JP2868324B2 - Workstation video frame buffer - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a frame buffer of a workstation such as an engineering workstation (interactive small computer for science and technology), and more particularly, to a method for placing a moving image such as a video image anywhere on a display screen. The present invention relates to a frame buffer for defining a window having an arbitrary shape / size and realizing a workstation capable of displaying the moving image in the window.

[0002]

2. Description of the Related Art Workstations, such as engineering workstations, have high / medium resolution graphic displays to achieve their high interactivity. In general, the graphic display displays a screen for a user to work, for example, a programming editor screen, or a graphic display screen of a calculation result of CAD (design support by a computer).

FIG. 4 shows a frame buffer and a color display of such a conventional workstation. Referring to FIG. 4, a frame buffer 4 stores image data D for one screen from a CPU (central processing unit) (not shown) of a workstation at a write address AD in synchronization with a write clock CL.
The image data written in the frame memory 4 is read out in synchronization with an image reading clock having a speed suitable for the color display 8 and an image reading address synchronized with the image reading clock, and is read out as R (red), G (green), B (blue). And a D / A (digital / analog) converter 7 for converting the image signal into a video signal. Resolution of color display 8 is 1280 horizontal and vertical
If the number of colors is 1024 and the number of colors is 8 bits for each of RGB, the frame memory 4 is accessed with a pixel address of 1280 × 1024, and each pixel stores 24-bit data. The images displayed on the screen of the color display 8 include, for example, an editor screen 102 for inputting characters and a graphic display screen 101 for CAD results. In addition, computer graphics images are calculated at high speed by a CPU or dedicated processor to generate animations, and the frame buffer 100
To view the moving image 103 in real time.

[0004] Some of these images are usually drawn in a square "window", such as an X window, to improve the user interface. This is the same for the animation (moving image) display.

[0005] In response to the demands of various applications in the future, a window having an arbitrary shape other than a square window will be required. Conventionally, displaying a non-moving image (a normal character input screen or a still image) in this arbitrary-shaped window is performed by software (program).
In practice, it does not matter much if the display speed is reduced.

[0006]

However, displaying a smooth (fast) moving image of about 10 to 30 frames / second, such as a video image, on a window of an arbitrary shape has never been realized.

SUMMARY OF THE INVENTION The present invention realizes a moving image frame which can be displayed on an arbitrary-shaped window, which cannot be realized by the above-mentioned conventional workstation, and can greatly expand the application range of the workstation. It is intended to provide a buffer.

[0008]

In order to solve the above-mentioned problems, a frame buffer according to the present invention comprises a first dual-port memory for buffering and storing non-moving image data;
A second dual-port memory having a smaller number of pixels than the dual-port memory and storing the moving image data in a buffer; In order to define a window for starting, a data "1" or "0" is written to a pixel address corresponding to a pixel inside the window, and a data "0" or "1" is written outside the window. , The read data from the first dual port memory to the A input, the read data from the second dual port memory to the B input, and the read data from the third dual port memory to the B input. Select input, and when the read data from the third dual port memory is “0” or “1”, the input data of the A input And a data selector for selecting and outputting the input data of the B input when the read data from the third dual port memory is "1" or "0", and a pixel for one screen to be displayed. A dual-port type memory for synthesizing data, a frame memory having an output of the data selector as an input,
Moving image position coordinate data indicating at which coordinate position in the frame memory the moving image data written in the second and third dual-port memories is written as CP
U, a moving image position coordinate register, an image transfer address for transferring image data from the first, second, and third dual port memories to the frame memory and an output of the moving image position coordinate register are input; An address converter that subtracts the output data of the moving image position coordinate register from the image transfer address and outputs the result to the second and third dual port memories as a read address; A D / A converter for converting the output pixel data from digital to analog and outputting it as an RGB (red, green, blue) image signal to be supplied to a color display, and using the first, second, and third image transfer clocks as the image transfer clock. The clock is supplied to the read clock input of the dual port memory and synchronized with the image transfer clock. A transfer address is supplied to a read address input of the first dual port memory and an input of the address converter, and the image transfer clock and the image transfer address are simultaneously input to a write clock input and a write address input of the frame memory. And the pixel data of the third dual-port memory corresponding to each of the pixels of the frame memory,
In the case of “0” or “1”, the corresponding pixel data of the first dual port memory is written, and the third
When the pixel data of the dual port memory is “1” or “0”, by writing the corresponding pixel data of the second dual port memory, the non-moving image data and the moving image data are written into the frame memory. To obtain image data obtained by combining

[0009]

With the above arrangement, the frame buffer of the present invention comprises a non-moving image dual port memory for buffering ordinary (non-moving image) image data, a moving image dual port memory for buffering moving image data such as a video image, and the like. A window definition dual port memory for defining a window (window) area for displaying a moving image and a dual port type frame memory for synthesizing a non-moving image and a moving image are defined in the window defining dual port memory. The moving image data of the moving image dual port memory area corresponding to the moving image window area and the image data of the non-moving image dual port memory area corresponding to the area other than the moving image window area are synthesized. Write to the frame memory, read it out at a predetermined cycle, and perform D / A conversion Te By supplying the display is converted into RGB video signals, it is possible to obtain a smooth (fast) moving the window of an arbitrary shape.

[0010]

An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a frame buffer according to one embodiment of the present invention. In FIG. 1, reference numeral 100 denotes a frame buffer, 8 denotes a color display having a resolution of 1280 × 1024, and has the same functions as those of the conventional example.

1 is equivalent to one screen (1280 pixels in width × 1024 pixels in height,
A first dual-port memory for buffering normal (non-moving image) image data of (pixel = 8 bits),
Write control is performed by a CPU (Central Processing Unit) (not shown) of the workstation, and first write data D1 is written to a first write address AD1 in synchronization with a first write clock CL1.

Reference numeral 2 denotes a second dual port memory capable of buffering pixel data for one screen of a moving image (640 × 480 pixels, 1 pixel = 24 bits), such as a magnetic disk or an optical disk. (NTSC standard television image, video image, etc.) Write control is performed directly (not through the CPU or main memory) from a storage device (not shown), and second write data (moving image data) is stored in a second write address AD2. ) D2 is the second write clock CL
2 is written synchronously.

Reference numeral 3 denotes moving image window pattern data (1 bit data corresponding to each pixel) which defines a pattern of a moving image window (window for displaying a moving image) for one screen (640 × 480). Is a third dual-port memory for writing data, has the same pixel configuration as the second dual-port memory, is write-controlled by the CPU (not shown), and has a third write address A
The third write data D3 is written to D3 in synchronization with the third write clock CL3.

The first, second and third dual port memories 1 to 3 have different data depths corresponding to one pixel (first = 8 bits, second = 24 bits, third = 1).
(Bit) is the same in the form of access (although it is the same in this embodiment, it is not limited to this). The right side of each is a read port, which can be read and written independently and asynchronously. Note that the left write port may be readable, and is not limited to this.

A color look 6 receives 8-bit pixel data read from the first dual-port memory and outputs corresponding 256 24-bit pixel data (8 bits for each of RGB). It is an up table.

Reference numeral 5 denotes 24-bit pixel data (A input) read from the color lookup table 6 and the second
And the 24-bit pixel data (B input) read from the dual port memory 2 is input, and the moving picture window pattern data (S input) read from the third dual port memory 3 is set to “0”. Is a data selector for selecting output data from the color look-up table 6 at the time of, and selecting output data from the second dual-port memory 2 at the time of "1" and outputting it to the Y output.

Reference numeral 4 denotes a frame memory which can be independently and asynchronously used for writing (left side) and reading (right side) at different ports, similarly to the first to third dual port memories. Pixel data of one pixel (24 bits) can be buffer-stored.

Reference numeral 7 denotes a 24-bit pixel data which is read in synchronization with an image reading clock and an image reading address synchronized with the image reading clock from the frame memory 4 and is digitally / analog-converted by 8 bits to R (red). , G
A D / A converter that outputs (green) and B (blue) video signals and supplies them to the color display 8.

Reference numeral 10 denotes moving image position coordinate data Xs, Y indicating the position in the frame memory 4 at which the moving image buffered in the second dual port memory 2 should be displayed.
s is a moving image position coordinate register written and stored by the CPU. The moving image position coordinate data is composed of 20 bits of a 10-bit Y coordinate Ys at a high order and an X coordinate Xs of 10 bits at a low order.

Reference numeral 9 denotes a subtraction of the 20-bit output data of the moving picture position coordinate register 10 from the 21-bit image transfer address input to the A input, and outputs 19-bit output data to the Y output. Of the dual port memories 2 and 3 as read addresses.

The frame memory 4, the D / A converter 7, and the color display 8 are exactly the same as those in the conventional example shown in FIG. 4, and can be easily realized by the conventional technology. next,
The operation of the embodiment of FIG. 1 will be described. 1 to 1 in FIG.
3 dual port memories 1 to 3 and frame memory 4
In order to make it easier to understand the meaning of the data, the data content of the memory is graphically displayed on the surface of each `` plate '' corresponding to the display screen of 1280 horizontal, 1024 vertical or 640 horizontal, 480 vertical. It is.

Now, the resolution is 1280 × 1024 and the number of colors is 2 to the 24th power (about
Consider a case where an image as shown in FIG. 1 is displayed on the color display 8 (16.7 million colors). On the screen, for example, several windows (display windows) such as an X window are displayed. 101 is DTP (Desktop Publishing) software, 102 is character
The base editor screen is displayed. These windows 101 and 102 are rectangular windows as in the prior art. In 103a, a moving image is displayed in a “heart” -shaped window as if it were a television image, as shown in FIG. Such a screen can be realized easily and at a sufficient speed by the following operations.

First, when information on the shape and position of a window (moving image window) for displaying a moving image is given by a user, the CPU (not shown) of the workstation transmits moving image window pattern data corresponding to the shape. From the left write port to the third dual-port memory 3. As illustrated in FIG. 1, the moving image window pattern data is defined by writing “1” inside the moving image window and writing “0” otherwise. As for the moving image position, assuming that the rectangular area of 640 × 480 shaded in FIG. It can be specified by setting it in the coordinate register 10.

Next, the CPU of the workstation operates as a moving image storage device (such as a magnetic disk device or an optical disk device).
(Not shown), the moving image data is directly transferred to the second dual port memory, and is controlled to be written from the left write port. The moving image data is written in a size and a position (address) covering the moving image window as shown in FIG. Any pixel data may be written at an address other than the size and position (shown in white in the second dual-port memory 2). In FIG. 1, the transfer and writing speed can realize a moving image. About 10
Set to ~ 30 frames / sec.

In addition, ordinary screens / windows (such as windows 101 and 102 in FIG. 1) other than a moving image are stored in the first dual port memory 1 by the CPU in the same manner as in the prior art (as in the frame memory 4 in FIG. 2). ) From the left write port. As a result, the first to third dual-port memories 1 to 3 buffer the pixel data expressed on the surface thereof in an "image-like" manner.

Next, an image transfer address (21
Bits: the upper 10 bits of Y data Yt and the lower 11 bits of X data Xt in (1) of FIG. 3) and the image transfer clock are the read address input and read clock of the read port on the right side of the first dual port memory 1. At the same time, the image transfer address is supplied to the A input of the address converter 9 at the timing of the image transfer clock. FIG.
As shown in (1), the image transfer address starts from the reading of the upper left pixel of the first dual port memory 1, and is incremented every image transfer clock. When the address reaches the upper right pixel, Yt = 0, Xt = At 1279, Yt = 1 and Xt = 0 at the next clock and return to the lower left by one line. This is repeated below. One frame is read out by the non-interlaced scanning method which ends with reading out the lower right pixel.

The address converter 9 converts a 21-bit image transfer address [Yt, Xt] shown in FIG. 3A input to the A input into a 20-bit image transfer address shown in FIG. 3 is subtracted from the moving image position coordinate data [Ys, Xs] of FIG.
19-bit output data [Yc, Xc] as shown in (3) is output at the timing of the image transfer clock. At this time, Yc = Yt−Ys and Xc = Xt−Xs, and the possible ranges are 0 ≦ Yc ≦ 479, 0 ≦ Xc ≦ 639, and “0” when the result of the operation exceeds this range. It is designed to output. FIG. 2 shows these coordinate relationships. The coordinates [Xc, Yc] on the moving image plane correspond to the coordinates [Xt, Yt] on the frame memory 4.

The one-frame reading cycle is a cycle in which one frame of moving image data is written from the moving image storage device (this is a period equal to or shorter than the frame period of the moving image itself).
It is set to a period less than or equal to this.

The 8-bit pixel data read from the first dual port memory 1 is stored in a color look-up memory.
The data is input to Table 6, and the corresponding 24-bit pixel data is output immediately. The 24-bit pixel data enters the A input of the data selector 5, and at the same time, the moving image data read from the second dual port memory 2 enters the B input. Also, the third input to the select S input of the data selector 5
Of the moving image window pattern data supplied from the dual port memory 3 is written to the first dual port memory 1 for the pixel corresponding to the address where the moving image window pattern data is “0”. The image data written to the second dual port memory 2 is output to the pixel corresponding to the address where the moving picture window pattern data is “1” in synchronization with the image transfer clock to the Y output. Is done.

Since the image transfer address and the image transfer clock are supplied to the write port of the frame memory 4, the same one-to-one display as that of the screen of the color display 8, which is expressed as "image" on its surface, is provided. The composite image data is written. Thereafter, the operation is performed in exactly the same manner as in the conventional example of FIG. 4, and a screen in which a moving image is displayed in a “heart” type window on the color display 8 is realized.

In the present embodiment, the resolution of the image (the size of the frame memory 4) is described as 1280 × 1024. However, the present invention is not limited to this, and any resolution can be realized. In FIG. 1, the color look-up table 6 is provided between the first dual-port memory 1 and the data selector 5, but this is not directly related to the present invention. A configuration without a table can be similarly realized.

[0032]

As described above, according to the present invention, an arbitrary-shaped window can be defined, and a moving image read from a moving image storage device such as a magnetic disk device or an optical disk device can be displayed on the window in real time. Thus, an extremely valuable workstation can be obtained, for example, when building an application using a moving image.

[Brief description of the drawings]

FIG. 1 is a block diagram of a moving image-compatible frame buffer according to an embodiment of the present invention.

FIG. 2 is a diagram illustrating the positional relationship of a moving image plane (second dual-port memory) in a frame memory.

FIG. 3 shows image transfer addresses Xt and Yt, moving image position coordinate data Xs and Ys, and output data Xc and Y of an address converter.
It is a figure explaining the relation with c.

FIG. 4 is a block diagram of a frame buffer of a conventional workstation.

[Explanation of symbols]

 1 First Dual Port Memory 2 Second Dual Port Memory 3 Third Dual Port Memory 4 Frame Memory 5 Data Selector 6 Color Lookup Table 7 D / A Converter 8 Color Display 9 Address Converter 10 Video Position Coordinate register 100 Frame buffer 101, 102 Normal non-video display window 103a Video window for displaying video

──────────────────────────────────────────────────の Continued on the front page (51) Int.Cl. ⁶ Identification code FI G09G 5/36 520 G06F 15/72 A

Claims (1)

(57) [Claims] 1. A first dual-port memory, which is write-controlled by a CPU (Central Processing Unit) of a workstation and buffers one screen of non-moving data to be displayed, and a moving picture storage that stores moving picture data A second dual-port memory, directly controlled by the device, having a smaller number of pixels than the first dual-port memory, and buffering the moving image data;
Has the same number of pixels as the second dual-port memory, and a pixel address corresponding to a pixel inside the window is defined in order to define a window for displaying the moving image on the screen. A third dual port memory for writing data "1" or "0" and writing data "0" or "1" outside the window, and read data from the first dual port memory for A input, The second
, And the read data from the third dual-port memory is input to the select input, and the read data from the third dual-port memory is "0" or "1". ", The input data of the A input is selected and output,
When the read data from the third dual port memory is "1" or "0", a data selector for selecting and outputting the input data of the B input and a data selector for synthesizing image data for one screen to be displayed. A frame memory which is a dual-port type memory, to which the output of the data selector is input, and to which coordinate position of the frame memory the moving image data written in the second and third dual-port memories is to be written A moving image position coordinate register written by the CPU, an image transfer address for transferring image data from the first, second, and third dual port memories to the frame memory; Input the output of the moving image position coordinate register and output the moving image position coordinate register from the image transfer address. Subtracted over data, the results of the above second
An address converter that outputs the read address to the third dual-port memory as a read address;
A D / A converter that outputs a B (red, green, blue) image signal, and supplies an image transfer clock to a read clock input of the first, second, and third dual port memories;
The image transfer address synchronized with the image transfer clock is supplied to a read address input of the first dual port memory and an input of the address converter, and at the same time, the image transfer clock and the image transfer address are stored in the frame memory. When the pixel data of the third dual port memory is "0" or "1" corresponding to each pixel of the frame memory, respectively, the first clock is supplied to the write clock input and the write address input.
And writing the corresponding pixel data of the second dual-port memory when the pixel data of the third dual-port memory is "1" or "0". A moving image corresponding frame buffer of a workstation configured to obtain image data obtained by combining the non-moving image data and the moving image data in the frame memory.