JPH029287A - Y/c separator for composite video signal - Google Patents

Abstract

PURPOSE:To realize picture reproduction with high quality and high definition by using two frame memories and using them for the control picture movement detection a noise reduction means and for mixing switch control for inter-line Y/C separation result and inter-frame Y/C separation result. CONSTITUTION:An output signal of the 2nd frame memory 20 is given to a noise reduction circuit 60 and output signals of the 1st, 2nd, 3rd and 4th field memories 15, 16, 23, 24 are given to a movement detection circuit 80 for movement detection. The circuit 80 detects the movement in the picture based on inter-1 frame difference signal and inter-2 frame difference signal, gives a coefficient KN to a noise reduction circuit 60 and coefficients KY, KC are given to a mixing changeover circuit 50. A luminance signal YL obtained from the inter- line Y/C separator circuit 30 and a luminance signal YF obtained from an inter-frame Y/C separator circuit 40 are mixed in a ratio in response to the coefficient KY or switched to generate a final luminance signal Y.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to a Y/C separation device for separating an NTSC color television composite video signal into a luminance signal and a carrier color signal.

Conventional technology NTSC horn type 1 υ video 1. ) The current Y/C separation circuit separates the luminance signal and carrier color signal from the signal.

Frequency filter or <L,' (by 2 filters,
There is. However, since the frequency band of the carrier chrominance signal and the frequency band of the luminance signal partially overlap, the separation method using the analog filter described above cannot perform complete Y/'C separation, and cross-color dot crawl interference occurs. Image quality deterioration could not be avoided.

Recently, the demand for high-quality, high-definition images has increased, and on the other hand, costs have decreased due to advances in semiconductor technology, so there is no hope that digital signal processing technology can be applied to practically solve the above problems. Don't get big
.

Therefore, one frame memory is used, and the digitally converted input composite video signal is put into this frame memory and delayed by one frame, and the delayed +11 composite video signal is input; Y using
/ CII) Consideration is being given to separating them.

Problems to be Solved by the Invention However, the digital Y/C female device described above is
If the k-image signal represents a still image, the signal has a strong relation to frame 111011, so it will operate properly.

In general, there is movement in images, and when there is movement, the correlation between frames becomes weaker, so Y/C separation remains insufficient.

On the other hand, 1 (co-produced) (the source signal processing circuit has a noise reduction circuit called a noise reducer for improving the S/N ratio, and this noise reduction circuit is also constructed digitally using frame memory. An attempt is being made.

Therefore, this invention not only introduces a frame memory for Y/C division and a frame memory for noise reduction, but also makes it possible to use the frame memory described in 1 for image motion detection. It provides a Y/C separation device.

Means for Solving the Problems A Y/C separation device for a composite video signal according to the present invention 1゛t(
J, the first and second field memories are connected in cascade to each other to receive one input composite video signal.
It is constructed by cascading a first frame memory, third and fourth field memories which are delayed by a frame.

The second frame memory is cascaded to the first frame memory.
a memory, a first subtraction means for calculating the difference between the input composite video signal and the output one-component video signal of the second frame memory, and outputting a two-frame difference signal; at least one means for subtracting between the output composite video signal of the first field memory of the first frame memory and the output subtraction video signal (+; sign) of the third field memory of the second frame memory; 1 by taking the difference
a second Φ reduction means for outputting an inter-frame difference signal of one field before the field; 1.22 inter-frame difference signals and at least one
a motion detecting means that receives two one-frame difference signals as input and detects [7 opposites of motion in an image represented by a half-coincidence video signal; converts the two-frame difference signals to a level corresponding to the degree of the detected motion; 11 stages of noise reduction subtracted from the heated video signal and the 1. input composite video signal and the adjacent line. sJ, line-to-line Y/C separation means 1 input 1 (co-production (using the source signal and the output composite video signal of the first frame memory) to separate the luminance signal and carrier color signal using the source signal inter-frame Y/C separation means for separating the lli common signal and the carrier color signal, and a luminance signal and carrier color signal by inter-line Y/C separation;
Mixing switching means that mixes or E7J-changes the luminance signal and carrier color signal resulting from inter-frame Y/C separation, depending on the degree of detected motion, and outputs the final luminance lχ (1:) and carrier color signal. As a one-frame difference signal used for motion detection, in addition to the one-frame difference signal 21 fields before, the input 1 (joint image signal and the first frame memory The one-frame difference signal that is the difference between the output one pleasure signal and the second
An inter-frame difference signal, which is the difference between the input video signal and the output video signal of the frame memory, may be used. Furthermore, 1
It is preferable to use two one-frame difference signals obtained by iLI from video signals that are shifted by one line from each other as the one-frame difference signals before the field.

Preferably, in the second motion detection means, motion detection used for at least mixing and switching of luminance signals and noise reduction is performed using both the one-frame difference signal and the two-frame difference signal. Further, motion detection for mixing and switching the carrier color f14 uses only the difference signal between two frames.

Operation: The degree of image movement is detected by the motion detecting means shown in FIG.

The detected degree of motion is utilized in noise reduction means and mixing switching means for Y/C separation.

The noise reduction means is a composite operation (the noise components included in the image signal have no relationship between frames, and since the phase of the carrier color signal is inverted between one frame, it becomes the same phase between two frames). Noise reduction is achieved by subtracting the difference signal between two frames from the input composite video signal, but if there is movement in one image, not only noise components but also noise components are added to the difference signal between two frames. A 1x component representing motion is also included.Therefore, the level of input 1 (the 2-frame difference C11 to be subtracted from the joint image signal) is changed depending on the degree of detected motion.

Inter-line Y/C separation means and inter-frame Y/C separation means are provided. In the case of images with no or little movement, the correlation between video signals is stronger between frames than between lines. On the other hand, in the case of images with a lot of movement, the more intense the movement, the weaker the correlation between frames of the video signal.
The first check between lines is relatively stronger. Therefore, the inter-line Y/C/)') separation results and the inter-frame Y/C separation results are mixed and switched in accordance with the degree of detected motion to obtain the final luminance signal and carrier color signal.

DESCRIPTION OF THE EMBODIMENTS N T S CIM is a multiple transmission signal of a combined edge image signal, a luminance signal Y, and a carrier color signal C. As shown in Figure 1, the luminance signal Y
has a frequency band of ~4.2 MHz R1 degree.

Transport (!! Jl; i No. C has frequency f (3,579
545MHz) C It is a signal obtained by balanced modulation of color subcarriers with the same /l'/ phase different by 90° by two color difference signals, and has a frequency of 9'S.
Centered on e, high frequency restriction 0.5 MHz, low frequency restriction 1.5
It has a frequency band of MHz. Y/C separation device (iN
This is to separate a luminance signal Y and a color signal (hereinafter simply referred to as a color signal) C from a TSC signal. Therefore, the input signal of this Y/C separator is NTS
The CI combined video signal and output signals are a luminance signal Y and a color signal C.

The Y/C separation device described here is a digital signal processing circuit. N T S Cm joint image signal is A/D converted 2:
'+-C is converted into, for example, an 8-bit digital signal and applied to a Y/C separator. A/D changelfi'A
rr clock (+i is a horizontal domei signal or color hearth +-f, ; has a frequency 4f8o that is four times the frequency rSo of the J-color scheme subcarrier with PLL locking, and the input j
The FJl image f5- is sampled at this frequency 4f and converted into a digital signal. Therefore, the digital Y/C separator basically operates in synchronization with the e clock signal of frequency 4f. When one cycle of this clock signal is 1'd-1/4f, as shown in Fig. 2C, one horizontal scanning period in one field is IH-910Td, and one vertical scanning period is l.
V-2G25H.

FIG. 3 shows the basic configuration of the Y/C separation device.

Y/C separation takes advantage of the fact that brightness If (=') has a strong correlation between frames or lines (between adjacent horizontal scanning lines). In some cases, the correlation between luminance signals is stronger between frames than between lines.On the other hand, in the case of moving images (this is called a video), the more intense the movement, the stronger the correlation between frames of luminance signals. Therefore, when there is intense movement, the correlation of luminance signals is 114 pairwise stronger between lines than between frames.

The same can be said about the strength of inter-line correlation and inter-frame correlation of color signals related to movement on the screen, but the variation in the strength of correlation is smaller for single color signals than for luminance signals.

Movement in an image is the reflection of a certain pixel at a certain point in time.
It is detected by taking the difference between the image signal and the video signal at the same pixel one frame or two frames before that point (the signal representing this difference is called the one-frame difference signal or the two-frame difference signal). . Generally, when the movement is relatively slow, the level of the 1-frame difference signal and the 2-frame difference signal (No. 3) is small; as the movement becomes more intense, the level of the 1-frame difference signal and the 2-frame difference signal also increases. Become.

On the other hand, the noise component included in the video signal is +1 between frames.
Therefore, noise can be reduced by multiplying the difference signal between two frames by an appropriate coefficient I smaller than 1 and then subtracting it from the original video signal. Since the phase is inverted by 1806, 1
The color signal C remains in the inter-frame difference signal, which is not desirable, and the two-frame difference signal is subtracted from the original video signal.

However, if there is movement in the image, a difference signal component based on the movement appears in the difference signal between two frames. Therefore, it is necessary to control the above-mentioned coefficients according to motion detection, and in some cases, set the coefficients to 0 so that no noise reduction operation is performed.

In FIG. 3, in order to obtain fIi signal delayed by one frame for inter-frame Y/C separation, to obtain a two-frame difference signal used for noise reduction operation, and to obtain a one-frame difference signal for motion detection. and two frame memories 10.20 are provided and connected in cascade to obtain a two-frame difference signal. The first frame memory January 0 is constructed by cascading a first field memory EIJ15 and a second field memory 1B. Similarly, the second frame memory 20 is formed by cascading a third field memory 23 and a fourth field memory 24 into h11s. More detailed information on these memories (:Xl or later will be described).

A noise reduction circuit 60 is connected between the NTS C+, U video signal output terminal and the first frame memory IO. The input composite video signal is sent to the line-to-line Y/C separation circuit 30. The signal is applied to an interframe Y/C separation circuit 40, a noise reduction circuit 60, and a motion detection circuit 80. First frame memory IO for interframe Y/C separation
The output signal is sent to the interframe Y/C separation circuit 40. The noise reduction circuit 60 includes a second frame memory 20.
An output signal of is given. Furthermore, the first . 2nd ° 3rd and 4th field memory 15.1
The output signals of G, 23°24 are input to the motion detection circuit 80.

The motion detection circuit 80 detects motion in the image based on [λ] of the one-frame difference signal and the two-frame difference signal, and calculates the motion into three types of coefficients each representing the culm of the motion.
occurs. The coefficient KN is given to the noise reduction circuit 60 as the coefficient to be multiplied by the above-mentioned YC two-frame difference signal. Other coefficients KY.

Ko is applied to a mixing switching circuit 50.

As described above, the strength of the line-to-line f-correlation and the frame-to-frame correlation of the luminance signal changes depending on the movement of the image. The luminance signal YL obtained from the inter-line Y/C separation circuit 30 and the inter-frame Y/C separation circuit 30
The luminance signal YF obtained from the C separation circuit 40 is converted into a coefficient K representing the degree of darkness of the motion given from the motion detection circuit 80.
The mixing QJ switching circuit 50 creates the final luminance signal Y by mixing or switching at a ratio according to . This mixing switching circuit 50 also applies to color signals as well.
The output color signal CL of the inter-line Y/C separation circuit 30 and the output color signal CF of the inter-frame Y/C separation circuit 40 are
The coefficient I(c) given from the motion detection circuit 80 is mixed or switched at a ratio corresponding to the final color to create and output the :sign C.

A part of the configuration shown in FIG. 3 is shown in more detail.

is shown in Figure 4.

NTSC composite video signal input terminal and noise reduction circuit GO
A line memory 31 for storing composite video signal data for one horizontal scanning line is provided between the two. The line memory 31 delays the input signal by one horizontal scanning period H and outputs the delayed signal. The first field memory 15 in the first frame memory lO.

A line memo IJII, a 2GI H memory 12 that stores video signal data for 261 horizontal lines, and a line memo IJII.
It is configured by cascade-connecting the memory I3.

The second field memory 16 stores input video signal data in two
Output with a delay of 62H. Two line memories 31
A subtraction circuit 61 of a noise reduction circuit 60 is connected between and 11. These line memories 31. Reference numeral 11 is for Y/C separation between lines and forms part of one circuit 30. The third field memory 23 of the second frame memory 20 is the field memory 2 of 262H.
1 and a line memory 22 are connected in cascade, and the fourth field memory 24 is a 262H field memory 22.

The line-to-line Y/C separation circuit 30 has an input composite video signal (
This is designated as L2), the output signal of the line memory 31 (this is designated as Lo), and the output signal of the in-line memo 'Jll (this is designated as Ll) are input. The signal L is the base, and the signal L1 is one line before, and the signal L2 is one line after.

The level of the signal L2 is multiplied by 1/4 by the coefficient multiplier 32 (signal L/4), and the signal is increased. is multiplied by 1/2 in the coefficient multiplier 33 (signal L/2), and the signal L1 is multiplied by 1/4 in the coefficient multiplier 34 ((;i, /4). Then, in the subtraction circuit 35, the signal L/ Signals L2/4 and L/4 are subtracted from 2. By this subtraction operation.

The luminance signals are canceled depending on the degree of inter-line correlation of the luminance signals. For example, if the correlation between the luminance signals in the signals t- and Lol L t is 1, the luminance signals are completely canceled out. Since the phase of the color signal is inverted between lines, the signal L2/4 and the signal L2/4 are the same. /2 and the color signal L1/4 are added in the subtraction circuit 35, so that the result is 1.
You can obtain twice the level of color signal. The output signal of the subtraction circuit 35 is:

A first chrominance signal bandpass filter with f as the center frequency and a passband approximately equal to the frequency C waveband of the chrominance signal shown in FIG.
, abbreviated as) 37, and is output as a color signal CL resulting from line-to-line Y/C separation.

-)i, the output signal of the subtraction circuit 35 is fed to the subtraction circuit 38 through a second chrominance signal band-pass filter (abbreviated as BPF2) 36 whose center frequency is "sc" and whose passband is narrower than that of BPF1. The subtraction circuit 38 is inputted with a signal indicating a cram school, and in this circuit 38,
The color signal, which is the output signal of BPF2, is the signal. By subtracting 3 from 3, the luminance signal remains, which is the Y between the lines.
/ C't> brightness due to distance (output as Ir number YL).

The interframe Y/C separation circuit 40 includes a line memory 31.
The output signal F (this is the same as the signal Lo described in section 1, but the symbol F is used because it is the basis for inter-frame Y/C separation s7) and the output signal F of the first frame memory 0. The output fli signal (1: F) from one frame before is input. The signal F has its level multiplied by 1/2 in the coefficient unit 4I (signal Fo/2), and the signal F is multiplied by the coefficient 8:v42. 1/
It is doubled (signal F, /2). Then, by subtracting the signal F1/2 from the signal Fo/2 in the subtraction circuit 43, the luminance signals are canceled out according to the degree of interframe correlation of the luminance signals. Since the color signal has a phase shift between frames, the subtraction circuit 43 converts the signal F o
/2 and Fl/2 color signals are added as a result, resulting in a 1x color signal. This color signal passes through the BPF 145 and is output as an inter-frame Y/C separated color signal CF.

On the other hand, the subtraction circuit 44 receives the signal F.

Since the color signal which is the output of the subtraction circuit 43 is subtracted from the chrominance signal, the luminance signal remains, and the luminance signal YF of the Y/C separation between frames is
is output as

The subtraction circuit 81 selects the image No. 1 (Lo and F) that is no ζ.

), the output signal of the second frame memory 20, which is the signal of two frames before, is subtracted from the two-frame difference signal D.
Output YC. This subtraction circuit 81 is a noise reduction circuit G.
0 and the motion detection circuit 80.

The inter-frame signal DY which is the output signal of this circuit 81
C is applied to the amplitude adjustment circuit 70 of the noise reduction circuit 60 and is also input to the main section 90 of the motion detection circuit 80.

The motion detection circuit 80 also includes subtraction circuits 82, 84, and 85. The subtraction circuit 82 receives the basic video signal (L
, Fo) is subtracted from the output signal of the first frame memory 0 (same as Fl), which is a signal one frame before, to output a one-frame difference signal DY2. This pre-stage one-frame difference signal DY2 is input to the main section 90 of the motion detection circuit 80.

The subtraction circuits 84 and 85 are for outputting four one-frame difference signals of one field before. The subtraction circuit 84 is the first
From the output video signal of field memory 5, the third
The output video signal of the field memory 23 is subtracted to output a one-frame difference signal DYOI. From the point of view of the above-mentioned base image (difference signal (L and F)), the output video signal of the first field memory 5 is the video signal of one field before, and this is the video signal of a different field in interlaced scanning. The signal DYOI, which represents the difference between the video signal of the previous frame and the video signal of the previous frame, is obtained using the video signal of the previous field as a basis. is ′y in the subtraction circuit 84.
The video signal one line after the video signal received is I-5.
,available. That is, the subtraction circuit 85 converts the output video signal of the 201H memory of the first field memory 10 to the 262H memory 21 of the third field memory 23.
The output video signal of is subtracted by 3 to obtain the 1-frame difference signal DYO2 of the previous field. 1 of these
DYOI and DYO2 are difference signals at 1'/position shifted by one line, and are input to the main section 90 of the motion detection circuit.

If necessary, further reduce 39 circuit 8 as shown by the dashed line:
3 is provided, and the inter-frame difference signal DYI of the previous frame is illuminated. That is, in the 2 subtraction circuit 83, the output 1 good image signal of the second frame memory 20 is subtracted from the output video signal of the first frame memory 10, and the resulting 1 frame difference signal DYI is sent to the motion detection circuit main section 90. Enter.

The main part 90 of the motion detection circuit 80 is a luminance signal YL.

A circuit for creating a coefficient 1<y (and 1-KY) for mixing 17J switching of YF and a coefficient 17K for mixing switching 9 fish signals CL, CY. (and a coefficient 1-Ko) and a circuit that creates a coefficient KN used for noise reduction.

An example of a circuit for creating a coefficient K, 1-I<Y for mixing and switching luminance signals is shown in FIG. This circuit has a two-frame difference signal DYC.

All 1-frame difference signals DYI, DY2゜DYOI
, DYO2 are input. The difference signal DY+ is input only when the subtraction circuit 83 is provided.

Between 2 frames? i (= DYC includes a luminance signal component and a color signal component that represent the movement between two frames.2
If there is no movement at all between frames, the level of this two-frame difference signal DYC is 0 (however, it includes noise).

In order to extract only the luminance signal component representing motion from this difference signal DYC, a first low-pass filter (this is called LPF1) is used.
) 101 is provided. The passband of this LPF is set to a band below the level that blocks the color (li).The signal passing through the LPF is transmitted to the absolute value circuit 1.
(12 is converted to absolute iIQ, and the secondary three comparators CIAC
It is given to IB and CIC respectively.

These comparisons 'A:'r CI A - CI C are for distributing the level of the luminance signal component representing motion into four levels. In this specification, the comparator generally outputs "0" when the level of the input signal is less than a threshold value (or threshold value) given to the comparator, as shown in FIG. If the value exceeds the value, a “1” output will be generated. Comparison & 4 CI A, CI B, C
IC niha'c reso huj' AICIA, IB, IC are given. These basic values have a relationship of LA<IB<IC. The outputs of these comparators CIA-CIC are applied to a decoder D).

The decoder D1 responds to input signals representing the three comparison results (1) to 2-bit, soto (MSB and L) as shown in Figure 1O.
SB) output. When the movement between two frames is intense, the output of the comparator CIC becomes "1", and in this case, the output of the decoder D1 becomes "11". On the other hand, if the movement between two frames is extremely small or
The output of A becomes "0", and the output of decoder D1 becomes "o".
In this way, the 2-bit output f, No. 1 of the decoder DI represents the degree of motion represented by the luminance signal between two frames.

Figure 1θ shows decoder D2 and decoder D4, which will be described later.
The operations are also commonly expressed.

The one-frame difference signal includes, in addition to a luminance signal component and a color signal component representing movement between one frame, a color signal component of a still image portion tL. The color signals have a phase difference of 180' between frames. 1-frame difference signals DYI, DY2.

DYOI and DYO2 are in charge of 1t! I device 111. 112.
113. 114, respectively A, B, C, D
Maximum KI after being multiplied KI KI
Kl (input to 1α circuit 115. Coefficient units 111 to 114 are for weighting four types of one-frame difference signals, and their coefficients "AKI=DKl" are preset to optimal values through experiments. The maximum value circuit 115 is a circuit that selects and outputs the one having the highest level among the input signals.It goes without saying that the coefficient unit 111 is not required when the above-mentioned subtraction circuit 83 is not provided. The signal representing the most intense movement selected by the maximum value circuit 115 is input to a second low-pass filter (referred to as LPF2) 11[i.The passband of LPF2 is narrower than the passband of LPF1 described above. It is designed to completely exclude the 1-fish signal component and pass only the luminance component with a lower frequency than it. This is because the luminance signal component representing the largest movement during one frame that has passed through the LPF2 is converted into an absolute value by the absolute value circuit +17, and is then converted into an absolute value by the three comparators C2A, C2B,
Enter into C2C.

The comparators C2A, C2B, and C2C also classify the level of the luminance signal component representing movement into four levels, and J! Direct vote value 2A2B, 2C (2A<2B<2C)
are given 0 These comparisons 2:'vC2A -C
The output of 2C (“1” or “0”) is input to decoder D2. The operation of decoder D2 is as shown in the same figure, and generates a 2-bit (N13B, LSB) output.

Each 2-bit output signal of the two decoders DI and D2 is input to a coefficient generation circuit 103. The coefficient generation circuit +03 generates the code (Table 4) of the input signals from the decoders DI and D2 (“
1" or "0"), it outputs a signal representing IF y to a predetermined coefficient, and its operation is the first
This is shown in Figure 2.

As mentioned above, decoder D1. ．． The output of D2 is 2 each
It shows the degree of movement represented by the luminance signal component in the inter-frame difference signal DYC and the i-frame difference signal, where output "11" indicates the most violent motion, and output "00" indicates the most gentle motion or still W image state. I'm playing 7 times.

The coefficient KY for the luminance signal takes a value between 0 and 1, and is set to a larger value as the movement becomes more intense. In this example, there are four possible values for the coefficient: 0. 0.375. 0
．． 625 and 1 in advance. 12th
In the figure, the magnitude relationship between the output of the decoder D1 and the output of the decoder D2 is defined as 'I'l+, and the coefficient KY is determined accordingly. Needless to say, the method of determining the coefficient KY is not limited to this, and various methods can be considered. The coefficient b1-Ky is determined as the value obtained by subtracting the coefficient KY from 1.

An example of a circuit for creating coefficients K, 1-K, for mixing and switching color signals is shown in FIG. Only the two-frame difference signal DYC is input to this circuit.

The difference signal DYC between two frames is passed through a third bandpass filter (
This is given to BPF3). BPF3 is BP
Although it has the same center frequency as F2, it has a passband that is narrower than F2, so that the color signal component representing motion in the two-frame difference signal DYC can be reliably extracted.

The output color signal component of the BPF 3121 is converted into an absolute value by an absolute value circuit 122 and input to two comparators C3A, C3B+, and so on. These comparators C3A and C3B have the basic monk ful1.
3A and 3B (3A < 3B) are given, respectively. The reason why the levels of color signal components are classified into three levels by two comparators is that mixing and switching of one color signal does not require detailed control due to the visual characteristics between words.

The output signals of comparators C3A and C3B are input to decoder D3. The operation of decoder D3 is shown in FIG.

This decoder D3 also outputs a 2-bit code representing a very large value with rapid movement.

The 2 and t output horns of the decoder D3 are next
Enter 3. The operation of the coefficient generation circuit 123 is shown in the thirteenth factor. The larger the value represented by the 2-bit code signal input from the decoder D3, the more intense the movement, and a larger value of the coefficient Kc is set.

Coefficient K. It also takes a value between θ and 1, and in this implementation ρ is 0, 0.5. and 1, and this value is selected according to the output of decoder D3. 1 to K
The coefficient 1-1 with the value of c subtracted is also determined.

An example of a circuit for creating a coefficient KN used for noise reduction is shown in FIG. This circuit includes a two-frame difference signal DYC and all one-frame difference signals DYI,
DY2 DYOI and DYO2 are input.

The two-frame difference signal DYC is converted into an absolute value by an absolute value circuit 131 and then inputted to a maximum value circuit 132.

The output signal of the absolute value circuit 13+ includes a luminance (+'+ sign component signal component) representing the movement between two frames.

etc.) J゛1 frame difference opening lid DYI, DY2DYOI
DYO2 is multiplied by A, B, C, and D by coefficient multipliers 141, 142, 143, and 144, respectively, and then input into the value circuit 145 to a maximum of 2 K2. These coefficient units are also used to give effective weights to the difference signal between frames, so the vi
, the number is optimal (directly set) by experiment.

The output signal of the LPF 214G is a luminance signal component representing the movement between one frame. This is converted into an absolute value by an absolute value circuit 147.

It is input to the maximum value circuit 132.

The maximum value circuit 132 selects and outputs one of the two input signals inputted from the absolute value circuits 131 and 147, which has a higher level. This circuit 132 output signal Lt base'<'vt4 A, 4 B, 4 C (4
A<4B<4C) is input to three comparators C4A, C4B, and C4C, respectively, and is divided into four stages. The outputs of these comparators C4A-C4C are sent to decoder D
Enter 4. The operation of decoder D4 is shown in FIG. 10 mentioned above. The 2-bit code output of decoder D4 is input to coefficient generation circuit 133.

The operation of coefficient generation circuit 133 is shown in FIG.

The larger the value represented by the output code of decoder D4, the more intense the movement. The coefficient KN is set to a smaller value as the movement becomes more intense. In this example, the coefficient I/N is 0.75, 0
, G25. It is fixedly set in four stages of 0.375 and O, and one of these is selected according to one input code. When there is little movement or it is a still image (input code is "00"), the coefficient KN is J'
N is also set to a large value of 0.75, and the coefficient is set to 0 when the movement is most intense (the input code is "11").

FIG. 8 shows an example of an amplitude one-week reduction circuit 70 included in the noise reduction circuit 60. As described above, this circuit 70 receives the two-frame difference signal DYC as input.

It outputs a wave subtraction signal to be added to the subtraction circuit 6.

The amplitude adjustment circuit 7a includes an I, 7J converter 72. This switch 72 connects any one of the three inputs a, b, and c to I and Ij according to the code signal output from the decoder D5, and outputs the resultant output to the subtraction circuit 61. be. As will be explained in detail later, depending on the degree of motion represented by the two-frame difference signal DYC, input a is selected when the motion is jjj (or gentle), input is intermediate, and input C is selected when it is severe. be done.

The two-frame difference signal DYC includes a luminance signal component and a color signal component (when there is no movement, these components are 0) representing movement between two frames, as well as a noise component. This difference signal DYC is input to the coefficient multiplier 71. The coefficients 2 and 71 are determined by -1-KN. The difference signal DYC multiplied by KN by the coefficient multiplier 7 is applied to the switch 72 as an input a.

The two-frame difference signal DYC is also input to the sign discrimination circuit 73, and its No. 71 (positive or negative) is discriminated, and this discrimination result is determined as a threshold value? ] Input to the signal generator circuit 4. Threshold? ] A threshold value A1 or A2 (inclusively represented by A for convenience) output from a threshold value generation circuit G3 or 64, which will be described later, is input to the encoder circuit 4. ? In the co-coding circuit 74, the code determined by the co-encoder circuit 73 is assigned to this input threshold value A, and the input threshold value A is given to the switch 72 as one input.

Threshold generation circuit G3. A threshold signal output from a threshold generation circuit provided separately from G4 may be input to the encoding circuit 74.

Input C of the switching circuit 72 is a 0 level signal.

The difference between the two frames (+i number DYC is also given to the absolute value circuit 75 in order to detect the degree of movement, and after being converted into an absolute value, it is input to the comparators CA and CB, respectively. Comparison 2
'j Different thresholds Al, A2, and Bl as base values for CA and CB.

B2 respectively generating circuits 63. G4 and G5.
H is provided. These threshold values are At<Bl.

The relationship is At <A2 and Bl <82. Threshold Al
.

Either one of A2 is selected by the changeover switch 7G,
Either one of the threshold values Bl and B2 is selected by the changeover switch 7.
7 and applied to comparators CA and CB, respectively. The changeover switches 76 and 77 are interlocked and can be switched manually. This allows the S/N correction rate to be changed depending on the image quality.

The output signals (“0” or “1”) of the comparators CA and CB are input to the decoder D5. FIG. 15 shows the operation of decoder D5 and the switching operation of switch 72 which is controlled by the output of decoder D5. 2 frame difference signal D
When the motion of the image detected based on YC is rapid, the output of the decoder D5 becomes "11", and the input C is selected by the switch 72. In this case, since the output of the switch 72 is O, the original video signal input to the subtraction circuit 61 passes through the subtraction circuit 61 as is. In other words, the noise reduction operation cannot be performed. When the movement is intermediate, the output of decoder L)5 becomes "10" and switching a:;72
Input mode is selected. Therefore, it is output from the threshold encoding circuit 74. Constant amplitude (equal to threshold A)
A threshold signal having the same sign as the difference signal DYC is applied to a subtraction circuit 61, and this threshold signal is subtracted from the original video signal. This reduces the noise reduction effect of a certain fl jfl by iF7. When the movement is gentle or absent, the output of the decoder D5 becomes "00", and the input a is selected by the switch 72. In this case, the difference signal between two frames that has been subtracted by KN (j'') by the coefficient unit 71 is given to the subtraction circuit 61, and this signal is subtracted from the original video signal, so that the noise reduction effect is adjusted according to the degree of movement. You can expect it.

FIG. 16 shows an example of a specific configuration of the mixing switching circuit 50. This circuit 50 includes a circuit for luminance signals and a circuit for color signals.

The circuit for the brightness If signal includes a coefficient unit 51 to which the brightness signal YL resulting from line-to-line Y/C separation is input, and an inter-frame Y/C circuit.
a coefficient unit 52 into which the luminance signal YF obtained by /C separation is input;
Add the output signals of these coefficient multipliers 51 and 52! , ), and the output signal of the adder circuit 53 becomes the luminance signal Y, which is the output of the Y/C=i device. The coefficient of coefficient a:i 51 is , and the coefficient of coefficient multiplier 52 is 1-KY. Signal YL is related to U'z a:i
51, the signal YF is multiplied by I(y).
It is multiplied by (1-KY) in 52. Coefficient 1: fIKY takes a larger value as the motion in the image (in the luminance signal) J, and the detected motion) becomes more intense, and the coefficient 1-Ky takes a larger value as the motion becomes smaller. Therefore, the more intense the movement, the more the inter-line Y/C separated luminance signal YL is mixed, and when , is 1, this signal YL is output as the luminance signal Y. Conversely, the smaller the movement, the more the inter-frame Y/C separated luminance signal YF is mixed, and in the case of an i'i'l' still image, the signal YF is output as the luminance signal Y.

The operation of the mixing switching circuit for color signals is exactly the same as for luminance signals. In other words, this circuit has a line-to-line Y/
A coefficient unit 54 that multiplies the color signal CL resulting from C separation by Kc, and multiplies the color signal CF resulting from interframe Y/C separation by (1-K.).
47. from a coefficient multiplier 55 and an adder circuit 5G that adds the output signals of these coefficient multipliers 54 and 55. The output signal of the adder circuit 56 becomes the color signal C which is the output of this Y/C separation device.

Coefficient K. takes a larger value as the movement in the image (movement detected based on the color signal) is more intense.

The coefficient 1-Kc takes a larger value as the movement is smaller. Therefore, the more intense the movement, the more the inter-line Y/C separation color signal CL
If the mixing ratio of 2Ko is 1, this signal CL is output as the color signal Y. Conversely, the smaller the movement, the more the inter-frame Y/C separated color signal CF is mixed, and in the case of a still image, the signal CF is output as the color signal C.

Effects of the Invention As described above, in this invention, two frame memories are used, and these frame memories are shared for generation of a difference signal between two frames for the noise reduction means and for motion detection. Therefore, the 1114 configuration can be simplified. In addition, the motion of the 1+-IIi image represented by the combined video signal is detected, and the degree of the detected motion is calculated based on the noise reduction means' noise reduction means' results for the noise reduction means, the Y/C between lines, and the Y/C between frames. Mixed switching between C minute #J knot 1. Since it is used for control, the movement f′,! Appropriate noise reduction (reduction and Y/C separation can be achieved, high quality, high definition images II) can be achieved. Furthermore, the first stage 1
Since the Y/C separation is performed using the video signal in the frame, there is no phase delay on the time axis of the system, and there is no need to consider the fηII difference between the two images and the sound.

[Brief explanation of the drawing]

FIG. 1 shows the frequency bands of the luminance signal and color signal, and FIG. 2 is a diagram for explaining the scanning of one field screen. FIG. 3 is a block diagram showing the basic (111 configuration) of the Y/C separation device. FIG. 4 is a block diagram showing the Y/C separation device in more detail. FIG. 5 shows a motion detection circuit for the luminance signal component, FIG. 6 shows a motion detection circuit for the color signal component, and FIG. 8 shows a motion detection circuit for noise reduction. FIG. 8 is a block diagram showing a specific configuration of an amplitude adjustment circuit in the noise reduction circuit. FIG. 9 is a block diagram for explaining the operation of the comparator. 10 and 11 are for explaining the operation of the decoder used in FIGS. 5 to 7. FIGS. Figures 5 to 7
It shows the operation of the coefficient generation circuits used in each figure. FIG. 15 shows the operation of the decoder used in FIG. 8. FIG. 16 is a circuit diagram showing the configuration of the mixing switching circuit. 10, 20...Frame memory +5. IG, 23.24...Field memory. 30...Y/C separation circuit between lines. 40...Inter-frame Y/C separation circuit 50...Mixing switching circuit 60...Noise reduction circuit 80...Motion detection circuit. 81, 82.83.84.85...Subtraction circuit. 90: Main part of the motion detection circuit. Above 'Jl,';'F Applicant's representative: 1 attorney, Denki Home Electronics Co., Ltd., Patent attorney: Kenji Ushiku (1 other person) Figure 9 Basics ≧ Input crab output basis Straw Naoku Hisakata; Output Fig. 0 Fig. 11 Decoder D3 (old king's name) Fig. 13 1' 4 signs of snoring (4 color words) Fig. 4, tVV product reading (), Itakui-san φ菟) Figure 15 Decoder D5 Figure 16 Y Candle engraving word C-8 signal

Claims (1)

[Scope of Claims] A first frame memory configured by cascade-connecting a first and a second field memory, the first frame memory delaying the input composite video signal by one frame, and the third and fourth field memories. a second frame memory cascaded to the first frame memory; configured by cascading memories; and a second frame memory cascaded to the first frame memory; at least one subtraction means for obtaining an inter-frame difference signal, the output composite image of the first field memory of the first frame memory; a second subtraction means for outputting a one-frame difference signal of one field before by taking the difference between the signal and the output composite video signal of the third field memory of the second frame memory; the two-frame difference signal; a motion detection means for detecting the degree of motion in an image represented by the composite video signal by receiving at least one one-frame difference signal as an input; converting the two-frame difference signal to a level corresponding to the detected degree of motion; noise reduction means for subtracting the input composite video signal from the input composite video signal; line-to-line Y/C separation means for separating the luminance signal and carrier color signal using the input composite video signal and the composite video signal of the line adjacent thereto; interframe Y/C separation means for separating a luminance signal and a carrier color signal using the signal and an output composite video signal of the first frame memory, and a luminance signal and a carrier color signal by line-to-line Y/C separation; A composite device comprising: mixing and switching means for mixing or switching a luminance signal and carrier color signal resulting from inter-frame Y/C separation, respectively, depending on the degree of detected motion, and outputting a final luminance signal and carrier color signal; Y/C separation device for video signals.