JPH03219372A - Method and device for coordinate transformation of image and one-dimensional frame buffer - Google Patents

Abstract

PURPOSE:To convert an address supplied in the longitudinal direction of coordinate of a two-dimensional coordinate system to a one-dimensional real address by setting the sum of functions of two components of two-dimensional coordinates as the coordinate of the address on a one-dimensional coordinate system. CONSTITUTION:A coordinate transformation device 1 is equipped with a two-dimensional address supplier 2, an (x) address offset reader 3, a (y) address offset reader 4, an offset table 5, an adder 6, a memory read/write part 7, and a one-dimensional memory medium 8. The address offset table 5 is equipped with an (x) direction address offset table 11, a (y) direction address offset table 12. When two-dimensional addresses (x) and (y) are supplied, function values f(x), g(y) set at the address offset table 5 are read out and the functions f(x) and g(y) are added, and the address (l)=f(x)+g(y) on the one-dimensional coordinate system is outputted, then, two-dimensional/one-dimensional address conversion is completed. In such a way, coordinate transformation between one- dimensional and two-dimensional coordinates can be performed at high speed.

Description

[Detailed Description of the Invention] [Industrial Application Field] This invention relates to coordinate transformation technology used in digital image processing, etc.
This invention relates to technology for accessing one-dimensional addresses on a dimensional coordinate system.

[Prior Art] Displaying a digital image on a display device, and at the time of displaying the digital image, processing the stored original image and displaying an enlarged image, a reduced image, a rotated image, and other processed images. is also known.

In such image processing, coordinate conversion of addresses is necessary.

Among these, conversion between one-dimensional coordinates and two-dimensional coordinates is often seen in image processing devices, and conventionally, conversion between one-dimensional coordinates and two-dimensional coordinates is often performed in image processing devices. A hardware technique that uses memory for this purpose or a software technique that uses a CPU have been adopted.

[Problem to be Solved by the Invention] However, with the above hardware-based conversion technology, when the image size is constant, the difference between the two-dimensional coordinates of the pixels of the image and the one-dimensional coordinates of the memory address where the actual data exists is It has a very simple relationship and has the advantage of being able to perform two-dimensional and one-dimensional coordinate transformation very easily. In this case, the capacity of the image memory becomes enormous, the relationship between the image coordinates and the actual one-dimensional memory address becomes complicated, and it becomes difficult to improve image quality, image resolution, and use memory effectively. Have difficulty.

Furthermore, the software conversion technique has the problem of increasing the amount of programs for address calculation and requiring additional processing time.

For this reason, it has been proposed to use a conversion circuit that converts an address given by coordinate values in a two-dimensional coordinate system into a one-dimensional real address (see Patent Application Publication No. 6645 of 1988). ), the specific details of the conversion circuit referred to here have not been developed, and in the end, this technology has not been realized.

This invention was made in view of the above circumstances, and has developed a technology that can convert an address given by coordinate values in a two-dimensional coordinate system into a one-dimensional real address. An object of the present invention is to provide an image conversion method and an image conversion apparatus that can shorten processing time, and a one-dimensional frame buffer using these techniques.

[Means for Solving the Problem 1] Corresponding to this object, the image coordinate transformation method of the present invention provides two-dimensional coordinate transformation method for determining an address on a two-dimensional coordinate system.
The (presence of) component function f(x).

It is characterized in that the sum Ω=f (x) + g (y) of o (y) is taken as the coordinate of an address on a one-dimensional coordinate system.

In addition, the image coordinate conversion method of the present invention converts data whose address is arbitrary coordinates (x, V) on a two-dimensional coordinate system into a
and a first-order existence x・
V Let bt be the data whose address is an arbitrary coordinate β on the one-dimensional coordinate system on the source medium, let f(x) be an arbitrary function of X, and let the value of f(x) corresponding to the 8 values of X be Let XT be the X-direction address offset table to store, let 9(y) be an arbitrary function of y, let YT be the y-direction address offset table to store the values of a (y) corresponding to the 8 values of y, and in advance, The value of f(x) corresponding to the 8 values of X is set in each element XT of XT, and the value o corresponding to the 8 value of y is set in advance.
By setting the value of (y) in each element YTy of YT, taking out XTx corresponding to X, taking out YT corresponding to y and adding XT and YTy, β=f (x) + g (y) Ω that satisfies the relationship
The feature is that a can be made to correspond to x, ■ by outputting .

Further, the uniformity conversion device of the present invention includes a two-dimensional address supply device that supplies two-dimensional addresses of the existence of , an x-direction address offset table XT that stores in advance XTx for 8 values of X, and 8-values of y. A y-direction address offset table YT that stores YTy for a given value in advance, an X-direction address offset reader XTR that reads XTx for a given X value, and a y-direction address offset table that reads YT for a given y value. Address offset reader YT
R, XTR, XTx and YT supplied from VTR
and an arithmetic unit that sets in advance the values of f(x) and a(y) corresponding to the 8 values of the presence of and in the address offset table for each direction of the presence of and. It is said that

Furthermore, the one-dimensional frame buffer of the present invention has elements a x in the two-dimensional data matrix a [x, Yl of the image.
, y are stored in a one-dimensional data matrix on a one-dimensional medium, and if α is an arbitrary constant, bCR+(y-1)X+(x-1 )=a
a two-dimensional address supply device configured such that a relationship between x and y is established, and supplying a two-dimensional address of the existence of the
An X-direction address offset table XT that stores in advance XT for 8 values of X, and Y for 8 values of y.
A y-direction address offset table YT that stores Ty in advance, an X-direction address offset reader XTR that reads XTx for a given X value, and a y-direction address offset that reads YTy for a given y value. XT supplied from the reader YTR and XTR, YTR
and an adder ADD that adds YTy to f(x)=x−i and Q(y)=(
y-1) An arithmetic device that sets the value of X+α.

a y-direction address offset reader YTR for reading YTy, and XT and Y supplied from XTR and VTR.
The presence of an adder ADD that adds Ty, and the presence of and in the address offset table in each direction in advance, f(x)-x-1 and Q corresponding to the 8 values of the presence of (y)-(y-1)X
The present invention is characterized by comprising an arithmetic device that sets the value of +α.

[Action 1 When the existence of a two-dimensional address is supplied, the function values f(x) and g(
y) is read out, and then z(x).

Q (y) is added and the address β on the one-dimensional coordinate system of f (x) + g (y) - Ω is output, thereby completing the two-dimensional to one-dimensional address conversion.

[Example] One embodiment of these inventions will be described below.

In FIG. 1, 1 is a coordinate conversion device. Coordinate conversion device! 1 is a two-dimensional address supply device 2, an X-direction address offset reader 3, a y-direction address offset readout 14,
It includes an offset table 5, an adder 6, a memory write/read section 7, and a one-dimensional memory medium 8.

The address offset table 5 includes an X-direction address offset table 11 and a yh-direction address offset table 12.

X direction, y direction address offset table 11.12
Before driving this coordinate transformation device 1, all coordinate components x-0, 1, 2, ..., y-0, 1, 2, ...
For ・XTo=f (0), XTx-f(1), XT2-f (2), ・XTx=f (n), YTo
-Q (0), YTy = Q (1), YT2-chi (2>, -YTy-q(m) (where n and m are the maximum values of each presence) to another device f
(microcomputer 13, etc.). The X-direction address offset table 11 and the Y-direction address offset table 12 are constructed of RAM, ROM, or the like.

The existence of an address corresponding to the existence of two-dimensional coordinates is supplied from the two-dimensional address supplier 2, and
TxYTy is the X direction address offset reader 3 and y
The direction address is read out by the offset reader 4, and
and y-direction data are carried to the adder 6 through offset paths 24 and 25, added, and converted into a one-dimensional address Ω.
The data is carried through the dimensional coordinate path 26 to a memory readout section which is an input/output section of the real memory, and actual data input/output is performed. As a result, mutual coordinate transformation between the one-dimensional and two-dimensional coordinates satisfying the transformation formula f(x)+a(y)·Q is performed.

Figure 2 is shown in Figure 1. This conceptually shows the relationship between one-dimensional and two-dimensional images of the invention. Assuming that the existence of the two-dimensional coordinates in FIG. 1 can take any integer value, the one-dimensional image 15 on the one-dimensional memory medium 8 containing the pixels 14 in FIG.
This means that the image has been converted into a virtual two-dimensional image (116), and a device using this invention can handle a two-dimensional image that exists on the plane of existence without being aware of the existence and state of one-dimensional real memory. Become.

FIG. 3 shows an example of an apparatus for transforming a two-dimensional image and creating a new synthesized two-dimensional image using two coordinate transformation apparatuses 1 of the present invention. 17 is an image address generator, in which X changes by 1 from 0 to the maximum value, and then y
By repeating the operations in which x increases by 1 and x changes from 0 to the maximum value, the coordinates of the entire image are generated. Accordingly, the image defined by the coordinate transformation device 11a (for image reading) is read, and the coordinate transformation device f1b (for image writing) is read.
is written on the image defined by . Each coordinate conversion device 1
Various modifications of the image are performed by setting f(x) and g(y) in a and lb in the address offset table 5, and a new image is generated. However, in the configuration shown in FIG.
If b can be shared, it may be shared.

Figure 4 shows an example of an address offset data set that sets f(x) and Q(y), so depending on the settings, the one-dimensional coordinate data of ■ is transformed into two-dimensional coordinate data such as ■ to ■. It can be performed.

Furthermore, if the present invention is applied to image input/output devices such as scanners,
It can be handled in the same way as converting a one-dimensional image on a memory, and even pixels that are originally serially input/output can be handled as a two-dimensional image on this device.

FIG. 5 shows an example in which the present invention is applied to a scanner.
In this example, it is assumed that the input port of the scanner is allocated on the virtual memory space (generally referred to as memory MacBook i10 system) and its address is 80.

Furthermore, by combining this coordinate conversion device with a plurality of image calculation chips, calculations between a plurality of two-dimensional images III can be performed at high speed.

Further, in the present invention, this is an example of effective use of the memory when a plurality of one-dimensional pixels exist on the memory.

As shown in FIG. 6, by using the apparatus of the present invention, continuous two-dimensional images as shown in ■ and ■ can be converted row by row (or column by column) as shown in ■ by enabling two-dimensional to one-dimensional address conversion.
The two-dimensional image can be placed at any position on the one-dimensional address, so one-dimensional meshing can be used effectively by considering the arrangement.

Next, the coordinate transformation WA system of the present invention can also be used as a 7-frame buffer 7 of a general image display device. A frame buffer usually refers to an image memory for one screen, but it can be configured by a one-dimensional address memory and the image conversion device of the present invention. That is, the two-dimensional data matrix a[x
, Y], (1) The actual data of elements a x , y is one-dimensional medium b[
(2) The fS element of a is an arbitrary element of b. In order, (3
)b=a,b α 11 α+1=a21・b=a...b α+2 3,1 0t+X-1""
All elements of the first column correspond as ax, 1, (4) Next, b, i+X=”12°b = a ・・・b −
aα÷X÷1 2,2 α→2X-I x
, 2 (all 11゛ of the second second column correspond, and (5) generally the relationship of the existence of ba + (y-1) X + (x-1) - a holds up to (6) the y-th column. do.

In this case, in the coordinate transformation device according to the present invention, (a)
Element of the y-th column and first row of matrix a and (b) element b of matrix S
(C) Q (y
) −(y−1) *X+α, (d) yth of matrix a
ba + (y-1)X + (x-1
) indicates (e) α÷(y 1)X (f) relative element number from b − corresponding to the first row of the y-th column of matrix a. 1, (h) set (i) XTxYT in the address offset table 5 in the device shown in FIG. If supplied, the value of Ω indicating (k) the element a x of the two-dimensional data matrix and the element s of the one-dimensional medium in which the actual data of y exists (1) is output.

So, set this x, y to 7 frames 7 (or image)
If the coordinate 1, Ω, is the memory address of a 7-frame buffer, the coordinate conversion device of the present invention can be configured as a 7-frame buffer of a general image display device.

Note that the embodiment described above is an example in which an address offset table in which f (x) and g (y) are set is used to derive one-dimensional coordinates, but f ( x).

g(y) may be derived by a calculation device.

[Effect of the invention 1 This invention provides two-dimensional W! Since the A-image information is stored, it is possible to have flexibility in the size and position of images to be handled.・Moreover, since the conversion of 1D and 2D addresses is realized using a hardware address offset table, the conversion of 1D and 2D addresses is fast.
Can perform interdimensional coordinate transformations. This results in particularly high-speed image recording processing. Furthermore, image transformation (rotation in 90° units, inversion, enlargement/reduction, etc.) can be performed extremely easily by selecting a coordinate transformation formula. Furthermore, the actual location of the image on the memory can be any location for each column or row of image information whose address is specified by two-dimensional coordinates, so when there are multiple images to be stored, it is difficult to use the one-dimensional memory. Effective use can be achieved.

[Brief explanation of drawings]

Fig. 1 is an explanatory diagram of the configuration of the coordinate conversion device, Fig. 2 is an explanatory diagram showing the virtual two-dimensional image and the pixel positions of the one-dimensional memory, and Fig. 3
Figure 4 is an explanatory diagram of the configuration of an image conversion device, Figure 4 is an explanatory diagram showing various images corresponding to various coordinate transformation formulas, and Figure 5 is a virtual memory when the coordinate transformation device is applied to the input section of a scanner. An explanatory diagram showing an image on an address, Fig. 6 is an explanatory diagram showing a one-dimensional memory storing a plurality of image information, and Fig. 7 is an explanatory diagram showing an address of image information on one screen and an image information address on the one-dimensional memory. FIG. DESCRIPTION OF SYMBOLS 1... Coordinate conversion device, 2... Two-dimensional address supply device, 3... X-direction address offset reader, 4...
- Y-direction address offset reader, 5... Offset table, 6... Adder, 7... Memory readout unit, 8... One-dimensional memory medium, 11... X-direction address offset table, 12・
...Y-direction address offset table, 13...Microcomputer, 14...Pixel, 15...1
Dimensional image, 16... Two-dimensional image, 17... Image address generator

Claims (5)

[Claims] (1) Sum of functions f(x) and g(y) of two components (x, y) of two-dimensional coordinates that determine the address on the two-dimensional coordinate system l =
An image coordinate conversion method characterized in that f(x)+g(y) is the coordinate of an address on a one-dimensional coordinate system. (2) Data whose addresses are arbitrary coordinates (x, y) on a two-dimensional coordinate system are a_x_,_y, and a_x_,_y
Let b_l be the data whose address is an arbitrary coordinate l on the one-dimensional coordinate system on the one-dimensional medium where exists, and let f(x) be x
Let XT be an x-direction address offset table that stores the value of f(x) corresponding to each value of x, let g(y) be an arbitrary function of y, and let g(y) be an arbitrary function of y, and let g(y) be an arbitrary function of y, and Let YT be the y-direction address offset table that stores the value of g(y), set the value of f(x) corresponding to each value of x to each element XT_x of XT in advance, and set the value of f(x) corresponding to each value of x in advance to each value of y. By setting the corresponding value of g(y) in each element YT_y of YT, taking out XT_x corresponding to x, taking out YT_y corresponding to y and adding XT_x and YT_y, l=f(x)+g( An image coordinate transformation method characterized by making a_x_,_y correspond to b_l by outputting l that satisfies the relationship y). (3) The functions f(x) and g(y) are determined based on the correspondence between the positions of the data string on the target two-dimensional coordinate system and the data string on the one-dimensional coordinate system of the recording medium. A method for converting coordinates of an image according to claim (2) (4) A two-dimensional address supplier that supplies two-dimensional addresses for x and y, an x-direction address offset table XT that stores XT_x for each value of x in advance, and YT_y for each value of y that stores in advance. A y-direction address offset table YT to be stored and an x-direction address offset reader XT that reads XT_x for a given x value.
R, a y-direction address offset reader YTR for reading YT_y for a given y value, XTR, YTR
Adder A that adds XT_x and YT_y supplied by
f(x) and g( corresponding to each value of x and y are stored in advance in the address offset table in the DD and x and y directions, respectively).
an arithmetic device for setting the value of y). (5) The actual data of elements a_x_,_y in the two-dimensional data matrix a[X, Y] of the image are stored in the one-dimensional data matrix b on the one-dimensional medium, and if α is an arbitrary constant, then the two b_α_+_(_y_-_1_)_X_+_(_x_-
_1_)=a_x_,_y relationship is established, and a two-dimensional address supplier supplies x and y two-dimensional addresses, and an x-direction address offset that stores XT_x for each value of x in advance. A table XT, a y-direction address offset table YT that stores YT_y for each value of y in advance, and an x-direction address offset reader X that reads XT_x for a given x value.
TR and read YT_y for the given value of y
Direction address offset reader YTR, XTR, YT
An adder ADD that adds XT_x and YT_y supplied from R, and an address offset table in each of the x and y directions are preliminarily filled with f(x)=x- corresponding to each value of x and y.
1 and an arithmetic device for setting the value of g(y)=(y-1)X+α.