JP3055024B2 - Image data transfer device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

2. Description of the Related Art Computer graphics (CG) and image processing have become active due to the rapid progress of information processing technology in recent years. The present invention relates to high-speed data transfer and coordinate transformation of images, which are frequently used in this field.

[0002]

2. Description of the Related Art Conventionally, in order to perform coordinate transformation (enlargement, reduction, rotation, translation, etc.), a calculation has been performed using a numerical processing means. For example, the image shown in FIG.
To convert the coordinates into the image shown in FIG. 8B, the data of the image shown in FIG.

Next, the x and y coordinate data of the image data stored in the memory A is subjected to a numerical calculation process to perform coordinate conversion to obtain x 'and y' coordinate data. The image data is transferred to the memory B based on the converted coordinate data x 'and y'.

[0004] Here, specific calculations of coordinate conversion will be described by taking enlargement, reduction, parallel movement, rotation around the origin, and the like as examples.

(A) Enlargement and reduction around the origin

[0006]

(Equation 1)

... First formula

Sx: magnification in x direction Sy: magnification in y direction

(B) Parallel movement

[0010]

(Equation 2)

... Equation 2

Tx: translation distance in the x direction Ty: translation distance in the y direction

(C) Rotation about the origin

[0014]

(Equation 3)

... Equation 3

(D) Combination of enlargement, reduction, rotation, and translation

[0017]

(Equation 4)

... Equation 4

[0019]

(Equation 5)

... Formula 5

[0021]

(Equation 6)

... Equation 6

Then, these combinations are represented by a matrix,
It can be represented by a combination of (4), (5) and (6). For example, when the image stored in the memory A is enlarged (or reduced), and further rotated, and then the parallel-translated image is taken into the memory B, for each image,

[X, y, 1] · Rs · Rθ · Rt = [
x ', y', 1]

... Equation 7

The desired coordinate transformation can be performed by performing the following calculations and transferring the respective image data to the memory. FIG. 9 shows a hardware configuration for causing these arithmetic processings to be performed, and these arithmetic processings are performed by a central processing unit (CPU) or a combination of the CPU and the arithmetic unit. That is, the CPU fetches and holds the image data of a certain coordinate in the memory A, and performs a numerical operation on the coordinate using addition / subtraction, multiplication, trigonometric functions, and the like, and stores the image held in the CPU based on the new coordinate value. The data is transferred to the memory B.

[0027]

However, in the above-described conventional numerical operation processing, the CPU needs to perform a relatively time-consuming function calculation such as multiplication or trigonometric function. There is a problem that high-speed processing cannot be performed because the processing must be performed for each pixel data via the CPU. For this reason, it has been attempted to improve the processing speed by mounting a CPU operating at high speed or a numerical processor. But C
There is a problem that the access time of the PU to the memory exists, which hinders the speeding up, and that these computers are expensive and large in scale. Also,
The computer had to be provided with numerical operation processing means, and it was difficult to mount processing functions on an integrated circuit such as a single chip.

[0028]

SUMMARY OF THE INVENTION The present invention has been devised in view of the above-mentioned problems, and a source memory as a source for storing image data to be subjected to projective transformation, and image data being transferred. A transfer destination memory, a bus line for performing data transfer between the transfer source memory and the transfer destination memory by direct memory access, and an adder for calculating at least an address of the transfer source or the transfer destination. The adder calculates the image data to be measured in the transfer source memory by the displacement amount of the X address and the Y address of the transfer source or the transfer destination, and converts the X address and the Y address into a plane including the images before and after the projective transformation. Scanning is performed by sequentially adding or subtracting in the direction parallel to the intersecting axis, and the image data of the transfer source memory is directly transferred via the bus line, thereby obtaining the transfer destination memory. Projective transformed image data is stored.

Further, the adder of the present invention can be configured to set the displacement of the address for each scanning direction of the address.

Further, the adder of the present invention has a first adder for calculating a source address and a second adder for calculating a destination address. Projection conversion in a direction not parallel to the address direction of the X and Y addresses or projection conversion accompanied by enlargement and reduction may be performed.

The adder according to the present invention converts the image data to be measured in the transfer source memory from the X address and the Y address by the displacement of either the X address or the Y address of either the transfer source or the transfer destination. Scanning is performed by sequentially adding or subtracting in the direction parallel to the axis where the planes including the preceding and following images intersect, and by directly transferring the image data of the transfer source memory via the bus line, X
It is also possible to configure so that image data subjected to projection transformation parallel to the axis or the Y axis is stored.

[0032]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention constructed as described above
The transfer destination memory stores the image data to be subjected to the projective transformation, the transfer destination memory transfers the image data, and the bus line performs direct memory access between the transfer source memory and the transfer destination memory. And the adder
At least one of the source and destination addresses is calculated, and the adder converts the image data to be measured in the source memory by the displacement of the X or Y address of the source or destination. , X address and Y address are sequentially added or subtracted in a direction parallel to the axis where the plane including the image before and after the projective transformation intersects, and scanning is performed, and the image data in the transfer source memory is directly transferred via the bus line. Thus, the image data subjected to the projective transformation in the transfer destination memory can be stored.

Further, the adder of the present invention can set the displacement of the address for each scanning direction of the address.

Further, in the adder according to the present invention, the first adder calculates the address of the transfer source, the second adder calculates the address of the transfer destination, and the X and Y addresses of the transfer source memory. Projection transformation in a direction that is not parallel to the direction or projection transformation with enlargement and reduction can also be performed.

The adder of the present invention converts the image data to be measured in the transfer source memory from the X address and the Y address by the displacement of either the X address or the Y address of either the transfer source or the transfer destination. Scanning is performed by sequentially adding or subtracting in the direction parallel to the axis where the planes including the preceding and following images intersect, and the image data in the transfer source memory is directly transferred via the bus line. Image data that has undergone a projective transformation parallel to the axis or the Y axis can also be stored.

[0036]

【Example】

"Principle of the invention"

A conversion example of a straight line as shown in FIGS. 4 and 5 will be described as an example.

FIG. 4 shows the two-dimensional coordinates x and y before conversion, and corresponds to the memory arrangement of the first memory (A) 5 which is the transfer source memory. For example, the x coordinate may correspond to a column address in the memory, and the y coordinate may correspond to a row address in the memory. Note that the positive and negative signs, the origin, and the like can be arbitrarily determined.

Here, the inclination angle of the straight line AB with respect to the x-axis is θ, and A (x ₀ , y ₀ ), B (x _n , y _n ), | x _n
−x ₀ | = n · Δx, | y _n −y ₀ | = n · Δy.

The converted (memory address after transfer) can also be represented as x'-y 'and two-dimensional coordinates as shown in FIG.

That is, the inclination of the straight line A'B 'is θ', and A '
_{(X 0 ', y 0'} ), B '(x n', y n '), | x n' -
x ₀ ′ | = n · Δx ′, | y _n ′ −y ₀ ′ | = n · Δy ′
Then, the straight line A'B 'in FIG. 5 is obtained by enlarging (or reducing), rotating, and translating the straight line AB in FIG.

Therefore, the above conversion is performed in the first memory (A) 5
To the second memory (B) 6, and the operation of this address is represented by the following equation.

Xn = sigma Δx + x ₀ ..12

Yn = sigma Δy + y ₀ ··· Expression 13

The above two equations represent the transfer source first memory (A)
5 is an address.

X _n ′ = sigma Δx ′ + x ₀ ′....

Y _n '= sigma Δy' + y ₀ 'Expression 15

The above two formulas are used for the transfer destination second memory (B).
6 address. Here, n represents the number of transfer data, and the first memory (A) 5 and the second memory (B)
6 is the same value. Furthermore, sigma is a summation symbol (i = 1
To n).

.DELTA.x and .DELTA.y represent displacements (incremental values) in one transfer in the x and y directions. Therefore,
When the transfer from the first memory (A) 5 to the second memory (B) 6 is performed n times, the straight line AB is converted, and the displacement amount of the AB in the x direction is n · Δx, and the displacement in the y direction is The quantity is n · Δ
y. Therefore, the tilt angle is

Θ = tan ⁻¹ (Δy / Δx) Equation 16

Θ ′ = tan ⁻¹ (Δy ′ / Δx ′) Expression 17

Is as follows.

Next, the coordinate transformation of translation, enlargement / reduction, and rotation will be specifically described.

(1) Parallel movement

The moving amounts to be moved in parallel are Δx ₀ and Δy ₀
given that,

X ₀ '= x ₀ + Δx ₀ ··· Expression 18

Y ₀ ′ = y ₀ + Δy ₀ ...

Further, Δx = Δx ′ (20)

Δy = Δy ′ Equation 21

By doing so, it is possible to translate the straight line AB into a straight line A′B ′.

(2) Enlargement / reduction

If the magnification to be enlarged / reduced is α,

Δx ′ / Δx = Δy ′ / Δy = α (22)

Δx, Δx ′, Δy, Δy ′ such that
If the ratio is set and data is transferred, enlargement and reduction can be realized.

(3) Rotation

As expressed by the sixteenth and seventeenth equations, x,
Displacement (incremental value) Δx, Δy, Δx ′, Δy ′ in y direction
Is appropriately set to obtain a desired angle | θ ′.
−θ | is realized.

By combining the above-described parallel movement, enlargement / reduction, and rotation, image conversion as shown in FIGS. 4 and 5 can be easily performed.

An embodiment of the present invention will be described with reference to the drawings.

An embodiment of the present invention will be described with reference to the drawings.
As shown in FIG. 1, the coordinate transformation processing device main body 1 of the present embodiment
Are a CPU 2, a first accumulator (A) 3, a second accumulator (B) 4, a first memory (A) 5, a second memory (B) 6, a data bus 7 It consists of

The main body 1 of the coordinate transformation processing device is built in an image processing device or the like, and enlarges input image data.
This is for performing coordinate conversion such as reduction and rotation. CP
U2 is a central processing unit. The first accumulator (A) 3 is
The second accumulator (B) 4 performs an address addition / subtraction process on the first memory (A) 5, and performs an address addition / subtraction process on the second memory (B) 6.

The first memory (A) 5 stores the image data of the transfer source, and the second memory (B) 6
Is for storing the image data of the transfer destination.
The data bus 7 transfers a DMA (direct memory access) from the first memory (A) 5 to the second memory (B) 6.
It is a bus line to do.

Next, the internal structure of the first accumulator (A) 3 will be described with reference to FIG. 2. The first accumulator (A) 3 includes a first latch 311 and a second latch 312. It comprises a fifth latch 315 and a sixth latch 316, a first multiplexer 321 and a second multiplexer 322, and a first adder 331 and a second adder 332. The second accumulator (B) 4 has the same configuration as the first accumulator (A) 3. That is, the third latch 313,
Fourth latch 314, seventh latch 317, eighth latch 318, third multiplexer 323, fourth multiplexer 324, third adder 333, fourth adder 3
34.

In addition, a transfer number counter 34 for controlling the cumulative number of the accumulators (A) 3 and (B) 4 and a comparator 35
Is provided.

Here, the two-dimensional coordinates x and y before the coordinate conversion correspond to the memory arrangement of the transfer source. For example, the x coordinate corresponds to the row address of the memory, and the y coordinate corresponds to the column address. Similarly, the two-dimensional coordinates after coordinate conversion are defined as x ′ and y ′.

According to this definition, the first latch 311
Is for setting x ₀ which is an initial setting x coordinate before coordinate conversion. Then, the second latch 312 sets y ₀ , which is the initial setting y coordinate before the coordinate conversion,
The latch 313 sets the initial set x 'coordinate x ₀ ' after the coordinate conversion, and the fourth latch 314 sets the initial set y 'coordinate y ₀ ' after the coordinate conversion. The fifth latch 315 is for setting the increment Δx of the x coordinate before the coordinate conversion, and the sixth latch 316 is for setting the increment Δy of the y coordinate before the coordinate conversion. is there. And the seventh latch 317
The eighth latch 318 sets an increment Δy ′ of the y ′ coordinate after the coordinate conversion, and sets an increment Δx ′ of the x ′ coordinate after the coordinate conversion.

The first multiplexer 321 is for selecting the coordinate data of the x coordinate before the coordinate conversion and the addition data of the first adder 331. The second multiplexer 322 selects the data of the y coordinate before the coordinate conversion and the added data of the second adder 332.

The first adder 331 is for adding the x-coordinate data before the coordinate conversion and the increment of the x-coordinate. The second adder 332 adds the data of the y coordinate before the coordinate conversion and the increment of the y coordinate.

Then, the third multiplexer 323 is
The fourth multiplexer 324 selects the x ′ coordinate data after the coordinate conversion and the data of the third adder 333. The fourth multiplexer 324 includes the y ′ coordinate data after the coordinate conversion and the fourth adder 334. Is used to select the addition data of.

Further, the third adder 333 is for adding the data of the x 'coordinate after the coordinate conversion and the increment of the x' coordinate, and the fourth adder 334 is for adding the data of the x 'coordinate after the coordinate conversion. This is for adding the data of the y 'coordinate and the increment of the y' coordinate.

Next, based on FIG. 3, the first accumulator (A) 3
And the operation of the accumulator (B) 4 will be described. First, it is necessary to perform an initial setting.
1, x ₀ , the second latch 312 has y ₀ , the third latch 31
3 is set to x ₀ ′ and the fourth latch 314 is set to y ₀ ′ to determine the transfer initial address.

Next, in step 2, the fifth latch 315
Δx, sixth latch 316 Δy, seventh latch 31
7 is set to Δx ′, and the eighth latch 318 is set to Δy ′. That is, the increment of each coordinate is set in the latch.

Next, in step 3, the number of transfers n is set in the comparator 35. Then, in step 4, if a start pulse is given to the transfer number counter 34, the first, second, third, and fourth multiplexers 321, 322, 32
3, 324 are activated. That is, according to the signal from the transfer number counter 34, the four multiplexers 321, 322,
323, 324 are first to fourth latches 311, 312,
The data set in 313 and 314 is selected.

Therefore, x ₀ , y ₀ , x ₀ ′, y ₀ ′ are selected. Next, in step 6, an addition operation is performed on each coordinate data and the increment value. That is, when i = 1

X _i = x _i−1 + Δx Equation 8

Y _i = y _i-1 + Δy Equation 9

X _i ′ = x _i−1 ′ + Δx ′ Expression 10

Y _i ′ = y _i−1 ′ + Δy ′ Equation 11

Is performed.

More specifically, on the input side of the first adder 331, the output x0 of the first multiplexer 321 is output.
And Δx set by the fifth latch 315 are input. Therefore, the calculation result of the above-described equation (8) is output to the output side of the first adder 331.

Similarly, on the input side of the second adder 332, the output y ₀ of the second multiplexer 322 is provided.
Δy set by the sixth latch 316 is input. The input side of the third adder 333 receives x ₀ ′, which is the output of the third multiplexer 323, and Δx ′ set by the seventh latch 317. On the input side, a fourth multiplexer 324
Y ₀ is a the output 'and, second 8 [Delta] y which is set by the latch 318' is input.

As a result, the first, second, third and fourth adders 33
According to 1, 332, 333, and 334, the above-described eighth expression to first expression
The operation of one equation is performed.

Next, in step 7, a control circuit (not shown) causes the first memory (A) 5 to perform a read operation and the second memory (B) 6 to perform a write operation.
X ₁ , y ₁ , x ₁ ′, which are the results of the calculations performed in the eleventh equation,
The address value of y ₁ ′ is sent to the first memory (A) 5 and the second memory (B) 6.

As a result, the address of the first memory (A) 5 is
(X₁, Y₁) Is read out, and
The image data is stored in the address of the second memory (B) 6
(X₁', Y ₁'). Therefore, the first memo
The image data stored in the file (A) 5 is converted into the converted second data.
To the address of the memory (B) 6.

Next, at step 8, the value of the transfer number counter 34 is incremented by +1 to set i = 2. (I =
i + 1). The transfer number counter 34 can be counted up by a signal from a memory control circuit (not shown) or the like.

The selection of the first, second, third, and fourth multiplexers 321, 322, 323, and 324 is performed by selecting the first, second, and third multiplexers.
The output signals are switched to the output signals from the adders 331, 332, 333, and 334. As a result, the output signal of the first multiplexer 321, x _1, and the same way the second and third,
The output signal of the multiplexer 322, 323, 324 of the 4, _{respectively, y 1, x 1 ',} y 1' becomes. Then, the four multiplexers 321, 322, 323, 32
4 is subsequently selected by the corresponding adder 331, 33
2, 333 and 334 are set on the output side. Next, in step 9, the value of the transfer number counter 34 is compared by a comparator. If i = n is not satisfied, steps 6 to 9 are repeated. Accordingly, the calculation of these addresses and the data transfer operation are repeated n times, and when the comparator 35 recognizes i = n, in step 10, the transfer number counter 34 and the adders 331, 33 of the first, second, third, and fourth,
The operations of 2, 333 and 334 are stopped, and the address calculation and transfer work are completed.

In the present embodiment configured as described above, at the time of coordinate conversion, address conversion is performed only by addition and data transfer between memories is performed without performing complicated calculations or function calculations. Furthermore, for data transfer between memories,
Since it does not pass through the CPU, extremely high-speed transfer can be performed.

In this embodiment, the first, second, third, and fourth adders 331, 332, 333, and 334 provide Δx, Δx
Although the operation of adding y, Δx ′, and Δy ′ is performed, it is also possible to equivalently subtract by setting these numerical values to two's complement.

Therefore, the addition / subtraction processing in this specification is achieved by an adder.

It should be noted that this processing may be constituted by a subtractor instead of the adder.

In this embodiment, the first accumulator (A)
3, two accumulators are used for the second accumulator (B) 4 respectively, but the operation can be performed by one adder.

Next, an example of transfer and coordinate conversion using the present transfer apparatus will be described.

Although FIGS. 4 and 5 show examples of straight line conversion, graphic conversion can be performed by a similar method.

That is, if the figure A shown in FIG. 6 is considered as a set of straight lines, it can be clearly understood. However, depending on the shape of the figure, the transfer number n and the displacement amount (incremental value) Δx, Δy, Δx ′, Δy ′, the starting point of the line, x ₀ , y
_{0, x 0 ', y 0} ' there is a need to set.

Further, without being limited to this method, figure conversion can be performed by performing area conversion as shown in FIG. In this case, the starting points x ₀ , y ₀ , x ₀ ′, y
₀ ′ is set for each straight line, and Δx, Δy, Δx ′, Δy ′, n
The image conversion can be performed only once by setting.

Further, it is also possible to convert the entire area of the first memory (A) 5 and transfer it to the second memory (B) 6. Note that the original image can be transferred as it is without converting the area of the first memory (A) 5.

In this embodiment configured as described above, the transfer number n, the displacement amount (incremental value) Δx, Δy, Δx ′, Δy ′, the starting point of the straight line, x ₀ , y ₀ , x ₀ ′, y Various coordinate transformations can be easily performed only by addition and subtraction of ₀ '. In other words, a variety of complicated coordinate transformations such as two-dimensional rotation, parallel movement of the origin, Helmert transformation for correcting the scale, affine transformation for correcting distortion, and projection transformation can be performed by simple methods such as Δx and Δy. By setting the parameters, the conversion operation can be easily performed.

In the above description, an adder is provided for each of the transfer source and the transfer destination, but an adder is provided for either the transfer destination or the transfer source. Can be achieved.

Next, the first accumulator (A) 3 of the first memory (A) 5 as the transfer source memory and the second accumulator of the second memory (B) 6 as the transfer destination memory A description will be given of an embodiment in which projective transformation in a direction not parallel to the X-axis and Y-axis of the memory or projective transformation involving rotation, enlargement, reduction, and the like is performed by using (B) 4.

That is, when an adder is used for one of the first memory (A) 5 and the second memory (B) 6, the projection is performed in a direction parallel to the X axis or the Y axis of the memory. Although only conversion can be performed, if an adder is provided in each memory, projection conversion in a direction not parallel to the X-axis and Y-axis, and projection conversion involving rotation and the like can be performed.

First, a case where the present embodiment is applied to projective transformation will be described with reference to FIGS.

FIG. 10 shows a plane image A'B 'including the axis γ.
The image ABCD (AB // γ //
A′B ′). In this conversion, it is assumed that the projection conversion image ABCD is stored in the first memory (A) 5 as shown in FIG.

Here, the image A'B'C'D 'before the projective transformation
Let's consider the case of converting to Line segment AB of the image,
Since CD, A′B ′ and C′D ′ are parallel to the axis γ, the axis γ
The image of the first memory (A) 5, which is the transfer source memory, is scanned in parallel with the above.

Then, as shown in FIG. 11B, if the second memory (B) 6 as the transfer destination memory is also scanned in the direction parallel to γ, the projection transformation that is not parallel to the X and Y axis directions is performed. Done.

When such scanning is performed, the scanning pitch of each line segment becomes equal (corresponding to the displacement amount). Therefore, the magnification α is determined for each line segment in accordance with Equation 22, and the displacement amounts △ x, △ y, △
x 'and △ y' are set, and the starting points x ₀ , y ₀ , x ₀ ′,
After determining y ₀ ′ and the number of transfers n, the first memory (A)
5 is transferred to the second memory (B) 6, FIG.
An image A'B'C'D 'before projective transformation shown in (b) can be obtained.

To obtain a rotated image before projective transformation,
First memory (A) 5 which is a transfer source memory in parallel with axis γ
Is scanned, and the second memory (B) 6 serving as the transfer destination memory is scanned in parallel to the axis γ ′ in the angular direction to be rotated as shown in FIG.

When such scanning is performed, the scanning pitch of each line segment becomes equal (corresponding to the amount of displacement), as in the case of conversion to an image before projective conversion. The magnification α is determined, and the displacement amounts △ x, △ y, △ x ', △ y'
Are set, and the starting point x ₀ , y ₀ , x ₀ ′, y ₀ ′ of the straight line and the number of transfers n are determined, and then the transfer is performed from the first memory (A) 5 to the second memory (B) 6. For example, the rotated images A "B" C "D" before the projective transformation shown in FIG. 11C can be obtained at once from the projective transform image ABCD. Then, the magnification α
By multiplying by the value you want to enlarge or reduce, you can also scale at the same time.

Further, by the same transfer method, an image can be projected and converted to be enlarged / reduced / rotated, or a projection-converted image can be converted into another projection-converted image, enlarged, reduced / rotated, and the like.

Next, an application example in which this embodiment is applied to pattern measurement will be described with reference to FIG. FIG. 12 shows a case in which an elliptic image is obtained. As shown in FIG. 12A, a radial scan of the pattern image is performed for each straight line, and at the same time, FIG.
The arrangement can be changed in the horizontal direction on the second memory (B) 6 shown in FIG.

By performing the pattern operation after such an arrangement change, there is an effect that image data can be easily obtained only by scanning the memory in the horizontal direction.

According to the conventional method, the scanning shown in FIG. 12A requires repeating the steps of calculating the coordinates of each point for each radiation one by one and taking in the data to perform the pattern operation. there were. As described above, the conventional method has a problem that as the number of scanning points increases, the time required for address calculation increases, and high-speed calculation processing becomes difficult.

On the other hand, in this embodiment, a scanning line segment is set for each line segment (x ₀ , y ₀ , x ₀ ′, y ₀ ′, △ x, △ y, △
By performing x ′, △ y ′, n), the arrangement can be changed only by the time required for the transfer, and the memory scan required for the pattern calculation can be performed simply and at high speed. There is an effect that processing can be performed.

Therefore, if this pattern measurement is used,
If the elliptical image is converted into a straight line that can be scanned horizontally, and this straight line is scanned and measured, the diameter of the elliptical image in an arbitrary direction can be measured.

As described above, in this embodiment applied to pattern measurement, necessary image components are scanned at high speed,
Since you can change the array to a position that is easy to process,
There is an effect that high-speed pattern measurement can be performed.

The pattern measurement is not limited to the above-mentioned application examples, but can be applied to any measurement.

Although the adder of this embodiment is composed of an integer part and a decimal part adder, it can be composed of at least a decimal part adder.

Further, image data is taken into the CPU from the address operation result of the first memory (A) 5, and the CPU
Image data can be written to a corresponding address in the second memory (B) 6. However, since image data must be transferred via the CPU, it is difficult to perform high-speed transfer.

[0128]

According to the present invention having the above-described configuration, a transfer source memory in which image data to be subjected to projection conversion is stored, a transfer destination memory in which image data is transferred, and a transfer source memory It comprises a bus line for performing data transfer with a destination memory by direct memory access, and an adder for calculating at least the address of either the source or the destination. The image data to be measured in the memory is sequentially transferred in the direction parallel to the axis at which the plane containing the image before and after the projective transformation intersects the X address and the Y address by the displacement of the X address and the Y address of the transfer source or the transfer destination. Scanning is performed by adding or subtracting, and the image data of the source memory is directly transferred via the bus line, so that the image data that has been projected and transformed by the destination memory is recorded. Since the multiplication of the coordinate data, do not require complicated math function calculation or the like, thus there is an effect that can provide a transfer device of the image data to be fast coordinate transformation an extremely simple structure.

In particular, when an adder for calculating an address is provided in each of the transfer source memory and the transfer destination memory,
There is an effect that projective transformation in a direction not parallel to the x-axis and y-axis directions and projective transformation involving rotation, enlargement, reduction, and the like can be realized.

[0130]

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of an embodiment of the present invention.

FIG. 2 is a diagram illustrating a configuration of an accumulator according to the present embodiment.

FIG. 3 is a flowchart of a calculation operation according to the embodiment.

FIG. 4 is a diagram showing an original image stored in a first memory (A) in coordinate conversion.

FIG. 5 is a diagram showing a converted image transferred to a second memory (B).

FIG. 6 is a diagram illustrating graphic conversion.

FIG. 7 is a diagram showing area conversion.

FIG. 8 is a diagram schematically illustrating coordinate conversion.

FIG. 9 is a diagram illustrating a conventional technique.

FIG. 10 is a diagram illustrating an example in which the present embodiment is applied to oblique transformation.

FIG. 11 is a diagram illustrating an example in which the present embodiment is applied to oblique transformation.

FIG. 12 is a diagram illustrating an example in which the present embodiment is applied to pattern measurement.

[Explanation of symbols]

 Reference Signs List 1 coordinate transformation processing device 2 CPU 3 first accumulator (A) 4 second accumulator (B) 5 first memory (A) 6 data bus 31 latch 32 multiplexer 33 adder 34 transfer number counter 35 comparator

Continuation of the front page (56) References JP-A-59-188773 (JP, A) JP-A-63-38986 (JP, A) JP-A-61-223986 (JP, A) JP-A-1-315845 (JP) , A) JP-A-61-214888 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) G06T 3/00-3/60 H04N 1/387

Claims (4)

(57) [Claims] 1. A transfer source memory in which image data to be subjected to projective transformation is stored, a transfer destination memory to which image data is transferred, and direct data transfer between the transfer source memory and the transfer destination memory. It comprises a bus line for memory access, and an adder for calculating at least one of the transfer source and transfer destination addresses.
In this adder, the image data to be measured in the transfer source memory is the displacement amount of the X address and the Y address of the transfer source or the transfer destination, and the planes including the images before and after the X and Y addresses are subjected to the projective transformation intersect. By performing scanning by sequentially adding or subtracting in the direction parallel to the axis, and directly transferring the image data of the source memory via the bus line,
An image data transfer device in which image data subjected to projection conversion in a transfer destination memory is stored. 2. The image data transfer device according to claim 1, wherein said adder is configured to set an address displacement amount for each address scanning direction. 3. The adder has a first adder for calculating a transfer source address and a second adder for calculating a transfer destination address. , Y
3. The image data transfer device according to claim 1, wherein the image data transfer device is configured to perform a projective transformation in a direction not parallel to the address direction of the address or a projective transformation accompanied by enlargement and reduction. 4. The image processing apparatus according to claim 1, wherein the adder converts the image data to be measured in the transfer source memory into one of a transfer source and a transfer destination.
The displacement amount of the X address and the Y address
Scanning is performed by sequentially adding or subtracting addresses in a direction parallel to an axis where a plane including an image before and after the projective transformation intersects,
By directly transferring the image data of the source memory via the bus line, the X-axis or Y-
3. The image data transfer device according to claim 1, wherein the image data subjected to the projective transformation parallel to the axis is stored.