JPH03204293A - Picture signal processor - Google Patents

Abstract

PURPOSE:To attain signal processing with high quality and simple constitution by using Y, R, B component signals for the picture signals and applying noninterlace processing attended with movement detection processing of the picture. CONSTITUTION:An input picture signal is converted into component signals as a red color signal and a blue color signal by a 1st conversion means 10 and a signal processing means applies signal processing attended with movement detection processing to the component signals obtained by the 1st conversion means 10. Then the component signal subject to the signal processing by the signal processing means is converted into component signals comprising a green signal, a red signal and a blue signal by a 2nd conversion means 32 and the result is outputted. Thus, the signal processing attended with the movement detection processing to the input picture signal is implemented with high quality and simple constitution.

Description

DETAILED DESCRIPTION OF THE INVENTION A. Industrial Field of Application The present invention provides motion detection processing for an input image signal.
In the case of a still image, a high-quality image can be obtained by eliminating flicker interference caused by interlaced scanning.

On the other hand, in the case of moving images, when an interlaced television signal is deinterlaced, the edges of the moving parts of the image become comb-like.
This results in unsightly image quality deterioration.

Therefore, non-interlacing processing has been attempted in which image motion information is extracted from the television signal and line interpolation is performed instead of field interpolation in the video portion, thereby preventing the above-mentioned image quality deterioration. ing.

As a motion detection circuit for extracting image motion information from the television signal, for example, as disclosed in Japanese Patent Application Laid-Open No. 6027287, the motion detection circuit extracts image motion information from an interlace-scanned television signal in two consecutive fields. There is a method that extracts the frame difference signal from the frame difference signal.

D Problems to be Solved by the Invention By the way, high-definition television signals generally consist of three primary color component signals: a red signal [R], a green signal (G), and a blue signal ((1)), and a luminance signal (Y) that is a red signal. Difference signal (P
r] and the red difference signal [P, ] are treated as component signals.

Here, each of the above color difference signals (p, ], CP,) is:
...■ As P, = B-Y / 1.826 ...■P,
= RY / 1.576 −・−■ It is shown.

In an image signal processing apparatus that performs deinterlacing processing accompanied by image motion detection processing as described above, Y-P, -P,
If component signals are handled, a matrix arithmetic processing circuit is required to convert the Y-P, -P, and component signals, which have double the data rate after non-interlacing processing, into R-G-B component signals. Also, R-G-B
When input image signals of component signals are supplied, the input stage converts the R-G-B component signals into Y-P,
- The signal is converted to a P, component signal, and then returned to an RGB component signal after deinterlacing processing, which tends to include calculation errors.

Furthermore, since the image motion detection process requires the luminance signal [Y], an image signal processing apparatus that performs non-interlacing processing that accompanies image motion detection
If non-interlacing processing is performed on the B component signal, a luminance signal (Y) must be formed from the R-G-B component signal and image motion detection processing must be performed, which increases the complexity of the circuit configuration. Problems such as an increase in the number of connection lines occur.

SUMMARY OF THE INVENTION In view of the above-mentioned problems, it is an object of the present invention to make it possible to perform signal processing involving motion detection processing on an input image signal with a simple configuration and high quality.

E If! Means for solving fi In order to achieve the above-mentioned object, the image signal processing device according to the present invention includes a first conversion means for converting a human image signal into component signals of a luminance signal, a red signal, and a blue signal. , signal processing means for performing at least motion detection processing on the component signal obtained by the first conversion means, and subjecting the component signal to signal processing based on the motion detection output; and signal processing by the signal processing means. The present invention is characterized by comprising a second converting means for converting the component signal into component signals of a green signal, a red signal, and a blue signal and outputting the component signals.

F action In the image signal processing device according to the present invention, the input image signal is converted into component signals of a luminance signal, a red signal, and a blue signal by the first conversion means, and for the component signals obtained by the first conversion means, The signal processing means performs signal processing accompanied by motion detection processing. The component signal processed by the signal processing means is converted into component signals of a green signal, a red signal, and a blue signal by a second conversion means and output.

G. Embodiment Hereinafter, one embodiment of the image signal processing device according to the present invention will be described in detail with reference to the drawings.

FIG. 1 shows the configuration of an embodiment in which the present invention is applied to a deinterlacing processing device for high-definition interlaced television signals.

This deinterlacing processing device has an analog input terminal (1) for high-definition interlaced television signals,
(2), (3) and digital input terminals (4), (
5) Equipped with (6).

Above analog input terminals (1), (2), (3)
Analog R-G-B component signals or Y
-P.

P, component signal is analog input image signal [5I
Nl], [5IN2], and [S, N3]. The above analog input image signal [S engineering,] (S+Nz)
, [5zNs) is analog/digital (A/D)
The data is digitized by a converter (8) and supplied as input image data rD+), CD2), [D3) to an input matrix calculation circuit (10).

The A/D converter (8) is operated by the sampling clock provided by the clock signal generator (8), and input image data CDI), (Dz), (D3) whose sampling frequency is 74.25 MHz.
form.

In addition, the digital input terminals (4), (5),
(6) includes an R-G-B component signal or a Y-
P.

・P, digital input image signals CDlN1), CDINZ), CDl83] whose sampling frequency is 74.25 MHz, which are obtained by digitizing component signals in advance, are supplied. Above digital input image signal [DIsl]
, CDINZ ), (DIsi ) are directly supplied to the input matrix calculation circuit (10) as input image data CD, ), [Dz ], (D3).

The input matrix calculation circuit (10) includes the R.G.
B component signal or Y-P, -P, input image data corresponding to component signal [D,), 'CDz
], CD3] are image data [DYl, CDR), (D
B).

Here, from the R・G-B component signal, Y =
0.212R+ 0.701G +0.0878B
...The luminance signal [Y] can be formed by the calculation process (■).

Also, from Y-P, -P, and component signals, R = Y + 1.576P, ----■B = Y
+ 1.826 P b . . . Red signal [R] and blue signal CB) can be formed by each calculation process.

The input matrix calculation circuit (10) includes the R.G.
Input image data corresponding to B component signal [:+
, ), CD, ], [D3), and Y-P.

・P, input image data corresponding to the component signal (
DI), (D2), CD! ) is provided with an arithmetic processing unit that performs arithmetic processing of the fifth and sixth equations above.

The arithmetic processing unit that performs the arithmetic processing of the fourth equation above is, as shown in the equalization circuit in FIG.
, CO,701], consists of three coefficient circuits (11), (12), (13) that give (0,087) and one adder (14). In addition, the arithmetic processing unit that performs the arithmetic processing of the fifth and sixth equations has the coefficients [1, 57
Two coefficient circuits (15) (16) giving 6'l, [1,826) and two adders (17), (18
).

Y obtained by the above input matrix calculation circuit (10)
- Image data corresponding to the R-B component signal (D
Y], (DR), (DB) are the first one-field delay circuit (21), the line interpolation circuit (22), and the first
The output matrix is supplied to the output matrix calculation circuit (31). Also, Y obtained from the input matrix calculation circuit (10)
- Image data corresponding to R-B component signal [D
Y), CDR], [DBl luminance data (DY
) is a motion vector detection circuit (23) and a motion detection circuit (
24).

Here, the interlaced television signal is the same in each field as shown in Figure 4, where each horizontal line where data actually exists is shown with a circle, and each horizontal line where no data actually exists is shown with an x mark. Within a field, data for each line exists every other horizontal line, and data for the same line exists for every other field.

The first field delay circuit (21) receives image data (
By giving a delay amount equivalent to one field period to DY), (DR), and [DB), the data of each line indicated by the X mark in Fig. 4 above, where no data actually exists in the current field, can be is the data one field before (DYF),
Interpolate with (DRF)8 (DBF). Data of one field before (DYF), (DRF), CD obtained through this first field delay circuit (21)
By) is supplied to the mixer circuit (25), and its luminance data (DYF) is supplied to the motion detection circuit (24).
).

A field memory is used for the first field delay circuit (2I).

The line interpolation circuit (22) also receives image data (DYICDR) from the input matrix calculation circuit (10).
), (DB), from the data of each line marked O in the above figure 4 where data actually exists in the current field, to the data of each horizontal line marked with an x mark in the above figure 4 which does not actually exist. Data [DYL), [DRL), l:
DBL] is formed by line interpolation calculation processing. Interpolated output data (DYL), (DRL) of each horizontal line formed by this line interpolation circuit (22),
(DBt) is supplied to the mixer circuit (25), and its luminance data [DYL] is passed through the second field delay circuit (26) to the motion detection circuit (24).
). The second field delay circuit (26
) consists of field memory and stores the above luminance data (
DYL) is given a delay amount corresponding to one field.

Further, the motion vector detection circuit (23) detects the motion vector of the input image with respect to the image data (DY) supplied from the input matrix calculation circuit (10). The detection output from this motion vector detection circuit (23) is used as control data for motion compensation using motion vectors and is transmitted to the first and second field delay circuits (21),
(26). The motion vector detection circuit (23) sends control data to the mixer circuit (23) indicating that the result of motion compensation using the motion vector corresponds to a line that does not exist in the previous field and was obtained by line interpolation processing. 25).

In addition, the motion detection circuit (24) includes luminance data CDY of the current field directly supplied from the input matrix calculation circuit (10) and the line interpolation circuit (22).
Line interpolated luminance data [DYL] of the current field directly supplied from the above-mentioned human matrix calculation circuit (10
) is supplied via the first field delay circuit (21), and the luminance data (DYF) of one field before is supplied from the line interpolation circuit (22) via the second field delay circuit (26). Motion detection is performed to determine whether the input image is a still image with no movement or a moving surface with movement by comparing the line interpolated luminance data [DYLF] of the previous l field. The motion detection output from the circuit (24) is sent to the mixer circuit (24).
25).

The mixer circuit (25) is connected to the motion detection circuit (24).
When it is determined that the input image is a static surface, the above-mentioned 4th field in which no data actually exists in the current field.
Interpolated output data (DY+p), (DR,), [DB,? ] as field interpolation output data CDYF), [DRF), by the first field delay circuit (21),
(DBF) is selected, and when the motion detection circuit (24) determines that the input image is a moving image, the interpolation output data (DY+r:l, (DRIP),
As (DRIP), line interpolation output data (DYL), (DRL), (D
BL].

Note that motion compensation based on the motion vector detected by the motion vector detection circuit (23), that is, each field delay circuit (21) based on the motion vector,
(26) is corrected, and the motion detection circuit (
Even if it is determined that the input image is a static surface according to 24), if the result of motion compensation using the motion vector corresponds to a line that does not exist in the previous field and was obtained by line interpolation processing, The above mixer circuit (2
5) is line interpolation output data [DYL], (DRL],
(DB,) is controlled to be selected.

Further, the mixer circuit (25) outputs the field interpolation data (DYF), (DRF), (DBF) and the line interpolation output data CDYL), (DRL] [D
This configuration prevents the interpolated output data from becoming unnatural by combining them at a predetermined ratio instead of simply switching between BL and BL.

The interpolated output data (DYlr), (DRIF), and [DBIP) output from the mixer circuit (25) are supplied to the second output matrix calculation circuit (32).

The first and second output matrix calculation circuits (31)
, (32) converts image data in the Y-P, -P, and component signal formats to image data in the R-G-B component signal format, respectively, and is output from the input matrix calculation circuit (lO). Y-P, ・P
, component signal format image data (DY), [D
R) and (DB) from the first output matrix calculation circuit (
31), image data (DRo) in the RG-8 component signal format.

CDG, ), [DB,) and output from the mixer circuit (25), Y-P, -P, interpolated output data [DY, P] in component signal format.

[DRIF) and [DBIP) to the second output matrix.

The interpolated image data (D Y I F.), [D
RIF. ), [DBIP. ] Convert to

Here, from Y-P, ・P, component signal, 0.701 0.701 0.701・
. . . The green signal (G3) can be formed by the arithmetic processing of (■).

Therefore, the first and second output matrix calculation circuits (
31) and (32) are each provided with an arithmetic processing unit that performs the arithmetic processing of the seventh equation above.

As shown in the equalization circuit in FIG.
/ 0.701) , (1,00010,701)
.

It is composed of three coefficient circuits (51), (52), (53) giving (-0,08710,701) and one adder (54).

The image data [DRO] obtained by the first output matrix calculation circuit (31), CDGo) (DBO
) and captured image data (DYIF.) in R-G/B component signal format obtained by the second output matrix calculation circuit (32), [DRIPo], (DBl,
. ] is the non-interlaced output image data [DYOII?] via the line buffer memory (40).
), (DROIIT), CDBOU,] as l
Each line is output alternately.

Non-interlaced output image data [DYour
], [DRout], (DBout) are analogized by a digital-to-analog (D/A) converter (41) operating according to a clock signal provided by a clock generator (42) to produce a non-interlaced television signal ( Yout ) , (ROLIT ) [
BOIIT) as signal output terminals (43), (4
4) is output from (45).

As in this embodiment, by handling the image signal as a Y-R-B component signal and performing de-interlacing processing accompanied by image motion detection processing as described above, motion vector detection processing and motion detection processing can be performed. The luminance signal [
There is no need to specially form Y). The input image signal is R-
Even if it is a G-B component signal, Y-P, -P, component signal, large Kama matrix calculation circuit (10)
can be easily converted into a Y-R-8 component signal by an arithmetic processing unit with a simple circuit configuration that performs the arithmetic processing of the fifth and sixth equations described above, and does not impose a special burden. In addition, each output matrix calculation circuit (31) (32) only needs to form the green signal (CG) by matrix calculation.
This is also not a burden, and since no arithmetic processing is performed on the red signal [R] and the blue signal CB), no conversion error occurs.

Note that if the image signal is treated as an R-G-B component signal and the non-interlacing process as described above is performed, each output matrix calculation circuit (31),
(32) is not necessary, but for motion detection processing, (a) R-G-B component signal of the current field (B) R-G-B component signal of one field before (c) For the interpolated output signal of the current field that has undergone intra-field interpolation processing (d) and the interpolated output signal of the previous field that has undergone intra-field interpolation processing, matrix calculation processing circuits are required to form luminance signals, respectively, and motion detection processing is required. The amount of hardware required and the number of connections becomes enormous.

H Effects of the Invention As described above, in the image signal processing device according to the present invention, the input image signal is converted into a Y/R-B component signal by the first conversion means, and the Y/R-B component signal obtained by the first conversion means is - Since the signal processing means performs signal processing involving motion detection processing on the R-B component signal, there is no need to specially form a luminance signal for motion detection processing, and the circuit configuration for motion detection processing is simple. becomes, and also,
The number of connection lines can also be reduced. Furthermore, in the image signal processing device according to the present invention, the component signal subjected to signal processing by the signal processing means is converted into an R-G-B component signal by the second conversion means, so that
In the second conversion means, only the green signal CG) needs to be formed by conversion processing, and no calculation error is included in the red signal CR) and the blue signal (B) in the second conversion means.

Therefore, according to the present invention, it is possible to provide an image signal processing device that can perform signal processing involving motion detection processing on an input image signal with a simple configuration and high quality.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of an embodiment in which the present invention is applied to a deinterlacing processing device for high-definition interlaced television signals, and FIG. FIG. 3 is a block diagram showing the configuration of an input matrix calculation circuit showing a matrix calculation circuit that converts into a signal, and FIG.
A block diagram showing the configuration of an input matrix calculation circuit that converts component signals into Y, R, and B component signals. Figure 4 shows lines where image data actually exists in an interlace-scanned television signal and lines where image data actually exists. Figure 5 shows the relationship with the line where
The figure is a block diagram showing an output matrix calculation circuit that converts Y, R, B component signals into R, G, B component signals. (1)～(6) ・・・・・・(10) ・・・・・・・・・(21),(26) ・・・・・・(22) ・・・・・・・・・・ (23) ・・・・・・・・・・・・ (24) ・・・・・・・・・・ (25) ・・・・・・・・・・ (31) (32),,,, . (40) ・・・・・・・・・・・・ (43) to (45) Input terminal Input matrix calculation circuit Field delay circuit Line interpolation circuit Motion hector detection circuit Motion detection circuit Mixer circuit Input matrix calculation circuit Line Buffer memory signal output terminal Fig. 3 l Fig. 4 Fig. 5

Claims (1)

[Claims] A first conversion means for converting an input image signal into component signals of a luminance signal, a red signal, and a blue signal; and at least motion detection processing is performed on the component signals obtained by the first conversion means. , a signal processing means for subjecting the component signal to signal processing based on the motion detection output, and converting the component signal subjected to the signal processing by the signal processing means into component signals of a green signal, a red signal, and a blue signal and outputting the converted signal. An image signal processing device comprising: second conversion means for converting.