JPH0385082A - Luminance signal chrominance signal separator - Google Patents

Abstract

PURPOSE:To attain accurate Y/C separation to a composite video signal including a region with rapid brightness and color change by selecting filters based on the distribution state of color and brightness. CONSTITUTION:When a change in color and brightness in a minute picture region is large in the horizontal direction and small in the vertical direction, a control signal outputted from a fixed storage device 8 controls an output of a vertical direction filter 6a to be outputted from a switch circuit 9. On the other hand, when a change in the color and brightness is large in the horizontal direction and small in the vertical direction, the control signal outputted from the device 8 controls the output of a horizontal direction filter 6b to be outputted from the circuit 9. As a result, in a color TV signal supplied to an input terminal 1, when a change in a luminance signal and a color signal is large, the color signal led out at a terminal 2 has less mixture with a luminance signal component and with a color signal of an adjacent sampling point.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a Y/C (luminance signal/chrominance signal) separation device that separates a luminance signal and a color signal contained in a composite video signal, and in particular, Y that can perform the separation according to the change state of the luminance signal and color signal in the signal.
/C separation device.

[Prior Art] Video equipment such as television receivers and VTRs (video tape recorders) that receive color television signals are
It includes a functional section called a /C separation circuit.

A color television signal is a composite video signal in which a horizontal/vertical synchronization signal and a color burst signal are added to a signal in which a luminance signal and a color signal modulated by a color subcarrier are superimposed. In video equipment such as television receivers and VTRs, luminance signals and color signals need to be processed separately. A circuit provided for this purpose is a Y/C separation circuit.

The Y/C separation circuit separates and extracts a luminance signal and a color signal contained in such a color television signal.

FIG. 5 is a schematic block diagram showing the general configuration of a conventional Y/C separation circuit that separates a luminance signal and a color signal from an NTSC color television signal.

Referring to the figure, this Y/C separation circuit has an input terminal 111
Signal output terminal 2. Luminance signal output terminal 3゜A/D converter 4
．． 1 line delay circuit 5g and 5h, compensation delay circuit 5L
, vertical filter 6h.

It includes a horizontal filter 61 and a subtractor 10.

An NTSC color television signal is input to the input terminal 1. After being digitized by the A/D converter 4, the color television signal is applied to a vertical filter 6h and also to a 1-line delay circuit 5g.

The one-line delay circuit 5g delays the input signal by one horizontal scanning period IH, that is, one line. The output of the 1-line delay circuit 5g is given as is to the vertical filter 6h, and is further processed by the 1-line delay circuit 5h.
After being delayed by a line, it is applied to the vertical filter 6h.

The vertical filter 6h is a filter that is usually called a two-line comb filter, and separates color signals by calculating the color television signals of three vertically adjacent lines. The output of the vertical filter 6h is given to a horizontal filter 61 which is a bandpass filter.

The horizontal filter 61 removes the luminance signal component included in the color signal provided from the vertical filter 6h.

The output of the horizontal filter 6i is sent to the output terminal 2 as a color signal.
is derived and given to the input of the subtracter 10. Further, the output of the one-line delay circuit 5g is applied to the subtracter 10 via a compensation delay circuit 51.

The compensation delay circuit 51 includes a horizontal filter 6 in the subtracter 10.
This is a delay circuit for further delaying the output of the 1-line delay circuit 5g by the delay time in the horizontal direction filter 61 in order to input the signals of the same portion on the image from the 1- and 1-line delay circuits 5g to the same B5. Then, the subtracter 10 obtains a luminance signal by subtracting the output (color signal) of the horizontal filter 6i from the output of the one-line delay circuit 5g provided via the compensation delay circuit 51, and obtains a luminance signal from the output terminal 3.
Derived as follows.

Next, the principle of separation of color signals and luminance signals in this Y/C separation circuit will be explained.

In the NTSC color television signal, the frequency band of the color signal is set inside the frequency bandwidth of the luminance signal, and the frequency tsc of the color subcarrier, the horizontal scanning frequency f,
There is a relationship between Esc-<455/2)fH. This means that the color subcarrier is shifted by one period in two lines in the vertical direction of the screen. In other words, the period of the color subcarrier is 1/f, C in the horizontal direction of the screen, and 2 lines in the vertical direction.

Here, in the A/D converter 4, the color television signal has a sampling frequency f which is four times the color subcarrier frequency fffc.
s ("4 fsc)" and is sampled in phase synchronization with the color subcarrier. Therefore, from the A/D converter 4, one (
The luminance signal components and color signal components of points on the screen (hereinafter referred to as sample points) picked up at intervals corresponding to IJjtll/f5 (-1/4fgc) are sequentially output as digital signals.

Here, since the sampling frequency f is four times the color subcarrier frequency tsc, the phases of the color signals at every other sampling point in the horizontal direction are shifted by 180° from each other, and are located at every third sampling point. The color signals of the sample points have the same phase. Also,
As described above, since the color subcarrier period in the vertical direction of the screen is two lines, the phases of the color signals of vertically adjacent sample points are also shifted by 180° from each other. Further, the color signal is a signal obtained by orthogonally modulating a color subcarrier with two different color signal components C1 and c2.

Therefore, the types of color signal components of horizontally adjacent sample points whose color signal phases differ by 90 degrees from each other are different from each other.

FIG. 6 shows the phase arrangement of the color signal and luminance signal in the digital color television signal output from the A/D converter 4 as the phase arrangement of the color signal and luminance signal on the screen together with the arrangement of sample points. It is a diagram.

Referring to the figure, as can be seen from the above, the color signal components of the sample points indicated by O in the figure and the sample points indicated by O are of the same type C1, but their phases differ by 180°. Similarly, the color signal components of the sample point indicated by Δ and the sample point indicated by M are of the same type C2, but their phases differ by 180°. In other words, the signal of the sample point indicated by ○ is the luminance signal component Y
and the sum of the color signal components C1 is expressed as Y+C1, then the signal at the sample point indicated by 0 is expressed as Y-C1.

Similarly, if the signal at the sampling point indicated by Δ is expressed as Y+C2, the signal at the sampling point indicated by m is expressed as Y-C2.

Thus, for example, n (
m(m”1.) on the nm2.3.4...)th line.
2. In order to extract the color signal component from the signal of the sample point S (m, n), the sample point S (m, n) is extracted so that only the color signal component of the sample point S (m, n) remains. vertically adjacent sampling points S(m, n-1) and S(
Appropriate calculations may be performed on the signal at sample point S (m, n+1) and the signal at sample point S (m, n). Specifically, the calculation expressed by the following equation is performed. In the following equation, the symbols indicating the sampling points are assumed to directly represent the signal components of the sampling points.

-1/4 [S (m, n-1) +S (m, n
+1)-2S (m, n)] In other words, in order to extract the color signal of the target sample point (hereinafter referred to as the sample point of interest) S (m, n), the luminance signal component of the sample point of interest is The sample point S (m) adjacent to the sample point of interest is canceled by the luminance signal components of the two sample points adjacent to this sample point.

The signal component of n) multiplied by an appropriate coefficient may be subtracted from the signal component of the sample point of interest S (m, n) multiplied by an appropriate coefficient. This calculation is performed in the vertical filter 6h in the Y/C separation circuit shown in FIG.

In FIG. 5, a signal in which the signal component of each sample point output from the A/D converter 4 is delayed by one line period, a signal delayed by two line periods, and a non-delayed signal are input to the vertical filter 6h. This means that the signal component of each sampling point is applied to the vertical filter 6h in parallel together with the signal component of the sampling point one line before and the signal component of the sampling point one line after. therefore,
In the vertical filter 6h, one line delay circuit 5g is always used.
The above calculation is performed using the signal component output from the sample point as the signal component of the sample point of interest, and the color signal component is extracted. As a result, from the vertical filter 6h, the A/D converter 4'
The color signal components extracted from the signal components of each sample point are sequentially output.

Transformations that delay the signal component of one sampling point by one sampling period 1/4f, C and one horizontal scanning period 1H are, respectively,
2 Using conversion symbols z-1 and Z-J (It-910
), the transfer function Hv (Z) indicating the transfer function of the vertical filter 6h is expressed by the following equation.

Hv (Z) = (-1/4) (1-Z-') 2Here, Z-' -EXP (j2πf/4fsc) (f
represents the signal frequency. ) Since the vertical filter 6h has the above characteristics, even if the vertical filter 6h removes high-frequency luminance signal components having frequencies that are odd multiples of the color subcarrier frequency, it does not remove low-frequency luminance signal components. Not done. A horizontal filter is used to remove low-frequency luminance signal components that cannot be removed by such a vertical filter.

The horizontal filter is a band-pass filter that passes only the band centered on the color subcarrier frequency fse by performing the same operation on the signal components of several horizontally adjacent sample points.

Therefore, considering the case where the signal component of each sample point includes the low-frequency luminance signal component as described above, the Y
In the /C separation circuit, the output signal of the vertical filter 6h is input to the horizontal filter 61 in order to further remove the remaining luminance signal component from the signal extracted by the vertical filter 6h. Horizontal filter 61
is the signal component of the sample point S (m, n) of interest and the color so that the luminance signal component of the sample point S (m, n) of interest is canceled by the luminance signal component of the sample point neighboring horizontally. The type of signal component is the sample point of interest S(m, n
), and the sample point of interest S (
m, n), for example, sample point S
Appropriate calculations are performed on the signal components of (m-2,n) and S (m+2.n).

Therefore, in reality, the signal at each sampling point input to the horizontal filter 61 is transmitted over one sampling period (1/4 f
s c ), and the above-mentioned calculations are performed on these delayed signals (including signals that are not delayed). The transfer characteristic of the horizontal filter 6i is set such that its transfer function Hh (Z) is expressed, for example, by the following equation.

Hh (Z)= (-1/32)(1-Z-2)2(
1+Z-')2 (1+Z-') The color signal of each sample point output from the horizontal filter 61 is given to the color signal output terminal 2, and the subtracter 10 outputs the color signal of each sample point from the compensation delay circuit 5i. The color signal of the point is subtracted from the signal before separation. That is, the subtracter 10 performs the calculation expressed by the following equation in order to separate the luminance signal Y(m, n) of the sample point of interest.

Y (m, n) -8(m, n) -C(m, n
) In the above equation, S (m, n) represents the signal before color signal separation of the sample point of interest, and C (m, n) represents the color signal of the sample point of interest.

As a result, only the luminance signal of each sample point is sequentially output from the subtracter 10 and applied to the luminance signal output terminal 3.

[Problems to be Solved by the Invention] As described above, in the conventional Y/C separation circuit, the signal component of each sample point is Extraction of color signal components from is uniformly performed using vertical filters and horizontal filters with preset characteristics. As a result, the following problems arose.

The vertical filter performs calculations to cancel luminance signal components of vertically adjacent sample points and obtain an average value of color signal components. Therefore, in order to accurately extract only the color signal component from the signal component of the sample point of interest, the luminance signal components and color signal components of the sample point of interest and the sample points vertically adjacent to it must be as similar as possible. need to be. Therefore, the smaller the change in brightness and color in the vertical direction of the image, the more accurate the Y/C separation by the vertical filter can be.

In other words, in areas where vertical brightness changes and color changes are severe in an image, the brightness signal components and color signal components of vertically adjacent sample points are not similar, so the signal component of the sample point of interest and the vertical direction The luminance signal components of sample points adjacent to the sample point do not completely cancel each other, and the average value of the color signal components of the sample points adjacent in the vertical direction is also quite different from the true color signal component of the sample point of interest. Become. As a result, a luminance signal component is mixed into the extracted color signal component, or a color signal component of a vertically adjacent sample point is mixed into the extracted color signal component. If the color signal components are not extracted accurately in this manner, the luminance signal components will not be accurately extracted from the signal components of the sample points in such areas during the subtraction process for extracting the luminance signals. That is, a color signal component may be mixed into the luminance signal component that is finally output from the Y/C separation circuit, or conversely, a luminance signal component may be mixed into the color signal component. This phenomenon occurs when a dot-like brightness change appears on the playback screen.
This causes so-called dot interference, or causes so-called cross color, in which irregular color patterns appear on the playback screen, resulting in deterioration of meat quality.

On the other hand, the horizontal filter performs calculations to cancel luminance signal components of horizontally adjacent sample points and obtain an average of color signal components.

Therefore, in order to accurately extract only the color signal component from the signal component of the spout sample point, it is necessary to make the luminance signal components and color signal components of the sample point of interest and sample points horizontally adjacent to it as similar as possible. Must be. Therefore, the smaller the change in horizontal brightness and color of the image, the more accurate the Y/C separation by the horizontal filter.

Therefore, in areas where horizontal brightness changes and color changes are large in the image, Y/C by the vertical filter is
A similar phenomenon occurs when separation occurs.

As can be seen from the above, a Y/C separation circuit using a horizontal filter is not suitable for a composite video signal derived from an image area where horizontal brightness and color changes are large and vertical brightness and color changes are small. Y/C separation by a vertical filter is inappropriate for composite video signals derived from image regions with small horizontal brightness and color changes and large vertical brightness and color changes.

However, in conventional Y/C separation circuits, Y/C separation is always performed by both horizontal and vertical filters having preset characteristics. For this reason, conventional Y/C separation circuits provide accurate Y/C separation for composite video signals derived from image areas with large brightness and color changes.
C separation is not performed. Therefore, the conventional Y/C separation circuit has the problem of causing image quality deterioration such as dot interference and cross color in the reproduced image.

An object of the present invention is to solve the above-mentioned problems, and to perform accurate Y/C separation even for composite video signals derived from images containing areas with large luminance and color changes. /C separation device.

[Means for solving: 1llIi] In order to achieve the above object, the luminance signal and chrominance signal separation device according to the present invention adaptively switches filters according to two-dimensional correlation of images to separate video signals. a sampling means configured to separate a luminance signal and a color signal component from each other, and for sampling a video signal including a luminance signal and a color signal to extract video signals of a plurality of points arranged in a grid on a screen; , a distribution state extraction means for extracting the distribution state of color and brightness in a minute area on the screen based on the video signals of the plurality of points extracted by the sampling means; and a filter according to the extraction output of the distribution state extraction means. a control signal generating means for generating a filter switching control signal for switching the filter; and a control signal generating means that switches the filter in response to filter switching control No. 13 generated by the control signal generating means, and images of a plurality of points extracted by the sampling means. and means for separating and extracting a color signal and a luminance signal from the signal.

[Operation] Since the luminance signal color signal separation device according to the present invention is configured as described above, the color and luminance distribution state within a minute area on the screen is extracted, and the color and luminance distribution state within this extracted minute area is extracted. A control signal for switching the filter is derived based on the luminance distribution state. Therefore, unlike the conventional method, the filter for separating and extracting the color signal and the luminance signal from the video signal is switched depending on the color and luminance distribution state of the image.

Embodiment FIG. 1 is a schematic block diagram showing the configuration of a Y/C separation circuit according to an embodiment of the present invention.

Referring to the figure, this Y/C separation circuit has input terminal 1 and
A color signal output terminal 2, a luminance signal output terminal 3, an A/D converter 4, one-line delay circuits 5a and 5b, a vertical filter 6a, a horizontal filter 6b, and a horizontal/vertical filter 6C. A concentration distribution extraction circuit 7 and compensation delay circuits 5c and 5d.

5e and 5f, a fixed storage device 8, a switch circuit 9, and a subtracter 10.

In the figure, an NTSC color television signal is applied to an input terminal 1. This color television signal is sampled by the A/D converter 4 at the same sampling frequency fs (-4fsc) as in the prior art. The color television signal sampled by the A/D converter 4 (hereinafter referred to as a sampled signal) is delayed by one line by a one-line delay circuit 5a, and further delayed by one line by a one-line delay circuit 5b. .

The sampled signal output from the A/D converter 4 is applied to the l-line delay circuit 5a, and at the same time, it is applied to the vertical filter 6.
g, horizontal and vertical directions, are also given to the filter 6C and the density distribution extraction circuit 7. 1 by the 1-line delay circuit 5a
The line-delayed sampling signal is given to a vertical filter 6a, a horizontal filter 6b, a horizontal/vertical filter 6c, and a concentration distribution extraction circuit 7 as a signal of the sample point of interest. Further, the output of the l-line delay circuit 5b, that is, the sampled signal delayed by two lines, is applied to the vertical filter 5a, the horizontal/vertical filter 6c, and the concentration distribution extraction circuit 7. That is, the vertical filter 6a,
Horizontal/vertical filter 6C1 and concentration distribution extraction circuit 7
, signals of three vertically adjacent sample points are input simultaneously.

The vertical filter 6a has a transfer function Cv, for example.
(Z) is Cv (Z)= (-1/4)(1-2-
j, ) 2, and as explained in "Prior Art", the color of each sample point is determined by performing a predetermined operation on the signals of three vertically adjacent sample points. Extract the signal.

The horizontal vertical filter 6C includes both a vertical filter and a horizontal filter. horizontal vertical filter 6
First, a predetermined calculation is performed on the three signals inputted to c by a vertical direction filter, and the color signal of each sample point is extracted from the vertical direction.

This extracted color signal is further passed through a horizontal filter and then given to the compensation delay circuit 5e as the output of the horizontal and vertical filter 6C. In other words, the horizontal/vertical filter 6C is a filter that has characteristics of both a vertical filter and a horizontal filter. The transfer function Chv(Z) representing the characteristics of the horizontal and vertical filter 6C is, for example, expressed by the following formula (%). Filter 6
The brightness and color distribution state in a minute region on the image centered on the sample point of interest in the horizontal and vertical filters 6C and 6C is extracted.

The horizontal filter 6b horizontally samples the signal input from the one-line delay circuit 5a at a sampling period (1/4 ftc).
A predetermined operation is performed on the signals of multiple sample points obtained by delaying by an integer multiple of the sample point, that is, the signals of the sample points horizontally adjacent to the sample point of interest, and the signals of the sample point of interest. Extract the color signal. A transfer function Ch (Z) representing the characteristics of the horizontal filter 6b is expressed, for example, by the following equation.

Ch (Z) = (-1/4) (1-Z-2) 2 The output of the vertical filter 6a is sent to the horizontal filter via the compensation delay circuit 5C, and the output of the horizontal filter 6b is sent to the horizontal The outputs of the vertical filters 6C are respectively applied to switch circuits 9 via compensation delay circuits 5e.

The switch circuit 9 includes a compensation delay circuit 5cs 5a.

The outputs of the vertical filter 6a, horizontal filter 6b, and horizontal/vertical filter 6C, which are provided through 2
and is applied to the subtracter 10.

The fixed memory bag g18 outputs a control signal to the switch circuit 9 according to the color and brightness distribution state in the micro image area centered on the spout sample point extracted by the concentration distribution extraction circuit 7 described above. The internal connection state of the switch circuit 9 is switched.

Specifically, when the change in color and brightness in the minute image area is large in the horizontal direction and small in the vertical direction, that is, when the vertical correlation is strong, the control signal output from the fixed storage device 8 is The connection state within the switch circuit 9 is set so that the output of the directional filter 6a is output from the switch circuit 9.

Conversely, if the changes in color and brightness in the minute image area are small in the horizontal direction and large in the vertical direction, that is, when the horizontal correlation is strong, the control signal output from the fixed storage device 8 is transmitted to the horizontal filter 6b. The internal connection state of the switch circuit 9 is set so that the output is output from the switch circuit 9.

Further, if there is no significant correlation in the color and brightness distribution states in the micro image area in either the horizontal or vertical directions, that is, if both the horizontal and vertical correlations are low, the output is output from the fixed storage device 8. The control signal set sets one internal connection state of the switch circuit 9 so that the output of the horizontal/vertical filter 6c is output from the switch circuit 9.

Therefore, the switch circuit 9 outputs the color signal of each sample point extracted by a filter suitable for changes in the luminance signal and color signal in a minute image area centered on each sample point. As a result, when there is a large change in the luminance signal and color signal in the color television signal applied to the input terminal 1, the color signal delivered to the color signal output terminal 2 is more likely to be mixed with luminance signal components than in the past. This results in less contamination of point color signals.

Now, the output of the switch circuit 9 and the output of the one-line delay circuit 5a, that is, the sampled signal delayed by one line, are input to the subtracter 10 via the compensation delay circuit 5f. Therefore, as in the prior art, the subtracter 10 subtracts the color signal output from the switch circuit 9 from the sampling signal delayed by one line input via the compensation delay circuit 5f, thereby increasing the luminance of each sample point. Extract the signal and output terminal 3
Derived as follows. However, as mentioned above, the color signal input to the subtracter 10 is separated more accurately than before, so the color signal component mixed into the luminance signal extracted by the subtraction process in the subtracter 10 is different from the conventional one. decrease compared to Therefore, the luminance signal derived to the output terminal 3 contains less color signal components than the conventional one.

As described above, according to this embodiment, the color signal and the luminance signal are accurately separated and extracted from the composite video signal. As a result,
The occurrence of dot interference and cross color in reproduced images is reduced compared to the conventional method.

Here, the compensation delay circuit 5f is arranged such that the two signals inputted to the subtracter 10 at the same time correspond to the same sampling point.
The sampled signal delayed by one line, that is, the signal of the sample point of interest, is passed through a vertical filter 6m and a horizontal filter 6m.
b, horizontal/vertical filter 6c, compensation delay circuit 5c,
The time is delayed according to the delay time caused by 5d and 5e.

Similarly, compensation delay circuits 5c, 5d, and 5e are each provided to input fJ numbers corresponding to the same sample point to switch circuit 9 at the same time. Therefore, the compensation delay circuits 5c, 5d, and 5e each have a vertical filter 6.
a, the outputs of the horizontal filter 6b and the horizontal/vertical filter 6C are delayed by a time corresponding to the delay time in the concentration distribution extraction circuit 7 and the fixed storage device 8.

Next, the configuration and operation of the concentration distribution extraction circuit 7 and the fixed storage device 8 will be explained in more detail.

FIG. 2 is a schematic block diagram showing an example of the configuration of the concentration distribution extraction circuit 7. As shown in FIG.

Referring to the figure, this density distribution extraction circuit 7 includes delay circuits 21a to 21f that delay the signal of each sampling point by 172 cycles of the color subcarrier, that is, 2 sampling cycles, and a minute An average value calculation circuit 22 that derives the average value of signals of nine sample points in the image area, and a compensation delay circuit 23a.
~ 231, and outputs a digital signal of “0” or “1” according to the magnitude relationship between each of the signals of the nine sample points and the average value derived by the average value calculation circuit 22. Comparators 24a to 24i are included.

The outputs of the comparators 24a to 24i are all stored in the fixed storage device i!
8 is input.

Next, the operation of this concentration distribution extraction circuit 7 will be explained in detail with reference to FIGS. 2 and 3. FIG. 3 is a diagram showing the arrangement of nine sample points in the minute image area on the screen.

Referring to FIG. 2, the signal manually inputted from the one-line delay circuit 5a, that is, the signal of the nth line including the sample point of interest, is delayed by two sampling periods by delay circuits 21c and 21d. Ru. Therefore, with reference to Figure 3,
The output of the delay circuit 21c, the output of the delay circuit 21c, and the output of the delay circuit 21d are respectively supplied with the signal of the sample point S23 located two sample points to the left of the sample point of interest S2□, the signal of the sample point of interest S22, and the sample of interest. The signals of the sample point S21 located two sample points to the right of the point S2□ appear simultaneously.

Similarly, the delay circuits 21a and 21b control the sample points S24, S2□, S2. The sample point SII%SI□ located on one line of S5. is obtained, and the delay circuit 21e,
21f, the sample points S24, S2□, S2. 1
Sample points 531qS32 and Si3 below the line are obtained.

That is, signals of nine sample points within the micro image area centered on the sample point S22 of interest are obtained simultaneously. The signals of the sample points Sll to SSZ thus obtained are given to the average value calculation circuit 22, and their average values S and 2 are derived by calculation. The calculation for determining the average value SAV is expressed, for example, by the following equation, when the densities of the nine sample points are each expressed by Sll to sag.

5Av-(4S22 +2 (S+ 2 + 32
++82 s +Ss 2) +S+ + 10s,
s+Ss++Sas)/-8 The average value SAv thus obtained is given to one input of each of the comparators 24a to 241.

On the other hand, the other input of each of the comparators 24a to 24i has
Signals from each of the nine sample points Sll to SSS are applied via compensation delay circuits 23a to 23i, respectively, and compared in magnitude with the average value SAY. Comparator 24a
~24L outputs a digital value "0" when the average value SAY is larger than the signal of each sample point, and otherwise outputs a digital value "1".

Note that the compensation delay circuits 23a to 23L each have a comparator 24.
This circuit is for compensating for the delay time in the average value calculation circuit 22 so that the signals of nine sample points in one minute image area are manually input from a to 24i at the same time as their average value SAv.

Since the signal of each sample point is a digital value (hereinafter referred to as density) representing the color and brightness, the average value SAY is a value representing the average color and brightness of the micro image area including the sample point of interest. . Therefore, each of the digital values output from the comparators 24a to 24L is divided by 2 with respect to the average value SAY.
It represents the density of each of the nine sample points in the micro image area that has been converted into a value. Therefore, the arrangement (pattern) of digital values simultaneously output from comparison 'a24a to 241 is as follows:
It represents the density distribution of a minute image area including the sample point of interest. That is, the color and brightness distribution of the minute image area including the sample point of interest are extracted as 9-bit digital data.

FIG. 4 is a diagram illustrating some patterns of digital values representing the densities of the nine sample points. In each table in the figure, the positional relationship of the nine digital values is the same as the positional relationship of the corresponding sample points on the screen.

'1fiJ diagram (a) is an example of a pattern that is considered to have a strong vertical correlation, Figure 4 (b) is an example of a pattern that is considered to be a strong horizontal correlation, and Figure 4 (C) is an example of a pattern that is considered to have a strong vertical correlation. Examples of patterns that are considered to have no significant correlation are shown below.

As can be seen from FIGS. 3 and 6, the types of color signal components are the same at the nine sample points, but the phases of the color signal components at vertically adjacent sample points are 180 degrees from each other.
The phases of the color signal components of horizontally adjacent sample points are also shifted by 180° from each other. Therefore, the above 9
When the vertical correlation between color and luminance is strong in a micro image region containing three sample points, the digital image of three sample points in each vertical direction is The value pattern has a regularity (0 and 1 are arranged alternately) reflecting the phase difference of the color signal.

). In addition, when the digital values of the three sample points in each vertical direction are the same with the above-mentioned regularity, as in some other patterns illustrated in FIG. 4(a), It is considered that almost no color is included in the vertical direction in the image area and the brightness is similar.

Similarly, if the horizontal correlation between color and brightness is strong in the micro image area, the pattern of digital values of the three sample points in each water direction will be as shown above, as illustrated in FIG. 4(b). It is thought that there is a certain regularity.

In addition, if the horizontal correlation and vertical correlation of color and brightness are not significant in the micro image area, as illustrated in FIG. It is considered that the above regularity cannot be found among the digital value patterns of the three sample points in any of the directions.

The determination of the two-dimensional correlation of the image with respect to the digital value pattern derived by the density distribution extraction circuit 7 is not limited to the above determination example, and a pattern for which Y/C separation using a vertical filter is considered to be preferable is determined to have a "strong vertical correlation". Y/C by horizontal filter.
Patterns for which separation is considered desirable are included in "patterns considered to have strong horizontal correlation," and patterns for which Y/C separation by both horizontal and vertical filters are considered to be desirable are included in "patterns with strong horizontal correlation and vertical correlation." It is sufficient if it is included in "patterns that are considered to be unremarkable."

Next, an example of the configuration and operation of the fixed storage device 8 that receives the output of the concentration distribution extraction circuit 7 having the above configuration will be described.

The fixed storage device 8 is, for example, a read-only memory such as a ROM (read-only memory), and stores the digital value pattern of the nine sample points output from the concentration distribution extraction circuit 7, that is, the 9-bit digital data. Receive it as an address signal. Each address of the fixed storage device 8 has a numerical value 0, 1 . . . expressed in 2 bits in advance. and 2 are written as data.

Specifically, in the micro image region containing the sample point of interest,
A value 0 is written in advance to the address corresponding to the 9-bit digital data when the vertical correlation between color and brightness is considered to be strong, and the address corresponding to the 9-bit digital data when the horizontal correlation is considered to be strong. The number 1 is written in advance in the address corresponding to 9-bit digital data when neither the horizontal correlation nor the vertical correlation is considered significant. 2 is written in advance.

The fixed storage device 8 then reads the data written therein from the address corresponding to the 9-bit digital data provided by the concentration distribution extraction circuit 7 and outputs it to the switch circuit 9.

Therefore, the fixed storage circuit 8 outputs 2-bit data indicating the type of color and brightness distribution state in the minute image area including the sample point of interest as a control signal for controlling the internal connection state of the switch circuit 9.

When the fixed storage device 8 is configured as described above, the switch circuit 9 receives the digital data “00” from the fixed storage device 8 indicating that the vertical correlation between brightness and color in the minute image area is strong, The directional filter 6a outputs the extracted color signal, and upon receiving the digital data "01" indicating that the horizontal correlation is strong, the horizontal filter 6b outputs the extracted color signal, and neither the horizontal correlation nor the vertical correlation is significant. It is configured to output the color signal extracted by the horizontal and vertical filter 6C in response to the digital data "10" indicating .Furthermore, it is also possible to use a so-called PLA (programmable logic array) as the fixed storage device 8. be.

In the density distribution extraction circuit 7 in the above embodiment, the average value SAY of the 9 sample point densities, which is the threshold value when binarizing the densities of the 9 sample points included in the micro image area, is calculated. However, the calculation method does not necessarily have to be the method used in the above embodiment. For example, the average value calculation expressed by the following formula, SAY '' (S+ + +S+ 2 +Fz s +
S2 + ten s22 + s2. ．． +s, 10s, 2+S
3 s ) /9 Alternatively, in consideration of ease of division in digital signal processing, a simplified average value calculation expressed by the following equation may be used.

SAY = (S+ + 10S+ 2 +S+ s 182 ++s,, 2 +s23 +S31 +S32
15si)/8 In addition, the threshold value when binarizing the signal of each sample point is set to a constant value SA for each of the nine sample points within one minute image area.
Instead of V, the threshold value may be changed for each sample point. For example, the m-th sample point S on the n-th line among the nine sample points. The threshold value sAV m n when converting into 21iff may be a value obtained as a result of the calculation expressed by the following equation.

SAvmn = (Sl l+S+ 2 +S1a +
S2 ++s22 +S23 +s, l +s, 2
+Sss Smn)/8 In the above example, the density of the sample point within the micro mountain image area was binarized and the density distribution within the micro image area was quantified.
The density distribution within the micro image area may be quantified by dividing the density of each sample point into three or more levels.

In the above embodiment, the number of sample points extracted from the micro image area was 9 in total, 3 horizontally and 3 vertically, but 1 sample point in total, 5 horizontally and 3 vertically.
The number of sample points to be extracted may be five, and the number of sample points to be extracted does not necessarily have to be nine.

In the above embodiment, the color television signal was sampled at a frequency four times that of the color subcarrier, but if the sampling points to be extracted are at a frequency that is arranged in a grid on the screen, then the frequency is four times that of the color subcarrier. Sampling may be performed at other sampling frequencies.

[Effects of the Invention] As described above, according to the present invention, the luminance signal component and the color signal component can be extracted based on the distribution of luminance and color within a minute area on the screen that includes the sampling point for which the luminance signal component and color signal component are to be separated and extracted. The two-dimensional correlation on the screen is determined, and filters are used to separate and extract the luminance signal and chrominance signal from the video signal, so that the mutual channel between the finally extracted luminance signal and chrominance signal is This reduces leakage and suppresses the occurrence of dot interference, which causes image quality deterioration.

[Brief explanation of drawings]

FIG. 1 is a schematic block diagram showing the configuration of a Y/C separation circuit according to an embodiment of the present invention, FIG. 2 is a schematic block diagram showing an example of a specific structure of the concentration distribution extraction circuit in FIG. 3
The figure shows the positions on the screen of the nine sample points extracted by the concentration distribution extraction circuit shown in Figure 2, and Figure 4 shows the nine sample points extracted by the concentration distribution extraction circuit shown in Figure 2. Figure 5 is a block diagram showing an example of a conventional Y/C separation circuit; Figure 6 is an NTSC color TV set. FIG. 3 is a diagram showing a row of signals on a screen of sampling points obtained by sampling a color subcarrier signal at a sampling frequency four times as high as the color subcarrier frequency. In the figure, 1 is an input terminal, 2 is a color signal output terminal, 3 is a luminance signal output terminal, 4 is an A/D converter, 5a, 5b, 5
g, 5h are one-line delay circuits, 6a and 6h are vertical filters, 6b is a horizontal filter, 6C is a horizontal/vertical filter, and 5c. 5d*5e, 5f, and 51 are compensation delay circuits, 7 is a concentration distribution extraction circuit, 8 is a fixed storage device,
9 is a switch circuit, and 10 is a subtracter. In addition, in the figures, the same reference numerals indicate the same or corresponding parts. Taku3 (2) (0) (b) Me40

Claims (1)

[Scope of Claims] A luminance signal/chrominance signal separation device that separates a luminance signal and a chrominance signal from a video signal sampled at a predetermined frequency by adaptively switching a filter according to the two-dimensional correlation of the image. a sampling means for sampling a video signal including a luminance signal and a color signal to extract video signals of a plurality of points arranged in a grid on a screen; and a video signal of a plurality of points extracted by the sampling means. distribution state extraction means for extracting the distribution state of color and brightness in a minute area on the screen based on the signal; and generating a filter switching control signal for switching the filter according to the extraction output of the distribution state extraction means. and a means for separating and extracting a color signal and a luminance signal from the video signal by switching the filter according to the extraction output of the control signal generating means. Device.