JPH0728986A - Picture synthesis processor - Google Patents

Abstract

PURPOSE:To reduce the transfer amount of data required for obtaining the picture element values of synthetic picture elements and to reduce the arithmetic operation processing amount of transparent synthesis in a picture processor for performing the transparent synthesis. CONSTITUTION:Foreground picture elements are classified into the three kinds of transparent/non-transparent/translucent depending on the values of the transparency information (Fa) of the foreground picture elements among inputted picture data (S11). Corresponding to the classified kinds, as the picture element value after the synthesis, the picture element value Rabgr of background picture elements is set when the foreground picture elements are transparent (S15 and S16), the picture element value Fbgr of the foreground picture elements is set at the time of the non-transparent (S17 and S18) and the picture element value after the synthesis is calculated in transparent synthesis when the foreground picture elements are translucent (S1D and S18). Thus, only the picture element value of the background picture elements is transferred in the case of the transparent, only the picture element value of the foreground picture elements is transferred in the case of the non-transparent and the entire data required for a transparent synthesis arithmetic operation are transferred only in the case of the translucent.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image synthesizing processing device capable of synthesizing watermarks of multi-valued images.

[0002]

2. Description of the Related Art As a conventionally known image compositing method, for example, Volu of Computer Graphics Volu
me 18, Number 3 (July 198
4) "Composing Digital Ima
As described in “ges”, by setting the pixel value and the transparency for each pixel of the image, when performing the compositing process between the images, the pixel value of each pixel is calculated according to the transparency value. There is a technique for watermark composition that changes the mixture ratio so that the background image can be seen through through the foreground image.Watermark composition is performed between the foreground image and the background image by editing the image. In this case, conventionally, the color information, the brightness information, and the like of the foreground image and the background image and the mask information for watermark composition are given to the respective processing units to perform the watermark composition.

[0003]

However, in the conventional technique, not only the color information, the brightness information and the like (hereinafter, image information) of the image but also the mask information is given to the processing unit in the same format as the image data. , When watermark composition is performed,
1.3 times (pixel information for each of the foreground and background) without watermark composition (for each of the foreground and background, three pieces of pixel information in the case of full color, one piece of pixel information in the case of a gray image) 2 from 3 and mask information 1)
There is a drawback in that about twice as much data needs to be transferred (one for pixel information and one for mask information for each foreground and background).

Therefore, an object of the present invention is to provide an image synthesizing processing apparatus capable of reducing the amount of data transfer when performing watermark synthesizing.

[0005]

The present invention (Claim 1) includes:
In an image processing device having a watermark composition function of a multi-valued image, foreground pixels are classified into three types of transparent, opaque, and semi-transparent according to the transparency information of the foreground pixels, and the content of the process is switched for each classification. Characterize. Here, “transparent” is an attribute of a pixel having a value of only a background pixel after combination, and “opaque” is an attribute of a pixel having a value of only a foreground pixel after combination, and “translucent” is an attribute of the pixel. It means the attribute of a pixel that has a value other than opaque.

According to the present invention (claim 2), in an image processing apparatus having a function of watermark composition of a multi-valued image, foreground pixels are classified into three types of transparent, opaque and semi-transparent according to the transparency information of the foreground pixels. For the semi-transparent foreground pixels, the background pixels are classified into three types of transparent, opaque, and semi-transparent according to the transparency information of the background pixels, thereby classifying the input pixel combinations into five types. However, the content of the processing is switched for each classification.

In each of the above inventions, the transparency information can be transferred to the processing unit in a compressed format.

[0008]

The pixel value (Dbgr) of the watermark-combined image of the image processing apparatus for watermark composition is Dbgr = {(1-Fa) * Ra * Rbgr + Fa * F as shown in equation (2) described later.
It is represented by bgr} ÷ Da. Here, Fa is the transparency of the foreground pixel, Ra is the transparency of the background pixel, Fbgr is the pixel value of the foreground pixel, and Rbg.
r is the pixel value of the background pixel, and Da is the transparency after combination.
When Fa or Ra is 0, it is transparent, when 1 is opaque, and 0
It is translucent when it is between 1 and 1. As shown in the table of FIG. 3 described later, when the foreground pixel is transparent, the pixel value D after composition is D.
bgr becomes the pixel value Rbgr of the background pixel, and when the foreground pixel is opaque, the combined pixel value Dbgr becomes the pixel value Fbgr of the foreground pixel. Only when the foreground pixel is semi-transparent, the calculation by the equation (2) is necessary. Therefore, the image processing apparatus of the present invention (Claim 1) classifies the foreground pixel into three types of transparent / opaque / semi-transparent according to the value of the transparency information (Fa) of the foreground pixel in the input image data, According to the classified type, the pixel value of the background pixel is used when the foreground pixel is transparent, and the pixel value of the foreground pixel is used when the foreground pixel is transparent, as the combined pixel value. The arithmetic processing is not performed, and the combined pixel value is obtained by the equation (2) only when the foreground pixel is semitransparent. By doing so, only the pixel value of the background pixel needs to be transferred in the case of transparent, and only the pixel value of the foreground pixel needs to be transferred in the case of opaque, and only in the case of semi-transparent formula (2). Since all the data required for the calculation of is transferred, it is possible to reduce the transfer amount of the data necessary for obtaining the pixel value of the composite pixel and also to reduce the calculation processing amount of the watermark composition.

The present invention (Claim 2) is a further refinement of the classification of the above-mentioned invention (Claim 1), in which foreground pixels are classified into three types of transparent / opaque / semi-transparent. If it is semi-transparent, it is transparent /
It is classified into three types, opaque and translucent. Then, as shown in the table of FIG. 4 which will be described later, it can be classified into five types depending on the combination of transparency information of the foreground pixels and the background pixels. Besides foreground pixels being transparent and opaque, foreground pixels are semi-transparent,
In the case of the classification in which the background pixel is transparent, the pixel value of the foreground pixel can be used as it is as the pixel value after combining, so that the amount of data transfer can be further reduced and the watermark combining can be performed as compared with the invention of claim 1. The processing can be simplified.

Although the transparency information of an image is given in the same format as the image information, the transparent and opaque portions of the transparency information of an image often appear continuously in many images. As in (Claim 3), the transparent and opaque portions can be further reduced in the transfer data by adopting a compression format such as run length compression.

[0011]

FIG. 2 is a block diagram showing the schematic construction of an embodiment of the present invention. This embodiment is a system having a configuration in which an image processing apparatus 20 according to the present invention is connected to a host-side computer in which a CPU 21, a memory 22, an input / output controller 23, etc. are connected by a bus 24. The image processing device 20 includes an input buffer 201, a pre-processing unit 202, an image processing unit 203, and a post-processing unit 204.
And an external output interface 209.

The input buffer 201 is an input buffer for transferring data from the host to the image processing apparatus 20. The pre-processing unit 202 is a unit that performs pre-processing corresponding to the processing performed by the data and the image processing unit 203 when passing data from the input buffer 201 to the next image processing unit 203. Specifically, the foreground image is classified into three types of transparent / opaque / semi-transparent according to the value of the transparency information from the input buffer 201. For a foreground image that is semi-transparent, the transparency information of the background image continues. A process of classifying the background image into three types of transparent / opaque / semi-transparent is performed according to the value.

The image processing unit 203 switches the processing according to the result classified by the pre-processing unit 202 and performs image processing on the transferred data. The post-processing unit 204 performs compression processing and data formatting on the data processed by the image processing unit 203. Output buffer 20
An output buffer 5 transfers data from the post-processing unit 204 to the host side or the external output interface 209. The external output interface 209 is an interface unit for outputting the processing result of the image processing apparatus to a device other than the host.

Next, in the image composition processing apparatus of this embodiment,
The details of the processing executed will be described. Below, the input image is a 32-bit image data in which the foreground image and the background image each have an 8-bit transparency and an 8-bit pixel value (for example, three types of blue component, green component, and red component). An embodiment of the present invention will be described by taking the case as an example. FIG. 1 is a flow chart showing the processing of the embodiment. It should be noted that in FIG. 1, a thin line frame represents processing performed by the preprocessing unit 202, and a thick line frame represents processing performed by the image processing unit 203. First,
The watermark synthesizing process will be described. The watermark synthesizing process is performed by changing the mixture ratio of the pixel values of the foreground image and the background image for two input images, the foreground image and the background image, according to the respective transparency information. Is a process for performing. In the watermark-combined image, the background image can be seen through the foreground image depending on the transparency value.

Hereinafter, Fa ... transparency of the foreground image (normalized to Fa / 255 in the case of 8 bits), Fbgr ... pixel value of the foreground image (in this case, blue component, green component, and red component are shown) , Ra ... Transparency of background image (Ra / 25 for 8-bit)
5), Rbgr ... Pixel value of background image (in this case, showing blue component, green component, red component) Da ... Transparency of image after watermark composition (in the case of 8 bits,
Da / 255 is normalized, Dbgr ... Pixel value of the image after watermark composition (in this case,
Abbreviated as blue component, green component, and red component)

The transparency Da of the image after watermark composition is expressed by the following equation. Da = (1−Fa) × Ra + Fa (Equation 1) Here, the transparency is transparent when the value is 0, opaque when the value is 1, and translucent when 0 <transparency <1. Show. Further, the pixel value Dbgr of the image after watermark composition is represented by the following equation. Dbgr = {(1-Fa) * Ra * Rbgr + Fa * Fbgr} / Da (Equation 2)

The classification for switching the processes, which is a feature of the present invention, will be described below. (Equation 1) and (Equation 2) are classified into three types of transparency 0 (that is, transparent), transparency 1 (that is, opaque), and other transparency (that is, semitransparent) for each foreground image and background image ( Steps S11 and S12) and nine combinations are obtained as shown in the table of FIG. The table of FIG. 3 shows the transparency Fa of the foreground image, the transparency Ra of the background image, the transparency Da after combination, and the pixel value Dbgr after combination for each combination.

The nine combinations shown in the table of FIG. 3 are summarized by the transparency after combination and the pixel value after combination, and can be divided into five cases shown in Table 2 of FIG. In this case, the foreground pixel is transparent having only the value of the background pixel after the composition, and opaque having the value of only the foreground pixel after the composition, depending on the transparency information of the foreground pixel.
And classified into three types of semi-transparent with values other than those,
And for the semi-transparent foreground pixel, the background pixel is
It is classified into three types of transparent, opaque, and semi-transparent according to the transparency information of the background pixel.

The data required for the process and the outline of the required process for the five cases shown in the table of FIG. 4 are shown in the table of FIG. When the foreground is transparent, the foreground is opaque, and the foreground is semi-transparent and the background is transparent, the process of watermark composition is only data transfer. Also, as shown in the table of FIG. 3 and the table of FIG. 4, the amount of calculation is smaller when the foreground is semitransparent and the background is opaque than when the foreground is semitransparent and the background is semitransparent. I'm sorry. When watermark composition is performed, only a part of the foreground image is semi-transparent, and many other parts are opaque or transparent. The reduction of the transfer amount of the data contributes to the improvement of the performance of the entire processing. An example thereof will be described in the second embodiment described later.

In the operation of the image composition processing apparatus of the first embodiment, first, the transparency information of the foreground pixel is sent from the input buffer 201 of FIG. 2 to the preprocessing section 202 (step S14 of FIG. 1).

The preprocessing unit 202 determines the transparency (step S11), and the transparency is 0 (that is, transparent) and 1
3 (ie opaque) and other (ie semi-transparent)
Divided into seeds. When the transparency information is represented by 8 bits, "transparency is 0" is a value of "00" in hexadecimal notation.
h ". Similarly," transparency is 1 "is" FFh ".
Is shown.

When the transparency Fa of the foreground pixel is 0, the transparency information of the background pixel and the pixel value (for example, the value of each component of red, blue and green) Rabgr is read (step S1).
5). Transparency information of the read background pixel and pixel value R
The abgr is sent to the image processing unit 203. Image processing unit 2
Reference numeral 03 denotes the input transparency information and the pixel value Rabgr, respectively, that is, the processed value, that is, the transparency of the composite pixel and the pixel value Dabgr.
(Step S16) (step S1)
E).

When the transparency Fa of the foreground pixel is 1, the pixel value of the foreground pixel (for example, the value of each component of red, blue, and green) Fb
gr is read (step S17). The read transparency information of the foreground pixel and the pixel value Fabgr are sent to the image processing unit 203. The image processing unit 203 uses the processed transparency value Dab as the transparency value and the pixel value Fabgr that have been input.
output as gr (step S18) (step S1)
E).

When the transparency of the foreground pixel is other than 0 or 1 (that is, semitransparent), the transparency information Ra of the background pixel is read (step S19), the transparency is determined (step S12), and the transparency is 0 ( That is, it is divided into three types: transparent), 1 (that is, opaque), and other (that is, semitransparent).

When the foreground pixel is semi-transparent and the background pixel is transparent, the pixel value Fbgr of the foreground pixel is read (step S17). The read transparency information of the foreground pixel and the pixel value Fabgr are sent to the image processing unit 203. The image processing unit 203 receives the input transparency information and the pixel value Fa
Each of bgr is output as a processed value Dabgr (step S18) (step S1E).

If the foreground pixel is semi-transparent and the background pixel is opaque, the pixel value F for each of the foreground pixel and the background pixel.
bgr ,. Rbgr is read (step S1)
A). The read transparency information and the pixel value of each pixel are sent to the image processing unit 203.

The image processing unit 203 determines the processed value Da of transparency.
1 (when the transparency information is represented by 8 bits, "
FFh ″), and the processed value Dbgr of the pixel is calculated according to (Equation 3) based on the input transparency information and the pixel value (step S1B). Dbgr = {(FFh−Fa) × Rbgr + Fa × Fbgr} ÷ FFh .. (Equation 3) Note that in Equation 3, Fa, Rbgr, and Fbgr are each 8b.
It is represented by it. Further, "FFh" is a hexadecimal notation. Processed value Da (=
FFh) and the calculated pixel value Dbgr is the processed value Dab.
It is output as gr. (Step S1E)

When the foreground pixel is semi-transparent and the background pixel is semi-transparent, the pixel value F is determined for each of the foreground pixel and the background pixel.
bgr and Rbgr are read (step S1C).
The read transparency information and the pixel value of each pixel are sent to the image processing unit 203. The image processing unit 203 receives the input transparency information Fa and Ra and the pixel values Fbgr and Rbg.
From r, the processed value Da of transparency is calculated according to (Equation 4) (step S1D). Further, the processed value Dbgr of the pixel is calculated according to (Equation 5) based on the input transparency information Fa and Ra and the pixel values Fbgr and Rbgr.

Da = {(FFh−Fa) × Ra + FFh × Fa} ÷ FFh (Equation 4), Dbgr = {(FFh−Fa) × Ra × Rbgr + FFh × Fa × Fbgr} ÷ {(FFh−Fa) × Ra + FFh × Fa} (Equation 5) Fa, Ra, Rbgr, F in (Equation 4) and (Equation 5)
Each bgr is represented by 8 bits. ”
FFh "is a hexadecimal notation.

The calculated transparency processing value Da and pixel value Dbgr are output as the processing values (step S1).
E). The above operation is performed until there are no more pixels to be processed (step S13).

(Second Embodiment) Next, another embodiment based on the above-mentioned first embodiment, that is, a second embodiment will be described. In the first embodiment, the transparency information of the image is given in the same format as the image information, but the transparent and opaque portions of the transparency information of the image appear continuously in many images, so that the transparency information is transparent. And the opaque part are run length compressed, and a header for transparent / opaque discrimination and the number of consecutive pixels are transferred. A semitransparent portion is a header for the semitransparent discrimination and consecutive pixels. The transfer data can be further reduced by transferring the number and the actual value of the transparency information of each pixel. The second embodiment is an example of the case of performing the run length compression. Similar to the first embodiment, its schematic configuration is shown in FIG. 2, and the processing functions shown in FIGS. 6 to 14 are added for that purpose.

The operation of the second embodiment when run length compression is performed on transparent and opaque pixels will be described below.
6 to 14 are flowcharts showing the processing of the second embodiment. 15A, 15B, and 15E show examples of the transparency information used in the second embodiment.

In the second embodiment, the compressed transparency information and the pixel values of the foreground pixel and the background pixel are all 8b.
It is represented by it. The compressed transparency information is 2b as shown in FIGS. 15 (a) (b) (C).
The transparency type information (it, Fm, Rm) of it and the run length length (Fl, Rl) of 6 bits are set to (Fm, Fm) for the foreground pixel.
In the form of l), the background pixel is assumed to be in the form of (Rm, Rl).

FIG. 6 shows the flow of the pixel classification process based on the transparency information in the preprocessing unit 202. From the input buffer 201 of FIG. 2, the transparency information F of the foreground pixel
m and Fl are sent to the preprocessing unit 202. Preprocessing unit 202
Then, the transparency type information Fm of the foreground pixel is stored in the foreground pixel type variable Fstate. Further, the run length length Fl of the foreground pixel is stored in the foreground unprocessed pixel run length length variable (Flength). The position information Fmaskpoint in the transparency information buffer of the foreground pixel is changed to indicate the position of the next transparency information (step S6).
1).

Transparency information Rm and Rl of the background pixel is sent from the input buffer 201 shown in FIG.
In the preprocessing unit 202, the background variable type variable Rstate
The background pixel transparency type information Rm is stored in. Also, the run length length Rl of the background pixel is stored in the background unprocessed pixel run length length variable Rlength. Position information Rmaskpoint in buffer of background pixel transparency information
Is changed to indicate the position of the next transparency information (step S62).

When the condition judgment shown in step S63 of FIG. 6 shows that the foreground pixel type variable is transparent and the background pixel type variable is also transparent (Fstate = 0, Rstate = In 0), the image processing unit 203 performs the processing shown in the flowcharts of FIGS. 7 and 14.

That is, the transparency processing value Da is set to 0 in the image processing unit 203 (step S64). The pixel value Rbgr (for example, the value of each component of red, blue, and green) of the background pixel is read out, and the position information Fpixpoint in the buffer of the pixel information of the foreground pixel indicates the position of the next pixel information. The position information Rpixpoint in the buffer of the pixel information of the background pixel is changed so as to indicate the position of the next pixel information (step S7).
1).

The pixel value of the background pixel is used as the processed value Dbgr of the pixel information (step S72), and the values of the transparency information and the pixel information are output as the processed value Dabgr. The values of the foreground unprocessed pixel run length variable Flength and the background unprocessed pixel run length variable Rlength are each decreased by one (step S73).

Foreground unprocessed pixel run length length variable Fle
It is determined whether ngth is 0 (step S
4) If the result of that determination is that Length is 0, the flow proceeds to step S62. Then, the transparency information Rm, Rl of the next background pixel is read, and the processing is continued.

Foreground unprocessed pixel run length length variable Fle
If ngth is not 0, it is determined whether or not the background unprocessed pixel run length length variable (Rlength) is 0 (step S5). If Rlength is not 0, the process proceeds to step S71 and the position Then, the pixel value Rbgr of the background pixel is read, and the process is continued (step S75).

Background unprocessed pixel run length length variable (Rl
If the processing of all pixels has been completed, the processing ends (step S141), and if not completed, the processing returns to step S61, and the transparency information Fm of the next foreground pixel, Fl is read and processing is continued.

In the condition judgment shown in step S61 of FIG. 6, when the type variable of the foreground pixel is transparent and the type variable of the background pixel is opaque (Fstate = 0, Rstate = In 1), the image processing unit 203 performs the processing shown in the flowcharts of FIGS. 7 and 14.

The image processing unit 203 uses the processed value of transparency (D
The value ab is set to FFh (step S65), and the pixel value Rbgr of the background pixel is read out. Hereinafter, step S in FIG.
71 to step S75 and step S141 are performed. First, Fstate = 0, Rstate
The processing is the same as that when = 0, and is as already described.

In the condition judgment shown in step S63 of FIG. 6, when the type variable of the foreground pixel is transparent and the type variable of the background pixel is semitransparent (Fstate = 0, Fstate) = Other)
The image processing unit 203 performs the processing shown in the flowcharts of FIGS. 8 and 14.

In the image processing unit 203, the transparency information Ra of the background pixel is read, and the position information Rmaskpoint in the buffer of the transparency information of the background pixel is changed to indicate the position of the next transparency information (step S8).
1). In the image processing unit 203, the processed value D of the transparency is
The transparency information Ra of the background pixel is used for a (step S82). The pixel value Rbgr of the background pixel is read. Position information Fpi in the buffer of pixel information of foreground pixels
The xpoint is changed to indicate the position of the next pixel information. Position information R in the pixel information buffer of background pixels
The pixpoint is changed to indicate the position of the next pixel information (step S83).

The pixel value Rbgr of the background pixel is used as the processed value Dbgr of the pixel information (step S84), and the processed value Dabgr of the transparency information and the pixel information is output. Also,
The values of the foreground unprocessed pixel run length length variable Flength and the background unprocessed pixel run length length variable Rlength are each decreased by one (step S85).

Next, it is determined whether or not the foreground unprocessed pixel run length length variable Length is 0 (step S86). If Length = 0, step S86 is executed.
62, the transparency information Rm of the background pixel located next,
Read Rl and continue processing.

As a result of the determination in step S86, Fleng
If th is not 0, it is determined whether the background unprocessed pixel run length length variable Rlength is 0 (step S87). If Rlength is not 0, the next background pixel of the background pixel The transparency value Ra is read and the processing is continued (step S81).

Background unprocessed pixel run length length variable Rle
When ngth becomes 0, the processing is ended if all the pixels have been processed (step S141), and if not completed, the process returns to step S602 and is positioned next.
The transparency information Fm and Fl of the foreground pixel is read, and the processing is continued.

In the condition judgment shown in step S63 of FIG. 6, when the type variable of the foreground pixel is opaque and the type variable of the background pixel is transparent or opaque (that is, Fstate = 1, Rstate =
In the case of 0 and in the case of Fstate = 1 and Rstate = 1), the processes shown in the flowcharts of FIGS. 9 and 14 are performed.

In the image processing unit 203, the processed value Da of transparency is set to FFh (step S66), and
The pixel value Fbgr of the foreground pixel is read out. Position information Fpixpoint in buffer of pixel information of foreground pixel
Is changed to indicate the position of the next pixel information. Position information Rpixpoi in the buffer of pixel information of background pixels
nt is changed to indicate the position of the next pixel information (step S91).

The pixel value Fbgr of the foreground pixel is used as the processed value Dbgr of the pixel information (step S92), and the processed value Dabgr of the transparency information and the pixel information is output as the processed value. Foreground unprocessed pixel run length length variable Flengt
h and background unprocessed pixel run length length variable Rlength
Is decreased by one (step S9).
3).

Foreground unprocessed pixel run length length variable Fle
It is determined whether ngth is 0 (step S9).
4). If the result of the determination is that Length = 0, the transparency information Rm, Rl of the background pixel located next is read in step S62, and the processing is continued.

As a result of the above judgment, when Length = 0, the background unprocessed pixel run length length variable Rlength
It is determined whether th is 0 (step S9).
5). If Rlength is not 0 as a result of the determination, the process returns to step S91, the pixel value Fbgr of the foreground pixel located next is read, and the process is continued.

Background unprocessed pixel run length length variable Rle
When ngth becomes 0, if the processing of all the pixels is completed, the processing is ended (step S141), and if not completed, the processing returns to step S61 of FIG. 6 and the transparency information Fm of the foreground pixel located next. , Fl, and continue processing.

In the condition judgment shown in step S63 of FIG. 6, when the type variable of the foreground pixel is opaque and the type variable of the background pixel is semitransparent (that is, Fstate = 1). , Rstate = Other
Processing), the processing shown in the flowcharts of FIGS. 10 and 14 is performed.

In the image processing unit 203, the transparency processing value D
a is set to FFh (step S66), and the position information Rmaskpoint in the buffer of the transparency information of the background pixel is set to
It is changed to indicate the position of the next transparency information (step S101). The pixel value Fbgr of the foreground pixel is read out. Then, the position information Fpixpoint in the pixel information buffer of the foreground pixel is changed to indicate the position of the next pixel information. Further, the position information Rpixpoint in the buffer of the pixel information of the background pixel is changed to indicate the position of the next pixel information (step S102).

The pixel value Fbgr of the foreground pixel is used as the processed value Dbgr of the pixel information (step S103), and the transparency information and the value Dabgr of the pixel information are output as the processed value. Foreground Unprocessed Pixel Run Length Length Variable Length
And the background unprocessed pixel run length length variable Rlength is decremented by one (step S104).

Foreground unprocessed pixel run length length variable Fle
It is determined whether ngth is 0 (step S10).
5) If the result of the determination is that Length = 0, the process moves to step S62 in FIG. 6, the transparency information Rm, Rl of the background pixel located next is read, and the processing is continued.

If the result of the above determination is that Length = 0, the background unprocessed pixel run length length variable Rlen
It is determined whether gth is 0 (step S10).
6) As a result of the determination, if Rlength is not 0, the process proceeds to step S101 again, the position information Rmaskpoint in the buffer of the transparency information of the background pixel is changed to indicate the position of the next transparency information, and the foreground pixel is changed. The pixel value Fbgr is read and the process is continued.

Background unprocessed pixel run length length variable Rle
When ngth becomes 0, the processing is completed if all pixels have been processed (step S141), and if not completed, the transparency information Fm of the next foreground pixel,
Fl is read and the process is continued (step S61).

When the condition judgment shown in step S63 of FIG. 6 indicates that the type variable of the foreground pixel is semi-transparent and the type variable of the background pixel is transparent (that is, Fstate = others). , Rstate = 0), the processing shown in the flowcharts of FIGS. 11 and 14 is performed.

In the image processing unit 203, the transparency information Fa of the foreground pixel is read, the position information Fmaskpoint in the buffer of the transparency information of the foreground pixel is changed to indicate the position of the next transparency information, and The pixel value Fbgr is read out. The position information Fpixpoint in the pixel information buffer of the foreground pixel is changed to indicate the position of the next pixel information (step S11).
1).

The image processing unit 203 determines the processed value Da of transparency.
Is used as the transparency value Fa of the foreground pixel, and the pixel value Fbgr of the foreground pixel is used as the processed value Dbgr of the pixel information (step S112), and the processed value Dabgr of the transparency information and the pixel information is used.
Is output as a processed value (step S113). Also,
The values of the foreground unprocessed pixel run length length variable Length and the background unprocessed pixel run length length variable Rlength are each decreased by one (step S113).

Foreground unprocessed pixel run length length variable Fle
It is determined whether ngth is 0 (step S11).
4) If the result of this determination is that Length is 0, the flow moves to step S62, the transparency information Rm, Rl of the background pixel located next is read, and processing continues.

If Flength is not 0, it is determined whether or not the background unprocessed pixel run length length variable Rlength is 0 (step S115), and Rlength is determined.
If is not 0, the process returns to step S111 again. Then, the transparency information Fa of the foreground pixel is read, and the position information Fmaskpoi in the buffer of the transparency information of the foreground pixel is read.
nt is changed to indicate the position of the next transparency information,
The pixel value Fbgr of the foreground pixel is read out. Position information Fpixpoint in buffer of pixel information of foreground pixel
Is changed to indicate the position of the next pixel information, and the processing is continued.

Background unprocessed pixel run length length variable Rle
When ngth becomes 0, the processing is completed if all pixels have been processed (step S141), and if not completed, the transparency information Fm of the next foreground pixel,
Read Fl and continue processing.

In the condition judgment shown in step S63 of FIG. 6, when the type variable of the foreground pixel is semi-transparent and the type variable of the background pixel is transparent (that is, Fstate = others) , Rstate = 1), processing according to the flowcharts shown in FIGS. 12 and 14 is performed. In the image processing unit 203, the pixel value Rbgr of the background pixel is read, and the position information Rpixpoint in the buffer of the pixel information of the background pixel is changed to indicate the position of the next pixel information (step S121).

Further, in the image processing unit 203, the transparency information Fa of the foreground pixel is read, and the position information Fmaskpoint in the buffer of the transparency information of the foreground pixel is changed to indicate the position of the next transparency information. The pixel value Fbgr of the pixel is read. The position information Fpixpoint in the pixel information buffer of the foreground pixel is changed to indicate the position of the next pixel information (step S12).
2).

Further, the image processing unit 203 sets the transparency processing value Da to FFh, and calculates the processing value of the pixel according to (Equation 6) based on the input transparency information and the pixel value (step S123). Dbgr = {(FFh−Fa) × Rbgr + Fa × Fbgr} ÷ FFh (Equation 6) (Fa, Rbgr, Fbgr are each 8 in (Equation 6))
It is assumed to be expressed in bits. "FFh" is 16
It is a notation. )

The processed value Da = FFh of the set transparency and the calculated pixel value Dbgr are output as the processed value Dabgr. Foreground unprocessed pixel run length length variable Fle
ngth) and the background unprocessed pixel run length length variable Rle
The ngth value is reduced by one (step S1).
24).

Foreground unprocessed pixel run length length variable Fle
It is determined whether ngth is 0 (step S12).
5) As a result of the determination, when the Length is 0, the process proceeds to step S62, the transparency information Rm and Rl of the background pixel located next is read, and the process is continued.

Foreground unprocessed pixel run length length variable Fle
If ngth is not 0, it is determined whether or not the background unprocessed pixel run length length variable Rlength is 0 (step S126). As a result of the determination, Rlength is determined.
If h is not 0, the process returns to step S121 again, the pixel value Rbgr of the background pixel is read, and the position information Rpixpoint in the buffer of the pixel information of the background pixel is
It is changed to indicate the position of the next pixel information, and the processing is continued.

Background unprocessed pixel run length length variable Rle
When ngth becomes 0, if the processing of all the pixels is completed, the processing is ended (step S141), and if not completed, the process proceeds to step S61, and the transparency information Fm, Fl of the next foreground pixel is set. Read and continue processing.

When the condition judgment shown in step S63 of FIG. 6 indicates that the type variable of the foreground pixel is semitransparent and the type variable of the background pixel is semitransparent (that is, Fstate = Others, Rstate = other cases), the processing according to the flowcharts shown in FIGS. 13 and 14 is performed.

The image processing unit 203 reads the transparency information Ra of the background pixel and changes the position information Rmaskpoint in the buffer of the transparency information of the background pixel so as to indicate the position of the next transparency information (step S131).

The image processing unit 203 reads the pixel value Rbgr of the background pixel, and modifies the position information Rpixpoint in the buffer of the pixel information of the background pixel so as to indicate the position of the next pixel information (step S132). .

The image processing unit 203 reads the transparency information Fa of the foreground pixel, changes the position information Fmaskpoint in the buffer of the transparency information of the foreground pixel so as to indicate the position of the next transparency information, and changes the pixel position of the pixel of the foreground pixel. Value of Fbgr
Read out. The position information Fpixpoint in the pixel information buffer of the foreground pixel is changed to indicate the position of the next pixel information (step S133).

The image processing unit 203 calculates the processed value Da of the transparency according to (Equation 7) based on the input transparency information and the pixel value. Further, the processed value Dbgr of the pixel is calculated according to (Equation 8) based on the input transparency information and the pixel value (step S1007).

Da = {(FFh−Fa) × Ra + FFh × Fa} ÷ FFh (Equation 7), Dbgr = {(FFh−Fa) × Ra × Rbgr + FFh × Fa × Fbgr} ÷ {(FFh−Fa) × Ra + FFh × Fa} (Equation 8) (Note that in (Equation 7) and (Equation 8), Fa, Ra, Rbgr,
Each Fbgr is represented by 8 bits. "FFh" is a hexadecimal notation. )

The calculated transparency processing value Da and the calculated pixel value Dbgr are output as the processing value Dbgar. The values of the foreground unprocessed pixel run length variable Flength and the background unprocessed pixel run length variable Rlength are each decreased by one (step S135).

Foreground unprocessed pixel run length length variable Fle
It is determined whether or not ngth is 0, and when Flenth is 0, transparency information Rm and Rl of the background pixel located next is read, and the processing is continued (step S13).
6).

Foreground unprocessed pixel run length length variable Fle
If ngth is not 0, it is determined whether or not the background unprocessed pixel run length length variable Rlength is 0 (step S137). If Rlength is not 0, the transparency information Ra of the background pixel is read again and the background pixel is read. Location information Rmaskpoint in the transparency information buffer
Is changed to indicate the position of the next transparency information, and the process is continued (step S131).

Background unprocessed pixel run length length variable Rle
When ngth becomes 0, the processing is ended if all pixels have been processed (step S141), and if not completed, the transparency information F of the next foreground pixel located next.
m and Fl are read and the processing is continued.

[0085]

As described above, according to the present invention (claims 1 and 2), the mask information for watermark composition is provided with a transparent part (a part of the background image only after composition) and an opaque part (after composition). Image information of the foreground image corresponding to the transparent part of the foreground image, and image information of the background image corresponding to the opaque part of the foreground image. In order to avoid unnecessary data transfer at the time of processing, for transparent and opaque parts of the foreground image, only the necessary part of the image information and the mask information are transferred. This has the effect of reducing the amount of data transfer and the calculation time in the case of performing it. Further, by further classifying, and by transferring the image information of the foreground image as a necessary part even for a part of the semi-transparent part (when the transparency of the background image is 0),
It is possible to further reduce the data transfer amount and the calculation time. Further, according to the present invention (Claim 3), the amount of data transfer can be further reduced by compressing the mask information of the transparent portion and the mask information of the opaque portion.

[Brief description of drawings]

FIG. 1 is a diagram showing a flow of processing performed in a first embodiment.

FIG. 2 is a diagram showing a schematic configuration of an embodiment of the present invention.

FIG. 3 is a diagram showing a combination of types of foreground pixels and background pixels classified by transparency.

FIG. 4 is a diagram showing five types of classification in which transparency after synthesis and pixel values after synthesis in the table of FIG. 3 are summarized.

FIG. 5 is a diagram showing data required for processing and an outline of processing.

FIG. 6 is a diagram showing a flow of processing performed by a preprocessing unit in the second embodiment.

FIG. 7 is a diagram showing a processing flow of an image processing unit when a foreground pixel is transparent and a background pixel is transparent or opaque in the second embodiment.

FIG. 8 is a diagram showing a processing flow of an image processing unit when a foreground pixel is transparent and a background pixel is semi-transparent in the second embodiment.

FIG. 9 is a diagram showing a processing flow of an image processing unit when a foreground pixel is opaque in the second embodiment.

FIG. 10 is a diagram showing a processing flow of the image processing unit when the foreground pixel is opaque and the background pixel is semi-transparent in the second embodiment.

FIG. 11 is a diagram showing a flow of processing of an image processing unit when a foreground pixel is semitransparent and a background pixel is transparent in the second embodiment.

FIG. 12 is a diagram showing a flow of processing of an image processing unit when a foreground pixel is translucent and a background pixel is opaque in the second embodiment.

FIG. 13 is a diagram showing a processing flow of the image processing unit when the foreground pixel is semitransparent and the background pixel is semitransparent in the second embodiment.

FIG. 14 is a diagram showing a process of determining the end of the process common to the processes of FIGS. 7 to 13;

FIG. 15 is a diagram showing an example of transparency information in the second embodiment.

[Explanation of symbols]

20 ... Image processing device 22 ... Memory 23 ... Input / output controller 201 ... Input buffer 202 ... Pre-processing unit 20
3 ... Image processing unit 204 ... Post-processing unit 205 ... Output buffer 20
6 ... CPU 209 ... External output interface

Claims (3)

[Claims] 1. An image processing device having a watermark compositing function of a multi-valued image, wherein a foreground pixel is transparent having a value of only a background pixel after composition, and only a foreground pixel is composed after composition according to transparency information of the foreground pixel. An image synthesizing processing device characterized by being classified into three types of opaque having a value and semi-transparent having a value other than those, and switching the content of processing for each classification. 2. An image processing apparatus having a watermark composition function of a multi-valued image, wherein a foreground pixel is transparent having a value of only a background pixel after composition and only a foreground pixel after composition, according to transparency information of the foreground pixel. The background pixels are classified into three types, that is, opaque having a value and semi-transparent having a value other than those, and the background pixels are classified into transparent, opaque, and semi-transparent according to the transparency information of the background pixels. An image synthesizing processing device characterized by classifying input pixel combinations into 5 types by classifying into 3 types and switching processing contents for each class. 3. The image composition processing apparatus according to claim 1 or 2, wherein the transparency information is transferred to the processing unit in a compressed format.