JPH07274198A - Image compositing device - Google Patents

Abstract

PURPOSE:To execute antialiasing by an inexpensive circuit. CONSTITUTION:A character and pattern video signal is stored in an image memory 1 and mixing ratio data corresponding to its each image element is stored in a mixing information memory 2. Then, a synchronization circuit 4 outputs a reading clock to a synchronization reading circuit 3. The synchronization reading circuit 3 generates the luminance signal, the chroma signal and the mixing ratio signal of the character and pattern video signal from the character and pattern video signal and mixing ratio data and outputs them to mixing circuits 5a and 5b. The mixing circuit 5a complementally varies gain and mixes the respective luminance signals of a background image and a character and pattern image in accordance with the mixing ratio signal and outputs it as the luminance signal of a mixed superimpose signal. On the other hand, the mixing circuit 5b complementally varies gain and mixes the respective chroma signals of the background image and the character and pattern image in accordance with the mixing ratio signal and outputs it as the chroma signal of the super impose signal.

Description

Detailed Description of the Invention

[0001]

[0001]

[0002]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image synthesizing apparatus for synthesizing and outputting a video signal such as a television signal, and particularly to a high quality character or figure digitally drawn or accumulated on a background image. Moreover, the present invention relates to a color superimposing device that superimposes inexpensively.

[0003]

[0002]

[0004]

2. Description of the Related Art Conventionally, when superimposing a character or a figure on a background image or performing chroma key processing, the background image, particularly the boundary between the background moving image and the inset image, is naturally displayed without a feeling of strangeness. It is important to make letters and figures look clear, and some proposals have been made for that purpose.

[0005]

As an analog synthesizing technique, a key signal, which is a trapezoidal wave gate signal in both horizontal and vertical directions, is used as an image signal in the vicinity of the boundary between two screens, and the image signal in the vicinity of the outside of the character graphic image is gated. Japanese Patent Publication No. 52-33933 is a special effect device for television screen composition that dissolves the image of, and changes the brightness and hue in a circle shape to perform inset composition to color the area around the soft chroma key and character figures. Is disclosed in.

[0006]

Further, by delaying the input signal, applying a weight and then performing non-additive mixing, a gate pulse wider than the input character signal and having a stepwise slope at the edge is formed. Then, the background signal is faded by this gate pulse, the character signal itself is delayed and stored in the gate pulse, and then by the inset synthesis, a sharp luminance change at the boundary between the edging portion of the character and the background image is caused. Japanese Patent Publication No. 54-6166 discloses a framed telop inserter for eliminating the invisible character.

[0007]

Both of these are related to edge processing when mixing a character graphic of an analog signal with a natural moving image. An analog image signal taken from a camera or the like and a key signal obtained by processing it are used. Was.

In addition to the need for a complicated analog circuit, there is a problem that jaggies occur when a digitally generated character / graphics signal is input.

[0009]

Further, the following technology is disclosed in the color superimposing apparatus of Japanese Patent Laid-Open No. 4-237778. An input composite signal that becomes a background image is separated into a luminance signal and a chroma signal (carrying color signal), and each of them is modulated and synthesized with a character figure whose band is limited to the luminance band and the chroma band. Then, after the modulated and combined luminance signal and chroma signal are combined into a composite signal (input character signal), the area / enlarged character / graphic signal is used as a key signal to superimpose the input character signal on the background image that is the input composite signal. Imposing

[0010]

This is because if the composite signal is switched and superimpose composition is performed, severe cross color and dot interference will be caused, and if luminance / chroma separation is performed and then superimpose composition is performed, the background will be generated at the luminance / chroma separation stage. The brightness /
After separating the chroma and performing superimposing composition, the background image other than the surrounding of the character is replaced with that of the composite signal to be used to prevent the deterioration of the image quality of the background image.

However, in this proposal, the image quality deterioration at the boundary between the character image and the character image that occurs even when the background image and the character image are separately combined with the luminance signal and the chroma signal (jaggies in the case of cross color or digitally generated character image). Etc.), nothing is considered. Further, in recent years, the number of devices that perform input / output while separating the luminance signal and the chroma signal is increasing, and it is no longer necessary to consider the image quality deterioration of the background image due to the luminance / chroma separation.

[0012]

As a device for preventing the occurrence of cross color, a video signal synthesizing device of Japanese Patent Laid-Open No. 5-161063 has been proposed. This is the part that superimposes the input composite signal (composite video signal) by using the key signal obtained by correcting the band by limiting the band of any input key signal according to the luminance signal component and the chroma signal component. By removing the chroma component, superimposing the luminance signal and then superimposing the chroma signal, you can prevent image quality deterioration (cross color) due to discontinuity of the chroma signal when switching the superimpose. There is.

[0013]

In this case, when the character / graphics signal is an analog signal such as that taken in by a camera, cross color or the like can be prevented, but since the background signal is subjected to chroma removal and two mixing processes, the image quality is deteriorated. Was prominent. Further, when digitally generated, there is a problem that jaggies, which will be described later, occur.

[0014]

A personal computer (PC)
Along with the development of, a device for analog superimposing a character graphic generated by a digital circuit on a background image has been proposed.

An example of this is the video signal conversion device disclosed in JP-A-61-134194. This first separates the composite input (background) signal into a luminance signal component and a chroma signal component. Then, R indicating the character graphic generated on the PC
The GB signal is encoded into a luminance signal and a color difference signal, the color difference signal is combined into a chroma signal, the luminance signals and the chroma signals are respectively switched, and then mixed and combined into a composite output signal. As a result, it is possible to obtain the effect on the cross color generated by the dot interference and the switching of the luminance signal.

However, since this is not a fundamental countermeasure, cross color is generated due to the discontinuity of the chroma signal, and the effect on flicker cannot be expected.

[0017]

Then, the problem of jaggies peculiar to a quantized image as described below remains unsolved. here,
Explain the causes of aliases and jaggies and how to deal with them.

Regarding the cause of the occurrence of alias or jaggies in digital images and anti-aliasing which is a countermeasure therefor, "Practical computer graphics--basic procedures and applications--" is a well-known document in Japan.
Supervised by Fujio Yamaguchi (Original author: David F. Rogers: 'Procedural
Elements for Computer Graphics'), etc., and frequently discussed in Japanese patent applications.

[0019]

In summary, the occurrence of aliasing is essentially due to the data being quantized. First
The problem of is that the number of scanning lines is finite, especially when the binarized graphic edge is oblique to the scanning direction.
It means that jagged fine steps (jaggies) occur. The second problem is that small figures and fine patterns are not accurately represented with respect to the quantization interval.

[0020]

The first problem is a major factor in the display quality of computer graphics and computer-synthesized characters. In addition, a second problem occurs in which thin lines and the like disappear, especially when the character graphic is reduced. In printers and the like, such a problem has been solved mainly by making the quantization interval fine (increasing the resolution).

[0021]

On the other hand, in the current television broadcasting system, the number of scanning lines and the bandwidth of luminance / chroma signal are regulated, and the quantization interval cannot be made fine. However, jaggies do not occur when a telop document written in black and white is captured by a camera or the like. It is considered that this is because the incident light of the camera pickup tube and the scanning beam have a certain spread, and thus this has a characteristic as a kind of spatial filter. Therefore, even in computer-synthesized characters and the like, aliasing can be removed by performing a spatial filter characteristic that a television camera has, in other words, by expressing a halftone in an edge portion.

[0022]

As described above, in the computer-synthesized image alone, by performing the halftone expression in the edge portion,
Aliasing can be eliminated. However, when this is embedded in the background image and combined, if the switching process by the binary key signal is performed, particularly when the background image is a moving image, an annoying aliasing occurs at the boundary portion.

When there is a thin line in the horizontal direction or a sharp edge in the horizontal direction, the flicker phenomenon that blinking occurs at half of the field frequency, that is, the frame frequency due to interlaced scanning of the CRT, causing flicker, is a similar problem. However, there have been few proposals that have been made by focusing on this.

[0024]

Therefore, a first object of the present invention is to eliminate aliasing and flicker that occur when superimposing a character graphic represented by a digital signal.

[0025]

On the other hand, both the character / graphics signal and the natural image signal have come to be digitally processed and transmitted, and the mixing and synthesis may be performed by a digital circuit. Also in this case, the jaggies as described above become a problem, and in order to solve this problem, the signal switching circuit of Japanese Patent Laid-Open No. 63-1261 and the image processing device of Japanese Patent Laid-Open No. 63-114472 are disclosed. There is.

[0026]

The signal switching circuit of Japanese Patent Laid-Open No. 63-1261 processes a video signal and an external signal by a control signal (key signal) representing a signal ratio formed from a switching signal digitized into several bits. The superimposing is performed by controlling the multipliers to be added and finally adding and mixing them. Then, by doing so, the problem that signals are not properly connected at the time of image switching by the quantized signal processing is solved and smooth connection is achieved.

[0027]

The image processing apparatus disclosed in Japanese Patent Laid-Open No. 63-114472 comprises an image data memory and a composite data memory corresponding to each pixel of the image data memory, and performs writing or synchronous reading in each pixel, The composition with the external video signal is performed in pixel units. In order to prevent jaggies from occurring, anti-aliasing processing is performed between the digital character / graphics image and the external (background) image.

[0028]

Both of these two methods are
If you do not generate the control signal and the information in the synthetic data memory properly, you cannot expect reliable anti-aliasing.
Regarding the generation of information, there is no specific disclosure, making proper anti-aliasing difficult. Further, in any case, since RGB processing is performed, many memories and mixers are required, and an RGB encoding / decoding circuit is used when inputting / outputting with a composite signal or a luminance / chroma (Y / C separated) signal. Was required, and it was costly.

[0029]

Therefore, a second object of the present invention is to enable reliable anti-aliasing by appropriately generating the key information, and further to superimpose in the state of the Y / C separation signal. The third object is to reduce the cost by performing

[0030]

In recent years, with the widespread use of a camera-integrated video tape recorder (camera-integrated VTR), a title screen imaged by a camera is temporarily stored in an image memory and the title screen is superposed on an image captured later. Camera-integrated VTR that records on tape while imposing
Have been proposed.

However, the title screen is quantized when it is loaded into the image memory, and the above-mentioned problems occur unless the anti-aliasing process is properly performed when superimposing the image data.

[0032]

As a proposal relating to a camera-integrated VTR having a superimpose function, there is a superimpose device disclosed in Japanese Patent Laid-Open No. 3-1776. This is a superimposing apparatus in which a color shift signal generated at an end of a constant brightness area (character pattern) is corrected so as to be generated within the constant brightness area. Then, color bleeding due to a color shift signal that occurs at the boundary with the background image whose brightness changes sequentially is made to be inconspicuous by causing it in a character graphic with a constant brightness.

[0033]

Further, in this embodiment, the camera integrated type V
An example of a TR superimpose circuit is shown. In the memo mode of this camera-integrated VTR, when the title image is captured by the camera and stored in the image memory,
The luminance signal is simply binarized by a comparator and then stored in the image memory. As described above, it is considered that the simple binarization is intended not only for the purpose of simplifying the circuit but also for easily removing the noise component included in the title image captured by the camera. To be
Then, for the binarized image data, inversion,
Coloring is applied and superimposing is performed by switching from the video signal input to the camera. Cross-color generated at that time is noticeable by causing color bleeding in character graphics with constant brightness. I'm losing it.

As described above, particularly in the case of binary-quantized data, the aliasing described above becomes remarkable, but 2
Without value quantization, it is difficult to remove noise components included in a title image by a simple method. Therefore, it is necessary to remove the alias generated by the binary quantization by some method.

[0035]

Therefore, a fourth object of the present invention is to eliminate a noise component generated when a title screen is taken into an image memory by a camera or the like and an alias generated by binary quantization.

[0036]

[0026]

[0037]

As described above, in the prior art, in the case of superimposing a character figure in the form of an analog signal in analog, when luminance / chroma separation is performed and then superimposing synthesis is performed. No consideration was given to the resulting image quality deterioration at the boundary between the character and the graphic (such as cross color and jaggies in the case of digitally generated character and graphic). Even if the generation of cross color is prevented, the problem of jaggies when digitally generating a character graphic is not solved.

In the case of analog superimposing a character graphic generated by a digital circuit on a background image, cross color, flicker and jaggies are generated due to discontinuous points of the chroma signal.

[0039]

Further, when the superimposing is also performed by the digital circuit, the information content of the key signal is unknown, which makes proper anti-aliasing processing difficult. And RG
Since B processing is performed, many memories and mixers are required, and RGB is used when inputting / outputting with Y / C separated signals.
It has been costly such as the need for an encoding / decoding circuit.

[0040]

Further, in the case of incorporating the title screen into the image memory in a camera-integrated VTR or the like, it is necessary to remove the alias generated by the binary quantization performed to remove the noise component contained in the title image. .

[0041]

Therefore, a first object of the present invention is to eliminate aliasing and flicker that occur when superimposing a character graphic represented by a digital signal.

The second purpose is to enable reliable anti-aliasing by appropriately generating the key information, and further to perform superimposing in the state of the Y / C separation signal. The third purpose is to reduce the cost.

A fourth object is to eliminate the noise component generated when the title screen is captured in the image memory by the camera or the like and the alias generated by the binary quantization.
An image synthesizing device that solves the above problems is provided.

[0044]

[0030]

[0045]

As means for achieving the above object, there is provided an image synthesizing device for superimposing a character and graphic signal on a background video signal composed of a luminance signal and a chroma signal, in which the character and graphic are stored. An image memory for outputting as a character / graphics video signal; a mixing information memory for storing a mixing ratio of the background video signal corresponding to each pixel of the image memory and the character / graphics video signal and outputting as a mixing ratio signal; The character / graphics video signal stored in the image memory and the mixing ratio signal stored in the mixed information memory are read in synchronization with the background video signal,
A synchronous readout circuit that separately outputs a luminance signal and a chroma signal of the character / graphics video signal and outputs the mixing ratio signal, and a luminance signal of the background video signal and the character / graphics video signal based on the mixing ratio signal. Of the luminance signal and the first and second mixing circuits for complementary gain changing and mixing to output a superimposed luminance signal, and a chroma signal of the background video signal and the character / graphics video signal based on the mixing ratio signal. It is an object of the present invention to provide an image synthesizing device including a second mixing circuit that outputs a superposed chroma signal by mixing and synthesizing the chroma signal by complementary gain change.

[0046]

[0031]

[0047]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Some embodiments of the image synthesizing apparatus of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing the first embodiment of the image synthesizing apparatus of the present invention. Then, the superimposing processing of the image synthesizing apparatus will be mainly described.

[0049]

As the background video signal which becomes the background image, the luminance signal and the chroma signal are supplied in a synchronous relationship. The signal form may be an analog signal or a digital signal.

The character / graphics video signal, which is a character / graphic superimposed on the background image, is in the form of an RGB3 primary color signal, a luminance signal / color difference signal, a known color palette or a luminance signal / chroma signal. And the mixing ratio data corresponding to each pixel is stored in the mixing information memory 2.

[0051]

Then, the synchronizing circuit 4 extracts the video synchronizing signal from the synchronizing signal included in the luminance signal or the chroma signal of the background video signal and outputs the reading clock to the synchronizing reading circuit 3. The synchronous reading circuit 3 keeps the synchronous relationship with each signal of the background video signal by this reading clock, and then the character and graphic video signal stored in the image memory 1 and each pixel stored in the mixed information memory 2 are held. Generates a luminance signal, a chroma signal and a mixing ratio signal of a character graphic video signal from the corresponding mixing ratio data,
It outputs to each of the mixing circuits 5a and 5b.

[0052]

When the mixing circuits 5a and 5b are digital circuits composed of multipliers, adders, etc., the luminance signal, the chroma signal and the mixing ratio signal are supplied to the mixing circuits 5a and 5b as digital signals. It When the mixing circuits 5a and 5b are analog circuits composed of fader circuits, mixers, etc., they are supplied to the mixing circuits 5a and 5b as D / A converted analog signals.

The mixing circuit 5a is a circuit for complementarily changing the gains of the luminance signals of the background image and the character / graphics image according to the mixing ratio signal, mixing them, and outputting as a luminance signal of the mixed superimpose signal.

[0054]

Here, an example of the mixing process by the mixing circuit 5a will be described with reference to FIGS. Same figure (A)
Shows the mixing ratio of the luminance of the character / graphics image (luminance stored in the image memory 1) and the character / graphics image stored in the mixed information memory 2 at each pixel. FIG. The example of the image of the performed result is shown.

As shown in FIG. 2A, when some pixels in the screen are taken out, there is a contour line a of the character graphic image shown by a dotted line on the pixel, and the lower left side of this contour line a. Is the area of the superimposed text image, and the upper right side is the area of the background image. Then, the value of the brightness and the value of the mixture ratio are represented as values that change between 0 and 100, and when the mixture ratio is 100, only the character graphic image is represented and the mixture ratio
In the case of, only the background image is represented. Further, here, the brightness of the character / graphics image is 100 (white) and the brightness of the background image is 0 (black).

[0056]

In such a case, each pixel completely included in the character / graphic area and not covered by the contour a corresponds to a luminance of 100 as the value of the image memory 1 and a mixing ratio of 100 as the value of the mixed information memory 2. The value is stored. Therefore, the mixing circuit 5a mixes and outputs the luminance signals of the character graphic image and the background image for these pixels at a ratio of 100: 0. As a result, as each of these pixels, white pixels with a luminance of 100 are displayed as a character graphic image on the display screen of FIG.

[0057]

For each pixel on the contour a, the mixture ratio is determined according to the area ratio between the character / graphic area and the background image area. In the case of this example, since the area ratios of the respective pixels divided by the contour line a are all 50:50, the mixture information memory stores the mixture ratio as 50. Therefore, the mixing circuit 5a mixes the luminance signals of the character / graphic image and the background image at a ratio of 50:50. As a result, as each of these pixels, a gray pixel having a luminance of 50 (pixels indicated by diagonal lines) is displayed on the display screen of FIG.

[0058]

Further, each pixel which is completely included in the background image area and does not cover the contour a is stored in the image memory 1 with a value corresponding to a luminance of 0 and in the mixing information memory 2 with a value corresponding to a mixing ratio of 0. There is. Therefore, the mixing circuit 5a mixes and outputs the luminance signals of the character graphic image and the background image at a ratio of 0: 100. In this case, since the luminance signal of the character / graphics image is not included in the processing result at all, a black pixel having a luminance of 0 is displayed as a background image on each of these pixels on the display screen of FIG.

When the area ratio of the character / graphic area and the background image area in one pixel is 30:70, the mixture ratio as the value of the mixture information memory 2 is 30.

[0060]

When the brightness of the character / graphics image is 80, the value of the image memory 1 is 80, and when it is mixed with the background image of 0 brightness at a mixing ratio of 50, it becomes a gray pixel of 40 brightness. Further, when the brightness of the character / graphics image is 80 and the brightness of the background image is 20, the value of the image memory 1 remains 80, and the result of mixing with the mixing ratio of 50 is a gray pixel of brightness 50.

[0061]

In this manner, the mixed circuit 5a outputs the luminance signals of the character / graphics image and the background image in a mixed manner, so that the luminance signal of the anti-brightness signal is detected at the boundary between the image signals of the two systems of the character / graphics image and the background image. Since the aliasing is performed, it can be displayed as a smooth superimposed image that is visually free from jaggies and flicker.

[0062]

On the other hand, the mixing circuit 5b complementarily changes the gains of the chroma signals (carrying color signals) of the background image and the character and graphic image according to the mixing ratio signal and mixes them, and the mixed chroma signal of the superimpose signal. Is a circuit that outputs as.

[0063]

As is known, the chroma signal has a frequency (NT
In the case of the SC system, the color subcarriers having the same 3.58 MHz but different phases by 90 degrees are balanced-modulated by two color difference signals and mixed. For example, it is described in "NHK Color Television Textbook" edited by the Japan Broadcasting Corporation. Is expressed by the following equation.

[0064]

(E _R -E _Y ) * cos (2πf _sc t) /1.14+ (E _{B-
E Y) * sin (2πf sc} t) /2.03 _{where, E R, E B, E} Y each red signal, blue signal, indicates the luminance signal, fsc is shows the color subcarrier frequency.

[0065]

Here, if, for example, the chroma signals of the two systems are simply switched at the boundary portion as in the conventional example, an abrupt waveform change will cause an unexpected color difference component in both phase and level. This may cause color, etc.

[0066]

Therefore, in the present invention, the mixing ratio is changed with an appropriate degree of graduality, and by performing the mixing with the complementary gain change, the unexpected color difference component is generated due to such a sharp waveform change. Can be avoided.

[0067]

The result of mixing when such mixing is performed is represented by the following equation.

{Α * (E _R1 -E _Y1 ) + (1 -α) * (E _R2 -E _Y2 )} * cos
_{(2πf sc t) /1.14 + {} α * (E B1 -E Y1) + (1 -α) * (E B2 -E Y2)} * sin (2πf sc t) /
2.03 However, E _R1 , E _B1 , and E _Y1 indicate the red signal, blue signal, and luminance signal of the background image, respectively, and E _R2 , E _B2 , and E _Y2 indicate the red signal, blue signal, and luminance signal of the character graphic image, respectively. Where α represents the mixing ratio.

[0069]

By performing such color mixing in the contour portion, the character graphic is not easily blended into the background and an unnatural image is not generated.

[0070]

In the above embodiment, the background image is input in the state of being separated into the luminance signal and the chroma signal, and is output while being separated. However, when the background video signal is the composite input signal, , Y / C separation circuit may separate the luminance signal and the chroma signal before inputting,
If you want to obtain a composite signal as the output signal,
It suffices to mix the luminance signal and the chroma signal output using the C mixing circuit before outputting.

[0071]

FIG. 3 is a block diagram showing a second embodiment of the image synthesizing apparatus of the present invention. The image synthesizer shown in FIG.
This is to superimpose a character graphic stored in the form of outline data on a background image.

[0072]

The outline data memory 6 stores the contour point coordinate values and the like of the superimposed graphic image. Then, the contour point coordinate values are supplied to the expansion / painting means 7a to be bit-mapped and stored in the expansion / painting memory 8a.

[0073]

The bit map character / graphic data stored in the expansion / painting memory 8a is information indicating the presence or absence of each pixel. This information is converted into multivalued luminance information by the filter means 9 such as a well-known spatial filter and stored in the gray scale memory 10a which is temporarily stored in character string units. This information is further colored by the image data generating means 11, that is, the RGB value and the like are determined, and then written in the address of the image memory 1 corresponding to a desired position on the screen. On the other hand, the brightness information stored in the gray scale memory 10a is selectively converted by the mixing information creating means 12 into mixing ratio data representing the mixing ratio of the character / graphic image and the background image, and corresponds to a desired position on the screen. Is written in the address of the mixed information memory 2. Note that this mixing ratio data corresponds to the area ratio between the character graphic image and the background image included in one pixel, as in the mixing circuits 5a and 5b in the first embodiment.

[0074]

When a natural moving image whose brightness and color change frequently is used as a background image, a method for improving the visibility when superimposing a character / graphic image on the natural moving image is one of the methods. As shown in FIG. 6, a method for superimposing a character graphic body c on a background image b with a relatively high brightness and a surrounding outline d with a relatively low brightness, and vice versa. In general, a method is used in which the character / graphic body c is made to have a relatively low brightness, a edging contour d having a relatively high brightness is provided around it, and the image is superimposed on the background image b.

[0075]

In this case, the image data generating means 11 and the mixed information creating means 12 mix (anti-aliasing) the character / graphic body c and the edging contour d at the boundary between the character / graphic body c and the edging contour d. Since then
The mixing circuits 5a and 5b are provided between the edging contour d and the background image b.
The same mixing as in the first embodiment is carried out. Then, this makes it possible to prevent jaggies and eliminate discomfort.

[0076]

Further, the structure of the portion A surrounded by the dotted line in FIG. 3 is the same as that of the first embodiment, and the superimposing processing after the image memory 1 and the mixed information memory 2 is the same as that of the first embodiment. The description is omitted because it is as described above.

[0077]

The third embodiment is shown in FIG. The third embodiment shown in the figure is configured to more effectively generate the mixing ratio information in the second embodiment.

First, the problem of antialiasing when a spatial filter is used as an example of the filter means 9 used in the second embodiment will be described.

[0079]

The binary character / graphic image once generated in display pixel units already has a noise component such as jaggies in the contour portion. Then, the processing by the spatial filter is to remove the jaggy component having a relatively high spatial frequency component by the spatial filter having the low-pass characteristic. However, it is impossible to completely remove the noise component once mixed in later, and some noise component remains.

[0080]

The anti-aliasing by two-dimensional averaging performed in this embodiment is intended to reduce the generation of noise components at the stage of generating a character / graphic image, rather than removing noise components afterwards.

As a relatively simple two-dimensional averaging method, a uniform averaging method and a weighted averaging method are known.
Here, the uniform averaging method will be described. A sub-pixel (sub-pixel) is a calculated pixel obtained by dividing one display pixel into a plurality of pixels in the X direction and the Y direction. In the case of the drawing shown in FIG. 5, one display pixel is divided into ½ in both the XY directions, which corresponds to a total of 4 sub-pixels. In the drawing, ● indicates the center of the display pixel and + indicates the sub-pixel. Shows the center of.

[0082]

In the uniform averaging method, the values (luminance) of these sub-pixels are uniformly added without weighting to calculate the value of the target display pixel.

By calculating the value of the display pixel in this way, the value corresponding to the area ratio in which each display pixel area includes the character graphic can be set as the value of the display pixel.

[0084]

The processing of this embodiment using such a two-dimensional averaging method will be described with reference to FIG.

First, similarly to the second embodiment, the outline data memory 6 stores the contour point coordinate values and the like of the superimposed graphic image. Then, the contour point coordinate values are supplied to the expanding / filling means 7a to be converted into a bit map. At this time, as described above, one display pixel is expanded into a plurality of sub-pixels, and each sub-pixel is expanded. Filling corresponding to the sub-pixel is performed,
It is stored in the sub-pixel developing / painting memory 8b.

[0086]

The bit map character / graphic data stored in the sub-pixel developing / painting memory 8b is binary information representing the presence / absence of a character / graphic image on each sub-pixel. This information is converted into multi-valued luminance information for each pixel by the filter means 9 that performs the above-described two-dimensional averaging method and stored in the gray scale memory 10a. Then, the brightness information is further colored by the image data generating means 11, that is, the RGB value or the like is determined, and then written in the image memory 1 corresponding to a desired position on the screen. On the other hand, the brightness information stored in the gray scale memory 10a is selectively converted by the mixing information creating means 12 into mixing ratio data representing the mixing ratio of the character / graphic image and the background image, and corresponds to a desired position on the screen. Is written in the mixed information memory 2. The mixture ratio data is determined by the area ratio between the character / graphic image and the background image included in one pixel, as in the first embodiment. The subsequent processing is performed in the same manner as in the first and second embodiments.

[0087]

In this way, since the value corresponding to the area ratio in which each display pixel area contains a character graphic becomes the value of the display pixel, it is possible to reduce the generation of noise components at the stage of generating the character graphic. Furthermore, in the present embodiment, since the character / figure image and the background image, which originally have a small amount of jaggies, are mixed based on the mixing ratio data, a visually very smooth character / figure image is superposed. Impose can be displayed.

[0088]

The fourth embodiment shown in FIG. 7 shows a case where the data of the character / graphics image is supplied not in the outline data but in the bitmap format. When outline data is used, it generally takes a long processing time to expand / fill, and it may be unsuitable for use in a device that requires high interactivity and high display speed. In such a case, the display speed can be increased by using the bitmap data, although there is a drawback that the enlargement / reduction, deformation, and various modifications of the character graphic cannot be performed freely.

[0089]

In FIG. 7, the data of the character graphic image is
Bit map data holding data as multi-valued luminance information represented by a plurality of bits per pixel is stored in the gray scale data memory 10b.

The multi-valued luminance information stored in the gray scale data memory 10b is colored by the image data generating means 11 and stored at a predetermined display position in the image memory 1 and at the same time mixed information is created. Means 12
Are supplied to the mixing information memory 2 and are subjected to a mixing ratio determination process and stored in a predetermined display position of the mixing information memory 2. Further, the determination of the mixing ratio data and the subsequent processing are performed in the same manner as in the first to third embodiments.

[0091]

Further, a fifth embodiment shown in FIG. 8 shows a case where character / graphic data is externally fetched in the form of an image signal. For example, it is intended to simultaneously solve the problems of aliasing and noise that occur when applied to a device such as a camera-integrated VTR described in the prior art.

[0092]

Such an image synthesizing device will be described with reference to FIG. The data of the character / graphic image is photoelectrically converted into a video signal by, for example, a video camera unit, AD-converted, and then supplied to the writing means 7b as an image input signal. The writing means 7b is a synchronous writing circuit or the like, which binarizes the brightness of the image input signal while keeping the synchronization, and writes it into the image memory 8c. At this time,
By binarizing the luminance, it is possible to almost eliminate the noise component due to the fine luminance change of the paper on which the handwritten title image is written. On the other hand, although a jaggy component is generated at the contour of the target character graphic, the mixture ratio information generated by supplying the image signal written in the fetched image memory 8c to the filter means 9 and performing the filtering process. Utilizing the mixture information creating means 12 in the latter stage
This jaggie component can be removed by controlling

[0093]

Further, in the mixture information creating means 12, pixels having a mixture ratio data value in which the mixture ratio of the character / graphics image is equal to or less than a predetermined threshold value are ignored so as not to be mixed, so that the area is small. It is possible to prevent the noise component from being superimposed.

Since the structure and operation of the portion indicated by A surrounded by the dotted line in the figure are the same as those of the first to fourth embodiments, the description thereof will be omitted.

[0095]

Next, some application examples will be described.

FIG. 9 is a block diagram showing a first application example, and shows an example of a superimpose function portion when the present invention is applied to a video titler.

This video titler uses a separate Y / C separation type or composite type analog video signal input from the outside as a background signal, and superimposes on this background signal a character / graphics signal which is digitally generated in this device. This is a device that pauses and outputs the processed color image signal to the outside as a separate Y / C separated format or composite format analog video signal (superimpose signal).

[0098]

The MPU is used to generate the character / graphic signal.
A (microprocessor unit) circuit 21 and a VDP (video display processor) circuit 22 are used.

The MPU circuit 21 includes key switches 23, M
PU24, program ROM (read only memory) 2
5, outline data ROM 26, dot data RO
M27, MPU-RAM (random access memory) 2
It is composed of 8 etc.

[0100]

The key switch 23 is an input device for instructing a character graphic to be displayed, its size, a display position, etc., and a series of processing sequences are executed by the MPU 24 in accordance with a program stored in the program ROM 25. Done. And the key switch 23
The shape of the character graphic designated by is stored in advance in the outline data ROM 26 in the form of two-dimensional contour vertex coordinates, and is stored in the MPU-RAM 28 by the MPU 24 reading this for each character, for example. Image data composed of binary sub-pixel information is created on the expanded / filled RAM area. Note that this image data is created, for example, by a method similar to that described with reference to FIG. 5 in the third embodiment so that 2 * 2 sub-pixels correspond to one display pixel.

[0101]

The image data thus developed / filled-out is processed by the software processing of the MPU 24 by the program of the program ROM 25, by the same processing as the filter means 9 described in the third embodiment (FIG. 4). The dimension averaging process is performed, and pixel information (gray scale image information) including multi-valued luminance information is created on the gray scale RAM area secured on the MPU-RAM 28. Then, the gray scale image information is subjected to coloring and mixing ratio determination processing.

[0102]

This coloring and mixing ratio determination processing will be described with reference to the example of characters in FIG. The grayscale image information of the character / graphic body c and the outline edging d is created in different area portions on the grayscale RAM area. And
Among the pixels, the pixel composed of only the character / graphic body c is
The grayscale image information of the character / graphic body c is colored in the designated color. In addition, the character graphic body c and the outline d
In the boundary part (pixels including both) of each of the colors specified by the character graphic body c and the edging outline d,
Complementary mixing is performed according to the grayscale image information of the character / graphic body c. A pixel composed of only the edging outline d is
Instructed coloring is performed including the contour portion of the edging contour d. Since the outline portion of the edging outline d is mixed with the background signal b, the grayscale image information is used as the mixture ratio information.

[0103]

The VDP circuit 22 has an MPU I / F (MPU interface) circuit connected to the MPU bus, a video I / F circuit for inputting a synchronizing signal and outputting a signal to the synchronizing data bus, and the like. Is a V-RAM that functions as the image memory 1 and the mixing ratio memory 2 of each of the above-described embodiments.
It is connected to the (video RAM) 29. And MP
V-RAM 29 from U24 via this VDP circuit 22
Can be written and transferred to the colored multi-valued image information (R) generated by the MPU circuit 21 described above.
The GB3 primary color signal) and the mixture ratio information are transferred to the V-RAM 29 via the VDP circuit 22.

[0104]

At this time, when the colored multi-valued image information and the mixture ratio information are stored in the same V-RAM 29, separate address areas may be prepared and stored, but for one pixel. It is also possible to allocate one data word and store the multivalued image information and the mixture ratio information at the same time. In this case, for example, a value of RGB three primary colors is assigned to each 4 bits and a value of the mixing ratio information is assigned to 4 bits and stored in one data word composed of 16 bits, whereby the V-RAM
The configuration of 29 can be simplified.

[0105]

On the other hand, in the synchronizing circuit 30, the synchronizing component is separated from the luminance signal of the input background signal, and the compensated synchronizing signal and master clock are generated by a PLL (phase locked loop) circuit or the like. These are VDP
It is supplied to the circuit 22 and the color encoder 33 and is used as a reference signal for keeping the synchronization of the output signals. In this example, the image data stored in the V-RAM 29 is
VD according to the above sync signal and master clock
The data is synchronously output to the synchronous data bus via the P circuit 22.
The latch circuit 31 latches the RGB three primary color signals and the mixing ratio information, and outputs them to each DAC (digital-analog converter) 32. The RGB three primary color signals DA-converted by the DAC 32 are supplied to the color encoder 33 and encoded into a luminance signal and a chroma signal according to the synchronization signal and the master clock supplied from the synchronization circuit 30.

[0106]

The superimpose processing is performed as shown in FIG.
The superimposing circuit shown in (A) or (B) is used. This superimposing circuit is a part corresponding to the part described in the first embodiment.

[0107]

In the configuration shown in FIG. 10A, the fader circuit 41 is supplied with the background brightness signal as an input signal, and the fader circuit 42 is supplied with the output brightness signal (character pattern brightness signal) of the color encoder as an input signal. Supplied. Further, the mixing ratio signal supplied from the latch circuit 31 is DA-converted by the DA conversion circuit 34, adjusted in level by the level adjustment circuit 35, distributed to two systems, and one of them is inverted so as to have a complementary function. After being supplied to the amplifier 36 and subjected to inversion processing, each fader circuit 37, 38
Supplied as a control input. Each fader circuit 37, 3
The luminance signal whose gain has been changed by 8 is added and mixed by a mixer (mixer) 39 and output as a superimpose luminance signal.

[0108]

Incidentally, each circuit 35 after the DA conversion circuit 34
9 to 39 are shown as the mixing circuit 5a in FIG. Further, when a general-purpose balanced modulation IC is used, the portion surrounded by the dotted line in the figure can be configured by one IC.
Then, although the superimposing process of the luminance signal has been described here, the same process is performed for the chroma signal.

[0109]

When the superimpose processing circuit is composed of a digital circuit, the DA conversion circuit 34
The configuration is as shown in FIG. The fade processing is performed by the two multipliers 41a and 41b,
The level range of the input mixing ratio signal is 0.0 ≦ α ≦ 1.
When it is set to 0, the control input to the multiplier 41a to which the background luminance signal is input is the input subjected to the processing (1-α) by the arithmetic circuit 40. Then, the mixing ratio signal is directly supplied as a control input to the multiplier 41b to which the character / graphic luminance signal is input, and the respective luminance signals whose gains have been changed by the multipliers 41a and 41b are mixed by the adder 42. After being DA-converted by a DA converter circuit (not shown), it is output as a superimpose luminance signal.

[0110]

In the above description of the first application example shown in FIG. 9, outline data stored in the outline data ROM 26 is used as a superimposing character graphic, but a menu display or the like is used. Grayscale character / graphic data stored in the dot data ROM 27 may be used in a portion having high interactivity and requiring a high response speed. In this case as well, by instructing the character graphic to be displayed, the size thereof, the display position, etc. by the key switch 23, the relatively time-consuming expansion / filling of outline data is performed according to the program stored in the program ROM 25. Instead, the MPU 24 executes the same processing as described above.

[0111]

FIG. 11 is a block diagram showing a second application example, and shows an example of a superimposing function portion when the present invention is applied to a camera-integrated VTR (video tape recorder).

This camera-integrated VTR uses an analog video signal of a format such as luminance / color difference generated in the camera section as a background signal and digitally displays a title screen etc. previously photographed by the camera section as shown in FIG. Are processed into a character / graphics signal, subjected to superimposing processing including anti-aliasing, and supplied to a recorder (recording) section as an analog video signal in a luminance / chroma signal format or the like.

[0113]

When the title screen photographed by the camera unit is used as a character / graphic signal as in this case, the title screen photographed by the camera unit is as shown in FIG.
In addition to the target character graphic e, extra information such as uneven brightness f and small stains g generated on the paper on which the title screen is drawn may be included. Therefore, in order to remove the extra information and superimpose (insert) the title screen composed of only the beautiful character / figure e on the background image, it is only necessary to remove the jaggies, cross colors, and the like that occur in the outline of the character / figure e. None, moreover, uneven brightness f and small dirt g
It is necessary to remove extra information such as to prevent unexpected parts from being superimposed and obscuring the necessary background screen.

[0114]

As shown in FIG. 11, the processing of the character / graphics signal of this camera-integrated VTR is carried out by the same method as in the first application example.
The PU circuit 51 and the VDP circuit 52 are used.

The MPU circuit 51 includes key switches 53, M
PU54, program ROM55, dot data ROM
57, MPU-RAM 58 and the like.

Then, the key switch 53 is instructed to take in the character / graphic, and the MPU 24 executes a series of processing sequences in accordance with the program stored in the program ROM 25.

[0117]

When a character screen is instructed to be captured by the key switch 53 while the title screen as shown in FIG. 12 is being photographed by the camera section, the luminance signal among the input signals from the camera section is the comparator 67. To the image capturing RAM 69 by the synchronous writing circuit 68. At this time, since an intermediate luminance area such as the luminance unevenness f in FIG. 12 is removed by the comparator 67, the image capturing RAM 69
Only the information of the character graphic e and the small stain g is stored in. Then, jaggies or the like may occur on the contour of the character graphic e due to the binary quantization by the comparator 67, but the binary quantized image data is filtered by software in the next stage, Grayscale R secured on MPU-RAM8 after jaggies are removed
Pixel information (gray scale image information) including multi-valued luminance information is created on the AM area.

[0118]

The grayscale image information is
The same coloring and mixing ratio determination processing as in the first application example is performed.

In general, an original image such as a title screen is often drawn in a low-luminance color such as black on a high-luminance background such as white paper. Therefore, the black character figure e is superimposed in white color. In order to
And so on so that the inversion process can be performed. On the other hand, the information of the small stain g cannot be removed by the processing of the comparator 67 because it is not an intermediate luminance,
Since this small stain g has a small area, the value of the mixing ratio becomes relatively small by performing the filtering process. As a result, for example, when it is superimposed on the background moving image, it becomes translucent and the like. It becomes unnoticeable. Furthermore, a threshold value is set for such mixture ratio data,
By preventing the mixture ratio data below a certain value from being stored in the mixture information memory, it is possible to remove the semitransparent small stain g, and to reduce the occurrence of unexpected superimposing processing. it can.

[0120]

Further, since the camera-integrated VTR uses the luminance / color difference from the camera section as the background signal, the reference signal for generating the synchronization signal and master clock compensated by the synchronization circuit 60 is the oscillation circuit. It is generated by 65. Then, the background luminance signal is input to the mixing circuit 5a by using the luminance signal input from the camera unit while capturing the background image, and the background chroma signal is capturing the background image. The luminance signal and the RY and BY color difference signals sometimes input from the camera unit are generated by being supplied to the color encoder 66, and are input to the mixing circuit 5b. The VDP circuit 52 and the superimposing circuits other than the above operate in the same manner as in the first application example, and therefore their explanations are omitted.

[0121]

When the superimposed graphic image has a fixed display content such as a date display and a response speed is required, the data captured from the camera unit is used as described above. Not the dot data R
Grayscale character / graphics data stored in the OM 57 may be used. The processing in this case is similar to that of the first application example.

[0122]

[0087]

[0123]

Since the image synthesizing apparatus of the present invention performs the superimposing processing in the state of the Y / C separated signal which has been widely used as the input / output signal in recent years, the image quality of the background image is prevented from being deteriorated. While eliminating aliasing, cross color and flicker, digital circuit configuration,
This can be realized at a low cost and with a simpler circuit configuration than performing superimposing processing in the state of RGB signals regardless of the analog circuit configuration.

[0124]

Further, the character / figure image is converted into multi-valued luminance information represented by a plurality of bits per display pixel, the coloring process and the mixture ratio are determined, and the background video signal and the character are calculated based on the mixture ratio. When the gains of the graphic video signals are changed complementarily and mixed and combined, the antialiasing can be performed more reliably.

[0125]

When the outline data of a character and a graphic is developed in the sub-pixel memory and the anti-aliasing is performed by generating the mixing ratio information, the quality is dramatically improved as compared with the conventional superimposed image by the switching method. Therefore, it is possible to improve the quality, and it is possible to display a relatively small character graphic with high visibility.

[0126]

Further, when the title screen is taken in the image memory by the camera or the like in the camera-integrated VTR or the like, the noise component generated at that time and the alias generated by the binary quantization can be removed, and the image quality is good. This has the effect of superimposing character graphics.

[Brief description of drawings]

FIG. 1 is a configuration diagram showing a first embodiment of an image synthesizing apparatus of the present invention.

FIG. 2A is a diagram for explaining a mixing process, and FIG.
FIG. 6 is a diagram showing an example of an image obtained as a result of performing a mixing process.

FIG. 3 is a configuration diagram showing a second embodiment of the image synthesizing apparatus of the present invention.

FIG. 4 is a configuration diagram showing a third embodiment of the image synthesizing apparatus of the present invention.

FIG. 5 is a diagram showing a relationship between display pixels and sub-pixels.

FIG. 6 is a diagram showing an example of a character with a edging outline.

FIG. 7 is a block diagram showing a fourth embodiment of the image synthesizing apparatus of the present invention.

FIG. 8 is a configuration diagram showing a fifth embodiment of the image synthesizing apparatus of the present invention.

FIG. 9 is a configuration diagram showing an example in which the image synthesizing apparatus of the present invention is applied to a video titler which is a first application example.

10A is a configuration diagram showing an example of an analog circuit of a superimposing circuit, and FIG. 10B is a configuration diagram showing an example of the same digital circuit.

FIG. 11 is a configuration diagram showing an example in which the image synthesizing device of the present invention is applied to a camera-integrated VTR which is a second application example.

FIG. 12 is a diagram showing an example of a title screen.

[Explanation of symbols]

1 Image Memory 2 Mixed Information Memory 3 Synchronous Readout Circuit 4, 30, 60 Synchronous Circuit 5a, 5b Mixing Circuit 6 Outline Data Memory 7a Expansion / Filling Means 7b Writing Means 8a Development / Filling Memory 8b Sub-Pixel Development / Filling Memory 8c Captured Image Memory 9 Filtering means 10a Gray scale memory 10b Gray scale data memory 11 Image data generating means 12 Mixing information generating means 21,51 MPU (microprocessor unit) circuit 22,52 VDP (video display processor)
Circuit 23,53 Key switch 24,54 MPU 25,55 Program ROM (read only memory) 26 Outline data ROM 27,57 Dot data ROM 28,58 MPU-RAM (random access memory) 29,59 V-RAM 31,61 Latch circuit 32, 34, 62, 64 DAC (digital-analog converter, DA converter circuit) 33, 63, 66 Color encoder 37, 38 Fader circuit 35 Level adjusting circuit 36 Inversion amplifier 39 Mixer (mixer) 40 Arithmetic circuit 41a, 41b Multiplier 42 Adder 65 Oscillation circuit 67 Comparator 68 Synchronous writing circuit 69 Image capture RAM a Contour line b Background image c Character figure body d Edged contour e Character figure f Brightness unevenness g Small dirt

Claims (6)

[Claims] 1. An image synthesizing device for superimposing a character / graphics signal on a background video signal composed of a luminance signal and a chroma signal, wherein the image memory stores the character / graphics and outputs it as a character / graphics video signal. A mixed information memory for storing a mixing ratio of the background video signal and the character / graphics video signal corresponding to each pixel and outputting as a mixing ratio signal; and the character / graphics video signal stored in the image memory. The mixed ratio signal stored in the mixed information memory is read in synchronization with the background video signal, and the luminance signal and the chroma signal of the character / graphics video signal are separately output and the mixed ratio signal is output. A synchronous reading circuit, and a gain complementary to the luminance signal of the background video signal and the luminance signal of the character / graphics video signal based on the mixing ratio signal. A first mixing circuit that changes and mixes and synthesizes, and outputs a superimposed luminance signal, and a gain change complementarily to the chroma signal of the background video signal and the chroma signal of the character / graphics video signal based on the mixing ratio signal. An image synthesizing apparatus comprising: a second mixing circuit for performing mixed synthesizing to output a superimposed chroma signal. 2. The image synthesizing apparatus according to claim 1, wherein an outline data memory for storing outline data of a character / graphic shape, and the outline data read from the outline data memory are expanded into dot data to perform a filling process. Expanding / filling means, filter means for converting the character / graphics image expanded and filled by this expanding / filling means into multi-valued luminance information represented by a plurality of bits per display pixel, and converted by this filter means. A gray scale memory for storing the multi-valued luminance information; an image data generating means for supplying at least the multi-valued luminance information to color-process the character graphic image; and a multi-valued luminance information for supplying the multi-valued luminance information. Mixing information creating means for determining a mixing ratio of a character graphic image to a background image Storing the character graphic data output from the image data generating means in a predetermined display position of the image memory, and storing the mixing ratio output from the mixing information generating means in a predetermined display position of the mixing information memory. An image synthesizing device characterized by: 3. The image synthesizing apparatus according to claim 1, wherein the character / graphic image is stored as multivalued brightness information represented by a plurality of bits per display pixel, and the multivalued brightness information is stored in the grayscale data memory. Image data generation means for supplying at least color processing to the character / graphics image, and mixed information creation means for supplying the multivalued luminance information to determine a mixing ratio of the character / graphics image to the background image, Characteristic data output from the image data generating means is stored in a predetermined display position of the image memory, and a mixing ratio output from the mixing information generating means is stored in a predetermined display position of the mixing information memory. Image synthesizer. 4. The image synthesizing apparatus according to claim 1, wherein a fetched image memory for storing the quantized image input signal as a character / graphic image, and a multi-bit image for representing the character / graphic image by a plurality of bits per display pixel. Filter means for converting the luminance information of values, a gray scale memory for storing the multivalued luminance information converted by the filter means, and the multivalued luminance information is supplied to color-process the character / graphic image. Image data generating means, and mixed information creating means for determining the mixing ratio of the multi-valued luminance information to the background image of the character / graphics image, and data of the character / graphics output from the image data generating means. Is stored in a predetermined display position of the image memory, and the mixing ratio output from the mixing information creating means is stored in a predetermined display position of the mixing information memory. Image synthesizing apparatus characterized by. 5. The image synthesizing apparatus according to claim 2, wherein a character / graphic image is formed for each sub-pixel obtained by multiplying at least one of the horizontal and vertical directions with respect to one display pixel by the expanding / filling means. Filling is performed according to the area ratio, and the filter means generates multi-valued luminance information of each display pixel from the calculation result including the two-dimensional averaging process for the values of the plurality of sub-pixels corresponding to the one display pixel. An image synthesizing device characterized by: 6. The image synthesizing apparatus according to claim 4, wherein the mixing information creating means does not mix the character and graphic images when the mixing ratio of the character and graphic images is equal to or less than a predetermined threshold value. Image synthesizer.