JPH04305779A - Image converter - Google Patents

Abstract

PURPOSE:To prevent output images from being intermittently generated in write address control and to easily execute also complex image conversion processing at real time. CONSTITUTION:This image converter for deforming an input image based upon a set conversion function and forming an output image is provided with a forming means 2 for forming plural output small areas S corresponding to respective sampling points on the input image based upon the difference value of the conversion function on each sampling point and a conversion means for outputting the image information signal on each sampling point of the input image as the image information signal of each sampling point in each corresponding output small area S.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image conversion apparatus that uses an image memory to perform geometric transformations such as rotation, enlargement, reduction, etc. of an image.

0002

BACKGROUND OF THE INVENTION Special effects can be brought to an image by implementing geometrical transformations of the image, such as rotation, enlargement, reduction, etc. In recent years, an image conversion device using an image memory has been proposed to realize this geometric transformation of an image.

[0003] Methods for realizing geometric deformation of an image using an image memory include methods based on write address control and methods based on read address control. Write address control is based on information about which point in the output image each sample point of the input image corresponds to, and the address is controlled when writing to the image memory. It was done as it was done. On the other hand, read address control is based on information about which sample point in the input image each point of the output image corresponds to, and performs address control when reading from the image memory. It was designed to be carried out in a realistic manner.

0004

[Problem to be Solved by the Invention] Conventionally, in the write address control, for one sample point of an input image,
Because only one sample point in the image memory was matched,
For example, there is a drawback in that the output image has gaps in the enlarged part. Therefore, in the past, this write address control has not been used except for simple reduction without interruption. On the other hand, with read address control, even when the image is enlarged, the same sample point in the image memory can be read out repeatedly, so there will be no interruption in the output image. However, this read address control does not provide information on which sample point in the input image each point of the output image corresponds to.
For example, calculations must be made by finding the inverse function of a predetermined transformation function, which is difficult for arbitrary geometric transformations. Therefore, conventionally, the type of geometrical deformation of an image that can be realized using read address control has been greatly limited.

[0005] After all, it is often more intuitive to provide information (calculated using a predetermined transformation function) about which point in the output image each sample point of the input image corresponds to.
If it is possible to prevent intermittent output images as described above, it would be more convenient to use write address control. In view of these points, the present invention is designed to prevent intermittent output images from occurring in the output image using write address control, and also to easily perform complex image conversion processing in real time.

[0006]

[Means for Solving the Problems] The image conversion device of the present invention is an image conversion device that transforms an input image based on a conversion function set as shown in FIG. 1 to form an output image. forming means 2 for forming a plurality of output small regions corresponding to each sample point on the input image based on the difference value at each sample point of the conversion function of each sample point; The converting means 3 outputs the image information signal as the image information signal of each sample point in the corresponding output small area.

[0007]

[Operation] According to the present invention, the difference value at each sample point of the conversion function is used to obtain a plurality of small output regions corresponding to each sample point, so the partial differential equation is obtained in advance. It is unnecessary, the time required for calculation is relatively short, and complex image conversion processing can be easily performed in real time.

[0008]

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of an image conversion apparatus according to the present invention will be described below with reference to the drawings. In Figure 2A, each sample point of the input image is indicated by a black circle "●",
For example, it is placed at a position on the orthogonal coordinates whose X1 and X2 components are integer values. In addition, in FIG. 2B, each sample point of the output image memory is indicated by a white circle "○", and Y1
Y2 is placed at a position corresponding to each sample point of the input image shown in FIG. 2A on the orthogonal coordinates. Furthermore, Figure 2
Although only a few sample points of the input image and sample points of the output image memory are shown in A and B, in reality, a large number of sample points exist.

Here, the conversion formula from the X1 X2 orthogonal coordinates to the Y1 Y2 orthogonal coordinates, which is used to obtain an output image that has undergone a predetermined geometric deformation, is as follows:

0010
]

[Math 1]

[0011]

[Math 2]

is given as sample point (x1
, x2 ) is the position (y1 , y2 ) (y1 = ψ1 (x1
, x2 ), y2 = ψ2 (x1 , x2 ). In this case, according to the conventional write address control method, positions near the position (y1, y2) (y1 and y2 generally consist of an integer part and a decimal part, but for example, values y'1 and y with the decimal part truncated) ′
Select only one sample point of the output image memory located at the position specified in step 2), and add one sample point of the input image to this one sample point.
The image information of the sample point (x1, x2) was written. Therefore, when enlarging an image, for example, sample points where no image information is written are created in the output image memory between sample points where image information is written, resulting in gaps in the output image. In this example, on the Y1 Y2 orthogonal coordinates, the partial differential value or difference value at the sample point (x1, x2) of the conversion equation is calculated corresponding to the unit area centered on the sample point (x1, using the predetermined area S((y1, y
2) as the center (shown with diagonal lines in FIG. 2B), and select all sample points of the output image memory included in this predetermined area S as corresponding to the sample points (x1, x2) of the input image. Then, the input image sample point (x
1, x2) is written therein. On this Y1 Y2 rectangular coordinate, the position (y1, y2) corresponding to the sample point (x1, x2) of the input image is
), four numerical values a11, a12, a21, and a22 are used to determine a predetermined area S centered on . And these four numbers a11, a12, a21 and a22
is the conversion formula mentioned above, Y1 = ψ1 (X1 ,X2
), Y2 = ψ2 (X1, X2), respectively,
It is determined using the partial differential value or difference value at the sample point (x1, x2) of the input image. When using partial differential values,

[0013]

[Math 3]

[0014]

[Math 4]

[0015]

[Math 5]

[0016]

[Math 6]

[0017] When the difference value is used,

[
0018

[Math 7]

[0019]

[Math. 8]

[0020]

[Math. 9]

[0021]

[Math. 10]

It is determined as follows. Here, α11, α12,
α21, α22 are coefficients for area correction, α11, α
12, α21, and α22 are all 1 or more. In other words, Y
1 Y2 When an interval occurs between the predetermined areas S defined on the orthogonal coordinates, this predetermined area is expanded to eliminate the interval between the predetermined areas S. Incidentally, the cause of the intermittency is that the predetermined area S is a11, a12, a21, a22.
There are calculation errors, etc.

At this time, on the Y1 Y2 orthogonal coordinates, a predetermined area (parallelogram area) S is determined. This parallelogram area S is centered at the position (y1, y2) corresponding to the sample point (x1, x2) (y1 and y2 consist of integer parts i1 and i2 and decimal parts s1 and s2, respectively). , Y1 component is a11 and Y2
It is defined by a vector whose component is a21 and a vector whose Y1 component is a12 and whose Y2 component is a22.

In the end, the input image sample point (x1, x
2) Sample points included within this parallelogram area S are selected as sample points of the output image memory to correspond to the above. Then, the sample points (x1, x2) of the input image are placed at the sample points of the selected output image memory.
The image information at is written, and image conversion processing is performed at the sample point (x1, x2) of the input image. By performing such processing over all sample points of the input image, image conversion processing is performed for the entire input image.

FIG. 1 is a block diagram showing the entire image conversion apparatus that performs such processing. In the figure, 1 indicates an input image memory, which is composed of, for example, a RAM, and the image information of the sample point (x1, x2) of the input image is
Sample points (x1, x2
) is written to. Further, in FIG. 1, 2 is a processing device, 3 is an address generator, and 4 is an output image memory.

The processing device 2 has the conversion formula Y1 = ψ1 (x1, x2), Y2 shown in Equations 1 and 2 above.
The information of =ψ2 (x1, x2) is input in advance, and address control and arithmetic processing for the image memories 1 and 4 are performed. Further, the processing device 2 sequentially specifies sample points of the input image to be subjected to image conversion processing.

Sample points (x1, x2) of input image
is specified, the processing device 2 immediately uses the conversion formula to calculate the position (y1, y2) of the output image corresponding to this sample point (x1, x2) (y1 = ψ1 (
x1 , x2 ), y2 = ψ2 (x1 , x2 )
) is calculated, and the above-mentioned numbers 3, 4, 5 and 6 are calculated.
, or the numerical value a1 according to number 7, number 8, number 9, and number 10
1, a12, a21 and a22 are calculated. In this case, a memory, for example, a ROM, is prepared in advance, in which the results of these calculations are written according to the conversion formula, and the processing device 2 reads y1, y2, a from this ROM.
The values of 11, a12, a21 and a22 may be read out. Information about the position (y1, y2) of the output image and the numerical values a11, a12, a21 and a22 calculated by the processing device 2 is supplied to the address generator 3.

The address generator 3 is configured as shown in FIG. 4, for example. In the figure, 30 indicates an arithmetic circuit, and information of numerical values a11, a12, a21, and a22 is supplied to its input terminals 30a, 30b, 30c, and 30d. In this arithmetic circuit 30, numerical values a11, a12,
Based on the information of a21 and a22, for example, Figure 3
Various numerical values w1, w2, l1, l2, p1 and p2 that determine the parallelogram area S as shown in B
is calculated. Among the numerical values for determining the parallelogram area S calculated by this arithmetic circuit 30, w1, l1, and p1 are supplied to the sample point calculation circuit 31A, and w2
, l2 and p2 are supplied to the sample point calculation circuit 31B.

Further, in FIG. 4, 32 indicates an integer/decimal part separation circuit, and its terminals 32a and 32b have numerical values y1 (=i1 (integer part) + s1 (decimal part)) and y2 (=i2 ( Integer part) + s2 (decimal part)) information is supplied. The fractional parts s1 and s2 separated by this separation circuit 32 are supplied to a sample point calculation circuit 31A and also to a sample point calculation circuit 31B.

In the sample point calculating circuit 31A, a parallelogram area SA is defined on the Y1 Y2 orthogonal coordinates as shown in FIG. 5 based on the information of the supplied values w1, l1, p1, s1 and s2. In the figure, white circles "○" are sample points of the output image memory 4. This area SA is at a position (s1
, s2) so that its center is located at. As a result, in this sample point calculation circuit 31A, this area SA
The sample points included in are calculated. Further, in the sample point calculation circuit 31B, the supplied numerical values w2, l2
, p2, s1 and s2, a region SB is defined on the Y1 Y2 orthogonal coordinates as shown in FIG. In the case of this area SB as well, its center is (s1, s2
). As a result, the sample point calculation circuit 31B calculates sample points included in this area SB. These sample point calculation circuits 31A
The information on the sample points calculated in 31B and 31B is supplied to the common sample point calculation circuit 33. This calculation circuit 33
Then, common sample points included in the areas SA and SB, that is, sample points to be included in the parallelogram area S in FIG. 3B are calculated. In this case, as is clear from the above description, the calculated sample points are indicated by relative positions with respect to the reference sample point (0, 0).

Information on this common sample point is supplied to the address generation circuit 34. This address generation circuit 34 is supplied with information on integer parts i1 and i2 of y1 and y2 separated by the separation circuit 32. This address generation circuit 34 converts the reference sample point (0,0) to (i1,i2) by being supplied with the information of i1 and i2.
2), and at the same time, the above-mentioned common sample points are specified on the output image memory 4. Based on this, the address generation circuit 34 generates an address signal ADo that specifies a common sample point on the output image memory 4, that is, a sample point that can be included in the parallelogram area S. , taken out to the output terminal 35. The address signal ADo generated by the address generator 3 is supplied to the output image memory 4 together with the control signal Sco via the processing device 2.

[0032] Also, the input image memory 1 is input from the processing device 2.
is supplied with an address signal ADi specifying the sample point (x1, x2) and a control signal Sci. Then, the image information ID of the sample point (x1, x2) of the input image written in the sample point (x1, x2) is read out, and the above-mentioned address signal ADo of the output image memory 4 is read out via the processing device 2. is supplied to the sample point specified by and written to. Like this,
According to the image conversion device shown in FIG. 1, sample points of an input image are designated by the processing device 2, and the above-described processing is performed on each of them, and image conversion processing is also performed on the entire input image. . As described above, according to the image conversion device according to this example, arbitrary geometric deformation is relatively easy because it is based on write address control. Moreover, the sample points of the input image are stored in a predetermined area (
This corresponds to the sample point of the output image memory included in the area (according to the rate of change in the image), and one sample point of the input image is simply made to correspond to one sample point of the output image memory, as in the past. Even when the rate of change in an image is large compared to that of a normal image, for example when enlarging an image, or when a transformation function other than linear transformation is given,
There is no sample point without image information in the output image memory, and there is no fear of intermittency in the output image. Also, according to this example, the difference value at each sample point of the conversion function is used to find multiple output small regions corresponding to each sample point, so the partial differential equation must be found in advance. There is no need for this, the time required for calculation is relatively short, and complex image deformation processing can be easily performed in real time.

Next, another embodiment of the image conversion apparatus according to the present invention will be described with reference to FIG. 6 and subsequent figures. In this example, each sample point (
A predetermined area S0 represented by an m×n rectangle centering on the sample point (x1, x2) of the input image and including other sample points on the X1 and X2 rectangular coordinates on which black circles (indicated by black circles “●”) are arranged. is determined, and this predetermined area S0
Accordingly, as shown in FIG. 6B, a predetermined area S is determined on the Y1 Y2 orthogonal coordinates where each sample point (indicated by a white circle "○") of the output image memory is arranged. This predetermined area S is centered at the position (y1, y2) corresponding to the sample point (x1, x2) of the input image. Thereafter, each of the sample points of the output image memory included in the predetermined area S is selectively associated with the sample points of the input image included in the predetermined area S0. The conversion formula is

[0034]

[Math. 11]

[0035]

[Math. 12]

When the predetermined area S0 on the X1 X2 orthogonal coordinates is defined as an m×n rectangle as shown in FIG. 7A, the predetermined area S defined on the Y1 Y2 orthogonal coordinates is as shown in FIG. 7B. As shown, the position (y1, y2
)(y1 = ψ1 (x1, x2), y2 = ψ2
(x1,x2)), four numerical values a1
It becomes a parallelogram area determined by 1, a12, a21, and a22. 4 numbers a11, a12, a21 and a22
is determined using the partial differential value or difference value at the sample point (x1, x2) of the input image for each of the conversion equations shown in Equations 11 and 12 above. When using partial differential values,

[0037]

[Math. 13]

[0038]

[Math. 14]

[0039]

[Math. 15]

[0040]

[Math. 16]

[0041] When the difference value is used,

[
0042

[Math. 17]

[0043]

[Math. 18]

[0044]

[Math. 19]

[0045]

[Math. 20]

It is determined as follows. Here, α11, α12,
α21 and α22 are coefficients for correcting the size of the parallelogram area, and α11, α12, α21, and α22 are all 1 or more. In other words, if an interval occurs between the parallelogram areas S defined on the Y1 Y2 orthogonal coordinates depending on the predetermined area S0 defined on the X1 X2 orthogonal coordinates, this parallelogram area S is enlarged to This eliminates gaps between the shape areas S. Incidentally, the cause of the intermittence is that the predetermined area S is a11,
Approximately finding using a12, a21, a22,
There are also calculation errors, etc. In this way, corresponding to the predetermined area S0 determined on the X1
1 Y2 A parallelogram area S is defined on the orthogonal coordinates,
Among the sample points of the output image memory, which correspond to the sample points of the input image included in the predetermined area S0,
Sample points included within the parallelogram area S are selected. However, it is unclear which of the sample points of the input image included in the predetermined area S0 each of the sample points of the output image memory included in this parallelogram area S corresponds to. Therefore, next, sample points of the input image included in the predetermined area S0 corresponding to each of the sample points of the output image memory included in the parallelogram area S are determined. That is, it is determined which sample point in the output image memory each sample point of the input image included in the predetermined area S0 should correspond to. Now,
The point (y10, y20) in the parallelogram area S that corresponds to the point (x10, x20) in the predetermined area S0 is

[0047]

[Math. 21]

It can be obtained by first-order approximation as follows. In this formula, . In this example, based on this, each sample point of the output image memory included in the parallelogram area S corresponds to which sample point among the sample points of the input image included in the predetermined area S0. For the determination, the inverse function of the linear approximation formula shown in Equation 21 above is used. The inverse function of this linear approximation is

[0049]

[Math. 22]

It is shown as follows. This number 22 is an inverse matrix. Here, among the sample points of the output image memory included in the parallelogram area S, the sample point (y11, y21)
Corresponding to the point (x11, x21),

[0051]

[Math. 23]

is determined within a predetermined area S0, and as a sample point of the input image to be made to correspond to the sample point (y11, y21), for example, a sample point (x'11 within the predetermined area S0 close to the point (x11, x21) , x'21) are determined. For example, from the image information of four sample points of the input image surrounding the point (x11, x21), the point (x11, x2
Image information corresponding to 1) may be interpolated. Regarding other sample points of the output image memory included in the parallelogram area S, sample points of the input image included in the corresponding predetermined area S0 are determined in the same manner. In this way, the sample points of the output image memory corresponding to the sample points of the input image included in the predetermined area S0 defined on the input image are determined. In this case, the sample points of the output image memory corresponding to the sample points of the input image are the sample points of the output image memory corresponding to the sample points of the input image, that is, the sample points of the output image memory corresponding to the sample points of the input image, in the above-described embodiment. This is approximately the same as the sample point of the output image memory included in the predetermined area S defined on the output image memory. Eventually, the image information at the sample point of the corresponding input image is written to the sample point of this output image memory, and the image conversion process in the predetermined area S0 is completed. Note that the predetermined areas S0 as described above are sequentially taken on the input image, and image conversion processing is performed on the entire input image.

FIG. 8 is a block diagram showing the entire image conversion apparatus that performs such processing. In this Figure 8,
Portions corresponding to those in FIG. 1 are designated by the same reference numerals. The processing device 2 has the conversion formulas Y1 = ψ1 (X1, X2), Y2 = ψ2 (
Information on values m and n that determine the size of a predetermined area S0 defined on the X1X2 orthogonal coordinates are input in advance, and address control and arithmetic processing for the image memories 1 and 4 are performed. Furthermore, the processing device 2 sequentially specifies sample points of the input image that should be the center of a predetermined area S0 sequentially taken on the X1 X2 orthogonal coordinates. Input image sample point (x1, x2
) is specified, this processing device 2 uses a conversion formula to calculate the position (y1, y2) on the output image (on the orthogonal coordinates Y1, Y2) corresponding to this sample point (x1, x2).
2 ) (y1 = ψ1 (x1 , x2 ), y2 =
ψ2 (x1,
, numbers a11, a12, a according to numbers 19 and 20
21 and a22 are calculated. The position (y1, y2) on the output image calculated by this processing device 2, the numerical value a11
, a12, a21 and a22 are stored in the address generator 3.
supplied to This address generator 3 is constructed as shown in FIG. In this address generator 3, various numerical values w1, w2, l1, l2, p1 and p2 are used to obtain a parallelogram area S as shown in FIG. 7B based on the information of numerical values a11, a12, a21 and a22.
is calculated. Then, from these values, the area SA is created on the output image (on the Y1 Y2 orthogonal coordinates) as shown in Figure 10.
and SB are defined, and these areas SA and SB
Common sample points included in , that is, sample points to be included in the parallelogram area S in FIG. 7B are calculated. After these sample points are specified as sample points (y11, y21) on the output image memory 4, the address generator 3 specifies the sample points (y11, y21) on the output image memory 4.
An address signal ADo is generated which specifies the sample point (y11, y21) on the top. This address signal ADo is supplied to the output image memory 4 via the processing device 2 along with the control signal Sco, and is also supplied to the address generator 5.

The address generator 5 is configured as shown in FIG. 9, for example. In the figure, 50 indicates an inverse matrix calculation circuit, and its terminals 50a, 50b, 50c, and 50d are
Numerical values a11, a12, a21 and a22 from the processing device 2
information is provided. In this arithmetic circuit 50,
is calculated. Numerical values b11, b12, b21 and b22
The information is supplied to the address generation circuit 51. The terminal 51a of the address generation circuit 51 is connected to the address signal ADo (sample point (y
11, y21). ) is supplied and terminal 5
Information on numerical values x1, x2, y1, and y2 is supplied from the processing device 2 to 1b, 51c, 51d, and 51e. In this address generation circuit 51,
According to the above-mentioned number 23, the sample point (
(x11, ADi is taken out.

This address signal ADi obtained by the address generator 5 is supplied to the input image memory 1 via the processing device 2 together with the control signal Sci. When the address signal ADi is supplied to the input image memory 1, the image information ID of the sample point of the input image is read from the sample point of the input image memory 1, and the sample point (x11 , x21) is obtained by interpolation, and the sample point (y11, y21) specified by the address ADo in the output image memory 4 is obtained.
is supplied to and written to. From this address ADo, calculate the sample point (x11, x21) of the input image,
Sample points (x11) of the input image from the input image memory 1
, x21), and after creating image information corresponding to the sample point (x11, x21), the sample point (y11, y21) of the image memory 4 for output is read The writing process is repeated for the number of sample points (y11, y21) included in the parallelogram area S. In this way, image conversion processing is performed for a predetermined area S0 centered on the sample point (x1, x2) of the input image.

As described above, according to the image conversion apparatus shown in FIG.
are sequentially determined and the above-described processing is performed corresponding to each of them, so that image conversion processing is performed over the entire input image.

As described above, according to this embodiment, the same effects as those of the above-mentioned embodiments are achieved, and it is not necessary to define a predetermined area on the output image for each sample point of the input image. This is convenient because the amount of processing can be reduced accordingly. According to the above embodiment, the image information obtained from the input image is once written to the input image memory, but the image information may be directly supplied from the input image to the output image memory. .

[0058]

According to the present invention, arbitrary geometrical deformation is relatively easy because it is based on write address control. Moreover, the sample points of the input image can be converted into samples of the output image memory included in a predetermined area (area according to the conversion ratio of the image) determined on the output image memory according to the sample points and the conversion formula. Compared to the conventional method that simply corresponds one sample point of the input image to one sample point of the output image memory, it can be used even when the rate of change in the image is large, for example when enlarging the image. , even when a transformation function other than linear transformation is given, there will be no sample points with no image information in the output image memory, and there is no fear of intermittency in the output image. Further, according to the present invention, since a plurality of small output regions corresponding to each sample point are obtained using the difference value at each sample point of the conversion function, the partial differential equation is obtained in advance. There is no need for this, the time required for calculation is relatively short, and complex image conversion processing can be easily performed in real time.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing an embodiment of an image conversion device of the present invention.

FIG. 2 is a diagram for explaining one embodiment of the present invention.

FIG. 3 is a diagram illustrating an embodiment of the present invention.

FIG. 4 is a block diagram showing a specific configuration example of an address generator.

FIG. 5 is a diagram for explaining one embodiment of the present invention.

FIG. 6 is a diagram for explaining another embodiment of the present invention.

FIG. 7 is a diagram for explaining another embodiment of the present invention.

FIG. 8 is a block diagram showing another embodiment of the present invention.

FIG. 9 is a block diagram showing a specific example of the configuration of an address generator.

FIG. 10 is a diagram for explaining another embodiment of the present invention.

[Explanation of symbols]

1 Image memory for input 2 Processing equipment 3,5 Address generator 4 Image memory for output

Claims (1)

[Claims] 1. An image conversion device configured to convert an input image based on a set conversion function to form an output image, comprising: a plurality of output small areas each corresponding to each sample point on the input image; a forming means for forming an image information signal of each sample point on the input image based on the difference value at each sample point of the conversion function of each sample point, and an image information signal of each sample point in the corresponding output small area. 1. An image conversion device comprising: conversion means for outputting an information signal.