JPH0360231B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to a color television receiver, and more particularly to a digital filter that separates a luminance signal and a color signal from a digitized color television signal.

[Prior art]

The NTSC composite color television signal S is
It is well known that it is a composite signal of a luminance signal Y and a color signal C, and that the color signal C is obtained by quadrature two-phase modulation of two color difference signals U and V at a color subcarrier frequency fsc. be. That is, S=Y+C=Y+Usin(2πfsct)+Vco
s (2πfsc t)...(1) If the frame frequency is f _F (=30Hz), the field frequency is f _V , and the horizontal scanning frequency is f _H , then fsc = 455/2f _H = 55/2525/2f _V = 455/ 2.525f _F
It is also well known that fsc=3.583MHz, which is determined by the relationship (2). Also, the signal S shown in equation (1)
It is also well known that when digitizing, the sampling frequency fs is set to fs = 4fsc.

f _S = 4f When the signal S is sampled in _SC , the signal at each sampling point near the signal S (m, n) at the sampling point corresponding to the n-th line and m-th column on the screen is expressed as the first The result will be as shown in the figure.

In Figure 1, a black triangle (for example, S(m,
n)) is Y+C [where C=1/√2 2(U+V)], and a hollow triangle (for example, S(m, n-1)) is Y
-C, a black square (e.g. S(m+1,n)) is Y+C' [where C'=1/√2 2(U-V)], a hollow square (e.g. S(m+1,n-1)) is Y-
It represents C′. Also, due to the relationship f _SC =455/2f _H , in the vertical columns, black and hollow are alternately arranged for each line.

Next, the PAL system composite color television signal P is P=Y+Usin(2πf _SC pt)±Vcos(2πf _SC pt)...(3), where Y, U, and V are each Y in equation (1). ,
Similar to U and V, f _SCp is the color subcarrier frequency in the PAL system, and the frame frequency of the PAL system is f _Fp
(=25Hz), field frequency is f _Vp , horizontal scanning frequency is f _Hp , f _SCp = (284-1/4 + 1/625) f _Hp = (284-1/4 + 1/625)・625/2f _Vp = ( 284−1/4+1/625)・625f _Fp ……(4). In equation (3), the sign ± is changed to - for even scan lines and + for odd scan lines. That is, the V component is inverted for each scanning line.

When the signal shown in equation (3) is digitized with a sampling frequency of f _Sp = 4f _SCp , the signal P (m, n) at the sampling point corresponding to the nth line and mth column on the screen is obtained.
The signals at each sampling point in the vicinity of are as shown in FIG. In other words, in the lines where ± in equation (3) is +, the equation is the same as in equation (1), so Figure 2 has the same arrangement as in Figure 1, and in the lines where ± is -, the black rectangle and the hollow are the same. Replace.

A conventional digital filter compatible with the NTSC system is constructed using the characteristics of the signal array shown in FIG. 1, and FIG. 3 is a block diagram showing an example of the conventional device. In the figure, 1 is a vertical filter, 2
is a bandpass filter, 3 is a subtracter, and the vertical filter 1 is composed of a vertical filter calculation circuit 10 and one-line delay circuits 11 and 12. 10
1 to 106 indicate each signal line.

Assume that a signal S expressed by equation (1) is input onto the signal line 101 as a digitized signal S(s) as shown in FIG. However, S in (s) is the number of the sampling point. S(m, n+1) in Figure 1
When a signal of 1 is input to the signal line 101, 1
The signal on the signal line 102, which is the output of the line delay circuit 11, is S(m, n), and the signal on the signal line 103, which is the output of the one-line delay circuit 12, is S(m,
n-1).

The arithmetic circuit 10 is H _C (m, n)=-1/4 {S(m, n-1)-2
S(m,n)+S(m,n+1)}...(5) is performed and H _C (m,n) is output. By the way, S(m, n-1), S(m, n), S(m, n+1)
are signals of lines adjacent to each other in the vertical direction, so generally the values of Y, U, and V in equation (1) should be extremely similar.

The values of Y and C are S (m, n-1), S (m, n), S
For (m, n+1), respectively (Y ₀ +△Y ₁ , C ₀
+△C ₁ ), (Y ₀ , C ₀ ), (Y ₀ +△Y ₂ , C ₀ +△C ₂ ), then △Y ₁ , △Y ₂ , △C ₁ , △C ₂ are generally This is an extremely small value. Therefore, the calculation result of equation (5) is H _C (m, n) = −1/4 {(△Y ₁ +△Y ₂ ) −4C ₀
+ (-△C ₁ -△C ₂ )} =C ₀ +1/4 (△C ₁ +△C ₂ ) -1/4 (△Y ₁ +△
Y ₂ )...(6).

The signal on the signal line 101 is S(m, n+
1) is followed by S (m+1, n+1), but at this point the output of the arithmetic circuit 10 H _C (m+1, n)
H _C (m+1, n)=C′+1/4(△C′ ₁ +△C
′ ₂ )−1/4 (△Y ₃ +△Y ₄ )……(7), and △C ₁ , △C ₂ , △C′ ₁ , △C′ ₂ , △Y ₃ ,
ΔY ₄ is generally an extremely small value.

That is, the signal S(s) input to the vertical filter 1 includes the luminance signal Y(s) (digitized signal Y in equation (1)) and the color signal C(s) (digitized signal Y in equation (1)). However, it can be seen from equations (6) and (7) that the luminance signal Y(s) is sufficiently attenuated at the output of the arithmetic circuit 10.

If the delay of one sampling period and the delay of one line are respectively Z ^-1 and Z ^-l using Z transformation, then Z ^-1 = exp{−j2πf/f _S } ...(8), and the formula ( From 2), l=910.

If the signal on signal line 101 is S(s), the signals on signal lines 102 and 103 are S(s) delayed by one line and S(s) delayed by two lines, respectively.
(s), and the transfer function Hv of vertical filter 1 is
(z) is Hv (z) = −1/4 (1−2Z ^-l +Z ^-2l ) = −1
/4(1−Z ^-l ) ² ...(10)

A bandpass filter 2 is used to provide further attenuation to the signal Y(s) remaining in the output of the vertical filter 1. The bandpass filter 2 is a bandpass digital filter designed to pass the band of the signals U and V at the center frequency f _SC .

For example, the transfer function H _B (z) of bandpass filter 2
is H _B (z)=-1/32(1-z ^-2 ) ² (
1+z ^-4 ) ² (1+z ^-8 )...(11)

The output of the band filter 2 can be regarded as the color signal C(s), and the subtracter 3 converts it into S(s)-
We can obtain C(s)=Y(s).

A conventional digital filter compatible with the PAL system is constructed using the characteristics of the signal arrangement shown in FIG. In other words, just as in the case of the NTSC system, formula (5) is calculated using one line above and below, in the PAL system, two lines above and below are used to calculate H _Cp (m, 2n) = -1/4 {S (m , 2n−2)−2S(
m, 2n) + S (m, 2n + 2)}... (12) and outputs the calculation result onto the signal line 104.

The conventional device is configured as described above, and in the vertical filter 1, △Y ₁ , △Y ₂ , △ in equations (6) and (7)
Since the filter is constructed on the assumption that the values of Y ₃ , △Y ₄ , △C ₁ , △C ₂ , △C′ ₁ , △C′ ₂ etc. are sufficiently small, the horizontal straight line In areas where the brightness and color of the image corresponding to adjacent horizontal scanning lines change drastically, the vertical filter 1 no longer functions as a filter, and as a result, the reproduced image becomes blurred due to color clouding or dot interference due to cross colors. It has the disadvantage of occurring.

[Summary of the invention]

This invention was made to eliminate the above-mentioned drawbacks of the conventional ones. In general, the image signals for pixels that are adjacent to each other in the direction of a line that is inclined to the horizontal direction (hereinafter referred to as a diagonal line) are also similar to each other. By focusing on the fact that it is similar to the above, and configuring separation filters for each of four predetermined directions, and by appropriately switching and using these four types of separation filters, the image signal can suddenly change between adjacent pixels. In particular, the reduction in filter effectiveness caused by this was prevented.

[Embodiments of the Invention] Examples of the invention will be described below with reference to the drawings. FIG. 4 is a block diagram showing an embodiment of the present invention, and corresponds to the input of the signals shown in FIG.
The same symbols as in Fig. 3 indicate the same or corresponding parts, and 4
are comparison judgment circuits, 5a, 5b, 5c, 5d, 5
Denoted by e and 5f are delay circuits each having two sampling periods, 6a and 6b are delay circuits each having (1 line - 4 sampling periods), 7 is a four-way separation filter, and 8 is a selector. Further, 111 to 125 are signal lines, respectively. Unlike the case in Figure 3,
An example is shown in which band filter 2 is omitted, but
Needless to say, the bandpass filter 2 may be inserted between the output of the selector 9 and the signal line 124.

The signal S (m, n) shown in FIG.
When the signal lines 101, 112, 11 appear on the
6,118 respectively have S(m+2,n+1) and S
(m-2, n+1), S(m+2, n-1), S(m
-2, n-1) appear, these signals have the same color subcarrier phase as the signal S(m, n), and the signal lines 111, 113, 115,
117 have S(m, n+1), S(m+2,
n), S(m-2, n), and S(m, n-1) appear, and these signals differ in color subcarrier phase by 180° from the signal S(m, n).

The four-way separation filter 7 separates the signal S(m,n), the signal S(m+2,n), S(m-2,n), S(m,n
-1), input the signal S(m, n+1) and C ^N ₁ (m, n)=-1/4{S(m-2, n)
-2S(m,n)+S(m,n-1)}...(13) C ^N ₂ (m,n)=-1/4{S(m+2,n)
-2S (m, n) + S (m, n-1)}... (14) C ^N ₃ (m, n) = -1/4 {S (m-2, n)
-2S(m,n)+S(m,n+1)}...(15) C ^N ₄ (m,n)=-1/4{S(m+2,n)
-2S(m,n)+S(m,n+1)}...(16) Perform the calculation.

C ^N ₁ (m,
n), C ^N ₂ (m, n), C ^N ₃ (m, n), C ^N ₄ (m, n)
The four values are input to the selector 9 through signal lines 120 to 123, respectively, and the signal selected by the control signal (outputted on the signal line 119) from the comparison/judgment circuit 4 is input on the signal line 124. is output to.

The comparison/judgment circuit 4 has signals S(m+2, n+1),
S(m-2, n+1), S(m, n), S(m+2,
n-1), S (m-2, n-1) are input, ε ^N ₁ (m, n) = | S (m, n)
-S (m-2, n-1) | ... (17) ε ^N ₂ (m, n) = | S (m, n)
-S (m+2, n-1) | ... (18) ε ^N ₃ (m, n) = | S (m, n)
-S (m-2, n+1) | ... (19) ε ^N ₄ (m, n) = | S (m, n)
-S(m+2, n+1) |...(20) The calculation is performed to determine the minimum value among ε ^N ₁ , ε ^N ₂ , ε ^N ₃ , and ε ^N ₄ , and the minimum value is determined as ε ^N ₁ , ε ^N ₂ , ε ^N ₃ , ε ^N ₄
Correspondingly, C ^N ₁ , C ^N ₂ , C ^N ₃ , and C ^N ₄ are output from the selector 9, respectively. In the case of the embodiment shown in FIG. 4, the signal output from the selector 9 is
Corresponds to C(s) in the figure, and S(s)− by subtractor 3
Obtaining C(s)=Y(s) is similar to the case of FIG.

FIG. 5 is a block diagram showing an embodiment of the present invention for the PAL composite color television signal shown in FIG. 2, in which the same reference numerals as in FIG. 4 indicate the same or corresponding parts and operate in the same manner. 6c, 6d
are each given a delay amount of (2 lines - 4 sampling periods) by a delay circuit.

Therefore, the signal P shown in FIG.
When (m, 2n) appears, the signal lines 101, 11
1,112,113,114,15,116,1
P(m+2, 2n+
2), P(m, 2n+2), P(m+2, 2n), P
(m-2, 2n), P(m, 2n), P(m-2, 2
n), P(m+2, 2n+2), P(m, 2n-2),
A signal of P(m-2, 2n-2) is output.

In the 4-way separation filter 7, C ^P ₁ (m, 2n) = -1/4 {P (m-2, 2n)
−2P(m, 2n)+P(m, 2n−2)}……(21) C ^P ₂ (m, 2n)=−1/4{P(m+2, 2n)
−2P(m, 2n)+P(m, 2n−2)}……(22) C ^P ₃ (m, 2n)=−1/4{P(m−2, 2n)
-2P (m, 2n) + P (m, 2n + 2)} ... (23) C ^P ₄ (m, 2n) = -1/4 {P (m + 2, 2n)
−2P(m, 2n)+P(m, 2n+2)}...(24) The calculation results are C ^P ₁ (m, 2n), C ^P ₂ (m,
2n), C ^P ₃ (m, 2n), and C ^P ₄ (m, 2n) are input to the selector 8.

In the comparison/judgment circuit 4, ε ^P ₁ (m, 2n) = |P(m, 2n)
-P (m-2, 2n-2) | ... (25) ε ^P ₂ (m, 2n) = | P (m, 2n)
-P(m+2, 2n-2) |...(26) ε ^P ₃ (m, 2n) = |P(m, 2n)
−P(m−2, 2n+2) |……(27) ε ^P ₄ (m, 2n)=|P(m, 2n)
−P(m+2, 2n+2)|...(28) to determine the one that gives the minimum value among ε ^P ₁ to ^{ε P} ₄ and to control the selector 9 according to this determination result; The subsequent operation of the circuit is the same as that shown in FIG.

Furthermore, the delay circuits 6a and 6b shown in FIG. 4 may be used as variable delay circuits that can be used in both the NTSC system and the PAL system.

〔Effect of the invention〕

As described above, according to the present invention, when separating a luminance signal and a color signal from a composite color television signal digitized using a sampling frequency that is four times the color subcarrier frequency, an image on a cathode ray tube is Since automatic switching is performed to use the filter in the direction in which the signal changes are the smallest, it is possible to provide a luminance signal separation filter that is effective for any image signal.

[Brief explanation of drawings]

Figure 1 shows the data array on a cathode ray tube when an NTSC composite color television signal is digitized at a sampling frequency four times the color subcarrier frequency, and Figure 2 shows a PAL composite color television signal. A diagram showing the data arrangement on a cathode ray tube when digitized at a sampling frequency four times the subcarrier frequency, FIG. 3 is a block diagram showing the configuration of a conventional filter, and FIGS. FIG. 2 is a block diagram showing one embodiment. 3...Subtractor, 4...Comparison/judgment circuit, 5a, 5
b, 5c, 5d, 5e, 5f, 6a, 6b, 6
c, 6d...Delay circuit, 7...4-way separation filter, 8...Selector. Note that the same reference numerals in the figures indicate the same or equivalent parts.

Claims (1)

[Claims] 1. For each data signal obtained by digitizing an NTSC composite color television signal at a sampling frequency four times the color subcarrier frequency, the data signal is Let the signal S(m,n) be arranged in the mth column and the nth column, and the data signal arranged in the column two samples earlier in the same row as the signal S(m,n) be the signal S(m-2,
n), and the data signal arranged in the same row as the signal S(m, n) and the column two samples later is the signal (m+
2, n), and the data signal arranged in the same column as the signal S(m, n) and one row before is the signal S(m, n-1).
Let S(m, n+1) be the data signal arranged in the same column and one row after the signal S(m, n), and arrange it in the column one row and two samples before the signal S(m, n). The data signal S(m-2, n-
1), one line before the signal S (m, n) and 2
The data signals arranged in columns after sampling are designated as signal S
(m+2,n-1), and from the signal S(m,n), 1
The data signal arranged in the column after the row and two samples before is defined as a signal S (m-2, n+1), and the signal S
The data signal arranged in the column one row and two samples after (m, n) is called signal S (m+2, n+1).
When, C ^N ₁ (m, n)=-1/4{S(m-2, n)-2S(m,
n)+S(m, n-1)} and calculates the signal C ^N ₁ (m,
n), C ^N ₂ (m, n)=-1/4{S(m+2,n)-2S(m,
n)+S(m, n-1)} and calculates the signal C ^N ₂ (m,
n), C ^N ₃ (m, n)=-1/4{S(m-2, n)-2S(m,
n)+S(m, n-1)} and calculates the signal C ^N ₃ (m,
A third direction separation filter outputs C ^N ₄ (m, n) = -1/4 {S (m + 2, n) - 2S (m,
n)+S(m, n+1)} and calculates the signal C ^N ₄ (m,
a fourth direction separation filter that outputs the signal S(m-2,n) for the signal S(m,n);
-1), signal S (m+2, n-1), signal S (m-
2, n+1), signal S(m+2, n+1), and one of the first, second, third, and fourth direction separation filters by determining which of the four signals S(m+2, n+1) is most similar. A luminance signal/chrominance signal separation filter comprising means for selecting the output of two direction separation filters. 2 For each data signal obtained by digitizing a PAL composite color television signal at a sampling frequency that is four times the color subcarrier frequency, the data signal is digitized in the mth column and 2nth row with respect to the arrangement position of the data signal on the screen. The signal P(m, 2n) arranged in
Then, the data signal arranged in the same row as the signal P(m, 2n) and the column two samples before is the signal P(m-
2, 2n), and the data signal arranged in the column two samples later in the same row as the signal P(m, 2n) is the signal P(m+2, 2n), and the data signal arranged in the same column as the signal P(m, 2n) is 2 samples. The data signal arranged before the row is the signal P.
(m, 2n-2), and the data signal arranged two rows later in the same column as the signal P(m, 2n) is P(m, 2n-2).
n+2), and the data signal arranged in the column two rows and two samples before the signal P(m, 2n) is the signal P(m-2, 2n-2), and the signal P(m, 2n) is
The data signal arranged in the column two rows before and two samples after the signal P (m+2, 2n-2) is the data signal arranged in the column two rows before and two samples after the signal P (m, 2n), and the data signal is arranged in the column two rows after and two samples before the signal P (m, 2n). The arranged data signals are expressed as signal P(m
-2, 2n+2), and when the data signal arranged in the column two rows and two samples after the signal P(m, 2n) is the signal P(m+2, 2n+2), C ^P ₁ (m, 2n )=-1/4{P(m-2, 2n)-2P(m,
2n)+P(m, 2n-2)} and calculate the signal C ^P ₁ (m,
2n), C ^P ₂ (m, 2n) = -1/4 {P (m + 2, 2n) - 2P (m,
2n)+P(m, 2n-2)} and calculates the signal C ^P ₂ (m,
2n), C ^P ₃ (m, 2n)=−1/4{P(m−2, 2n)−2P(m,
2n)+P(m, 2n+2)} and calculates the signal C ^P ₃ (m,
2n), C ^P ₄ (m, 2n) = -1/4 {P (m + 2, 2n) - 2P (m,
2n)+P(m, 2n+2)} and calculates the signal C ^P ₄ (m,
A fourth direction separation filter outputs the signal P(m-2, 2n) for the signal P(m, 2n);
2n-2), signal P(m+2,2n-2), signal P
(m-2, 2n+2), signal P (m+2, 2n+2)
means for selecting the output of one of the first, second, third, and fourth direction separation filters by determining which of the four signals is most similar; Luminance signal color signal separation filter.