JP2853116B2 - Image conversion device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION   The present invention will be described in the following order. A Industrial application fields Summary of invention B C Conventional technology Problems to be solved by invention D Means for solving the E problem (FIGS. 1 to 5) F action (Figs. 1 to 5) G Example (FIGS. 1 to 11) Effect of H invention A Industrial application fields   The present invention relates to an image conversion device, and particularly to a display device on a two-dimensional plane.
By applying geometric transformation to the input image
After attaching the image to an arbitrary three-dimensional surface,
To obtain the same converted image as seen through the original plane
It is. Summary of invention B   The present invention relates to an input image data stored in an input image memory.
Data corresponding to the position converted by the conversion processing unit.
By writing to the memory area of the input image memory,
In an image conversion device that converts an input image into a converted image,
Four vertices of the first minute rectangular area forming the input image
Converts image data into a converted image,
Each image included in the second minute rectangular area formed on the image
The element position to the position corresponding to the position of the first minute rectangular area
By interpolation, the second fine pattern formed on the converted image is obtained.
A gap can be prevented from being generated between the small rectangular areas. C Conventional technology   This type of image converter is used for special effects devices for broadcasting and
It is used in the creation of the animation etc.
For example, JP-A-58-19975 and JP-A-60-59474
Transform a small rectangular area of the input image as disclosed
There has been proposed a method of converting an image into a minute rectangular area.
This image conversion method is used when converting an input image to a converted image.
Can be approximated by a linear expression (that is,
Linear approximation) to reduce the amount of data processing
Can be significantly reduced, which in practice allows
Image conversion can be performed (one frame for 1/30 seconds or less)
Image data can be processed). Problems to be solved by invention D   However, according to this method, fine patterns formed on the input image
When converting a small rectangular area to a small rectangular area on the converted image
In addition, there is a possibility that a gap may be formed between adjacent minute rectangular areas.
You. This gap gives extreme deformation, especially to the input image.
This is likely to occur for parts that need to be obtained. This problem
To solve the problem, JP-A-58-19975 and JP-A-60-594
In Japanese Patent Publication No. 74, two adjacent small rectangular areas overlap each other.
Parameter to control the image conversion coefficient.
Data in the conversion equation,
Some ideas such as slightly increasing the size of the small rectangular area
Had been done.   In this specification, the term “small rectangular area” is used.
Means that the shape of the small area is square, rectangle, parallelogram,
It shall include other squares.   These methods do not create gaps in the converted image.
Can be said to be an effective means in the
In terms of the quality of the resulting converted image,
If the degree of approximation is insufficient, image quality will deteriorate
May cause.   The present invention has been made in view of the above points, and
Calculation can be performed simply and at high speed using a linear conversion equation.
Without deteriorating the image quality without losing the advantage
Attempt to propose an image conversion device that can be avoided
It is. Means for solving problem E   In order to solve such a problem, in the present invention,   Input image data IM_INInput image memo to read and store
3 and the input image data stored in the input image memory 3.
Ta IM_INIs divided into a first minute rectangular area ER1, and the first
Of the four vertices of the small rectangular area ER1 (x₀, Y₀), (X
₁, Y₀), (X₀, Y₁), (X₁, Y₁) To each conversion function
A conversion processing unit 2 that performs conversion based on
Memory areas corresponding to the positions of the four vertices
A. Image data of four vertices of the first minute rectangular area ER1
By reading and storing the data, the second minute rectangular area ER
For output to form image data corresponding to the four vertices 2
Image memory 4, conversion function and second minute rectangular area ER
Count that increases by the value corresponding to the pitch of each pixel in 2.
Of each pixel in the second minute rectangular area ER2 based on the
Position f₀₀ ^*, F₀₁ ^*, F_Ten ^*, F₁₁ ^*And the first minute rectangular area
Position information data indicating the relative positional relationship with each pixel in ER1
And a conversion destination interpolation means 1 for generating the
The conversion processing unit 2 performs the first fine processing based on the position information data.
The pixel data of each pixel in the small rectangular area ER1 is
Output image corresponding to the corresponding pixel position in the rectangular area ER2
The data is written in the memory area of the image memory 4. F action   Input image IM_INFour small rectangular areas ER1 formed above
Vertex (x₀, Y₀), (X₁, Y₀), (X₀, Y₁), (X₁,
y₁) Is converted by the conversion processing unit 2 and converted image IM
_OUTTop four positions f₀₀ ^*, F₀₁ ^*, F_Ten ^*, F₁₁ ^*Conversion to
Is done.   And input image IM_INEach pixel included in is converted to
According to the interpolation address conversion data IAD of the information interpolation means 1
Converted image IM_OUTInterpolated above. Here the converted image IM_OUTUp
The position of the pixel interpolated in the small rectangular area ER2 of
Statue IM_INCorresponding pixels included in the small rectangular area ER1 of
Relative to the first minute rectangular area ER1
Is determined by the positional relationship.   Thus, input image IM_INUpper first small rectangular area ER1
Four vertices (x₀, Y₀), (X₁, Y₀), (X₀, Y₁),
(X₁, Y₁Convert image IM_OUTConverted above, converted 4
Image data position f₀₀ ^*, F₀₁ ^*, F_Ten ^*, F₁₁ ^*By
To form the second minute rectangular area ER2
The adjacent small rectangular area ER2 has a common border line
And as a result, the adjacent second minute rectangular area ER2
It is possible to effectively avoid the possibility that a gap is formed between the two. G Example   An embodiment of the present invention will be described below in detail with reference to the drawings. Second
The figure shows the input image IM_INShows a two-dimensional plane (this is the xy coordinate system
Above) by a line parallel to the y-axis and a line parallel to the x-axis.
Thus, a minute rectangular area ER1 is formed.   Input image IM_INThe intersection of the x-axis and y-axis lines is
The coordinates of the four vertices of the small rectangular area ER1 (x₀, Y₀),
(X₁, Y₀), (X₀, Y₁), (X₁  y₁), This rank
Is the conversion formula   Output image expressed in st coordinate system on two-dimensional plane by
IM_OUTConverted to above. Therefore, the output image IM_OUTAbove,
Input image IM_INFour vertices (x₀,
y₀), (X₁, Y₀), (X₀, Y₁), (X₁, Y₁Corresponding to)
Four points are obtained by calculation based on equation (1).
Therefore, as shown in FIG. 3, on the st plane,
These four points f^*(X₀, Y₀), F^*(X₁,
y₀), F^*(X₀, Y₁), F^*(X₁, Y₁) In a straight line
The small rectangular area ER on the xy plane
A small rectangular area ER2 corresponding to 1 can be formed.   Input image IM_INEach small rectangular area ER1 is shown in FIG.
Thus, a predetermined number of pixels are included. One small rectangular area
The pixel position in the x-axis direction in ER1 is represented by coordinates u,
When the pitch between two pixels is Δu, the pixel
X-axis from reference point (u = 0), which is one vertex of region ER1
To the other vertex in the direction (u = 1) at intervals of Δu in the u direction.
Are lined up.   Also, the position of the pixel in the y-axis direction in the minute rectangular area ER1
And the pitch between pixels is represented by Δv.
The pixel is y from the reference point (v = 0) of the minute rectangular area ER1.
To the other vertex (v = 1) in the y-axis direction at Δv intervals
Are lined up.   Thus, each small rectangular area ER1 has its reference point (x₀, Y₀)
Can be specified using the xy coordinate system, and the small rectangular area ER
Each pixel in 1 is represented by Δu by the coordinates (u, v) on the uv plane.
Interval (u = 0, Δu, 2Δu, 3Δu...) And Δv
Can be specified in intervals (v = 0, Δv, 2Δv, 3Δv ...)
Will be.   In FIG. 4, a point P (u, v) of the minute rectangular area ER1
Considering this, this point P (u, v) is
Direction is divided at a ratio of u: (1-u).
One point in the direction v is divided at a ratio of v: (1-v).
You can see that   Now convert image IM_OUT4 of the corresponding small rectangular area ER2
One vertex f₀₀ ^*= F^*(X₀, Y₀…… (2) f_Ten ^*= F^*(X₁, Y₀) …… (3) f₀₁ ^*= F^*(X₀, Y₁) …… (4) f₁₁ ^*= F^*(X₁, Y₁) …… (5) If it is represented by, input image IM_INUpper small rectangular area ER1 (fourth
The pixel corresponding to each pixel included in FIG.
Output image IM having square vertices represented by equation (5)_OUTupper
It should be distributed in the small rectangular area ER2.   Therefore, in the present invention, the position of each pixel in the region ER2
To four vertices f₀₀ ^*, F_Ten ^*, F₀₁ ^*, F₁₁ ^*On the basis of the,
It is obtained by performing an interpolation operation.   That is, the input image IM_INUpper small rectangular area ER1 (fourth
The point P (u, v) on the figure) is a small rectangular area when viewed in the u direction.
Two opposite sides of the region ER1 (that is, u = 0 and u =
U: (1-u) between two opposing sides passing through the point 1)
It is represented as a divided point. Also, relative to v direction
Two sides (that is, relative through the points v = 0 and v = 1)
Between two sides facing each other) as a point divided by v: (1-v)
Is done.   This relationship is converted image IM_OUTAlso holds for
Corresponding to the point P (u, v) on the minute rectangular area ER1
G on the small rectangular area ER2 (Fig. 5)^*(U, v)
The position is obtained by calculation. Thus input image IM_INFine above
For the small rectangular area ER1, v =
0 to v = 1 are sequentially set, and
Pixels are sequentially scanned in the u direction from u = 0 to u = 1.
When the position of the corresponding pixel is_OUTUp
Can be converted onto the small rectangular area ER2.   The interpolation operation is performed by the method described with reference to FIG.
It is. That is, four vertices of the transformed image ER2 on the st plane
Is, as described above for equations (2) to (5),
Vector f from the origin₀₀ ^*, F_Ten ^*, F₀₁ ^*, F₁₁ ^*Represented by
It is. Therefore, first, the position vector f₀₀ ^*And f_Ten ^*Tie
On the line, a position vector p with a ratio of u: (1-u)₁ ^*Ask for
Then P₁ ^*= U (f_Ten ^*−f₀₀ ^*) + F₀₀ ^*     = (1-u) f₀₀ ^*+ Uf_Ten ^*         ...... (6) Becomes The position vector f_Ten ^*From the position vector f₀₀ ^*
To obtain the vector (f_Ten ^*−
f₀₀ ^*) U: 1 vector component to position vector f₀₀ ^*Join
If possible, position vector P₁ ^*Can be requested.   Similarly, position vector f₀₁ ^*And f₁₁ ^*One side connecting between
, A position vector P representing the position of u: (1-u)_Two ^*
Is P_Two ^*= U (f₁₁ ^*−f₀₁ ^*) + F₀₁ ^*     = Uf₁₁ ^*+ (1-u) f₀₁ ^*         …… (7) Becomes Therefore, the position vector P₁ ^*And P_Two ^*On the line connecting
The position internally divided by the ratio of v: (1−v) is expressed by the following equation. P_Three ^*= V (P_Two ^*−P₁ ^*) + P₁ ^*          …… (8) The position vector P represented by_Three ^*Can be asked as
You.   Therefore, the above equations (6) and (7) are substituted into equation (8).
And rearranging the coefficients u and v, P_Three ^*= V ｛u (f₁₁ ^*−f₀₁ ^*) + F₀₁ ^*−u (f_Ten ^*−f
₀₀ ^*) + F₀₀ ^*｝ + U (f_Ten ^*−f₀₀ ^*) + F₀₀ ^*     = (1-v) ｛(1-u) f₀₀ ^*+ Uf_Ten ^*｝ + V
｛(1-u) f₀₁ ^*+ Uf₁₁ ^*｝     = F₀₀ ^*+ (F_Ten ^*−f₀₀ ^*) U + (f₀₁ ^*−f₀₀ ^*)
v + (f₀₀ ^*−f_Ten ^*−f₀₁ ^*+ F₁₁ ^*) Uv …… (9) It can be seen that there is a relationship.   Here, in equation (9), a₀ ^*≡f₀₀ ^*                           ……(Ten) a₁ ^*≡−f₀₀ ^*+ F_Ten ^*                  …… (11) a_Two ^*≡−f₀₀ ^*+ F₀₁ ^*                  …… (12) a_Three ^*≡f₀₀ ^*−f₀₁ ^*−f_Ten ^*+ F₁₁ ^*      ……(13) Equation (9) becomes the following equation. P_Three ^*= A₀ ^*+ A₁ ^*u + a_Two ^*v + a_Threeuv …… (14) Can be expressed by a linear expression of coefficients u and v as follows:
Will be.   This means that one of the coefficients u and v is fixed.
If the other is variable in the state where
Le P_Three ^*Is obtained by performing the operation of the linear equation.
And can be done.   Now, if the coefficient u is fixed and the coefficient v is varied,
If the position vector P_Three ^*Is P_Three ^*= (A₀ ^*+ A₁ ^*u) + (a_Two ^*+ A_Three ^*u) v                                       …… (15) , A constant (a₀ ^*+ A₁ ^*u) and (a)_Two ^*+ A_Three ^*
It becomes a linear expression for v that includes u).   Therefore, as described above with reference to FIG._IN
In the small rectangular area ER1 in, v is 0, Δv, 2
By changing Δv, 3Δv, etc., u
When the scanning line for the direction is specified in order,
Converted image IM_OUTIn the small rectangular area ER2 (FIG. 5)
Indicates that the corresponding scan line can be specified
ing.   In this way, the coefficient v is set at every pixel position Δv in the v direction.
Are sequentially specified, so that the scan line in the u direction is
The coefficient v is fixed in a state where one is specified one by one.
From this, if equation (14) is rearranged for coefficient u, P_Three ^*= (A₀ ^*+ A_Two ^*v) + (a₁ ^*+ A_Three ^*v) u                                       …… (16) Can be represented by   In this equation (16), the value of the coefficient u is 0, Δu, 2Δu,
U at every pitch Δu of pixels in the u direction, such as 3Δu.
By specifying the pixels in the direction sequentially, on each scan line
Of the converted image IM_OUTSmall on
In the rectangular area ER2, the image on the corresponding scan line
This means that elements can be sequentially scanned.   Therefore, input image IM_INOf small rectangular area ER1 in
Elementary position P^*If (u, v) is specified,
Interpolation by coefficients u and v based on the conversion formula
Can be calculated. This interpolation formula is It expresses.   Converted image IM_OUTOf each pixel included in the small rectangular area ER2 of
Scanning is performed as follows.   First, input image IM_INBy setting u = 0
The coefficient u is fixed, and the coefficient v is sequentially set to 0, Δv, 2Δ
.., 3Δv.... Dark
If v = 0, the converted image IM_OUTUpper point P_Three ^*Is the formula (17)
From Position. Subsequently, when v = Δv, from equation (17), The position vector P_Three ^*Is the value of equation (18)
From a_Two ^*Move to the position incremented by Δv.   When the coefficient v becomes 2Δv, the position vector P_Three ^*Is
From equation (17) From the state of equation (19), a_Two ^*Δv
Move to the position incremented only. Furthermore, the coefficient v is 3
In the same way as ΔvFrom the state of equation (20), a_Two ^*Δv
Move to the incremented position.   The same applies to the input image IM_INThe value of the above coefficient v is not Δv
As the number increases, the converted image IM_OUTTop position
Vector p_Three ^*Is a_Two ^*At the position incremented by Δv
Will be moved.   Thus, in the small rectangular area ER1, the value of the coefficient v is v
= 0, Δv, 2Δv, 3Δv...
When you go, the converted image IM_OUTTop position
Tor P_Three ^*Are sequentially specified.   And the coefficient v (= 0, Δv, 2Δv, 3Δv...)
When the value of the coefficient u is 0, Δu, 2Δu, 3Δu,
……, the conversion image IM_OUTUp
Position vector P_Three ^*Changes as follows.   First, when v = 0, u = 0, Δu, 2Δu, 3Δu ...
, The position vector P_Three ^*Is Like IM converted image_OUTReference point a₀ ^*From position a₁ ^*
It moves to a position that is sequentially incremented by Δu.   On the other hand, when the coefficient v is Δv, the coefficient u is 0, Δ
u, 2Δu, 3Δu ...
Image IM_OUTUpper position P_Three ^*Is the same as equation (22) or (25)
Like As in the previous pixel position (a₁ ^*+ A_Three ^*Δ
v) Move to a position to sequentially increment by Δu
Good.   Similarly, when the coefficient v is 2Δv, the coefficient u is 0, Δ
u, 2Δu, 3Δu...
Vector p_Three ^*Is For the previous pixel like (a₁ ^*+ A_Three ^*2Δv) Δ
It moves to the position to increment by u in order.   Thus converted image IM_OUTThe small rectangular area ER2 of
Statue IM_INOf the small rectangular area ER1
It is inserted.   Therefore, according to the above method, the input image IM_INSurely change
Replacement image IM_OUTCan be converted to   Converted image IM_OUTFine formed in
The small rectangular area ER2 is a straight line connecting the four specified vertices.
Is formed to form a rectangle.
One side of each rectangular area ER2 is one side of the adjacent minute rectangular area ER2.
It will be used in common, so that the converted image IM
_OUTThere is a gap between the small rectangular areas converted above
The fear can be effectively avoided.   The above operation can be represented by the following general formula.   That is, the input image IM_INScan with reference points u = 0, v = 0
When scanning starts from, the change corresponding to the scanning start position
Replacement image IM_OUTIs located at And then responds to the selection of the scan line in the v direction
Converted image IM_OUTThe scanning start position selected in isIt is represented by Then, the pixels included in each scanning line are scanned.
Going forward, the corresponding converted image IM_OUTAt
The scanned pixel is Is represented by   The calculations of equations (34) to (36) are performed using the image shown in FIG.
Calculated by conversion destination information interpolation circuit 1 in conversion device PD
The calculation result is supplied to the conversion processing unit 2.   In FIG. 6, the image conversion device PD includes an input image IM._INso
Is written into the input image memory 3 and
Image IM_INEach pixel data included in each minute rectangular area ER1
Are sequentially read out and sent to the conversion processing unit 2. Each small rectangle
The image data of each pixel included in the region ER1 is converted by the conversion processing unit.
The converted address is added to the output image
It is written to memory 4.   The input image IM is input to the conversion processing unit 2._INEach small rectangular area ER1
The addresses for the four vertices of
Converted image IM_OUT(Fig. 2)
Position f of four vertices of small rectangular area ER2^*(X₀, Y₀), F
^*(X₁, Y₀), F^*(X₀, Y₁), F^*(X₁, Y₁Equivalent to)
Is assigned. And in each small rectangular area ER2
The included pixels are supplied from the conversion destination information interpolation circuit 1.
A small rectangular area E
Address representing the position interpolated with respect to the four vertex positions of R2
The input pixel IM_INEach small rectangular area ER
1 is attached to the small rectangular area ER2 on the
Converted image IM similar to seeing it on a two-dimensional plane_OUTTo
The corresponding converted image data can be output.   As shown in FIG. 7, the conversion destination information interpolation circuit 1
The interpolation described above, which is generalized by Equation (36)
It has a configuration for realizing calculations.   In FIG. 7, reference numeral 11 denotes a processor, which is an image memo for input.
The pixel data of each small rectangular area ER1 is read from the memory 3
, The position vector f representing the four vertices₀₀ ^*, F_Ten
^*, F₀₁ ^*, F₁₁ ^*To the conversion circuits 12, 13, 14, and 15.
These conversion circuits 12 to 15 are constituted by registers, and (34)
Equations (36)
Generate parameters.   That is, in the first conversion circuit 12, the expression on the right side of Expression (34) is used.
a₀ ^*To the position vector f using the relationship of equation (10).₀₀ ^*Based on
Get it.   In addition, the second conversion circuit 13 calculates the expression of the second term on the right side of Expression (35).
a_Two ^*Let Δv be the position vector f using the relationship of equation (12).₀₀ ^*
And f₀₁ ^*And the pixel pitch Δv in the v direction.
You.   The third conversion circuit 14 calculates the second term of the right-hand side of the equation (36).
Equation a_Three ^*Of vΔu, the expression a_Three ^*Use Δu for the relationship in equation (13)
And the position vector f₀₀ ^*, F₀₁ ^*, F_Ten ^*, F₁₁ ^*And u
Direction pixel pitch Δu.   Further, the fourth conversion circuit 15 calculates the second term on the right side of the equation (36).
a₁ ^*The expression for Δu is calculated using the relationship of equation (11) as a position vector.
f₀₀ ^*, F_Ten ^*And the pixel pitch Δu in the u direction.
You.   Output a of first conversion circuit 12₀ ^*Consists of multiplexers
Is directly supplied to the combining circuit 21 and the second conversion
Output a of circuit 13_Two ^*Δv is added to the combining circuit 21 through the adding circuit 22.
Given to. The synthesis circuit 21 executes the operation of the expression (35).
Circuit, the v clock of the v counter 23_CLK1In sync with
Perform arithmetic operation. Equation (35) is one before in the v direction
For the first term on the right-hand side representing the scan line_Two
^*It is a recurrence formula that adds Δv. Operate on this recurrence formula
Therefore, the synthesis circuit 21 outputs the operation result output to the next scan line.
The operation is fed back to the adder circuit 22 for the operation and the expression a_Two ^*Δ
and v.   Coefficient a given from conversion circuit 12 to synthesis circuit 21₀ ^*No
Specifies the reference point at the start of scanning of the small rectangular area ER1
It is used as an initial condition to perform.   Thus, the data obtained by the operation in the synthesis circuit 21 is obtained.
The data is used to specify each scan line in the minute rectangular area ER1.
It is supplied to the synthesizing circuit 24. The synthesis circuit 24 is minute
Pixels included in each scan line of the rectangular area ER1 are interpolated
In order to output as each pixel of the minute rectangular area ER2, (3
6) u-direction clock obtained from u counter 25
U_CLK1The arithmetic operation is performed by.   The u counter 25 scans one scan line of the minute rectangular area ER1.
Clock signal u_CLK1Read at the timing of
At the same time, scanning of the pixels for the one scanning line
Is completed, the v clock signal v_CLK0The v
To the counter 23 as a count pulse, and
Direction clock u_CLK1V direction clock synchronized with_CLK1Synthesize
The circuit 21 can be supplied.   Output a of the third conversion circuit 14_Three ^*Δu is supplied to the multiplication circuit 31.
Multiplied by the coefficient data v given from the v counter 23
The result of the multiplication is added to the fourth conversion circuit
15 outputs a₁ ^*Δu, and thus the output of the adder circuit 32
At the force end, an output that satisfies the expression of the second term on the right side of expression (36) is obtained.
You.   Here, equation (36) expresses the data for each pixel in the u direction.
It indicates that it can be obtained by the recurrence formula.
The conversion data of the previous pixel is output to the output terminal of the synthesis circuit 24.
Is fed back to the adder circuit 33 and the output of the adder circuit 32 is output.
And the result of the addition is supplied to the synthesis circuit 24.
The combining circuit 24 is thus obtained at the output of the adding circuit 33
The conversion data represented by equation (36) is supplied from the synthesis circuit 21
To the output image memory 4
It is transmitted as a dress signal.   Here, the data given from the synthesizing circuit 21
The initial position of the line is given by the adder 33.
The resulting data represents the location of each pixel on each scan line.
The synthesizing circuit 21 uses these data to scan each scanning line.
Output data that sequentially specifies the pixels contained in the
And In the case of this embodiment, the u counter 25
Pulse input u_CLK0Is counted, and the count content changes.
V clock_CLK1Is sent. Soon cow
The scan content in the u direction in the small rectangular area ER1
Each time the value matches the number of pixels contained in the
Clock signal v_CLK0To the v counter 23
This is counted up by "1". v cow
Each time the counter 23 performs the counting operation, the v clock v_CLK1To
At the same time, the count value v (this is the pitch Δ
(obtained by repeatedly adding v)
I do.   Thus, the output image memory 4 has the contents shown in FIGS.
As described above, the conversion image IM_OUTMake up the micro
While specifying the rectangular area ER2,
Address with contents that specify the interpolation position of the included pixel
Signal is applied, so that at the corresponding address location,
Input image IM_INPixel data included in the small rectangular area ER1
Data can be read.   The converted image thus written in the output image memory 4.
Data is read sequentially in the order that a raster screen can be obtained.
As we came out, we mentioned above in Figure 2.
As shown in the figure, the input image IM_INSmall rectangular area ER
1 is attached, and this is a transformed image I which is seen through the st plane.
M_OUTCan be obtained.   In this regard, according to the configuration of FIG.
Uses one multiplication circuit 31 and three addition circuits 22, 32, and 33
And can be constructed short enough for practical use
Easy image conversion device that can perform image conversion processing in time
Can be realized.   FIG. 8 shows another embodiment of the conversion destination information interpolation circuit 1.
Therefore, the parts corresponding to those in FIG.
In the configuration shown in FIG. 7, the same device is used without using the multiplication circuit 31.
It is intended to realize the function.   That is, in the case of FIG. 8, the expression of the second term on the right side of the expression (36)
Is realized by the following configuration.   First, instead of the third conversion circuit 14 in FIG.
A fifth conversion circuit 14X having a road configuration is used. This conversion circuit 14X
Is the expression a in the second term on the right side of the expression (36)_Three ^*Equivalent to ΔvΔu
The output data to be output to the multiplexer
To the synthesis circuit 42.   The output a of the fourth conversion circuit 15 is output to the synthesis circuit 42.₁ ^*Δu is
Output a₁ ^*Δu + a_Three ^*ΔvΔu
Send out.   The output of the synthesis circuit 42 is fed back to the addition circuit 41.
The combining circuit 42 is thus supplied from the v counter 23.
Clock signal v_CLK2According to a_Three ^*ΔvΔu
Lock signal v_CLK2Is added to the adder circuit 41 each time
I do.   Thus, the synthesizing circuit 42 has a function of one scanning line.
When the scanning is completed, the next scanning line is designated from the v counter 23.
V clock signal v_CLK2Is added repeatedly when
Perform the operation, thereby substantially the second term on the right side of the equation (36)
Expression a_Three ^*The calculation of ΔvΔu can be realized.   With the configuration shown in FIG. 8, a multiplication circuit is not required.
Therefore, the configuration can be further simplified.   In the above embodiment, the present invention is applied to the case where the input image IM_IN
Converts each pixel included in the small rectangular area ER1 into a converted image IM_OUTof
When obtaining conversion destination information when converting to a small rectangular area ER2
The embodiment applied to FIG. 9 has been described. In addition, FIG.
When adding depth information as other image information,
Similarly, the interpolation method according to the present invention can be applied.   In FIG. 9, reference numeral 45 denotes a depth information interpolation circuit.
In equation (9), the positions of the four vertices of the small rectangular area ER1
Placement vector f₀₀ ^*, F₀₁ ^*, F_Ten ^*, F₁₁ ^*Instead of each
Converted image IM by using depth information at vertices
_OUTThe depth information in the small rectangular area ER2
Can be. This depth information is sequentially stored in the depth information buffer.
Stored in the memory area corresponding to each pixel on the output screen of the memory 46
And the interpolation operation of the depth information interpolation circuit 45.
The obtained depth information is stored in the depth information buffer memory 46.
Should be compared with the line information and written to the output image memory 4
Pixels with depth information corresponding to the hidden surface are obtained as data.
When the output image memory 4 is
By supplying the write control information WCN to the
Make it work.   In this way, the input image IM_INSmall rectangular area ER1
IM converts the depth information at the four vertices_OUTMinute
After conversion to the rectangular area ER2,
The depth information is interpolated and the
Statue IM_OUTHidden surface treatment without gaps
An image conversion device that can be executed can be obtained.   FIG. 10 shows another example using the interpolation method according to the present invention.
Other cases when performing shading control as image information
An example is shown.   In the case of the embodiment shown in FIG. 10, the same reference numerals are used for the portions corresponding to FIG.
As shown with a symbol, a shading information interpolation circuit 47
The shading information obtained in the above is supplied to the conversion processing unit 2.
Be paid. The shading information interpolation circuit 47 is shown in FIG.
In the same manner as described above for the depth information interpolation circuit 45.
Thus, the position vector f of the above equation (9)₀₀ ^*, F₀₁ ^*, F_Ten
^*, F₁₁ ^*Instead of the four vertices of the small rectangular area
Performs interpolation using shading information in
You. Here, the shading process refers to the direction of the surface or the light source position.
Performs a smooth shadow on the converted image based on the location
Say processing.   Thus, the shading interpolation obtained in the interpolation operation.
Information, converted image IM_OUTEach pixel included in the small rectangular area of
For the converted image information obtained for
By adding shadowing information,
Image information can be obtained.   As a circuit for realizing such a shading process,
As shown in FIG. 11, the image information VID is multiplied by a multiplication circuit 50.
Is given shading information SHD to the
To send out as imaged image information VIDX
I do.   With the configuration as shown in FIGS. 10 and 11, the converted image IM
_OUTImages that can be shadowed without creating gaps
An image conversion device can be easily realized. Effect of H invention   According to the present invention as described above, the input image IM_INFormed on
Convert the four vertices of the small rectangular area ER1
Image IM_OUTConverted to the converted image IM_OUTThe top four points
Input image IM_INFine above
Interpolated to correspond to each pixel included in the small rectangular area ER1
By calculating, the input image IM_INThe adjacent small rectangle
Converted image IM without gaps between shape regions_OUTTo
An image conversion device capable of performing image conversion can be easily realized.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a schematic diagram for explaining the basic principle of an image conversion apparatus according to the present invention, and FIGS. 2 and 3 are schematic lines showing an input image IM _IN and a converted image IM _OUT thereof. FIG. 4, FIG. 4 and FIG. 5 show small rectangular areas ER1 and E
FIG. 6 is a block diagram showing an embodiment of an image conversion apparatus according to the present invention, FIG. 7 is a block diagram showing a detailed configuration of a conversion destination information interpolation circuit, and FIG. FIG. 10 is a block diagram showing another embodiment of the conversion destination information interpolation circuit 1, FIG. 9 is a block diagram showing another embodiment of an image conversion apparatus capable of performing depth information interpolation processing, and FIGS. FIG. 13 is a block diagram showing another embodiment of an image conversion apparatus capable of performing shading information processing. 1... Conversion destination image information interpolation circuit, 2.
…… Input image memory, 4 …… Output image memory, 45…
… Depth information interpolation circuit, 46 …… depth information buffer memory,
47… Shading information interpolation circuit, IM _IN …… Input image, IM _OUT … Converted image, ER1, ER2 …… Small rectangular area.

────────────────────────────────────────────────── ─── Continued on the front page (56) References Andrew S.M. glassner, "COMPUTER GRAPHICS USER'S GUIDE" FI RST EDITION (1984), Howard W. Sams (US) p101-110 (58) Field surveyed (Int. Cl. ⁶ , DB name) G06T 15/00-15/20

Claims (1)

(57) [Claims] An input image memory for reading and storing input image data; and dividing the input image data stored in the input image memory into a first minute rectangular area, and dividing four first vertices of the first minute rectangular area. A conversion processing unit that converts each position based on the conversion function; and a memory area corresponding to the positions of the four vertices converted by the conversion processing unit.
An output image memory for forming image data corresponding to the four vertices of the second minute rectangular area by reading and storing the image data of the two vertices; Position information indicating the relative positional relationship between the position of each pixel in the second minute rectangular area and each pixel in the first minute rectangular area based on a count value that increases by a value corresponding to the pixel pitch. Conversion destination interpolation means for generating data by interpolation operation, wherein the conversion processing unit converts pixel data of each pixel in the first minute rectangular area into the second An image conversion apparatus for writing data in a memory area of the output image memory corresponding to a corresponding pixel position in a small rectangular area of the above.