method in which each sub-pixel is treated with a different weight.

Further, for a continuous tone image whose density changes continuously (multi-tone image converted for digital processing), the anti-aliased density is subjected to read-modify-write processing. Writing to memory is performed. That is, when determining the density of a pixel on the frame memory that stores multi-valued data, not only the density data of the area to which the pixel belongs, but also the density data of the pixel on the frame memory is read out. Then, new density data is determined while referencing (modifying) and writing (writing) on the frame memory.

In this read-modify-write process, the density of a pixel at the edge portion, which is the boundary between two regions, is set to the density of one region by the area ratio K (0≤K) of the pixel in the region. ≤1) and the sum of the density on the frame memory and the value multiplied by (1-K) is taken to determine the value, in order to keep the gradation continuity and to make the aliasing of edges inconspicuous. It is an effective method and is widely used.

[0010]

However, as shown in FIG. 6, the density D = 64, 50, 35, 15 respectively.
When the conventional read-modify-write process is performed on the case where the areas A, B, C, and D are arranged side by side, the pixels PX1, PX2, and PX3 applied to the edge part that divides the areas A, B, C, and D, respectively. The concentration is D = 61,
Although it is desirable that the values are 41 and 33, as shown in FIG.
It becomes 51, 29, 30.

The pixel PX1 with D = 51 and the pixel PX3 with D = 30 are not conspicuous because they are still between the densities 64 and 50 of the areas A and B and the densities 35 and 15 of the areas C and D, respectively. = 29, the pixel PX2 has a density of 5 in each of the areas B and C.
Since the density is lower than 0 or 35, even if the difference between the density of the area C and the density of the area C is small, it is conspicuous as a bright thin line, and there is a problem that the image quality of the entire image is deteriorated.

Further, when a white background (D = 0) figure (D = 64) as shown in FIG. 10A is processed, the density of each pixel on the edge portion depends on each area ratio. The density as shown in FIG. 10B is stored in the frame memory. In order to erase this figure, if the same figure is painted white (D = 0) and the conventional read-modify-write processing is performed, a light outline remains, as shown in FIG. 11B. However, there was also a problem that the image quality was significantly impaired.

In the past, there has been no method for accurately processing the density of each pixel on the edge portion so as not to cause such a deterioration in image quality, but in any case, the processing time becomes complicated and the processing time becomes long. It had the drawback of being long.

The present invention has been made in view of the above points, and has been achieved by the conventional anti-aliasing processing and read
An object of the present invention is to determine the pixel density without deteriorating the image quality in the processing time equivalent to the modification write processing.

[0015]

In order to achieve the above object, the present invention fills a region surrounded by edges of an image formed by edges displayed by vectors,
An image processing apparatus for determining the density of the target pixel to be written in the frame memory by performing the read-modify-write processing on the density of the target pixel subjected to the anti-aliasing processing is as follows.

According to a first aspect of the present invention, the area ratio of the target pixel subjected to the read-modify-write process is less than 1, and the density of the target pixel on the left and right of the target pixel on the frame memory is already read-modify. If the density is lower than any of the three pixels adjacent to the target pixel on the scan line subjected to the light processing, the density of any of the five adjacent pixels is set to the frame memory of the target pixel. A means for determining the concentration to be written in is provided.

In the first invention, among the target pixel and the five adjacent pixels, a position storage means for storing the position of the lowest density pixel of the pixels adjacent to the pixel to be processed next is provided, When the determining means processes the pixel to be processed next, the density comparison may be started from the pixel at the position stored in the position storing means.

According to a second aspect of the present invention, the area ratio of the read-modify-write processed target pixel is less than 1, and its density is higher than that of any of the pixels adjacent to the left and right of the target pixel on the frame memory. If it is low, the density determining means is provided for determining the density to be written in the frame memory of the target pixel according to the density of one or both of the pixels adjacent to each other on the left and right.

According to a third aspect of the present invention, the buffer memory for temporarily storing the density of the read-modify-write processed pixel and the area ratio of the read-modify-write processed target pixel are less than 1. When the density of the target pixel read out from the frame memory is not 0, the read-modify-write processing of the density of the read-modify-write processed target pixel on the left and right sides of the target pixel on the frame memory When the density is lower than the specified density, density determining means is provided for setting the intermediate value of the densities of the pixels adjacent to the left and right as the density to be written in the frame memory of the target pixel.

[0020]

The image processing apparatus according to the first aspect of the invention has an area ratio K <
1, that is, if the density of the target pixel subjected to the read-modify-write processing on the edge portion is lower than the density of any of the five pixels subjected to the read-modify-write processing adjacent to the target pixel, Since the determining means sets the density of any of the five pixels adjacent to the target pixel as the density of the target pixel, bright thin lines that deteriorate the image quality of the edge portion are not generated.

Further, when the density determination means processes the next pixel, if the density comparison is started from the pixel having the lowest density at the position stored in the position storage means, unnecessary density comparison can be avoided. Processing time can be shortened.

In the image processing apparatus according to the second aspect of the present invention, the area ratio K <1, that is, the read-modify-write process is performed so that the read-modify-write processed density of the target pixel on the edge portion is adjacent to the left and right of the target pixel. When the density is lower than the density of any of the pixels that have been selected, the density determining unit determines the density of the target pixel according to the density of one or both of the pixels adjacent to each other on the left and right, so that a bright thin line that deteriorates the image quality of the edge portion Does not occur.

In the image processing apparatus according to the third aspect of the present invention, the area ratio K <1 and the density read from the frame memory is not 0, that is, the read modification of the target pixel in which the density has already been written to the edge portion. When the density subjected to the write processing is lower than the density of any of the read-modify-write processed pixels adjacent to the left and right of the target pixel, the density determination means determines the intermediate value of the densities of the pixels adjacent to the left and right. Since is the density of the target pixel, a bright thin line that deteriorates the image quality of the edge portion does not occur.

[0024]

FIG. 2 is a block diagram showing the relationship between the host machine 10 and the printer 15. A host machine 10 such as a computer is provided with a keyboard 11 which is an input device and a CRT display 12 which is a display device. The image information described in the PDL language input from the keyboard 11 is displayed on the CRT display 12 and edited. Output to the printer 15 for each frame.

The printer 15 is composed of a controller 20 and an engine 30. The controller 20 performs anti-aliasing processing on image information written in the PDL language input from the host machine 10 and multivalued image data for each page. , And once stored in the frame memory, it is output to the engine 30 as a video signal. The engine 30, which is an output device, converts each line into an image in accordance with an input video signal, prints the formed image of one page on a sheet, and outputs the image.

FIG. 3 is a circuit diagram showing the configuration of the controller 20 which is an embodiment of the image processing apparatus according to the present invention shown in FIG. The controller 20 includes a receiving device 21,
CPU 22, font ROM 23, RAM 24, ROM
25, a frame memory 26, a transmission device 27, and a straight line drawing device 28, which are connected to each other by bus lines such as a data bus, an address bus, a control bus, and the like.

Image data and pixel density described in the PDL language which are respectively input from the host machine 10 via the receiving device 21, multi-valued level (indicating pixel density), page size, output image enlargement or reduction ratio, etc. The image information for one page consisting of is temporarily stored in the RAM 24.

Next, a CPU (central processing unit) 22
The RAM 24 refers to the vector font stored in the font ROM 23 for the character code in accordance with the programs stored in the ROM 25 in advance.
The image data stored in (1) is decomposed into straight line components in the horizontal direction (X axis direction), anti-aliasing processing is performed, and output to the straight line drawing device 28.

The straight line drawing device 28 creates new density data from the density data of each pixel in the X-axis direction which is sequentially input from the CPU 22 and the density data of the pixels read from the frame memory 26 corresponding to each pixel. By determining and sequentially writing to the frame memory 26, the frame memory 2
Draw an image of the straight line component on 6. That is, read / modify / write processing is executed to develop multi-tone image data on the frame memory 26.

In accordance with a command from the CPU 22 in accordance with the designated pixel density, the transmitter 27 sends the multi-level image data for one page developed in the frame memory 26 to a video signal synchronized line by line with the image clock. The image formed by the engine 30 is printed on a sheet by converting the image into a sheet and outputting it to the engine 30.

FIG. 4 is a circuit diagram showing an example of the configuration of the straight line drawing device 28 shown in FIG. 3. FIG. 4A shows the entire configuration of the straight line drawing device 28, and FIG. The configuration of a read-modify-write control unit (hereinafter referred to as "RMW control unit") 34, which is one of the constituent elements and executes read-modify-write processing, is shown.

The straight line drawing device 28 shown in FIG.
Is an FI including five FIFO memories 32a to 32e.
An FO memory unit 32, a memory synchronization control unit 33 that outputs a signal that controls the operation timing of each memory, an RMW control unit 34, an oscillation circuit (OSC) 35 that outputs a clock CLK that serves as a synchronization reference, and a clock. An X counter 36 that counts CLK and outputs coordinates in the X-axis direction;
It is composed of a comparator 37 which detects the match between the contents of the counter 36 and the data X2.

The RMW control section 34 shown in FIG. 4B.
Is a LUT (look-up table) circuit 42 and 3
Number of buffer memories (hereinafter simply referred to as "buffer") 4
It is composed of a latch unit 43 composed of 3a to 43c, a calculation unit 44 which is a density determining unit for comparing the density data held in the latch unit 43 to determine new density data of the target pixel, and a buffer 45. .

The CPU 22 sends the write signal WR together with the FIF
To each of the FIFO memories 32a to 32e of the O memory unit 32, the density value DATA of each pixel, the area ratio K, the vertical coordinate Y of the scan line, the drawing start position X1, and the end position X2 are input, and the respective FIFO memories are input. When stored in 32a to 32e, the signal E indicating the memory empty output from the FIFO memory unit 32 to the memory synchronization control unit 33.
M is negated to signal that data has been entered.

When the read signal RD is input from the memory synchronization control unit 33, the FIFO memory unit 32 retains the density value DATA, the area ratio K, and the ordinate Y of the data stored in each of the FIFO memories 32a to 32e. It is output to the memory synchronization control unit 33 and latched in the memory synchronization control unit 33. The start position X1 is loaded into the X counter 36,
The X counter 36 outputs its contents to the memory synchronization control unit 33 and the comparator 37.

The X counter 36 uses the loaded X1 as a start value to input the clock CL from the oscillation circuit 35.
Count up K. The comparator 37 is an X counter 36.
Contents and end position X input from the FIFO memory 32e
2 is compared, and when they are equal, an end signal is output to the memory synchronization control unit 33.

The memory synchronization control unit 33 includes an X counter 36.
While the content of is counting up from X1 to X2, the density value DATA and the area ratio K are output to the RMW control unit 34, and the content of the X counter 36 is sent to the frame memory 26 as an address signal ADR. It outputs a signal MWR for instructing writing and reading of the memory.

The RMW control unit 34 uses the frame memory 26.
The density value Dm (readout data) of the pixel corresponding to the pixel having the coordinates Y and X to be processed is read and latched by the buffer 45, and the LUT circuit 42 latches the density value Dm (on the frame memory) and the memory. The density value DATA and the area ratio K input from the synchronization control unit 33 (instructed by the CPU) are input, and a new density value, which is the calculation result of the read-modify-write process, is latched from the data. To the buffer 43c in the first stage.

Since the three buffers 43a to 43c of the latch section 43 connected in series with each other act as a shift register for multi-valued data, the data latched in the buffer 43c by the clock CLK is transferred to the next clock CLK.
To the buffer 43b, and then to the buffer 43a at the next clock CLK to be held respectively and output to the arithmetic unit 44.

Therefore, considering the target pixel whose output B output from the buffer 43b is the density value as a reference, the output A output from the buffer 43a is the density value of the pixel immediately before (left adjacent), The output C output from 43c is the pixel immediately after that (on the right side). The calculation unit 44 compares the three outputs A, B, and C, and determines the new density value of the target pixel whose output value is the output B as the density value held in the buffer 43b according to the result, and outputs the density value Dn. By outputting the (drawing data) to the frame memory 26, each pixel on the line Y is drawn.

FIG. 5 is an explanatory diagram showing the positional relationship between the target pixel and pixels adjacent thereto. Target pixel P indicated by a thick frame
Is at the position of the coordinate (X) of the scan line (Y), the process generally advances the scan line from top to bottom, and pixels on the same scan line are advanced from left to right.
The scan line (Y-1) has been subjected to the read-modify-write processing and is drawn in the frame memory 26, and the scan line (Y + 1) included on the lower side with hatching is unprocessed.

The pixel L on the left side of the scan line, the target pixel P, and the pixel R on the right side are subjected to read-modify-write processing, and their densities are outputs A, B shown in FIG. It is C. Therefore, when comparing the density of the target pixel P (output B) with the density of the adjacent pixels, it is meaningless to compare with the lower three pixels LL, L, RL, and there are two pixels on the left and right. L, R (output A,
C), or five pixels obtained by adding the upper three pixels LU, U, and RU that can be read from the frame memory 26 are compared.

Before explaining the routine of the embodiment according to the present invention, in order to clarify the main part of the present invention, FIG.
A flow chart of the routine of the conventional example shown in 2 will be described. In the following description of each flow, for example, “step 3” is abbreviated as “S3”.

When the routine of the conventional example shown in FIG. 12 is started, a large number of vectors for displaying edges forming an image are input in S1, and it is determined in S2 whether or not the input vector is a curve. If not, that is, if it is a straight line, it jumps to S4 as it is, and if it is a curve, it proceeds to S3 and executes linear approximation to decompose it into a plurality of approximate straight lines, and then S4.
Proceed to.

In S4, data indicating the starting point, the ending point (each coordinate value) of each straight line vector, etc.
(T)). For vector registration, one of the both ends of the straight line vector having the smaller Y coordinate value is the starting point, and if the Y coordinate values are the same, the one having the smaller X coordinate value is the starting point.
When all the vectors have been registered, start point Y
Perform vector sorting in ascending order of coordinates,
Organize the data in ET.

Next, in order to determine the pixel density for each line (eg, for line Y), the lines and pixels (pixels) are subdivided and divided into sub-lines and sub-pixels. In S6, the active edge table (hereinafter referred to as "AET") is updated for each subline to be filled.

At S7, the intersections with all the vectors (V1, V2) on the sub-line are obtained, arranged in order from the one having the smallest X-coordinate value to form two edge pairs, and the sub-pixels between them are filled. When the filling on one sub-line is completed, the process is repeated until one line is completed.

In step S8, the area ratio K of pixels is determined by counting the number of sub-pixels filled in for each pixel. For example, if W subpixels out of 25 are filled, the pixel area ratio is K = W / 25.

In S9, it is determined whether or not the area ratio K = 1. If K = 1, the process proceeds to S10 and the input density DA
TA is determined as the new density Dn of the target pixel and the process jumps to S12. If not (K <1), the process proceeds to S11, where the read / modify / write processed density is changed to a new density D.
n.

The new density Dn determined in S12 is written in the frame memory 26. When the writing of one line is completed, the process proceeds to S13, and it is determined whether the writing of all the scan lines, that is, the writing of the image of one page is finished,
If not, the process returns to S6, and if finished, the process ends.

FIG. 1 is a flow chart showing the routine of the first embodiment of the image processing apparatus (controller 20) according to the first invention, and the same parts as those in the conventional example shown in FIG. And the description is omitted. The flow chart of the first embodiment shown in FIG. 1 is different from the conventional example in that S21 to S26 are added instead of S11.

If it is determined in S9 that the interview rate K <1, S2
The process proceeds to step 1 to calculate the read-modify-write processed density, but unlike S11 of the conventional example, the new density Dn is not immediately determined, and the process proceeds to step S22 where the processed density of the target pixel P is the adjacent pixel. Density comparison is performed to determine whether or not the density is lower than one of the pixels (pixels L, R, LU, U, and RU shown in FIG. 5).

If the density of the target pixel P is low, the process proceeds to S23 and S is performed until the comparison with the five adjacent pixels is completed.
22 is repeated, and if all are completed, that is, if it is lower than the density of any of the adjacent pixels, the process proceeds to S24 and the density of any of the adjacent pixels is changed to the new density Dn of the target pixel P.
Then, S12 jump is performed.

If the density of the target pixel P is not small in S22, that is, if even one of the adjacent pixels has a pixel density equal to or smaller than the density of the target pixel P, the process proceeds to S25.
The density after the treatment of is determined as a new density Dn. Then S2
In step S6, the position of the pixel having the lowest density among the pixels U and RU adjacent to the pixel R to be processed next among the adjacent pixels and the target pixel P is stored in the position storage means, for example, the RAM 24, and then in step S12. move on.

In this way, if the minimum density of the adjacent pixels is MIN and the maximum density is MAX, the new density is Dn = MIN even if the minimum density is selected in S24, and the maximum density is selected. It is MAX. Further, the density determined in S25 is Dn ≧ MIN, and the density subjected to the read-modify-write processing does not exceed the higher density of the two regions which are in contact with each other at the edge. Therefore, the new density Dn of the determined target pixel falls within the density range of the two regions, and no brighter or darker edge occurs.

Further, by storing the position of the pixel having the lowest density among the pixels U, RU and P adjacent to the pixel R to be processed next in S26, when the pixel R is processed as the target pixel, the other two pixels are processed. Since it is only necessary to compare the pixel at the stored position with the pixels on the right and upper right of the pixel R without comparing the density with each pixel, the processing speed is increased.

FIG. 6 is an explanatory diagram showing an example of an image in which the density continuously changes, and FIG. 7 shows each pixel in the vicinity of each edge portion of the image shown in FIG. 6 expanded in the frame memory 26 as a density value. It is explanatory drawing which shows each state. The actual read-modify-write processing will be described below with reference to FIGS. 6 and 7.

As shown in FIG. 6, the regions A, B, C, and D having different densities have X coordinate values X1, X2, and X, respectively.
Boundary is bounded by three edges X1, X2, X3 of 3,
Their concentrations are Da = 64, Db = 50, Dc, respectively.
= 35 and Dd = 15. For convenience of explanation, the edges are both vertical, so that each edge X1, X
Pixels PX1, PX2 and PX3 related to 2 and X3 are common to each line.

In this example, the area ratio Ka of the region A in the pixel PX1 is 80%, and the area ratio Kb of the region B is Kb = (1-K
a) = 20%, the area ratio Kb of the region B in the pixel PX2 = 40%, and the area ratio Kc of the region C = (1-Kb) =
60%, and the area ratio Kc of the region C in the pixel PX3
= 90%, area ratio of region D Kd = (1-Kc) = 10%
Suppose

Each pixel PX1, PX2, which is related to the edge part
The concentrations D1, D2 and D3 of PX3 are originally obtained as the sum of the products of the concentrations on both sides and their area ratios, as shown in FIG.
It is desirable that = 61, D2 = 41, and D3 = 33.

[0061]

## EQU1 ## D1 = Da * Ka + Db * Kb = 64 × 0.8 + 50 × 0.2 = 61.2≈61 D2 = Db * Kb + Dc * Kc = 50 × 0.4 + 35 × 0.6 = 41 D3 = Dc * Kc + Dd * Kd = 35 × 0.9 + 15 × 0.1 = 33

However, in order to calculate from the densities of the regions on both sides of the edge, it is necessary to input the densities DATA of the regions on both sides in addition to the densities DATA of the region being processed from the CPU 22, which complicates the process. Therefore, in the conventional read-modify-write processing, the density data Dm of the corresponding pixel on the frame memory 26 is read and processed as the density outside the area being processed.

That is, the conventional read modify
In the write processing, if the density of the region being processed is D, the area ratio of the pixel at the edge is K, and the density of the corresponding read pixel is Dm, the new density Dn to be written.
Was calculated and determined as shown in Formula 2.

[0064]

(2) Dn = D × K + Dm × (1-K)

Therefore, the frame memory 26 is initially cleared to Dm = 0, and the area A (Da = 6) is set.
During the process of 4), the area ratio of the pixel on the left side of the pixel PX1 is K
= 1 and the area ratio of the pixel PX1 is Ka = 0.8. Therefore, as shown in Equation 3, the density Dn of the pixel on the left side of the pixel PX1 is 64, and the density Dn1 of the pixel PX1 is 51.2. It is written in the frame memory 26.

[0066]

## EQU3 ## Dn = 64 × 1 + 0 × 0 = 64 Dn1 = 64 × 0.8 + 0 × 0.2 = 51.2

Next, when the area B (Db = 50) is processed, the area ratio of the pixel PX1 is Kb = 0.2 and the density Dm1 = 51.2 to be read out, and the pixels PX1 and Px1.
The area ratio of the pixels sandwiched by X2 is K = 1 and Dm = 0, and for the pixel PX2, the area ratio is Kb = 0.
4 and Dm = 0, the pixel PX is
1 modified density D1 = Dn1 = 51, pixel PX1,
The density Dn of the pixel sandwiched by PX2 = 50, and the pixel PX
The density of 2 becomes Dn2 = 20.

[0068]

Dn1 = 50 × 0.2 + 51.2 × 0.8 = 50.96≈51 = D1 Dn = 50 × 1 + 0 × 0 = 50 Dn2 = 50 × 0.4 + 0 × 0.6 = 20

Similarly for the area C (Dc = 35),
The area ratio of the pixel PX2 is Kc = 0.6 and Dm2 = 20, and the area ratio of the pixel sandwiched between the pixels PX2 and PX3 is K =
1, Dm = 0, and the area ratio of the pixel PX3 is Kc = 0.
Since Dm = 0 at 9, the pixel PX is
2 modified densities D2 = Dn2 = 29, pixels PX2,
The density Dn of the pixel sandwiched by PX3 = 35, and the pixel PX
The density of 3 becomes Dn3 = 31.5.

[0070]

Dn2 = 35 × 0.6 + 20 × 0.4 = 29 = D2 Dn = 35 × 1 + 0 × 0 = 35 Dn3 = 35 × 0.9 + 0 × 0.1 = 31.5

Finally, when the area D (Dd = 15) is processed, the area ratio of the pixel PX3 is Kd = 0.1 and Dm3 =
Since the area ratio of the pixel on the right side of the pixel PX3 is K = 1 and Dm = 0, the corrected density D3 = Dn3 = 30 of the pixel PX3 and the pixel PX are obtained as shown in Expression 6.
The density of the pixel on the right side of 3 is Dn = 15.

[0072]

## EQU6 ## Dn3 = 15 × 0.1 + 31.5 × 0.9 = 29.85≈30 = D3 Dn = 15 × 1 + 0 × 0 = 15

The frame memory 2 is processed as described above.
The density data written in No. 6 is shown in FIG. Figure 7
In the example of the conventional processing shown in FIG. 7C, the pixels PX1, PX2 and PX which are related to the edge are compared with the example shown in FIG.
The values of the densities D1, D2 and D3 of No. 3 are both low.
Nevertheless, the density D1 = 51 of the pixel PX1 is Da>D1>
There is a relationship of Db, and the density D3 = 30 of the pixel PX3 is also Dc.
Since>D3> Dd, even if the density errors are 10 and 3, they are not conspicuous and there is no fear that the image quality will be deteriorated.

However, the density D2 of the pixel PX2 = 2
9 does not mean that the density error is 12, but Db>
Due to the relationship of D2 <Dc, even if the density difference is only 6 lower than Dc, it is conspicuous as a bright thin line that should not be there, and the image quality is significantly impaired. In general, if the area ratio of the higher density region is less than (low density / high density), bright thin lines appear.

Also in this embodiment, the read-modify-write processing described above is performed by the LUT circuit 42 of the RMW control section 34 shown in FIG. 4B. The LUT circuit 42 is an LU that stores the results calculated in advance.
Concentration DATA using T (look-up table)
(D), the area ratio K, and the read density Dm are input and a new density Dn is immediately output to the latch section 43, so that the processing is instantaneously performed.

It is apparent from the above description that the density D input from the CPU 22 is output when K = 1, and the read density Dm is output as a new density Dn when K = 0.

The embodiment according to the present invention is different from the prior art in that a latch section 43 including buffers 43a to 43c,
That is, the calculation unit 44, which is a concentration determination unit, is provided.
For example, when the density Dn of the pixel PX1 output from the LUT circuit 42 is stored in the buffer 43b via the buffer 43c of the latch unit 43, the buffers 43a, 43b, 4
The outputs A, B, and C output from the 3c to the calculation unit 44 are the densities obtained by the read-modify-write process from the pixels preceding and succeeding the pixel PX1 that are continuous, that is, from the left adjacent pixel to the right adjacent pixel.

That is, in the example shown in FIGS. 6 and 7,
The pixels adjacent to each of the pixels PX1, PX2, and PX3 and the pixels on the line above them (pixels L, LU, R, and RU shown in FIG. 5) do not cover the edge portion, and thus the area ratio K = 1. Therefore, the output A, C is the density of the regions A and B or the regions B and C or the regions C and D, respectively.

The operation unit 44 of the first embodiment, whose flow chart is shown in FIG. 1, reads the target pixel (pixel P in FIG. 5).
The density (output B) subjected to the modify write processing is adjacent (read as Dm from the frame memory 26) on two pixels (L, R) adjacent to the left and right and a line (Y-1) already processed. If the density is lower than any of the three pixels (LU, U, RU), the density of one of the adjacent pixels is set as the density Dn to be written in the frame memory 26, and the other is output B. Is the concentration Dn.

When the density is lower than any of the former ones, it is sufficient to determine in advance which pixel of the adjacent pixels has the density of Dn. Since no darker edge than the high-density area or brighter edge than the low-density area does not appear in the area, the image quality is not deteriorated. here,
The simplest example will be described in which the density (output A) of the pixel L on the left is Dn.

In the example shown in FIGS. 6 and 7, the pixel P
Since X1 (D = 51) and pixel PX3 (D = 30) are both larger than the pixel densities D = 50 and D = 15 on the right side, the output B is written in the frame memory 26 as a new density Dn.

In the case of the pixel PX2 (D = 29), the pixels L and LU (D = 50), the pixels R and RU (D = 35), and the pixel U which has already become D = 50 by the processing described below. 7B, the density D = 50 (output A) of the pixel L on the left is written as a new density Dn in the frame memory 26, as shown in FIG. 7B. The density D = 50 of the pixel U is similarly determined when the line (Y-1) is processed.

Even if it has been decided in advance that the density (output C) of the pixel R on the right side is Dn, since the density of the pixel U is D = 35, the density D of the pixel PX2 is D = 2.
The relationship in which 9 is lower than the densities of all adjacent pixels remains unchanged.

For example, in the case of the pixel PX1 or PX3, the processing in S26 of the flow chart shown in FIG. 1 is performed for the pixels U, RU and the target pixel P which are also adjacent to the pixel R to be processed next among the compared adjacent pixels. If the position of the pixel RU having the lowest density (D = 50 or 15) among them is stored in the RAM 24, the pixel P and the pixel P when the pixel R becomes the target pixel next time.
As described above, the processing time can be shortened without the need to compare with.

FIG. 8 is a flow chart showing a routine of the second embodiment of the image processing apparatus according to the second invention. The same parts as those of the first embodiment shown in FIG. 1 are designated by the same step numbers. The description is omitted. The second embodiment shown in FIG. 8 is different from the first embodiment in that S32 to S35 are provided instead of S22 to S26.

The densities of the target pixel P calculated in S21, that is, the output B, are compared in S32 and S33 with the left and right pixels L,
If the output B is smaller than either the output A or C as compared with each concentration of R, that is, the output A or C, the process proceeds to S35.
The output B is determined as the new density Dn of the target pixel P and S12 is determined.
Proceed to. If the output B is smaller than the outputs A and C, the process proceeds to S34, a new density Dn is determined according to the densities of one or both of the left and right pixels L and R, and the process proceeds to S12.

New density D of the target pixel P in S34
n is determined by (1) the density of the left pixel L, that is, the output A, and (2) the density of the right pixel R, that is the output C, (3).
Any one of the larger density of the outputs A and C, (4) the smaller density of the outputs A and C, and (5) the average value of the outputs A and C may be adopted. ) Either of them should be decided.

As described above, in the second embodiment, since the density of the target pixel is only compared with the two left and right pixels, it is more than the case of comparing with the five adjacent pixels as in the first embodiment. Also reduces processing time. Further, by providing the three buffers 43a to 43c as shown in FIG. 4B, it is not necessary to read from the frame memory 26 each time a comparison is made, so that the processing time can be further shortened. .

The same effect can be obtained by reducing the number of the three buffers 43a to 43c. That is, in FIG. 4B, the buffer 43c is removed and the LUT circuit 4
Even if the data is directly output from the buffer 2 to the buffer 43b and is also input as the output C to the arithmetic unit 44, the operation does not change. Further, if the buffer 43a is also eliminated and only the buffer 43b is provided and the output A, which is the density of the pixel L on the left side, is read from the frame memory 26 and latched in the arithmetic unit 44, only one buffer is required.

FIG. 9 is a flow chart showing the routine of the third embodiment of the image processing apparatus according to the third invention, which is the same part as the first and second embodiments shown in FIGS. 1 and 8. “Start” through S8 and S13 and “end” are omitted. In addition, S that is the same part and has the same step number
Detailed description of 9, S10, S12, and S21 is omitted.
Needless to say, the third embodiment can be implemented when there are two or three buffers, but the flow chart shown in FIG. 9 is an example when there is one buffer.

When the area ratio K of the target pixel P is determined and the process goes from S8 to S9, the interview ratio K is determined, and if K = 1, S
The density DATA input in 10 is set as the pixel density Dn, and S1 is set.
Jump to 2. If it is determined in step S9 that K ≠ 0 (K <1), the process proceeds to step S21, the density by the read-modify-write process is calculated, and then the process proceeds to step S41.

In S41, it is determined whether or not the density (read value) of the target pixel P read from the frame memory 26 for the read / modify / write processing is 0. If it is 0, the process jumps to S45. That is, if not 0, S4
In step S43, the calculated density of the target pixel P is stored in the buffer (for example, 43b), and the LUT circuit 42 is caused to execute the read-modify-write process for the pixel R in S43. As a result, the outputs B and C indicate the densities of the target pixel P and the next pixel R, respectively. Therefore, if the densities of the pixel L are read from the frame memory 26, comparison with the left and right pixels becomes possible.

In S44 (corresponding to S32 and S33 in FIG. 8), it is determined whether or not the density of the target pixel P is within the range of the density of the left and right pixels L and R, that is, an intermediate value. If it is a value, the process proceeds to S45, and the read / modify / write processed density of the target pixel P, that is, the output B is set as a new density Dn, and the process proceeds to S12.

No, that is, if the density of the target pixel P is higher or lower than the density of any of the left and right pixels L and R, S
In step S46, one of the intermediate values of the left and right pixel densities is set as the new density Dn, and the process proceeds to step S12. The intermediate value of the densities is a density value that is less than or equal to the density of the high-density pixel of either of the left and right pixels and is greater than or equal to the density of the low-density pixel, and the intermediate value to be adopted may be a high-density value or a low-density value. However, in general, if the average value of both is adopted, the change in density becomes smoother.

When the density Dn determined in S12 is written in the frame memory 26, it is determined in S47 whether or not the density of the next pixel R has been calculated, that is, whether or not the density has passed through S43. If the former, that is, the output of the LUT circuit 42 has the density of the pixel R, the process returns to S42 to store the density value in the buffer, and if not, the process goes to S13 and repeats until all scan lines are completed. .

An example of processing a graphic according to the flow chart of the third embodiment shown in FIG. 9 will be described. FIG. 10 (A)
Is an illustration showing the relative positions of the hatched vector graphic and each pixel, and FIG. 10B is an explanatory diagram showing the density values of some of the pixels of the frame memory 26 in which the vector graphic is written. The illustrated portion is composed of seven pixels P0 to P6 in each of three lines Y-1, Y, Y + 1.

When the figure with the density D = 64 shown in FIG. 10A is written in the frame memory 26 which was initially cleared so that all the pixels are 0, the density value of each pixel is as shown in FIG.
The value becomes 0 (B). For example, in the case of the line Y, the densities of the pixels P0 and P6 outside the figure remain 0, and the area ratio K of the pixels P1 and P5 on the edge portion is
50% and 47%, the pixels P2 to P4 in the figure have an area ratio K
= 1 and the input concentration DATA is D = 64, respectively.
Is.

Therefore, the pixels P1 and P5 are K
= 0.5, 0.47, that is, K ≠ 1, so S9 to S2
Go to 1. Since Dm = 0 in the calculation formula shown in Equation 2, the densities of the pixels P1 and P5 are 32 and 3 respectively.
0 is obtained. Next, since the read value (Dm) = 0, the process proceeds from S41 to S45, sets the obtained concentration to Dn, and then proceeds to S12.

Since the pixels P2 to P4 have K = 1 respectively, the process proceeds from S9 to S10, sets Dn = 64, and sets S12.
Proceed to. In S12, each Dn is set to the frame memory 26
Therefore, as shown in FIG. 10B, the density values 32, 64, 64, 64, 3 of P1 to P5 are recorded in the line Y.
0 is formed. The same applies to the other lines Y-1 and Y + 1.

In the case of erasing the figure formed in this way, the method of overwriting the figure shown in FIG. 10A with the density of 0 is performed. In the conventional example based on only, the pixels P2 to P4 of the line Y with K = 1 are 0, but K <1 applied to the edge portion.
11 does not become zero because the density is the product of the respective densities Dm and (1-K) shown in FIG.
As shown in (B) of No. 3, although the frame has a light density, it remains.

According to the third embodiment, the pixel having K = 1 advances to S9, S10 and S12 and becomes Dn = DATA = 0. Further, the pixel of K <1 applied to the edge portion, for example, the pixel P1 of the line Y proceeds from S9 to S21 and DATA =
From 0, K = 0.5 and Dm = 32, the density value 16 is obtained as in the conventional example. Next, since the read value Dm ≠ 0, the process proceeds from S41 to S42 to store the density 16 of the pixel P1 in the buffer, and then the density of the next pixel P2 is calculated in S43, and DATA = 0, K = 1 is calculated. A density value of 0 is obtained regardless of Dm (= 64).

In S44, the density of the pixel P1 (16) is compared with the densities of the left and right pixels P0 and P2 (both are 0).
Then, the intermediate value of the densities of the pixels P0 and P2 is set to Dn.
Since either the right or left pixel P0 or P2 is taken as an intermediate value or the average value thereof is 0, Dn =
It becomes 0. The same applies to the pixel P5.

Therefore, according to the third embodiment, as shown in FIG.
(A), the density of all the pixels related to the figure becomes 0, and the figure is completely erased. That is, the image quality is not deteriorated not only when the figure is drawn but also when the figure is erased.

[0104]

As described above, the image processing apparatus according to the present invention determines the pixel density without deteriorating the image quality in the same processing time as the conventional anti-aliasing processing and read-modify-write processing. You can do it.

[Brief description of drawings]

FIG. 1 is a flowchart showing a routine of a first embodiment of an image processing apparatus according to the present invention.

FIG. 2 is a block diagram showing a relationship between a host machine and a printer including an image processing apparatus.

3 is a circuit diagram showing a configuration of a controller which is an embodiment of the image processing apparatus shown in FIG.

4 is a circuit diagram showing an example of a configuration of the straight line drawing device shown in FIG.

FIG. 5 is an explanatory diagram showing a positional relationship between a target pixel and pixels adjacent thereto.

FIG. 6 is an explanatory diagram showing an example of an image whose density continuously changes.

7 is an explanatory diagram showing an example of density values in a state where the image shown in FIG. 6 is expanded on a frame memory.

FIG. 8 is a flowchart showing a routine of a second embodiment.

FIG. 9 is a flowchart showing a routine of a third embodiment.

FIG. 10 is an explanatory diagram showing an example of a state in which a figure and its image are expanded.

11 is an explanatory diagram showing an example of a state in which the image shown in FIG. 10 is erased.

FIG. 12 is a flowchart showing a routine of a conventional example.

[Explanation of symbols]

10 Host Machine 15 Printer 20 Controller (Image Processing Device) 22 CPU 24 RAM (Position Storage Means) 43a to 43c
Buffer memory 44 Calculation unit (concentration determining means)

Claims (4)

[Claims] 1. In order to fill a region surrounded by the edge of an image formed by edges displayed as a vector, an anti-aliased density of a target pixel is subjected to a read-modify-write process to thereby perform a read-modify-write process on a frame memory. In the image processing device for determining the density of the target pixel to be written in, the area ratio of the target pixel subjected to the read-modify-write processing is less than 1, and the density is on the left and right of the target pixel on the frame memory. If the density is lower than the density of any of the two adjacent pixels and the three pixels adjacent to the target pixel on the scan line that has already undergone the read-modify-write processing, then the density of the adjacent five pixels A density determining means for setting any of the densities as a density to be written in the frame memory of the target pixel. That image processing apparatus. 2. The image processing apparatus according to claim 1,
Position storage means for storing the position of the lowest density pixel of the pixels adjacent to the pixel to be processed next among the target pixel and the five adjacent pixels is provided, and the density determination means is configured to An image processing apparatus, characterized in that when processing a pixel to be processed, the density comparison is started from the pixel at the position stored in the position storage means. 3. In order to fill an area surrounded by the edges of an image formed by the edges displayed as vectors, the anti-aliased density of the target pixel is subjected to read-modify-write processing so that it is stored in the frame memory. In the image processing device for determining the density of the target pixel to be written in, the area ratio of the target pixel subjected to the read-modify-write processing is less than 1, and the density is on the left and right of the target pixel on the frame memory. When the density is lower than that of any of the adjacent pixels, the density determining means for determining the density to be written in the frame memory of the target pixel according to the density of one or both of the pixels adjacent to the left and right is provided. A characteristic image processing device. 4. In order to fill a region surrounded by the edges of an image formed by the edges displayed as a vector, the anti-aliased density of the target pixel is subjected to a read-modify-write process so that it is stored in the frame memory. In an image processing apparatus for determining the density of a target pixel to be written in, a buffer memory for temporarily storing the density of a read-modify-write processed pixel, and an area ratio of the read-modify-write processed target pixel Is less than 1 and the density of the target pixel read from the frame memory is not 0, the density of the target pixel subjected to the read-modify-write processing is adjacent to the right and left of the target pixel on the frame memory. Lower than any of the read-modify-write processed densities of Case, the image processing apparatus characterized by intermediate values of the density of pixels adjacent to the right and left are provided and concentration determination means for the writing density in a frame memory of the target pixel.