JPH057816Y2 - - Google Patents

Description

[Detailed explanation of the idea] The present invention will be explained in the following order.

A. Industrial field of application B. Overview of the invention C. Conventional technology D. Problems that the invention attempts to solve (Figures 6 and 7) E. Means for solving the problems (Figure 1) F. Effect (Figure 1) (Figure 1) G Example (Figures 1 to 5) H Effect A Industrial field of application The present invention relates to an image conversion device, and is particularly suitable for application to a special effects device.

B. Summary of the invention The present invention is an image conversion device that adds and outputs weighted and distributed image data of an input image for each pixel of an output image, and then normalizes and outputs it, thereby simplifying the overall process. It is possible to obtain an image conversion device in which deterioration of image quality is prevented by the configuration.

C. Prior Art Conventionally, there are two types of image conversion devices of this type: a reading method and a writing method.

In other words, the writing method enlarges, reduces, and
When converting images in various ways such as movement and transformation,
This method directly maps the image data of each pixel that makes up the input image to the output image in sequence according to the mode of image conversion, and has the advantage of being able to freely transform the input image compared to the writing method. are doing.

However, in order to directly map image data to the output image, the display position of the pixels of the input image converted to the output image side may not match the display position of the pixels of the output image. There is a problem that errors occur in the images and the image quality deteriorates.

For this reason, in the writing method, there was a problem that it was difficult to freely enlarge the input image.

On the other hand, in the reading method, contrary to the writing method, each pixel of the output image is inversely converted to the input image side in advance and the display position on the input image is determined, so that each pixel of the output image and the input image are The image data of each pixel of the output image is obtained by interpolation calculation based on the correspondence relationship.

According to this readout method, it is relatively easy to enlarge the input image and there is less deterioration in image quality compared to the writing method, so the readout method has traditionally been used in this type of special effects device. is being done.

D. Problems to be solved by the invention However, in read-out image conversion devices, each pixel of the output image must be inversely converted to the input image side in advance, so when transforming the input image into a complex shape, The arithmetic processing required for such inverse transformation becomes complicated, and as a result, there is a problem that the input image cannot be transformed as freely as compared to the writing method.

In order to solve this problem, the computing power of the image conversion device must be increased, which inevitably makes the overall configuration of the image conversion device complicated and expensive.

Furthermore, in such a read-out type image conversion device, when an input image is reduced and displayed while being moved, as the input image moves,
There is a possibility that the output image may change periodically.

That is, as shown in FIGS. 6 and 7, when the input image IM _IN is reduced and displayed, the display position of each pixel of the output image that is inversely converted to the input image IM _IN side is changed by that amount P _x1,y1 , P _{x2 ,y2} ,..., P _xn,yn , P _xo,yo ... become wider with respect to the pixels of the input image IM _IN .

Therefore, for example, if a small black display portion I _B is displayed on a white background in the input image IM _IN , as the input image IM _IN moves, the display position of the inversely transformed pixel P _{x1, y1} ,P _x2,y2 ,...
When one of P _xn,yn , P _xo,yo ... is applied to the black display part I _B (Fig. 7) and the display position of either
The cases in which P _x1,y1 , P _x2,y2 , . . . , P _xn,yn , P _xo,yo .

In this case, the inversely transformed display positions P _x1,y1 , P _x2,y2 ,
..., P _xn,yn , P _xo,yo ..., it is possible to obtain the image data of the black display portion I _B if even one of them falls on the black display portion I _B ; Otherwise, you will only be able to obtain image data with a white background. Therefore, in the output image, a black display portion that is reduced in size as the input image IM _IN moves is displayed in a blinking manner periodically (hereinafter referred to as aliasing).

For this reason, in the past, the image data of the input image IM _IN was filtered to remove high frequency components of the image data representing small display areas, thereby preventing the occurrence of aliasing and deterioration of image quality. It is designed to prevent

However, when attempting to filter image data, the frequency characteristics of the spatial filter circuit must be switched depending on the rate at which the input image is reduced. This complicates the configuration of the spatial filter circuit and also reduces the overall image conversion device. There was a problem that the configuration was complicated.

This idea was created taking the above points into consideration.
The present invention attempts to propose an image conversion device that has an overall simple configuration and can prevent deterioration of image quality.

E Means for Solving the Problem In order to solve this problem, in the present invention, the input image is converted into an image by mapping the image data D _I,J of each pixel of the input image onto the output image IM _OUT . In the image conversion device 1 configured to output an output image IM _OUT , the display position of each pixel of the input image on the output image IM _OUT
An address conversion circuit 5 that outputs address information representing Q _I,J , and image data D _I,J of each pixel of the input image,
Output image represented by the address information of the relevant pixel
To multiple pixels Pw I, _J , Pw I+1, _J , Pw I, _{J+1 ,} Pw _I+1 , _{J+1 of the output image IM OUT surrounding the display position Q I,J on IM} OUT,
Weighting circuits 5, 6, 7, which sequentially weight and distribute
8, 9 and the distributed image data obtained through the weighting circuits 5, 6, 7, 8, 9 as output images.
Each pixel of IM _OUT Pw _I,J , Pw _I+1,J , Pw _I,J+1 , Pw _I+1,J+1
The first adding circuits 5 and 10 add the weights K _I,J , K _I+1,J K _I,J+1 , K _I+1,J+1 required for weighting and output the images. Each pixel of IM _OUT Pw _I,J , Pw _I+1,J , Pw _I,J+1 ,
Second addition circuit 5, 12 that adds each Pw _{I+1, J+1}
and the image data Dw _I,J obtained via the first addition circuits 5 and 10 are weighted by the weighting amount obtained via the second addition circuits 5 and 12 corresponding to the image data Dw _I,J. A division circuit 15 is provided which divides and outputs the result of addition of K _I,J , K _{I+1, J K I,J +} 1 and K _{I+1, J+1} .

F Effect Image data DI _,J of the input image is converted to surrounding pixels Pw _I,J , Pw _I+1,J , Pw _I,J+1 , in the output image IM _OUT
By weighting and distributing Pw _{I+1,J+1, the} display position Q _I,J of the pixels of the input image converted to the output image IM _OUT side and the pixels Pw _I,J , Pw _I+ of the output image IM _{OUT 1,J} ,
Even if the display positions of Pw _I,J+1 and Pw _I+1,J+1 do not match, it is possible to map all image data D _I,J .

Furthermore, the pixels Pw _I,J , Pw _I+1,J , of the output image IM _OUT ,
Pw _I,J+1 , Pw _I+1,J+1 weighted image data and weighting amount K _I,J , K _I+1,J K _I,J+1 , K _I+1,J By adding and dividing by ₊₁ , it is possible to obtain image data Dw _I,J with a predetermined signal level even if the enlargement or reduction ratio changes, thus preventing deterioration of image quality. can.

G. Embodiment An embodiment of the present invention will be described in detail below with reference to the drawings.

In FIG. 1, reference numeral 1 designates a writing-type image processing apparatus as a whole, which supplies input image information VD _IN in which image data DI _{, J} of each pixel constituting an input image are continuous to an input image memory circuit 2.

The write address information generation circuit 3 generates image data
The value changes corresponding to D _I,J , and the corresponding image data D _I,J
Write address information AD _WI representing the horizontal and vertical display position of each pixel on the input image with
is output to the input image memory circuit 2.

Therefore, the input image memory circuit 2 sequentially writes image data according to the write address information AD _WI .
D _I,J is stored.

On the other hand, the read address generation circuit 4 generates the image data D _I,J as well as the write address information AD _WI .
Read address information AD _RI representing the display position in the horizontal and vertical directions on the input image of the pixel having
It is output to the input image memory circuit 2 and address conversion circuit 5.

As a result, image data is sequentially read from the input image memory circuit 2 based on the read address information AD _RI .
D _I,J is outputted to the output image memory circuit 10 via multiplication circuits 6, 7, 8 and 9.

The control circuit 11 outputs a control signal DC to the address conversion circuit 5 according to the mode of image conversion set and input by the operator.

As shown in FIG. 2, the address conversion circuit 5 converts pixels of the input image represented by the read address information AD _RI in the horizontal direction on the output image IM _OUT based on the control signal DC and the read address information AD _RI . After obtaining the conversion address information representing the display position in the vertical direction and the display position in the vertical direction, the output image IM _OUT that surrounds the display position QI _,J on the output image IM _OUT , which is represented by the conversion address information, in the horizontal and vertical directions is calculated. 4 pixels
Write address information of Pw _I,J , Pw _I+1,J , Pw _I,J+1 and Pw _I+1,J+1 Aw _I,J , Aw _I+1,J , Aw _I,J+1 and Aw _{I+1, J+1} are output to the weighting coefficient memory circuit 12 and the output image memory circuit 10.

Therefore, as shown in Fig. 3, for example, when obtaining an output image IM _OUT obtained by converting the input image to the same size, an area surrounded by four adjacent pixels (hereinafter referred to as the output screen) is determined according to the conversion address information. The output screen grid (referred to as a grid) is selected, and as the read address information AD _RI changes, the output screen grid is sequentially moved and selected in the horizontal and vertical directions, so that one image is displayed for one output screen grid. Data D _I,J will be allocated.

On the other hand, as shown in Fig. 4, for example, when obtaining an output image IM _OUT that is reduced by 1/2 in the horizontal direction only, two image data D are generated in the horizontal direction for one output screen grid. _{I and J} will be assigned.

Conversely, as shown in FIG. 5, when obtaining an output image IM _OUT that has been enlarged twice in the horizontal and vertical directions, for example, for every other output screen grid in the horizontal and vertical directions, one image data
D _I,J will be assigned.

Further, the address conversion circuit 5 includes a multiplication circuit 6,
7, 8 and 9, and write address information Aw _I,J , Aw _I+1,J , Aw _I,J+1 and weighting coefficient memory circuit 12
Weighting coefficients K _{I ,J} , K _I+1,J K _{I,J+1 corresponding to Aw I+1,J+1}
and K _I+1,J+1 , and each pixel Pw _I,J , Pw _{I+1, J , which surrounds the display position Q I,J} from the display position Q _I ,J represented by the conversion address information. The image data D _I, J is weighted according to the distances r ₁ , r ₂ , r ₃ and r ₄ to the points Pw _{I,J+1 and} Pw _I+1,J+1 .

In other words, the display position represented by the translated address information
From Q _I,J, each pixel Pw _I,J , Pw _I+1,J , Pw _I,J+1 and
When the distances r ₁ , r ₂ , r ₃ and r ₄ to Pw _I+1,J+ 1 are equal, each weighting coefficient K _I,J , K _I+1,J K _I,J+1 and K _I+1,J+1
Output so that the values of are equal and the sum is 1.

On the other hand, when the values of distances r ₁ , r ₂ , r ₃ and r ₄ are different, the distance r ₁ , r ₂ , r ₃ with the smaller value among the distances r ₁ , r ₂ , r _{3 and} r ₄ or the weighting factor corresponding to r ₄
K _I,J ~ K _I+1 , weighting coefficients K _I ,J corresponding to distances r ₁ , r ₂ , r ₃ or r ₄ with large values as the value of _J+1 is increased
~ Decrease the value of K _I+1,J+1 and set the weighting coefficient K _I,J ~
Weighting is done so that the sum of the values of K _{I+1 and J+1} becomes 1.

Thus, the address conversion circuit 5 converts the output image
It operates as an address conversion circuit that obtains address information representing the display position of a pixel of an input image on IM _OUT , and also functions as a weighting circuit that weights and distributes image data D _{I, J} together with multiplier circuits 6, 7, 8, and 9. Configure the circuit.

Therefore, in the output image memory circuit 10,
As read address information AD _RI changes, write address information Aw _I,J , Aw _I+1,J , Aw _I,J+1 and
Each pixel represented by Aw _I+1,J+1 Pw _I,J , Pw _I+1,J ,
Weighted image data can be obtained corresponding to Pw _{I,J +1 and Pw I+1,J+1} .

As a result, as shown in Figure 3, when performing image conversion to the same size on the input image, the output image
For one pixel Pw _I,J of IM _OUT , the display position Q _I-1,J-1 , Q _I,J-1 , in the four output screen grids surrounding it
Read out four weighted image data from the image data of Q _I-1,J and Q _I,J Address information
It can be obtained sequentially as AD _RI changes.

On the other hand, as shown in Figure 4, in the case of image conversion that is reduced by 1/2 in the horizontal direction, for one pixel Pw _I,J of the output image IM _OUT , the four surrounding pixels 8 display positions in the output screen grid
Q _I-1,J-1 1, Q _I-1,J-1 2, Q _I,J-1 1, Q _I,J-1 2, Q _I-1,J
Eight weighted image data can be obtained from the image data of 1, Q _I-1,J 2, Q _I,J 1, and Q _I,J 2, respectively.

On the other hand, as shown in Figure 5, in the case of image conversion that is doubled in the horizontal and vertical directions,
One display position in the four output screen grids surrounding one image Pw _I,J of output image IM _OUT
One piece of weighted image data can be sequentially obtained from the image data of Q _{I-1 and J-1} as the read address information AD _RI changes.

In this way, by weighting and distributing the image data D _I,J, even if the display position of the pixel of the input image converted to the output image side does not match the display position of the pixel of the output image, the output image IM Image data can be mapped to each _OUT pixel without exception, and deterioration in image quality can be prevented.

Therefore, the input image can be enlarged and transformed as desired.

The output image memory circuit 10 stores write address information.
Based on the weighted image data ADw _I,J , ADw _I+1,J , ADw _I,J+1 , ADw _I+1,J+1 , the corresponding write address information ADw _I,J , ADw _{I+ 1,J} , ADw _I,J+1 ,
Each pixel of output image IM _OUT corresponding to ADw _I+1,J+1
The image data of Pw _I,J , Pw _I+1,J , Pw _I,J+1 , and Pw _I+1,J+1 are stored sequentially.

At this time, as shown in FIGS. 3 and 4, 1
Multiple display positions Q _I-1,J-1 for one pixel Pw _I ,J,
Q _I,J-1 , J, Q _I-1,J , Q _I,J or Q _I-1,J-1 1, Q _I-1,J-1
2, Q _I,J-1 1, Q _I,J-1 2, Q _I-1,J 1, Q _I-1,J 2, Q _I,J
1, Q _I,J When weighted image data is obtained from 2, the image data of each pixel Pw _I,J already stored in the output image memory circuit 10 is sequentially added and stored.

In this way, for one pixel Pw _I,J of the output image IM _OUT , the plurality of image data assigned to the four output screen grids surrounding it are weighted,
The sum information is sequentially stored in the output image memory circuit 10.

Incidentally, the output image memory circuit 10 and the address conversion circuit 5 constitute a first addition circuit that adds weighted and distributed image data for each pixel of the output image.

On the other hand, the weighting coefficient memory circuit 12 uses the write address information ADw _I,J , ADw _I+1,J , ADw _I,J+1 and ADw _I+1,J+1 to Included address information
Each pixel Pw _I,J , Pw _I+1,J of the output image IM _OUT corresponding to ADw _{I,J , ADw I+1,J} , ADw I, _J+ 1 and ADw _{I+ 1,J} +1 ,
Pw _I,J+1 and Pw _I+1,J+1 weighting coefficient K _I,J , K _I+1,J
K _I,J+1 and K _I+1,J+1 are stored sequentially.

At this time, similarly to the output image memory circuit 10,
Multiple display positions Q _{I -1, J-1} for one pixel Pw I, J
〜Q _I,J or Q _I-1,J-1 1〜Q _I,J 2 from weighting coefficient K _I,J 〜
If K _{I+1, J+1} are obtained, the weighting coefficients of each corresponding pixel are sequentially added and stored.

Thus, the weighting memory circuit 12 and the address conversion circuit 5 constitute a second adding circuit that adds weighting coefficients required for weighting image data for each pixel of the output image.

In practice, in the case of reduced image conversion, the sum information Dw _I,J of weighted image data obtained from multiple display positions surrounding one pixel of the output image IM _OUT is calculated as shown in Figure 4. By asking for
Image data of high frequency components that change between the plurality of display positions (that is, corresponding to adjacent pixels on the input image) can be canceled out.

Furthermore, the smaller the reduction ratio, the more image data of the input image will be mapped to one output screen grid, and as a result, the frequency components that cancel out will also be mapped to the image data of the input image. low frequency components are canceled out.

Therefore, the frequency characteristics of the image data of the output image can be corrected without using a spatial filter circuit, and it is possible to obtain an image conversion device with a simpler overall configuration and less deterioration in image quality. can.

The read address generation circuit 13 outputs an output image IM _OUT
The read address information AD _RO corresponding to each pixel is output to the output image memory circuit 10 and the weighting coefficient memory circuit 12.

As a result, based on the read address information AD _RO , the sum information of the weighting coefficients of each pixel of the output image IM _OUT and the sum information of the image data Dw _I,J are sequentially output.

The weighting coefficient detection circuit 14 receives the sum information of the weighting coefficients, and outputs the sum information to the division circuit 15 when the value of the sum information is other than 0.

The division circuit 14 divides the sum information Dw _I,J of the image data into
It is divided by the value of the sum information of the weighting coefficients, and the image data obtained as a result is sequentially output as output image information VD _OUT .

By dividing in this way, 1 of the output image
Normalizes the signal level of image data to a predetermined signal level, even when image data of pixels of multiple input images is mapped to one pixel, or conversely, even when image data of only one pixel is mapped to one pixel. can do.

Thus, the sum information of the weighted image data
By normalizing and outputting Dw _I,J , even in the case of image conversion that locally enlarges or reduces,
It is possible to obtain output image information VD _OUT consisting of image data whose signal level is normalized to a constant value according to the expansion or reduction ratio of the image data of the input image.

Furthermore, when the value of the sum information of the weighting coefficients is 0 (for example, when the input image is reduced, a portion of the output image where the image data of the input image is not mapped), the weighting coefficient detection circuit 14 detects switching information.
S _K is output, and image data representing the background of the output screen is output via the division circuit 15 as the image data of the pixel.

In this way, the sum information of the weighting coefficients can be used as the switching information S _K for switching between the image data obtained by changing the input image and the background image data, so that the overall configuration can be simplified accordingly.

In the above configuration, the image data of the input image is converted into conversion address information representing the display position Q _I,J on the output image based on the control signal SC output from the control circuit 11 according to the mode of image conversion. is,
Based on the changed address information, the pixels around the display position Pw _I,J , Pw _I+1,J , Pw _I,J+1 and Pw _I+1,J+1
The image data D _I,J of the input image is weighted and distributed.

After the distributed image data is added for each pixel, it is normalized by being divided by the value of the sum information of the weighting coefficients and output from the division circuit 15.

According to the above configuration, it is possible to map the image data of each pixel of the input image to each pixel of the output image, and also correct the frequency characteristics and normalize the signal level without providing a spatial filter circuit. image data can be obtained.

Therefore, it is possible to obtain an image conversion device with a simpler overall configuration and less deterioration in image quality.

In the above-described embodiment, a case has been described in which the image data of the input image is assigned by weighting the four pixels of the output image, but the present invention is not limited to this.
The image data of the input image may be weighted and assigned to pixels of points, 16 points, etc.

In fact, in the above-described embodiment, if the input image is enlarged by a factor of two or more, some pixels of the output image will be
There is a possibility that the image data of the input image may not be assigned.

However, if the input image is assigned to 16 pixels, for example, even if the input image is enlarged by that much, all image data can be mapped.

Furthermore, in the above embodiment, a case was described in which the weighting coefficient detection circuit 14 displays a background image when the value of the sum information of the weighting coefficients is 0, but the present invention is not limited to this, and for example, The background image may be displayed when the value of the sum information is less than or equal to a predetermined value.

In this way, aliasing that may occur on the output image at the edge of the converted image can be prevented with a simple configuration.

H. Effect of the invention As described above, according to the invention, after weighting image data and assigning it to each pixel of the output image, adding and normalizing each pixel,
As a whole, it is possible to obtain an image conversion device that has a simple configuration and prevents deterioration of image quality.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of the image conversion device according to the present invention, FIGS. 2, 3, 4, and 5 are schematic diagrams for explaining its operation, and FIGS. FIG. 7 is a schematic diagram for explaining the operation of a conventional image conversion device. 1... Image conversion device, 2... Input image memory circuit, 5... Address conversion circuit, 6, 7, 8, 9...
...Multiplication circuit, 10...Output image memory circuit, 12
... Weighting coefficient memory circuit, 15 ... Division circuit.

Claims (1)

[Claims for Utility Model Registration] An image conversion device configured to output the output image obtained by converting the input image by mapping the image data of each pixel of the input image onto the output image, an address conversion circuit that outputs address information representing the display position of each pixel of the image on the output image; and an address conversion circuit that outputs address information representing the display position of each pixel of the image on the output image; For multiple pixels of the above output image surrounding the display position of
a weighting circuit that sequentially weights and distributes; a first addition circuit that adds the distributed image data obtained via the weighting circuit for each pixel of the output image; a second addition circuit that adds weighting amounts for each pixel of the output image; and a second addition circuit that adds the image data obtained through the first addition circuit in correspondence to the image data. 1. An image conversion device comprising: a division circuit that divides and outputs the result of the addition of weighted amounts obtained through the method.