JP2002225381A - Image processor and image processing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To eliminate moire pattern and rough feeling from an output image at the time of semitransparent processing of an object. SOLUTION: The image processor 1 for interpreting a page description language delivered from a host computer H and converting it into print data in print output format comprises a section 12 for making a decision whether the page description language performs semitransparent processing of a foreground image and a background image by logical operation or not, and a semitransparent processing section 16 for calculating the average value of pixels corresponding to the foreground image and background image when a decision is made that semitransparent processing is performed by logical operation to obtain semitransparent image data.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to an image processing apparatus and an image processing method for performing translucent processing when two images overlap each other and outputting print data.

[0002]

[Prior Art] Microsoft Office (a trademark of Microsoft Corporation)
Application software, it is possible to specify a semi-transparent effect on an overlapping portion between objects in an object in a document.

[0003] The translucent effect is used to obtain an effect that 50% of a background object and 50% of a foreground object are displayed. Particularly, it is widely used in a business document or the like to give an effect of a character portion. I have.

The translucent process is originally defined by an intermediate color between two objects called a background (Destination) and a foreground (Pen) as shown in FIG. Here, the background and the foreground are usually R (red), G (green), and B (blue) colors, and each channel has an 8-bit value.

For example, if the background value is (R, G, B) = (3
0, 160, 85) and the foreground value is (R, G, B) = (12
In the case of (8, 128, 64), the semi-transparent color is obtained by calculating the average value of each channel (R, G, B) = (79, 144, 7).
5) is obtained. However, the average value calculation for obtaining the intermediate color cannot describe the logical operation.

Therefore, conventionally, this semi-transparent object is divided into a foreground, a background, and an image mask by using a fixed mask pattern called an image mask as shown in FIG.
Processing is performed as a result of the processing.

[0007] ROP processing is a logical operation processing method in which a logical operation method can be selected by an ROP code number, and various image processing can be performed by designating an image overlapping portion between objects having 8-bit values. It is.

In the translucent processing, a background value is displayed in a portion where an image mask dot is OFF, and a dot is turned ON.
A logical operation for displaying the foreground value in the portion is given by the ROP code. Specifically, NOTXORPEN, MERGEPE
Three types of ROP codes, N and NOTXORPEN, are used.

Here, a pattern called a checkerboard pattern, in which white (OFF) dots and black (ON) dots appear alternately in the horizontal and vertical directions, is usually used as an image mask in the translucent processing.

When a translucent image obtained by such a ROP operation is displayed on a CRT, the dot spacing of the checkerboard pattern is not perceived by the eye because the resolution of the CRT is sufficiently high. As a result, a translucent color is generated. You can get the effect as if you were.

When an image including a translucent object created by the checkerboard pattern and the ROP operation processing is output to a printer, the application uses Windows GDI (Graph
The background, the image mask, the foreground and the respective ROP operation codes are transmitted through the ical display interface.

The image forming apparatus of the printer performs ROP calculation processing based on the transmitted ROP code, and writes an image in the drawing memory. However, RGB8bi
In order to perform the ROP operation with the t or CMY 8 bits, an 8-bit image after the ROP must be stored, which requires a huge memory.

An image forming apparatus such as a printer usually requires 1-bit CMYK data.
Each object having an RGB 8-bit value sent from the DI is finally screen-processed and CMYK1
It is converted into bit image data. Therefore, the image written in the drawing memory may be a CMYK 1-bit image.

In view of this, for example, Japanese Patent Application Laid-Open No. 11-16 / 1999
ROP processing itself is performed by 1 bi
A method performed at t is known. Using this method, G
The object sent from DI is immediately 1 bit
Since the image is converted into an image, the ROP operation is performed at 1 bit based on the ROP code, and the image is stored, the data amount becomes 1/8 of each color as compared with the 8-bit image.

However, it is difficult to completely prevent image quality deterioration such as color shift and roughness by performing ROP processing originally defined by 8 bits with 1 bit. Therefore, in order to prevent roughness and accurately perform color matching, even if the processing speed is reduced, the ROP processing is performed in 8-bit CP.
If the processing is performed on the U side and priority is given to the processing speed, 1
A method of converting the data into bits and performing ROP processing in hardware on the printer side may be adopted.

[0016]

As described above, Microsof
t In applications such as Office, when translucent parts appear, they are decomposed into background, image mask, and foreground, and RO
A P code is added.

In this operation, as described above, the background is masked with an image mask such as a checkered pattern, the background is drawn on a transparent portion (a portion where no dot of the image mask is placed), and a non-transparent portion (the image mask of the image mask) is drawn. This has the effect of drawing a new foreground in the area where the dots are on).

However, in this ROP processing,
There is a problem that an image mask and a screen matrix pattern used for binarization interfere with each other to generate a moire pattern that is difficult to see.

For example, when a checkered pattern is used for an image mask, moire always occurs unless each screen angle is 45 degrees. Even if each screen angle is 45 degrees and the number of lines is the same as that of the checkerboard pattern, an extreme phenomenon occurs in which the dots of the image mask and the screen matrix completely overlap or do not overlap completely. However, another problem that the color after the ROP operation is completely changed occurs.

The above problem occurs remarkably when the ROP is processed with 1 bit. However, even when the ROP is processed with 8 bits, an object already masked by the solid pattern is screened. Moire and color misregistration may occur.

[0021]

SUMMARY OF THE INVENTION The present invention has been made to solve such problems. That is, the present invention is an image processing apparatus that interprets a page description language sent from a host computer and converts the page description language into print data in a print output format. Determining means for determining whether or not the page description language is to be subjected to translucent processing by a logical operation, when the determining means determines that the page description language is to be subjected to translucent processing, the average value of the corresponding pixels of the foreground image and the background image And translucent processing means for calculating translucent image data.

According to another aspect of the present invention, there is provided an image processing method for interpreting a page description language sent from a host computer and converting the page description language into print data in a print output format, wherein the page description language performs a logical operation on a foreground image and a background image. A step of determining whether or not to perform translucent processing, and, if it is determined that the page description language is to perform translucent processing by a logical operation, calculating an average value of corresponding pixels of a foreground image and a background image And making the image data translucent image data.

According to the present invention, when the page description language sent from the host computer performs a translucent process on the foreground image and the background image by a logical operation, the correspondence between the foreground image and the background image is obtained. Since translucent image data is generated by calculating the average value of pixels, translucent image data can be created without performing logical operations using an image mask, and the occurrence of roughness and moiré after translucent processing is suppressed. become able to.

[0024]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram illustrating an image processing apparatus according to the present embodiment. That is, the image processing apparatus 1 of the present embodiment interprets the PDL (page description language) sent from the host computer H,
A PDL interpretation unit 11, a determination unit 12, a binarization processing unit 13, a ROP processing unit 14, a drawing memory 15, a translucent processing unit 16, a multi-valued A restoring unit 17 and an average processing unit 18 are provided.

The PDL interpretation section 11 is a section for interpreting the PDL sent from the host computer H. This PD
The PDL in which the image data is expressed by RGB each having 8 bits is input to the L interpretation unit 11.

The determining unit 12 adds, to the PDL interpreted by the PDL interpreting unit 11, a ROP code (NOTX
ORPEN, MERGEPEN, NOTXORPEN) are determined.

C (cyan), M
(Magenta), Y (yellow), K (black) 8b each
It data is input and binarized using a predetermined screen matrix (threshold matrix).

The ROP processing unit 14 includes a binarization processing unit 13
The CMYK 1-bit data that has been binarized is input, and the specified ROP processing (excluding translucent processing) is performed. 1 bit for each CMYK after ROP processing
Are stored in the drawing memory 15.

When the determining unit 12 determines that the ROP process specifies the translucent process, the translucent processing unit 16 reads the binarized image data from the drawing memory 15 and performs multi-value restoration. The multi-value data is generated by the unit 17,
The average processing unit 18 performs an average processing corresponding to a translucent processing. Further, the multi-valued image data subjected to the average value processing by the binarization processing unit 18a is converted into binary data, and stored in the drawing memory 15 again.

Next, the multi-value restoration section 17 of the translucent processing section 16
The multi-value data generation and the binarization processing by the binarization processing unit 18a will be described. In the present embodiment, two processes will be described. First, as a first method, a method of relatively easily generating multi-value data will be described.

Here, the pixel values in the 8-bit foreground image input from the PDL interpretation unit 11 are P8, 1b
image mask of it and IM
The pixel value of the background image stored in 5 is D1, and the threshold value of the screen matrix used for binarization is S8.

First, a threshold value S8 of the corresponding screen matrix is selected for a portion where the dot is ON in the pixel value D1 of the background image. That the dot of the pixel value D1 of the background image is ON means that the background value restored by at least 8 bits is equal to or larger than the threshold value S8.
Therefore, assuming that the 8-bit restoration approximate value is D8, D8 = (2
56 + S8) / 2.

In the pixel value D1 of the background image, the portion where the dot is OFF means that the background value restored by 8 bits is less than the threshold value S8. Therefore, 8
The bit restoration approximate value is set to D8, and D8 = S8 / 2 is given. As a result, an 8-bit approximate restored value is obtained.

Next, an average value V8 is obtained between the restoration approximate value D8 and the pixel value P8 of the foreground picture Zao to obtain a binarized value V1. Note that this 8-bit restoration value corresponds to the image mask I
The calculation is performed only for the portion that overlaps the M region.

In summary, the ON dot portion: V8 = (D8 + P8) / 2 = ((25
6 + S8) / 2 + P8) / 2 OFF dot portion: V8 = (D8 + P8) / 2 = (S8
/ 2 + P8) / 2, which is compared with the threshold value S8. If V8> S8, the dot is turned on, and if V8 ≦ S8, the dot is turned off.

Therefore, if the calculation is simplified, ON dot portion: comparison of (256 + 2 * P8) / 3 with S8 OFF dot portion: comparison of 2 * P8 / 3 with S8, the ON of the dot in the translucent area, OFF can be determined.

The image obtained in this manner is used as a background image D
By re-entering 1, the translucent image can be obtained.

In this first method, 8b from one point of the pixel of interest
It is possible to calculate the it restoration value, and there is an advantage that the resulting translucent image is less likely to be rough.

Next, the second method will be described. The second method is characterized in that pixels around the pixel of interest and the threshold values of the screen matrix corresponding to the pixels are used, and the multi-value inversely transformed image is obtained again after obtaining a translucent value with 8 bits. The binarization is also considered.

Hereinafter, a specific 8-bit conversion method of the background value will be described with reference to FIGS. First, a certain range of dots is scanned with each dot of the background value as the center. At this time, w that determines the maximum value of the dot scanning range
m and hm are given (step S101 in FIG. 2).

That is, as shown in FIG. 3, four points (x−wm, y−hm), (x
+ Wm, y-hm), (x-mw, y + mh), (x +
mw, y + mh) are all pixels within the square having the vertex as the scanning target.

When inversely converted to an 8-bit value, wm, hm
The value of 3 to 7 is appropriate. The larger the value, the longer the processing time, but the better the gradation reproducibility. Therefore, it is often sufficient to take a value of about 4.

Next, the initial scanning value is set to w = h = 1 (step S102 in FIG. 2), and the pixel of interest (x,
y) (step S103 in FIG. 2) and four points (x-
w, y-h), (x + w, y-h), (x-w, y +
h), (x + w, y + h), all pixels in a square (3 × 3 pixels in this case) having vertices are scanned.

At this time, among the 1-bit background values, all the threshold values of the screen matrix corresponding to the portion where the ON dots are laid are selected, and the maximum value of the threshold values is determined and set as maxv. 2 are selected, and the minimum value of the thresholds is set to minv (FIG. 2).
Step S104).

Here, when maxv ≦ minv,
All the 8-bit values in the square around the pixel of interest are maxv
Even if the approximation is made with the same value between to and minv, there will be no inconsistency with the screen matrix (that is, if the approximate 8-bit value is re-binarized, the same background value will be obtained again).

Therefore, in this case, scanning in a wider range is performed to improve the approximation accuracy. That is, 1 is added to w and h, and four points (xw, y-
h), (x + w, yh), (xw, y + h), (x
The same operation is performed by scanning all the pixels in a square (5 × 5 pixels in this case) having the vertices of + w, y + h). This operation is repeated, and if w = wm or h = hm, ma
xv is adopted as the 8-bit restoration value of the target pixel (steps S105 to S108 in FIG. 2).

When the relation of maxv> minv is satisfied on the way, four points (x
−w, y−h), (x + w, y−h), (x−w, y +)
h), the maximum value of the value obtained by multiplying the lighting pixel ratio in the square having the vertices of (x + w, y + h) by 255 and the maxv value one step before is calculated as the 8-bit restored value of the pixel of interest (step in FIG. 2). S110 to S112).

For example, regarding the binary background value and the threshold value of the corresponding screen matrix shown in FIG. 4, the maximum value maxv and the OFF value of the threshold value corresponding to the ON dot in the 3 × 3 range are first set for the target pixel. The minimum value minv of the threshold values corresponding to the dots is obtained.

In this case, maxv = 48, minv = 8
3, and since maxv ≦ minv, the scanning range is expanded, and the same process is performed in a 5 × 5 range.
65, minv = 77 are obtained. 8 bits of the pixel of interest
Assuming that the value is between maxv and minv, it is possible to narrow down the 8-bit value by continuing this process while expanding the range.

Next, when a similar process is performed in a 7 × 7 range,
maxv = 110, minv = 77, and maxv> m
Since inv holds, the maxv value 65 one step before, and
Approximate 8 calculated from the dot lighting ratio 5/25 at that time
Approximate the maximum value with bit value 5/25 * 255 = 51 8b
Adopted as the it value. In this case, since 65 and 51 are used, 65 is adopted as an approximate value.

The above steps are repeated while moving the pixel of interest, and if an 8-bit restored value of the background is obtained, a translucent color can be obtained through the average value calculation step and the binarization processing step. The 8 bit restored image in this way is
It is excellent in gradation reproducibility in that approximation of the threshold is performed while increasing the range as long as the threshold relation is correct.

Actually, in a gradation portion or the like, the gradation for the maximum scanning range can be ensured. That is, if wm = hm = 4, (4 * 2 + 1) * (4
* 2 + 1) = 81 gradations can be secured.

Further, when the threshold value relationship is not satisfied, the threshold value of the target pixel is immediately calculated from the lighting ratio of the peripheral pixels, so that the reproducibility of the sharp portion of the original image is excellent.

When the multi-value restored image is subjected to the translucency processing, the same screen is used for the second screen again.
Since the image is converted into a value image, roughness does not occur, and actual reduction in resolution is hardly noticeable.

Further, by using the second method, it is possible to reduce the conventional roughness in the translucent processing even if a blue noise mask or the like is used for the binarized screen in addition to the halftone dots.

Note that the 8-bit value calculation step is not limited to the above-described embodiment, and it is also possible to reduce the increment of the scanning range. Also, a form other than a square may be used. It is also possible to use the second method in combination with the first method.

For example, when the foreground value is a dark color, the influence of the background color is reduced, so that it is possible to use the method of performing multi-value restoration by the first method. Alternatively, when the foreground color is dark, the values of wm and hm can be set to a small value to speed up multi-value restoration.

The above description has been made for the case where the ROP processing is performed with one bit. However, when the ROP processing is performed with eight bits, the configuration is as shown in FIG. That is, this configuration includes a PDL interpretation unit 11, a determination unit 12, a ROP processing unit 13, a drawing memory 15, a color conversion unit 19, and a binarization processing unit 1.
PDL is processed in the order of 3.

The translucent processing section 16 determines the ROP code (NOTXORPE
N, MERGEPEN, NOTXORPEN), ROP
The processing of reading the RGB 8-bit foreground sent to the processing unit 13 and reading the RGB 8-bit background already stored in the drawing memory 15 and taking an average value is performed (average value processing 18). Thereafter, the obtained image is input to the drawing memory 15 (the image mask input from the PDL interpretation unit 11 is not used). By performing this process, the translucent image can be accurately reproduced.

[0060]

As described above, the present invention has the following effects. In other words, when a translucent object is drawn by the ROP process, it is detected and the translucent color is created by calculating the average value of the background and the foreground. This makes it possible to suppress hard-to-see moiré patterns and roughness. As a result, an arbitrary screen matrix can be used, and it is not necessary to create a screen matrix in consideration of interference with an image mask, and it is possible to design a screen with normal image quality giving priority.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating an embodiment.

FIG. 2 is a flowchart illustrating the flow of the present embodiment.

FIG. 3 is a diagram illustrating a square scanning target including a target pixel.

FIG. 4 is a diagram illustrating a specific example of multi-level restoration.

FIG. 5 is a block diagram illustrating another embodiment.

FIG. 6 is a diagram illustrating a translucent process using an average value.

FIG. 7 is a diagram illustrating translucent processing by a logical operation.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Image processing apparatus, 2 ... Image output part, 11 ... PDL interpretation part, 12 ... Judgment part, 13 ... Binarization processing part, 14 ... ROP
Processing unit, 15: drawing memory, 16: translucent processing unit, 17
... Multi-value restoration unit, 18 ... Average value processing unit

Continued on the front page F term (reference) 2C087 AA16 AB05 BA03 BA12 BD05 BD07 2C262 AA24 AB01 AB13 AC07 BB03 BB15 CA04 DA06 GA23 5B057 AA11 CA01 CA07 CA12 CA16 CB01 CB07 CB12 CB16 CE08 CE17 5B080 DA01 FA17 GA25 5C076 AA

Claims (8)

[Claims] 1. An image processing apparatus for interpreting a page description language sent from a host computer and converting the page description language into print data in a print output format, wherein the page description language performs a semi-transparent process on a foreground image and a background image by a logical operation. Determining means for determining whether the page description language is to be subjected to semi-transparent processing by a logical operation by the determining means; And a translucent processing means for calculating an average value of the translucent image data to generate translucent image data. 2. The translucent processing means restores a binary image of the foreground image and the background image by a conversion process to print data before the translucent process into multivalue image data. 2. The image processing apparatus according to claim 1, further comprising a value restoring unit. 3. The multi-level restoration means according to claim 2, wherein a multi-level pixel value is obtained based on a threshold value of a binarization threshold value matrix corresponding to the binarized pixel and binary data of the pixel. The image processing device according to claim 2. 4. The multi-level restoring means converts a threshold value of a binarization threshold matrix corresponding to a peripheral pixel centered on one of the binarized pixels and binary data of the center pixel. 3. The image processing apparatus according to claim 2, wherein a multi-valued pixel value is determined based on the multi-valued pixel value. 5. An image processing method for interpreting a page description language sent from a host computer and converting the page description language into print data in a print output format, wherein the page description language performs a semi-transparent process on a foreground image and a background image by a logical operation. Determining whether the page description language is to be subjected to translucent processing by a logical operation, and calculating an average value of corresponding pixels of the foreground image and the background image. Converting the image data into translucent image data. 6. When performing the translucent processing, of the foreground image and the background image, those that have been binarized by a conversion process to print data before the translucent processing are restored to multi-valued image data. 6. The image processing method according to claim 5, wherein: 7. The multi-valued restoration data is calculated based on a threshold value of a binarization threshold value matrix corresponding to a binarized pixel and the binary data of the pixel. Image processing method. 8. The multi-level restoration data is converted into a threshold value of a binarization threshold matrix corresponding to peripheral pixels centered on one of the binarized pixels and binary data of a pixel at the center thereof. 7. The image processing method according to claim 6, wherein the calculation is performed based on: