JPS63291088A - Display circuit - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed description of the invention] (Industrial Application Field J The present invention relates to a display circuit.

[Summary of the invention]

In a display circuit for computer graphics, the present invention allocates display data in accordance with a predetermined rule. It has improved color display ability.

[Prior Art J] In computer graphics, it is generally said that 256 colors (=2') per pixel is sufficient for color display ability. In other words, ll! It is sufficient to allocate 8 bits to 11 elements.

However, recently there has been a demand for processing and displaying natural images, such as images captured with a video camera.
In this case, the ability to display 256 colors per pixel is insufficient.

However, in computer xifix,
2000ii! ! Since it has a display capacity of about I element x 2000 pixels, the color table inability per pixel can be increased (
As a result, the size of the display memory becomes too large (too large).For example, if 8 bits are allocated to each of red, green, and blue in one pixel, 24 bits per pixel will require 4 bits. Therefore, as a whole, 24 hits x 2000il] li element x 2000i!!
This results in a size of 11.4 megabytes (1 element = 960000000 bits), which means that processing the image will take a lot of time, or even if you use a hard disk drive, you will only be able to store a small number of images ( (Save) Yes, but in any case it would be impractical.

Therefore, a method is being considered in which the number of hits per pixel is set at 8, but the color source capability is expanded by using a color look amplifier table.

Figure 6 shows an example of this, and +11 shows the clock from the address corresponding to the horizontal scanning and vertical scanning of the color picture tube (not shown). In synchronization with CK, the image display data is taken out to the 1] version. In this case, for example, the size of data D^1 for one image # (91) is 8 as shown in Figure 7.
It is a human, but the 8 Hiso I- is Hiso) bv, bi
,... Red front moon is used) bs~b7...
. . . Green 72 Bo river bottle) bl, bO . . . Assigned for blue display. That is, data IIATA includes bits lb7. beJ,l
'b5~b2'', l'b1. boJ power tweet/1 response! I
It shows each density (color saturation) of 11 elements of red, green, and blue.

In addition, (21) to (23) are color look up and
The mapping circuit for the chiful, in this example a memory, in which memories (21) and (23) each have 4 addresses, memory (22) has 16 addresses, and all memories (21) to (23) ) has a capacity of 8 hints per l address, and data corresponding to the color shown in table 1' is stored in the valley address.

Furthermore, (31) to (33) are 8-bit latches, (4
1) to (43) are 8-bit D/A converters.

Then, pixel data 11 A from the memo January 1)
1' When A is extracted, the red bit b7.
b.

is supplied to the memory (21) as its address signal, 8-bit data row ATR is taken out from the corresponding address (address of the value indicated by pinto b?+bli), and this data DATR is latched by clock CK (31).
After being launched, it is supplied to the D/A converter (41) and converted into an analog red signal R.

Similarly, green bins bs to b2 and blue bins bl and bo of data DATA are memories (22).

(23) as their address signals, and each 8-bit data 11ATG is supplied from the corresponding address.
, DATB are retrieved and these data 1lA1'G
, 1181'B are supplied to D/A converters (42) and (43) through latches (23) and (33), and are converted into analog green signal G and blue signal B.

These signals R, G, and B are then supplied to the color picture tube, so that a color image corresponding to the data DATA in the memory ill is displayed.

In this case, data 1) AT chuyo 2' = 256i11 becomes the address signal of memories (21) to (23).
Rino (can take direct, memory (21) ~ (23
) output data DATR-DATB has a total of 24 bits and can take on 224==16777216 types of values, so 256 out of 16777216 colors can be displayed simultaneously.

In this way, by using the color look amplifier table, many hues and gradations (densities) can be expressed even if the capacity of the display memo (January 1) is small, and a colorful image can be displayed.

[Problem that the invention seeks to solve]

However, in the above circuit, only 256 colors out of 16777216 colors are selected and these 256 colors are displayed at the same time, which lacks a variety of colors for natural images, and it is difficult to achieve naturalness in color gloss. Sometimes it's missing. For example, when displaying a shiny fruit, only 2 bits are assigned to red and blue, and there are 4 gradations, so the quantization is noticeable and the color changes become fluffy.

Furthermore, considering the colors red, green, and blue, on average only 8 bits/3W2.7 bits are allocated, which results in a lack of gradation, a decrease in resolution, and a significant deterioration in image quality. .

This invention attempts to solve these problems.

[Means for solving problems]

The present invention provides a display circuit that retrieves display data for each pixel from a display memory and performs 1)/A conversion on data based on the retrieved display data to obtain three primary color signals of a color image. Among the three primary colors and luminance, the display data is data about a color or luminance that has the largest density gradient with an adjacent inner element, and is data having an identification code indicating the color or luminance,
The above display data taken out from the above display memory is
A display circuit that retains display data of the same color or brightness according to the identification code until it is next retrieved from the display memory, and converts it into data with a larger number of bits than the original data by ma-noving. It is.

[Effect]

Natural images are displayed without any problems with 11wm or resolution.

Embodiment] In FIG. 1, the display memory 11 has the same capacity as the one described above, and image display data DATA is input from addresses corresponding to the horizontal scanning and vertical scanning of the color picture tube in synchronization with the clock CK. A gun is fired.

In this case, one i! Data IIA for i element (9)
T^ is 8 bits as mentioned above, but this data DAT
^ is data compressed according to the flowchart shown in FIG. 2, for example, and has a format shown in FIG. 3, for example.

That is, in FIG. 2, Rs + G s lH1
rYi is the original data before compression, for example, data captured by a video camera, and is the red and green color of the first pixel (9) on one screen.

This is digital data that indicates blue color and brightness levels in 8 bits.

Then, in step (101), Ho = 128, Go = 128, Bo = 1 as initial values.
28Yo -0,3RQ +0.59GO+0.11B
O,i=1 is set, and then in step (102), the luminance data Yi of the first pixel is changed to the data R4.
= Bi to Y i -0,3Ri +0.59Gi +0.11H
i, and the data of the lfk-th pixel Ri”=Yi and the data of the previous pixel Ri-t~Y1-
The absolute value of the difference from 1 ΔR~ΔY ΔR-l Ri Rt-t l ΔG=lGi Gi-t 1 ΔB=lBt Bt-i l ΔY= l Yi Yi-x l is calculated, and then the Ste-knob ( In step 103), the maximum value among the values ΔR to ΔY is determined.

Then, the value ΔR is the maximum (ΔR>ΔG, ΔR>ΔB.

When ΔR>ΔY), in step (104), data Ri is converted from 8-bit data (original data) to 6-bit data (compressed data) Rc, and as shown in FIG. 00'' is added to the upper part of the data Rc as an identification code indicating that it is red data, and the data Rc as a whole has 8 bits.

Therefore, data Rc has the upper two bits bv, t)r
, is the identification code “00゛°” indicating that it is red data, and the remaining 6 bits) b5 to b0 are the first pixels (9
) has information on the density of red color.

Further, at this time, the 6-bit nodes b5 to bbo of the data Rc are obtained by allocating 6 bits to the valid section of the frequency distribution of the original data, as shown in FIG. 4, for example. Therefore, data Rc has a sufficient quantization level even if only the 6 bits b5 to bo have density information.

Further, in step (103), the value ΔG is the maximum (Δ
B〉ΔR3ΔG〉ΔB, ΔG〉ΔY), in step (105), the data Gi is set to step (1
The data G is processed in the same way as the data Ri in 04).
c. However, in the data Gc, as shown in FIG.
1", and the remaining six hits b5 to bo are density information.

Further, in step (103), the value ΔB is the maximum (ΔB
〉ΔR1ΔB〉ΔG, ΔB〉ΔY), in step (106), data Hi is in step (104).
) is processed in the same manner as the data Ri and becomes data Bc. However, in the data Hc, the upper two bits bv+b6 are used as identification code "10", and the remaining six bits b5 to bo are used as density information.

Further, in step (103), the value ΔY is the maximum (Δ
When Y〉ΔR1ΔY〉ΔG, ΔY〉ΔB), in step (107), data Yi is changed to step (10
Data Yc is processed in the same way as data R+ in 4).
It is said that However, in the case of data Yc, the upper two bits b7. b6 is the identification code "11", and the remaining 6 bits b5 to bO are density information.

Then, the data He, Gc, Bc or Yc formed in any one of steps (1(ll) to (10'/))
is stored as data DATA of the i-th pixel (9) at the i-th address of the storage means (for example, memo January 1) in step (10B).

Then, in step (109), the value l is incremented by "1" and the next pixel 191 is targeted for processing, and in step (110) it is determined whether the above processing has been performed for all pixels (9). If there is a pixel (9) that has been processed and has not been processed, step (102)
Subsequent processing is performed in the same manner.

Therefore, for each pixel (9), the memory ill q is
Information on the density of the color or luminance that has the largest level change with respect to the previous pixel (9) is stored as data DATA.

Furthermore, the memories (21) to (23) for color look amplifier tables have 256 addresses, and each address has a capacity of 9 bits. In the memory (21), the upper two bits Av and As of addresses A7 to Ao are the red identification code value "00".
” or when the brightness identification code value becomes “11”,
A predetermined value is set so that the most significant bit D8 of the data D8 to Do becomes "1" and data at the level of the red component of the desired color is output as the lower bits D7 to Do (red data DATII). Data is stored. In other words, in memory (21), OOH~3
1"H address and COH-FFH address (H indicates hexadecimal), De-"1", D7~Do=[red data lo^1
"R" is stored.

In addition, in the memory (22), addresses A7 to A
The upper two bits of o, Av and As, are the values of the green identification code.”
01" or the brightness identification code value "11", the most significant bit IJ8 of data D++ to Do becomes "I
”, and the data of the level of the green component of the color you want to display is at the lower pips 1-D7 to D.

Predetermined data is stored so as to be output as (green data RO^TG). That is, memory (22)
In, 40H to 7FH7 and 8 and COH-1'FH
At the address, De-“l”, Dv-Do, =r green data RO^TGJ is stored. Furthermore, memory (23)
In, the upper 2 bits Av of addresses A7 to Ao,
/'s becomes the value "10" of the blue identification coat or the value "II" of the brightness identification coat, the data D8 to Do
Dv~
Predetermined data is stored so as to be output as Do (blue data DATB). In other words, in memory (22), address 80H-BFH and GOI
At addresses I~FFH, Dll-1'', Dv~Do
= 1-Blue data row ATBJ is stored.

Note that the upper 2 bits b7. to data DA1'. b6 is”
11”, memory (21) to (23) COH
The -FFH address is addressed simultaneously for these memories (21) to (23), thus memory (2
Data DAI″R-11ATB are simultaneously output from 1) to (23), but at this time, data DATR-DAi
The value indicated by 'H is 1'G: D＾゛1
“B= 0.3: 0.59: 0.11...
・・・These data DAT) so that it becomes fi+
The ratio of l-DATB is set. However, data DA'
The absolute value of l'R~υAi'll is a value according to the lower six positions b5~bo of the data IIAT^.

In addition, the memory (21) for lunch (31) to (33)
) to (23), the most significant bits Da-1)8 are supplied as a latch failure signal.

According to such a configuration, data 11 from memo January 1)
When AT8 is taken out, this data υA1'A is supplied to memories (21) to (23) and stored as some data D.
This data is converted to ATR~0ATB, and this data D A 1'
li to 11A1'H are supplied to latches (31) to (33).

In this case, if the data υ revision from the memo January 1) is, for example, red data Rc, its upper two bits bv and be are 00'', so the most significant bit of the output of the memory (21) is Only Do becomes “1” and this bit D8
The data DAT)l is launched at launch (31) by 5
Therefore, the D/A converter (41) outputs a red signal l? is taken out.

In addition, at this time, the data row 8 TA is memory (22)
.

Since it is also supplied to (23), the memory (22).

Data O^1”G and DATB are also output from (23), but the upper bits bv and be of data DATA are “
OO'', the most significant hits Dm and De of the outputs of memories (22) and (23) will never become “1”, and therefore data DATG and DATB will be launched at launches (32) and (33). Without a doubt,
Data DATG, DA when latched before that
TII remains held.

This state in the latch (31) is maintained until the next red data Rc is taken out from the memo tube 1), that is, until the change in density of red becomes larger than the change in density of other colors.

Also, if it is green data Gc or blue data Bc, the most significant bit D8 of the output of memory (22) or (23) according to the upper 2 bits bv and be from data DATA6 (memo January 1) becomes 1'', so data lo^TG or IIATB is latched (32
) or (33), and the D/A converter (
The green signal G or the blue signal B is extracted from 42) or (43).

Furthermore, if the data DAT^ from the memo January 1) is luminance data Yc, its upper 2 bits b? , be “
11'', data IIAT)l-DATB is output from all memories (21) to (23), and all the most significant bits D8 to D8 become 1'', so latches (31) to (33) ) has data DAnl
~DATB are latched at the same time, but at this time, data DATR-DATB is set to the ratio as shown in equation (1), so at this time, the level ratio of signals RNB obtained from converters (41) to (43) Also R: G: H= Q, 3: 0.59: 0.11
Therefore, the signal R-B becomes a luminance signal (brightness information).

Therefore, for each pixel (9), the data DATA is red ~
The identification code determines whether it is blue or brightness data. In other words, for each pixel (9), the color or brightness that has the largest change in density from the previous pixel (9) is determined, and the change is determined. The data in the latches (31) to (33) are updated in accordance with the color or brightness that is the largest, and the signal R-B is taken out.

In this way, 11 signals R to B can be obtained for 3J, but in this case, according to the present invention, 6 bits are assigned to each of red to blue, so 18 bits can be used for 1 signal.
Therefore, 2 & =
Since 262,144 colors can be generated simultaneously, even natural images can be displayed with sufficient gradation, and natural and glossy colors can be projected.

In addition, priority is given to displaying colors with large density changes after the change, and the idea of brightness is also adopted.When the brightness change of the original data is large, the signal R-B is set to 18 bits according to the identification code "11". Since the brightness of the display screen is changed at the same time, it is possible to obtain extremely high resolution.

Furthermore, since the idea of brightness is adopted, for example, the boundary between a black part and a white part is not colored. That is, at the boundary between the black part and the white part, for example, as shown in FIG. 5, the red to blue data Hi-Bi all change significantly at the same time. If the signals Yi and ΔY are not adopted, the value ΔR9ΔG is determined in step (103).
, the largest of ΔB, for example ΔG〉ΔR1ΔG〉
If ΔB, the value ΔG is determined to be the maximum, and step (1
05), the data Gc is regarded as the data DAT of the first pixel (9), so when displaying, the i-th pixel (
9) for red and blue, for example, (
The s degree of the i-1)th pixel (9) is taken as the degree of s, and the density of green is taken as the density after the change.

Therefore, in the i-th pixel (9), the density of green is greater than the density of red and blue, so it is colored green. And the next (i+1)
The same thing happens for the th pixel (9).

However, according to this invention, since the idea of brightness is adopted and the signals Y+ and ΔY are formed, the density of each color in the first pixel (9) changes at the same time, so that it is not colored. .

Furthermore, since an identification code indicating the color or brightness is added and used only for color or brightness data that has a large density change, the memo 1) does not have a large capacity. Furthermore, there is no need for a special color compression circuit, and the problem can be handled simply by mapping using memories (21) to (23).

In addition, in the above, the upper bit b7 to data enclosure 1'
．． When forming the latch enable signal from b6 to the latches (31) to (33), if it is formed by an AND circuit and an inverter, the memories (21) to (23) will be 1
The address may have a capacity of 8 bits. In that case, memory (21) to (23) and latch (31)
-(33) can be reversed.

Furthermore, in step 61 above, it is assumed that the data 0^T^ is stored in the memory (1) in step (108), but in reality, it is stored in the hard disk or other memory and the necessary processing is performed. After finishing, it may be transferred to the memory 11). Further, the processing shown in FIG. 2 can be realized by either software or hardware.

〔Effect of the invention〕

According to this invention, 6 bits are allocated to each of red to blue, so one color is expressed with 18 bits, so 21' = 262144.
Since colors can be generated simultaneously, even natural images can be displayed with sufficient gradation, and natural and glossy colors can be projected.

In addition, priority is given to displaying colors with large density changes after the change, and the idea of brightness is also adopted.When the brightness change of coolness data is large, the signal R-8 is set to 18 bits according to the identification code "11". Since the brightness of the display screen is changed simultaneously at a considerable rate, extremely high resolution can be obtained.

Furthermore, since the idea of brightness is adopted, for example, the boundary between a black part and a white part will not be colored. That is, at the boundary between the black part and the white part, for example, as shown in Fig. 5, the red to blue data Ri''Bi all change significantly at the same time. , in step (103) the value ΔR5ΔG
, the largest of ΔB, for example ΔG〉ΔR1ΔG〉
If ΔB, the value ΔG is determined to be the maximum, and step (1
05), the data Gc is regarded as the data 0^T^ of the i-th pixel (9), so when displaying, the data Gc of the i-th pixel (
9) for red and blue, for example, (
The density of the i-1)th pixel (91) is set, and the density of green is set to the density after the change.

Therefore, in the first pixel (9), the density of green is greater than the density of red and blue, so that attention is drawn to green. And the next (1+1)
th ii! The same thing happens with the ii element (9).

However, according to I9J, the idea of brightness is adopted,
Signal Y1. Since it forms ΔY, the first pixel (9
) the density of each color changes simultaneously and is therefore not colored.

Moreover, since an identification coat indicating the color or brightness is added and used only for color or brightness data that has a large change in density, the capacity of the memory file does not increase. Furthermore, there is no need for a special color compression circuit, and the problem can be handled simply by mapping using memories (21) to (23).

[Brief explanation of drawings]

FIG. 1 is a system diagram of an example of the present invention, and FIGS. 2 to 7 are diagrams for explaining the same. +11 is display memory, (21) to (23) is mapping memory, (31) to (33) are latches, (4
1) to (43) are D/A converters.

Claims (1)

[Scope of Claims] A display circuit that retrieves display data for each pixel from a display memory and performs D/A conversion on data based on the retrieved display data to obtain three primary color signals of a color image, comprising: Among the three primary colors and luminance, the display data is data on the color or luminance that has the largest density change with adjacent pixels, and is data that has an identification code indicating that color or luminance, and is used for the above display. The above display data retrieved from memory is
The display circuit retains display data of the same color or brightness according to the identification code until the next time it is retrieved from the display memory, and converts the display data into data having a larger number of bits than the original data by mapping.