JPH0417594B2 - - Google Patents

Description

[Detailed Description of the Invention] [Technical Field of the Invention] The present invention relates to a Y, C signal for extracting a luminance signal (hereinafter referred to as Y signal) or a color signal (hereinafter referred to as C signal) from a PAL system composite video signal. The present invention relates to a separation filter, and particularly to one that digitally performs Y and C separation after converting a PAL composite video signal, which is an analog signal, into a digital signal.

[Prior art]

FIG. 1 shows the configuration of a conventional digital Y and C separation filter. In the figure, 1 is analog
A/D converter that digitizes PAL signals, 2
3 is a vertical filter that extracts a line alternating signal (the vertical high frequency component and the horizontal low frequency component of the Y signal and the C signal) from the output of the A/D converter; 3 is the vertical filter 4 is a delay element for compensating for the time delay in the vertical filter 2 and the bandpass filter 3; This subtracter calculates the difference between the output of the bandpass filter 3 and the output of the delay element 4.

Next, the operation of the YC separation filter having the configuration shown in FIG. 1 will be explained.

The PAL analog composite video signal is digitized by the A/D converter 1 at a predetermined sampling frequency s, resulting in a digital signal sequence S(i, j).
(i=1, 2,..., m, j=1, 2,..., n). This digital signal sequence S(i, j)
is applied to the vertical filter 2 and waved. The vertical filter 2 is a filter for outputting a line alternating signal, and generally consists of two 1H delay elements that delay the input signal by one line.
When focusing on a certain sample, the vertical low frequency components of the Y signal are removed using sample values located one line above and below that sample.

Next, the output signal of this vertical filter 2 is input to a bandpass filter 3, where the horizontal low frequency component of the Y signal is removed, and as a result, the C signal in the sample of interest is obtained. Further, the Y signal can be obtained as the difference between the output of the delay element 4 and the obtained C signal. That is, the C signal is obtained at the output of the band pass filter 3, and the Y signal is obtained at the output of the subtracter 5.

Since the conventional YC separation filter is configured as follows, when the video signal changes in the horizontal direction, the energy of the horizontal high frequency component of the Y signal increases. Since it is generally difficult to provide the bandpass filter 3 with sharp cutoff characteristics, horizontal high frequency components remain. Therefore, as a result, the characteristics of YC separation are impaired, and the output Y signal becomes C
The result is a mixture of signal components. Furthermore, since the horizontal high-frequency components of the Y signal that have not been removed affect the alternating signal of the next line, this YC separation filter has the disadvantage that the response to changes in the vertical video signal is also poor. It was hot.

[Summary of the invention]

This invention was made in order to eliminate the drawbacks of the conventional ones as described above. When focusing on a certain sample point, the correlation with neighboring samples in the vertical and horizontal directions of the video signal is detected, and this information is used to control the filter. By adaptively selecting the characteristics, the objective is to improve the response to local changes in the video signal.

[Embodiments of the invention]

The PAL video signal has a color subcarrier frequency of 4
When sampling at twice the frequency, the sample values within the field are distributed as shown in FIG. That is, the sample value at the i-th horizontal sample point (i, j) on the horizontal scanning line j is expressed as S(i,
j), sample points (i-2, j), (i
+2, j), (i, j-2), sample values at (i, j+2) S(i-2, j), S(i+2, j),
S(i, j-2) and S(i, j+2) all have the same color subcarrier phase, and differ from S(i, j) in color subcarrier phase by 180°. Further, a normal video signal has a property that there is a strong correlation between adjacent samples in the horizontal and vertical directions. Therefore, the above S
(i, j), the above S(i-2, j), S(i+
2,j), S(i,j-2)S(i,j+2) can be used to perform YC separation.

Embodiments of the present invention will be described in detail below with reference to the drawings.

In Fig. 3, 1 is an A/D converter that converts an analog signal into a digital signal, and 11 is this A/D converter.
Vertical signal that removes the Y signal component from the output of converter 1.
4 is a delay element for compensating the delay in the vertical and horizontal selection filter 11; 5 is the difference between the output of the vertical and horizontal selection filter 11 and the output of the delay element 4; This is a subtracter that calculates .

In the vertical and horizontal direction selection type filter 11 in FIG. 3, 6-1 and 6-2 are the first and second two
7-1, 7-2, 7- are line buffer memories that delay input digital signals by two horizontal scanning periods (hereinafter referred to as two lines).
Reference numeral 3 denotes first, second, and third sample buffer memories, which are memories that delay the input digital signal by two samples. 8-1, 8-2 are the first,
In the second adder circuit, the first adder circuit 8-1 is configured to add the output of the second two-line buffer memory 6-2 and the output of the first sample buffer memory 7-1. , the second adder circuit 8-2 is configured to add the output of the first two-line buffer memory 6-1 and the third sample buffer memory 7-3. 9-1, 9-2, 9
-3 are first, second, and third multipliers, and the first multiplier 9-1 and the second multiplier 9-2 are the adder circuit 8.
It is configured to multiply the outputs of -1 and 8-2 by 1/4, respectively. Further, the third multiplier 9-8 multiplies the output of the second sample buffer memory 7-2 by 1/2. 10-1, 10-2, 10-
3 and 10-4 are first, second, third, and fourth subtracters, and the first subtracter 10-1 is the third multiplier 9-4.
3 and the output of the first multiplier 9-1, and the second subtracter 10-2 calculates the difference between the output of the third multiplier 9-3 and the output of the second multiplier 9-2. This is to find the difference between the outputs. Third subtractor 10-3
is the difference between the output of the first sample buffer memory 7-1 and the output of the second two-line buffer memory 6-2, and the fourth tripler 10-4 is the difference between the output of the first two-line buffer memory 7-1 and the second two-line buffer memory 6-2. Line buffer memory 6-1 and the third
and the output of the sample buffer memory 7-3. 14-1 and 14-2 are the first and second absolute value calculators, respectively;
The absolute value of the output of the subtracter 10-3 and the fourth subtracter 10-4 is determined. A comparator 15 compares the outputs of the first absolute value calculator 14-1 and the second absolute value calculator 14-2 and applies a signal to a changeover switch 16, which will be described below. Reference numeral 16 denotes a changeover switch, which selects the output signal of the first subtractor 10-1 and the output signal of the second subtractor 10-2 connected to this switch 16, according to the signal from the comparator 15. , is configured to output.

In order to obtain the C signal C (i, j) at the sample value S (i, j) of a certain sample point (i, j) of the signal series digitized by the A/D converter 1, sample points (i, j-2), (i, j+2) two lines apart from each other.
sample values S(i, j-2), S(i, j+2)
And the sample values S(i
-2, j) Using the four sample values of S(i+2, j), the high frequency components T _V and T _H of the Y signal in the vertical and horizontal directions are calculated according to the following equations.

T _V = S (i, j + 2) - S (i, j - 2) T _H = S (i + 2, j) - S (i - 2, j) As is clear from Fig. 3, the above _TV The output signal of the subtractor 10-3 is the output signal of the fourth subtractor 10-4, and the above T _H is the output signal of the fourth subtractor 10-4.

Next, the high frequency components T _V and T _H of the Y signal obtained by the above calculation are converted into absolute values |T _V | and |T _H | by the absolute value calculator 14, respectively. . These signals are further input to a comparator 15,
Comparator 15 switches switch 1 according to the following conditions.
6, the output signals of the first and second subtracters 10-1 and 10-2 are selected and the C signal is taken out.

When |T _V |<|T _H |, the output of the switch 16 becomes the output signal of the first subtracter 10-1.

When |T _V | |T _H |, the output of the switch 16 becomes the output signal of the second subtracter 10-2.

That is, sample values S(i, j-2), S(i,
j+2), S(i-2, j)S(i+2, j), find the vertical high frequency component and horizontal high frequency component of the Y signal, and calculate the two values in the direction where these values are smaller. Using the sample values, the following calculation is performed to perform adaptive control to remove the low frequency component of the Y signal.

H _c (i, j) = -1/4S (i, j-2) + 1/2S
(i, j)-1/4S (i, j+2) V _c (i, j)=-1/4S (i-2, j)+1/2S
(i, j) - 1/4S (i+2, j) H _c (i, j) shown in the above formula is the first subtractor 10-1, and V _c (i, j) is the second subtractor 10-1. This is the output signal of the subtracter 10-2.

As a result, the sample position (i,
The vertical or horizontal Y signal component in j) is removed, and the H _c signal or V _c signal is immediately obtained as the C signal. Also, at this time Y
The signal Y(i,j) is determined by the following calculation as the difference between the output signal of the delay element 4 and the C signal.

Y (i, j) = S (i, j) - C (i, j) Furthermore, for sample point (i, j), this sample value S (i, j) and sample point (i-2, j
-2) (i+2, j-2), (i-2, j+2),
(i+2, j+2), the sample value S(i-
2, j-2), S(i+2, j-2), S(i-2,
j+2) and S(i+2, j+2). For example, if |S(i+2, j+2)−S(i, j)| is the smallest, then C(i, j)=-1/4{S(i+2 , j) −2S
(i, j) + S (i, j + 2)} Y (i, j) = -1/4 {S (i + 2, j) + 2S
(i, j)+S(i, j+2)} YC separation can also be performed.

〔Effect of the invention〕

As described above, according to the present invention, by utilizing the two-dimensional characteristics of the image signal, the vertical and horizontal selection filters operate corresponding to directions with little change in the vertical or horizontal direction, and Separation is performed.

As a result, it is possible to construct a Y and C separation filter with improved responsiveness to local changes in video signals and with less image deterioration.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the structure of a conventional digital Y and C separation filter. Figure 2 is
FIG. 2 is an explanatory diagram showing an arrangement of sampling pixels of a PAL video signal on one field screen; FIG. 3 is a block diagram showing an embodiment of the invention. In the figure, 1 is an A/D converter, 2 is a vertical filter, 3 is a bandpass filter, 4 is a delay element, 5 is a subtracter, and 6-1 and 6-2 are first and second two-line switches. Formemory, 7-1, 7-2, 7-
3 is the first, second and third sample buffer memory; 8-1 and 8-2 are the first and second adder circuits; 9
-1, 9-2, 9-3 are the first, second, and third multipliers; 10-1, 10-2, 10-3, and 10-4 are the first, second, third, and fourth multipliers; subtractor, 11 is vertical,
Horizontal selection filter; 14-1 and 14-2 are first and second absolute value calculators; 15 is a comparator; 16
is a switch. In the drawings, the same or corresponding parts are designated by the same reference numerals.

Claims (1)

[Claims] 1. Adaptive luminance that samples a PAL system analog composite video signal at a sampling frequency that is four times the color subcarrier frequency and separates the composite video signal into a luminance signal and a chrominance signal. It is a signal color signal separation filter, and the sample value S(i, j) of the sample point (i, j) is
j) and the sample point (i, j-
2), (i, j+2) sample value S(i, j−
2), from S(i, j+2), calculate the first color signal value V _C (i, j) of the sample point (i, j) as V _c (i, j)=-(1/4)・S(i,j-2)+(1/2
)・S(i,j)−(1/4)・S(i,j+2)
j) and a sample point (i-2,
j), (i+2, j) sample value S(i-2,
j), S(i+2, j), the second color signal value H _c (i, j) of the sample point (i, j) is calculated as H _c (i, j)=-(1/4)・S(i-2,j)+(1/2
)・S(i,j)−(1/4)・S(i+2,j)
Sample points (i, j-2) separated by a line,
Sample value S(i, j-2) of (i, j+2), S
(i, j+2), the sample point (i, j)
The high frequency component T _V (i, j) of the luminance signal _in the vertical direction on the screen at
2) Vertical brightness signal high frequency component extraction means that calculates
Sample points (i-2, j) separated by samples,
Sample value S (i-2, j) of (i+2, j), S
From (i+2, j), the sample point (i, j)
_The high frequency component T _H (i, j) of the luminance signal in the vertical direction on the screen at
j), and the absolute value of TV obtained by the vertical brightness signal high frequency component extraction means, |T _{V |} , at the sample point S(i, j). ), and the absolute value of _{T H} | is the high frequency component energy of the horizontal brightness signal at the sample point S (i, j); horizontal brightness signal high frequency component energy extraction means; By comparing the magnitude of |T _V |, which is the output of the means, and |T _H |, which is the output of the horizontal direction luminance signal high frequency component energy extraction means,
Image correlation detection means detects the correlation of images at the sample point (i, j), and according to the output of the image correlation detection means, |T _V |<
|T _H | then V _c (i, j), |T _V |≧|T _H | then H _c (i, j), the sample point (i, j)
a color signal extracting means that outputs a color signal C(i, j), and C(i, j) which is an output of the color signal extracting means.
Adaptive luminance characterized by comprising: a luminance signal extracting means for subtracting the value of S(i, j) from S(i, j) as the luminance signal Y(i, j) of the sample point (i, j); Signal color signal separation filter. 2 An adaptive luminance/chrominance signal separation filter samples a PAL analog composite video signal at a sampling frequency four times the color subcarrier frequency and separates the composite video signal into a luminance signal and a chrominance signal. Then, the sample value S(i, j) of the sample point (i, j) is
j) and the sample value S(i,
The sample point (i, j) is obtained from the sample value S(i-2, j) of the sample point (i-2, j) two samples to the left of the sample point (i, j).
The first color signal value LU _c (i, j) of LU _c (i, j)=-(1/4)・S(i-2, j)+(1/2
)・S(i,j)−(1/4)・S(i,j−2)
j) and the sample value S(i,
From the sample value S(i+2, j) of the sample point (i+2, j) that is two samples to the right of the sample point (i,
The second color signal value RU _{c (} i, j) of
)・S(i,j)−(1/4)・S(i,j−2)
j) and the sample value S(i, j+2) of the sample point (i, j+2) two lines below the sample point (i, j)
j+2) and the sample value S of the sample point (i-2, j) two samples to the left of the sample point (i, j)
(i-2, j), the sample point (i, j)
The third color signal value LL _c (i, j) of LL _c (i, j) = - (1/4) · S (i - 2, j) + (1/2
)・S(i,j)−(1/4)・S(i,j+2)
j) and the sample value S(i, j+2) of the sample point (i, j+2) two lines below the sample point (i, j)
From the sample value S(i+2, j) of the sample point (i+2, j) two samples to the right of the sample point (i, j),
_{RL c} (i, j) = - (1/4) S (i + 2, j) + (1/2
)・S(i,j)−(1/4)・S(i,j+2)
j) and a sample point (i-2, j-) two lines above and two samples to the left of the sample point (i, j).
From the sample value S(i-2, j-2) of 2),
High frequency component T _LU (i, j) of the luminance signal in the upper left direction on the screen at the sample point (i, j)
Upper left luminance signal high frequency component extraction means for calculating T _LU (i, j) = S (i-2, j) - S (i, j-2); Sample value S(i,
j) and a sample point (i+2, j-) two lines above and two samples to the right of the sample point (i, j).
From the sample value S(i+2, j-2) of 2),
High frequency component of the luminance signal in the upper right direction on the screen at the sample point (i, j) T _RU (i, j)
Upper right luminance signal high frequency component extracting means that calculates T _RU (i, j) = S (i + 2, j) - S (i, j - 2), and a sample value at the sample point (i, j). S(i,
j) and a sample point (i-2, j+) two lines below and two samples to the left of the sample point (i, j).
From the sample value S(i-2, j+2) of 2),
High frequency component of the luminance signal toward the lower left on the screen at the sample point (i, j) T _LL (i, j)
lower left luminance signal high frequency component extraction means that calculates T _LL (i, j) = S (i-2, j) - S (i, j + 2), and a sample value at the sample point (i, j). S(i,
j) and a sample point (i+2, j+) two lines below and two samples to the right of the sample point (i, j).
From the sample value S(i+2, j+2) of 2),
High frequency component T _RL (i, j) of the luminance signal in the lower right direction on the screen at the sample point (i, j)
The lower right luminance signal high frequency component extraction means calculates T _RL (i, j) = S (i + 2, j) - S (i, j + 2), and the upper left luminance signal high frequency component extraction means upper left direction luminance signal high frequency component energy extraction means for determining the obtained absolute value |T _{LU |} of the upper left direction luminance signal high frequency component energy at the sample point S (i, j); Upper right luminance where the absolute value of T _RU obtained by the upper right luminance signal high frequency component extraction means |T _RU | is the high frequency component energy of the upper right luminance signal at the sample point S (i, j) The signal high frequency component energy extraction means and the absolute value of T _LL obtained by the lower left direction luminance signal high frequency component extraction means are calculated as the lower left direction luminance signal at the sample point S (i, j) _. and the absolute value of T _{RL obtained by the lower left luminance signal high frequency component energy extracting means, which is the high frequency component energy of the lower left luminance signal, and the lower right luminance signal high frequency component energy extraction means, and the absolute value of T RL} |T _RL | A lower right direction luminance signal high frequency component energy extraction means which uses the lower right direction luminance signal high frequency component energy in S(i, j), and an output of the upper left direction luminance signal high frequency component energy extraction means. |T _LU | is the output of the upper right luminance signal high frequency component energy extraction means |T _RU | is the output of the lower left luminance signal high frequency component energy extraction means |
By comparing the magnitude of T _LL | and |T _RL |, which is the output of the lower right luminance signal high frequency component energy extracting means, the sample point (i,
According to the output of the image correlation detection means for detecting the image correlation in j), and the output of the above image correlation detection means, MIN(|T _LU |, |T _RU |, |T _LL |, |T _RL |)=
｜T _LU ｜ then LU _c (i, j) is MIN(｜T _LU ｜, ｜
T _RU ｜, ｜T _LL ｜, ｜T _RL ｜=｜T _RU ｜then RU _c
(i, j) is MIN(|T _LU |, |T _RU |, |T _LL |, |
T _RL ｜=｜T _LL ｜then LL _c (i, j) is MIN(｜
T _LU |, |T _RU |, |T _LL |, |T _RL |) = |T _RL |, then RL _c (i, j) (where MIN(S, T,
U, V) indicate the minimum value among S, T, U, and V. Also, if two or more conditions are met,
Priority is given in the order of LU _c , RU _c , LL _c , and RL _c . ) a color signal extracting means that outputs a color signal C(i, j) from the sample point (i, j); and C(i, j), which is the output of the color signal extracting means.
Adaptive luminance characterized by comprising: a luminance signal extracting means for subtracting the value of S(i, j) from S(i, j) as the luminance signal Y(i, j) of the sample point (i, j); Signal color signal separation filter.