JPH01280970A - A ghost removal device, a television receiver, a tuner and a video tape recorder equipped with the same, and a ghost detection device included in the ghost removal device - Google Patents

Abstract

PURPOSE:To eliminate ghost before detection by applying Kepstrum analysis to a signal prior to video detection so as to obtain a delay time of a ghost component and generating a signal cancelling the ghost component. CONSTITUTION:An output signal of a transversal filter 501 is converted into a digital signal for an optional period by an A/D conversion circuit 502 and stored in a waveform memory 503. A tap gain data in a transversal filter 501 corresponding to the ghost parameter obtained by a subtraction circuit 505 is read from a tap gain memory 506 to correct it and the tap gain data after correction is inputted to a tap gain memory 506. Each tap gain data stored in the tap gain memory 506 is converted into an analog signal by a D/A conversion circuit 507 and fed to a tap gain adjustment terminal of a tap amplifier corresponding to the transversal filter 501. Thus, ghost component is reduced or eliminated.

Description

Detailed Description of the Invention [Field of Industrial Application] The present invention relates to a ghost removal device, a television receiver, a tuner, and a video tape recorder equipped with the same, and a ghost detection device included in the ghost removal device. It is.

[Conventional technology]

In a television receiver, when radio waves coming directly from the transmitting antenna (desired waves) and radio waves reflected from buildings, etc. (reflected waves) are simultaneously received by the receiving antenna, an image of the desired waves appears on the screen. In addition, the reflected waves cause images to shift and appear as ghosts.

Ghosts are a major factor in deteriorating picture quality in television receivers, and various methods have been used to remove and prevent ghosts. One example is a ghost removal device using a transversal filter in the video band.

The basic operation of such a conventional ghost removal device using a transversal filter will be explained using FIG. 10 and FIG. 10A.

FIG. 10 is an explanatory diagram of the operating principle of a conventional coast removal device using a transversal filter.

In the figure, 1100 is an input terminal, 1101 is a tapped delay line, 1102 is a tap amplifier, 1103 is an adder, 1104 is a subtracter, 1105 is an output terminal, and τ is the delay amount as the tap interval of the tapped delay line 1101. (delay time), which is determined by the highest frequency component included in the input video signal.

FIG. 10A is a signal waveform diagram of the main part of FIG. 10.

Please refer to FIG. 10 and FIG. 10A. When a video signal in which an in-phase ghost component with a delay time of 3τ and an amplitude A is added to the desired signal as shown by a is input to the input terminal 1100.
Tap amplifier (C3) 11 connected to the third tap from the input terminal 1100 side of the tapped delay line 1101
02 (hereinafter referred to as tap gain) is assumed to be A. At this time, since the ghost compensation signal indicated by a is obtained at the output of the adder 1103, by subtracting the ghost compensation signal a from the input signal a at the subtracter 1104, the ghost indicated by C is output to the output terminal 1105. A signal with components removed can be obtained.

There are various algorithms for automatically determining the tap gain of a tap amplifier to cancel the ghost component by detecting the location of the ghost component (delay time from the desired signal), but from the viewpoint of stability, IE・E.E. Transactions on Consumer Electronics Volume C.E. 26. August 1980,
Pages 629 to 635 (IEEE, Tra
Consumer Elect
ronics, Vol, CE-26.

August, 1980. p, 629-p, 63
The correlation algorithm discussed in 5) is commonly used.

FIG. 11 is a block diagram showing a configuration example of a conventional ghost removal device employing a correlation algorithm.

In the figure, 100 is an input terminal, 104 is a transversal filter, 105 is a D/A conversion circuit, 106 is a tap gain memory, 107 is a subtraction circuit, 108 is an arithmetic circuit, 109 is a differentiation circuit, and 110 is an A/D conversion circuit. circuit, 111
1 is an output waveform memory, 112 is an output terminal, and 113 is a reference waveform memory.

A video signal input from the input terminal 100 is output from the output terminal 112 after ghosts are removed by the transversal filter 104 .

The output signal of the transversal filter 104 is differentiated by a differentiating circuit 109, then converted to a digital signal by an A/D converter circuit 110, and stored in an output waveform memory 111. In 10 days, the arithmetic circuit performs the correlation calculation shown in the following equation using the data (y,) in the output waveform memory 111 and the reference waveform data (r,) that does not include ghost components and is stored in the reference waveform memory 113 in advance. I do.

Zi=tx Σ r K41 ・)' K
--(1) Here, α is the positive constant subtraction circuit 107, and old data (data representing the tap gain) stored in the tap gain memory 106 C,
Output signal Z of the arithmetic circuit 108 from OLa.

to obtain new tap gain data C,new (H designates the i-th tap amplifier with respect to the order as shown as 01 to C4 in FIG. 10).

C・“IIW=C,ota Z,
(2) After the D/A conversion circuit 105 converts the new tap gain data C, n- into an analog signal, the transversal filter 104 (more specifically, the first
Gain adjustment terminals of tap amplifiers C6 to C4 in Figure 0)
By supplying the transversal filter 10 to
4. The ghost component in the output signal that has passed through step 4 is reduced. By repeating this operation, the ghost component is gradually reduced and finally removed.

Here, in FIG. 11, the meaning of differentiating the output signal of the transversal filter 104 using the differentiating circuit 109 will be explained.

FIG. 11A is a signal waveform diagram of the main part in FIG. 11. Please refer to FIG. 11 and FIG. 11A.

Generally, ghost component detection involves avoiding signal parts that may constantly fluctuate, such as pictures, and picking up signals such as the vertical synchronization signal that always appear at a constant period, and then detecting the ghost component for the leading edge of this vertical synchronization signal. This is done using the leading edge of the vertical synchronization signal where it is technically easy to do so. (IEE Transactions on Consumer Electronics Volume CEE 24, August 1978, pp. 267-271)
ionon Consumer Electronic
s, Vol, CE-24.

1978, p. 267-p. 271)) Figure 11A (a
), the vertical sync signal and the ghost component superimposed on it (
The leading edge of the vertical sync signal and its ghost component)
It is assumed that such a signal is output from the transversal filter 104. Then, in order to detect the delay time τg and amplitude A of the ghost component, the vertical synchronization signal and the ghost component are differentiated by a differentiating circuit 109, and the result is shown in FIG.
With the pulse waveform shown in b), if the correlation with a similar pulse waveform without ghost components (FIG. 11A (C)) stored in advance in the reference waveform memory 113 is determined,
The correlation value shown in FIG. 3(d) is obtained, and in the process of calculating the correlation value, the delay time τg of the ghost component is determined in the arithmetic circuit 108.

Once the delay time τg is determined, the transversal filter 6
Since we know which order position in 01 to adjust the gain of the tap amplifier, we can slightly modify the gain of that amplifier accordingly. This process is repeated for each incoming vertical synchronization signal to gradually reduce the ghost component and eventually reduce it to zero.

Furthermore, as a specific example of the ghost removal device, there is
The one described in JP-A-90620 is known.

[Problem to be solved by the invention]

As mentioned above, conventional ghost removal devices focus on the leading edge portion of the vertical synchronization signal in the video signal, detect the delay time of the ghost component, and correct the tap gain accordingly. The frequency of correction of the tap gain is limited to once per field (1/60 second), and there is a problem in that it takes a long time to completely remove ghosts.

In addition, in conventional ghost removal devices, ghost components are detected in the video signal after detection, so as stated in [Television Society of Japan Technical Report, TEBS 73-4, 1981, pp. 77 to 82]. However, in order to obtain a sufficient ghost component removal effect, it was necessary to employ ideal synchronous detection (
In envelope detection and pseudo-synchronous detection, the performance against orthogonal and anti-phase ghosts deteriorates due to distortion during detection. ) The purpose of the present invention is to significantly reduce the length of time required to complete ghost removal, and to provide a synchronous detection method that is ideal as a detection method in order to obtain a sufficient ghost component removal effect. An object of the present invention is to provide a ghost removal device that does not require the adoption of a ghost removal method, a television receiver, a tuner, and a video tape recorder equipped with the ghost removal device, and a ghost detection device included in the ghost removal device.

[Means for Solving the Problems] In order to achieve the above object, the present invention provides a ghost removal device,
A transversal filter receives and outputs a video signal, a tap gain memory stores the tap gain of each tap amplifier included in the filter, and performs cepstrum analysis on the video signal that has passed through the filter. cepstrum analysis means for detecting a delay time of a ghost component of picture information included in a video signal with respect to the picture information; and means for correcting tap gain data stored in the tap gain memory according to the detected delay time; means for removing ghost components from the video signal passing through the filter by controlling the gain of each tap amplifier in the filter according to the modified data.

In other words, in order to achieve the above object, the ghost removal device according to the present invention performs cepstrum analysis on a signal before video detection (for example, a video intermediate frequency signal), determines the delay time of the ghost component, and removes the ghost component accordingly. By generating a canceling signal, ghosts are removed before detection.

Furthermore, the above purpose is to perform cepstral analysis on the video signal after video detection, rather than the video signal in the IF stage before video detection, to determine the delay time, amplitude, and phase of the ghost, as a video signal containing a ghost, and to calculate the delay time, amplitude, and phase of the ghost. This is accomplished by generating a ghost compensation signal and eliminating ghosts.

When performing cepstrum analysis on a video signal at the 1F stage before image detection, even though the delay time and amplitude of the ghost can be determined, the phase cannot be determined, so detailed ghost compensation cannot be expected. .

[Effect]

In short, the first point of the present invention is to detect the delay time (and magnitude) of the ghost component of the picture signal included in the video signal by performing cepstrum analysis on the video signal before image detection at the IF stage. Therefore, a method for detecting this delay time will be described below.

A signal X(t) as a desired wave that has passed through a certain route,
A signal (
Let us consider a method for measuring the time difference T when a reflected wave (reflected wave) reaches a certain point with a time difference T. This can be thought of as a case where the signal x(t) is reflected by a surface with a reflectance r and is mixed with a delayed signal T, so the signal y(t) mixed with both can be expressed as follows. be done.

y(t)=x(t)+r−x(t−T) ・−(3
) If x(t) is a pulse-like signal with a short duration, the delay time T can be easily determined as the time interval between two pulses appearing in the waveform, but if x(t) is a continuous signal such as a video signal, That doesn't happen when it comes to traffic lights. However, if we pay attention to each frequency component that forms the waveform, the reflected wave undergoes a phase rotation proportional to the frequency based on the time delay.
Since this overlaps with the original waveform, the frequency spectrum of the overlapped waves must undergo periodic changes due to phase rotation. Therefore, by examining the frequency spectrum of the observed signal y (t), the delay time T can be found.

Therefore, by performing Fourier transform on both sides of 2(3), the following equation is obtained.

Y(f) =X(f)(1+ r exp (-j 2
tt f T)) (4) Here, Y(f) and X(f> are respectively y(t
).

xD). Calculating the power spectrum of y(L) from this results in the following.

l Y(f) l” = lX(f)l”((1+r CO3(2πfT))”
+r"sin" (27Cf T))=lX(
f) l” (1+r”+2 r cos (2πf T
) )...(5) In other words, l YCf'') l'' makes the magnitude of the power spectrum lX(f) l'' of the original signal wave along the frequency axis with a period of 1/T. If r and I
If X(f)l'' does not vary with frequency, then I
T can be found by measuring the period of change of Y(f)12 depending on the frequency. However, in this case IX(f)
I'' may vary with frequency.

If the first and second terms on the right side of equation (5) are separated, T can be found since the second term is not related to the original signal x (t). This separation is possible by taking the logarithm of equation (5). That is, by taking the logarithm,
The product on the right side has the form of a sum, as shown in the following equation.

1 logl Y(f) l ”= logl X(f)
l ”+ log (1+r”+2 r cos (2
πf T) )・・・・・・(6) Therefore, looking at this equation (6), the second term on the right side is a periodic function, and the first term is the power spectrum of the signal source x(t). Therefore, if we consider the logarithmic power spectrum expressed by equation (6) as a waveform along the frequency axis and perform the Fourier transform of that waveform, the Fourier transform of the second term becomes a line spectrum, whereas the first term becomes a line spectrum. The Fourier transform of the term results in another form representing the shape of the power spectrum of x(t). That is, if we write the Fourier transform of equation (6) as an equation, C(q)=F(logl Y(f)l 2)=F(lo
gl X(f)l" ) +F'(log(1+r"
+2 r cos(2tt f T) ])...
・(7) This is the spectrum of a periodic function whose second term on the right side has a period of 1/T.

C(q) defined by equation (7) is the Fourier transform of the logarithmic power spectrum, and the square of this absolute value can be called the power spectrum of the logarithmic power spectrum, but it is not the power spectrum of a time function. , since this cannot be called a spectrum, it is called a cepstrum.

In this way, even when the power spectrum IX(f)l'' of the desired wave X(1) changes depending on the frequency like a television signal, after taking the logarithm of the above equation (5),
When the frequency f is Fourier transformed, the above equation (7) is obtained, and the second term on the right side of the equation (7) becomes the spectrum of a periodic function whose period is l/T, so the delay time T of the reflected wave is
(For more details, see "Introduction to Digital Signal Processing" by Kido, Marukubi, p. 152-p. 156).

As described above, if the delay time T of the ghost component of the picture information included in the video signal is determined in this manner, the ghost component can be removed.

The second point of the present invention is to perform cepstrum analysis on the video signal after video detection, rather than the video signal before video detection at the IF stage, to obtain a fine-grained ghost compensation signal using a transversal filter.
The objective is to detect the following three parameters related to ghosts.

(1) Delay time (T) (2) Size (D/U ratio, amplitude ratio of reflected wave and desired wave) (3) Phase (φ) Below, by cepstral analysis of the video signal,
It will be briefly explained that the above three parameters can be detected.

(1) Delay time (T) This explanation has many points that overlap with the previous explanation, but for convenience of explanation, the explanation will be repeated.

FIG. 12 is an explanatory diagram of the principle of ghost delay time detection using cepstrum analysis. For simplicity, Figure 12(a)
The explanation will first be made using a pulse (sin x / x) having a flat amplitude frequency characteristic within the band (fc) shown in -(1).

(sin x /
If y(t) is y(t), y(t) is expressed by the following equation.

)'(t)=x(t) 1r −x(mu−T)
...-(3) Fourier transformation 11jY(
f) and the power spectrum IY(f)l'', Y(f)=X(f) ・ (1+r e−””)...
...(4) IY(f)l" = IX(f)l"(1+r"+2
r cos 2 πf T) --(5)(5)
From the formula, the power spectrum IX(f)12 of x(t) is flat within the band, as shown in Figures 12(a)-(2), so the power spectrum IX(f)12 of 1(1) is
f)l'' becomes a periodic function with a period of 1/T within the band due to the second factor of equation (5) (Figure 12(b)-(2))
.

Therefore, by Fourier transforming the equation (5) with respect to the frequency f, a line spectrum is generated at the position of the horizontal axis T, so that the delay time T can be detected.

(Fig. 12(b)-(3)) The above describes the case where a ghost is detected using a (sin x / x) pulse, but when using an image signal of an arbitrary pattern, (3) x(t) in the equation becomes an image signal of an arbitrary picture, and its power spectrum IX(f
)I” is also not flat within the band.

As a result, if we just perform Fourier transform on equation (5), (
5) It is generally difficult to detect the periodicity of two ghostwritten terms.

Therefore, in order to separate the two factors in equation (5), (5)
Taking the logarithm of the equation, togl Y(f)l”= logl X(f)l”+
log(10r”+2 r cos2 πf T)...
...(6), and the second term of equation (6) has a period of 1 with respect to the frequency f.
Since it is a periodic function of /T, the ghost delay time T can be detected by Fourier transforming equation (6).

(2) Size (D/U ratio) FIG. 13 shows the relationship between the cepstrum C(q) obtained by Fourier transforming the above equation (6) and the ghost amplitude r.

From this figure, the value of cepstrum C(q) is proportional to the ghost amplitude r, so the size of the ghost (D/U ratio)
can also be detected.

(3) In phase television broadcasting, video signals are transmitted using the vestigial sideband method shown in FIG. 14(a).

(a) 0 to 1.25 MHz... Double sideband (b) 1.25 to 4.2 MHz... In a single sideband TV receiver, it is sent as a vestigial sideband. The video signal is passed through the IF filter shown in FIG. 14(b) at the IF stage, and then detected and reproduced.

Therefore, the power spectrum of the image signal when a ghost is added will be considered below. (Because N is simple, (s
This will be explained using the in x / x ) pulse. ) 1st
FIG. 5 is a characteristic diagram showing the power spectrum of a video signal including ghosts. Figure 15(a) shows the in-phase ghost, (
b) When an orthogonal ghost is added to (sin x /
x) Shows the power spectrum of the pulse.

(a) In-phase ghost: A periodic function with a constant amplitude of period 1/T within the band (b) Orthogonal ghost: The amplitude increases monotonically from 0 to 1.25 MHz, and above 1.25 MHz, A periodic function with a period of 1/T whose amplitude is constant (however, there is a phase difference of 90' from the case of an in-phase ghost).
This will be explained using figures. FIG. 16 is an explanatory diagram of the detection output of orthogonal ghosts.

(a) Near DC The signal near DC of the video signal is detected using the upper and lower waves having almost the same amplitude even after passing through the IF filter, so as shown in Figure 16(a), In the case of orthogonal ghosts, no output is produced when fully synchronous detection is performed.

(b) 1.25M7 or higher T For components up to 1.25MHz, the IF filter detects them using upper and lower waves with different amplitudes, so as shown in Figure 16(b), orthogonal A detection output is generated even in the case of a ghost, and its magnitude is proportional to the amplitude difference between the upper side wave and the lower side wave, so the closer it is to 1.25 MHz, the larger it becomes.

(c) 1.25 MHz or higher For components of 1.25 MHz or higher, only the upper side wave is detected, so an output with a constant amplitude is generated after detection.

Synchronous ghost and orthogonal ghost have been added (sinX
/X) The power spectra of the pulses both have a period of 1/
It is a periodic function of T, but since it has a phase difference of 90°, when calculating each cepstrum, as shown in Figure 17, the real part Re(C(q)) for in-phase ghosts and the imaginary part for orthogonal ghosts. An output occurs at Im(C(q)).

This generally means that the real part of the cepstrum Re(C(q
)) and the imaginary part 1m(C(q)), it is shown that the phase φ of the ghost can be found from the following equation.

As described above, by cepstrum analysis of the image signal, the delay time (T) and size (D/U ratio) of the ghost can be determined.
7 phases (φ) can be obtained.

FIG. 18 shows a cepstrum waveform when a synchronous ghost with a D/U ratio of 6 dB and a delay time of 7 μs is added to the image signal.

From the figure, it can be seen that a pulse-like cepstrum appears at a position of 7 μs on the horizontal axis, and a ghost is detected. Also,
A pulse-like waveform also occurs at the 4 μs position on the horizontal axis, but
This is caused by the first term of equation (6), that is, the power spectrum of the image signal. However, this can be removed by obtaining similar cepstrums using a plurality of different image signals and calculating the average of the cepstrums.

This is because the cepstrum indicated by the ghost component occurs as long as the ghost does not change, but unnecessary components due to the image signal are thought to appear at different positions depending on the content of the image signal.

〔Example] An embodiment of the present invention will be described below with reference to FIG.

FIG. 1 is a block diagram showing one embodiment of the present invention. In the figure, 500 is an input terminal, 501 is a transversal filter, 502 is an A/Di conversion circuit, 503 is a waveform memory, 504 is a cepstrum analysis circuit, 505 is a subtraction circuit, 506 is a tap gain memory, and 507 is a D/A The conversion circuit 508 is an output terminal.

In this ghost removal device 510, a video signal inputted from an input terminal has ghost components reduced and removed by a transversal filter 501, and then the video signal is sent to an output terminal 508.
is output from.

The output signal (video signal) of the transversal filter 501 is converted into a digital signal for an arbitrary period by the A/D conversion circuit 502 and stored in the waveform memory 503. A cepstrum analysis circuit 504 obtains ghost parameters from the signal stored in the waveform memory 503. The subtraction circuit 505 reads out the gain data of the tap in the transversal filter 501 corresponding to the obtained ghost parameter from the tap gain memory 506, corrects it, and stores the corrected tap gain data in the tap gain memory 506.
Enter. Each tap gain data stored in the tap gain memory 506 is converted into an analog signal by the D/A conversion circuit 507, and then supplied to the tap gain adjustment terminal of the corresponding tap amplifier in the transversal filter 501, and the ghost data is Enables reduction and removal of components.

FIG. 2 shows an example of the configuration of the cepstrum analysis circuit 504. 503 is an input terminal, 1002.1005 is a Fourier transform circuit, 1003 is a power spectrum calculation circuit, 10
04 is a logarithmic calculation circuit, 1006 is an averaging circuit, 100
7 is a tap gain correction calculation circuit, and 505 is an output terminal.

Fourier transform circuit 1002. Power spectrum calculation circuit 1003. Using the logarithm calculation circuit 1004 and the Fourier transform circuit 1005, calculate equations (4)2 to (6),
The cepstrum of the signal input from the input terminal 503 is determined. This operation is repeated and the averaging circuit 1006
The average of the cepstrum obtained in step 100 is calculated, and the tap gain correction amount of the tap corresponding to the delay time T of the ghost detected by the tap gain correction calculation circuit 1007 is calculated and output from the output terminal 505.

When the averaging circuit 1006 uses an arbitrary pattern image signal, an unnecessary component of the cepstrum is generally generated due to (6) the first term of ghostwriting (loglX(f)M), but this unnecessary component occurs at a different location depending on the pattern. Therefore, the purpose is to calculate the average of a plurality of cepstrums and suppress unnecessary components.

Here, since the output of the logarithm calculation circuit 1004 is a real function and an even function, a similar result can be obtained even if the Fourier transform circuit 1005 is a circuit that performs inverse Fourier transform. The block diagram in this case is shown again in FIG. 2A. It will be appreciated that in FIG. 2A, instead of Fourier transform circuit 1005, inverse Fourier transform circuit 1005A is connected.

Cepstral analysis circuit 504 shown in FIGS. 2 and 2A.
Although the 504A is constructed entirely of hardware, a microcomputer or the like is used to perform the above (4)
The calculation of the equation 7) and the calculation of the tap gain correction amount may be performed by software.

In addition, in the above equation (5), r and IX(f)+2 do not change with frequency, and by measuring the period of change of IY(f)l'' with frequency, T can be found, and the tap gain is adjusted accordingly. When calculating quantities, the logarithm calculation circuit 1004 is not necessary.

FIG. 3 shows another embodiment of the ghost removal device according to the present invention. In this figure, blocks with the same numbers as in FIG. 1 indicate the same functional blocks. Others 6
00 is an A/D conversion circuit, 601 is a transversal filter, and 602 is a D/A conversion circuit.

In the ghost removal device 510 shown in FIG. 1, the transversal filter 501 is made up of analog circuits, but the transversal filter 601 making up the ghost removal device 511 shown in FIG. 3 is all made up of digital circuits. be done.

Therefore, in the present ghost removal device 511, the signal inputted from the input terminal 500 is converted into a digital signal by the A/D conversion circuit 600, and then inputted to the transversal filter 601. Further, the output signal of the transversal filter 601 is converted into an analog signal by the D/A conversion circuit 602 and then outputted from the output terminal 508. The waveform memory 503 includes a transversal filter 6
The output signal of tap gain memory 506 is stored for an arbitrary period of time, and the output signal of tap gain memory 506 is directly input to transversal filter 601.

The ghost removal operation is similar to that of the ghost removal device 510 shown in FIG.

FIG. 4 shows a ghost removal device 510 or 5 according to the present invention.
11 shows a configuration example of a television receiver equipped with 11.

In the figure, 200 is an antenna, 201 is a high frequency amplification circuit, 202 is a frequency conversion circuit, 203 is an intermediate frequency amplification circuit, 204 is a video detection circuit, 205 is a video amplification circuit,
206 is a color difference signal demodulation circuit, 207 is an original signal demodulation circuit,
208 is a cathode ray tube.

A television signal radio wave received by an antenna 200 is amplified by a high-frequency amplification circuit 201, frequency-converted by a frequency conversion circuit 202 to an intermediate frequency for demodulation, and then transmitted to an intermediate amplification circuit 203.
Amplify with. Here, channel selection is performed by changing the local oscillation frequency of the frequency conversion circuit 202.

The signal amplified by the intermediate frequency amplification circuit 203 is input to the ghost removal device 510 or 511 to remove the ghost component, and then detected by the video detection circuit 204 to produce a luminance signal and a color difference signal which are the outputs of the video signal amplification circuit 205. Demodulation circuit 20
6 and the primary color signal demodulation circuit 20.
R,G at 7.

The three primary color signals of B are obtained and projected onto the cathode ray tube 208.

FIG. 5 shows another configuration example of a television receiver equipped with the ghost removal device 510 shown in FIG. 1.

In this figure, blocks with the same numbers as in FIG. 4 indicate the same functional blocks. In addition, 700
701 is a frequency conversion circuit, 701 is an intermediate frequency amplification circuit, and 702 is a frequency conversion circuit.
．． ．． 703 is a switch circuit.

In the television receiver shown in FIG. 4, the output signal of the frequency conversion circuit 202 is usually frequency-converted so that the carrier wave of the television signal radio wave is 58.75M)Iz, so the ghost removal device 510 or The A/D conversion circuit 502 or 600 constituting 511 has a 58.75MH
It is necessary to operate at a frequency at which the z signal is sufficiently sampled (a frequency that satisfies the sampling theorem). Therefore, in the television receiver shown in FIG. 5, the frequency conversion circuit 700 converts the carrier wave of the television signal radio wave into 58
．． Frequency smaller than 75MHz (e.g. 10MHz)
At the start of ghost component removal, the switch circuits 702 and 703 are closed to the b side, the ghost component is detected using the output signal of the intermediate frequency amplification circuit 701, and the transversal filter 501 Determine each tap gain. During this period, a screen with ghost components added is displayed on the cathode ray tube 208 because it is better than not displaying any screen at all.

After determining each tap gain of the transversal filter 501, switch circuits 702 and 703 are closed to the a side, the output signal of the intermediate frequency amplification circuit 203 is input to the transversal filter 501, ghost components are removed, and then video detection is performed. By supplying the circuit 204, the cathode ray tube 2
The screen from which ghost components have been removed is displayed on 08.

After the switch circuits 702 and 703 are closed to the a side, correction of the tap gain is stopped.

In this television receiver, the signal supplied to the A/D conversion circuit 502 that constitutes the ghost removal device 510 is
Since the frequency can be lower than 8.75 MHz, its operating frequency can also be lower than in the case of the receiver shown in FIG.

In the television receiver shown in FIG. 5, after the switch circuits 702 and 703 are closed to the a side, the correction of each tap gain of the transversal filter 501 is stopped, so that the phase etc. of the ghost component changes thereafter. In this case, the transversal filter 501 cannot be controlled accordingly.

FIG. 6 shows an embodiment of a ghost removal device that can cope with changes in the phase of ghost components, and FIG. 7 shows an example of the configuration of a television receiver using the device.

In FIG. 6, blocks with the same numbers as in FIG. 1 indicate the same functional blocks. 8O1 is a transversal filter.

In this ghost removal device 800, the output signal from the intermediate frequency amplification circuit 701 shown in FIG. 7 is converted into a digital signal by the A/D conversion circuit 502, and then
Stored in 03. The signal stored in the waveform memory 503 is input to the transversal filter 801, and the output signal of the transversal filter 801 is used to adjust each tap of the transversal filter 501 in the same manner as the ghost removal device 510 shown in FIG. Determine the gain. At the same time, each tap gain stored in the tap gain memory 506 is also supplied via a D/A conversion circuit 507 to a transversal filter 501 to which the output signal from the intermediate frequency amplification circuit 203 is input, and is used to amplify the intermediate frequency. circuit 20
The ghost component added to the output signal from 3 is removed, and the removed signal is supplied to the video detection circuit 204 shown in FIG.

In this ghost removal device 800, the A/D conversion circuit 502
Since the output signal from the intermediate frequency amplification circuit 701 can be taken into the waveform memory 503 at any time, it is possible to deal with phase fluctuations of ghost components, etc.

In the present coast removal device 800, the transversal filter 501 is configured with an analog circuit, but the same effect can be obtained even if the transversal filter 501 is configured with a digital circuit as shown in FIG.

FIG. 9 shows another configuration example of a television receiver equipped with the ghost removal device 510 or 511 according to the present invention shown in FIGS. 1 and 3.

In this figure, blocks with the same numbers as in FIG. 5 indicate the same functional blocks. 2100 sets the carrier frequency of the television high frequency signal to 58.75M) [z
The frequency conversion circuit 2200 converts the carrier frequency converted to a lower frequency by the frequency conversion circuit 2100 to a lower frequency (for example, lOMHz) to a higher frequency of 58.75M.
This is a frequency conversion circuit that converts the frequency back to Hz.

The feature of this television receiver is that the ghost removal device 51
Transversal filter 50 configuring 0 or 511
1 or 601, the frequency of the television signal input to
By setting the frequency lower than 58.75 MHz, the operating frequency can be lowered and the circuit scale of the transversal filter 501 or 6010 can be reduced (10th
As explained in the figure, the tap interval of the transversal filter is determined by the highest frequency of the input signal.The higher the frequency, the smaller the tap interval.If the delay time of ghosts that can be handled is constant, the number of taps increases. ).

FIG. 19 shows still another embodiment of the ghost removal device according to the present invention. In this figure, blocks with the same numbers as in FIG. 3 indicate the same functional blocks. 6
01 is a transversal filter, and 602 is a subtraction circuit.

In the ghost removal device 512 shown in the figure, the difference between the input signal and the output signal of the transversal filter 601 is calculated by the subtraction circuit 601.
02 and inputs it again to the transversal filter 601, it is possible to obtain the same ghost removal effect as in FIG. 3.

FIG. 20 shows still another embodiment of the ghost removal device according to the present invention.

In this figure, blocks with the same numbers as in FIG. 3 indicate the same functional blocks. 141 is a variable delay line, 601a to 601c are transversal filters,
143 is an adder circuit, 144 degrees 505a to 505c are subtraction circuits, and 506a to 506C are tap gain memories.

In the ghost removal device 147 shown in the figure, the variable delay line 141
and one or several transversal filters 601
A ghost compensation signal is generated by a to 601c. Each of the transversal filters 601a to 601C corresponds to a single ghost, and by using a plurality of transversal filters, it is possible to correspond to a plurality of ghosts.

In the figure, the signal input from the input terminal 500 is
After being converted into a digital signal by the A/D conversion circuit 600, the variable delay line 141. Transversal filter 601
After ghosts are removed by the signals a to 601c, the addition circuit 143 and the subtraction circuit 144, the signals are converted into analog signals by the D/A conversion circuit 602 and output from the output terminal 508.

The output signal of the subtraction circuit 144 is also input to the waveform memory 503, and a signal of an arbitrary period is stored therein.

The cepstrum analysis circuit 504 obtains ghost parameters from the signal stored in the waveform memory 503 and inputs them to the variable delay line 141 and each subtraction circuit 505a to 505c. The variable delay line 141 delays the output signal of the A/D conversion circuit 600 by the delay time T of each ghost in accordance with the signal supplied from the cepstrum analysis circuit 504, and outputs the output signal to each transversal filter 601a to 601a.
Supply to 01C. In addition, subtraction circuits 505a to 505C
Then, the tap gain is corrected according to the ghost parameters corresponding to each of the transversal filters 601a to 601C obtained by the cepstrum analysis circuit 504, and the corrected tap gain is stored in each tap gain memory 506.
Enter a to 506C. Each tap gain memory 50
Each tap gain stored in 6a-506C is supplied to each tap of the corresponding transversal filter 601a-601c.

In the ghost removal device 147 shown in FIG.
Each transversal filter 601 is delayed by
a to 601c.

The tap gain of each of the transversal filters 601a to 601c is determined according to the parameters regarding the amplitude and two phases of each ghost output from the cepstrum analysis circuit 504.
Compensation signals for each ghost can be obtained from the outputs of 1a to 601c. By subtracting these sums from the input signal in the subtraction circuit 144, ghosts can be removed.

The transversal filters 601a to 601c have a configuration as shown in FIG. 2, for example. In the figure, the operation of each block is similar to the operation of the blocks with the same numbers in FIG. 10.

Ghost removal device 51 in FIG. 1, FIG. 3, or FIG. 19
At 0.511 or 512, the delay line constituting the transversal filters 501 and 601 needs to have several hundred taps, and furthermore, the same number of tap amplifiers as the number of taps are needed. However, the ghost removal device 147 in FIG.
The tapped delay line required for the transversal filter 506 may have about 10 taps. Therefore, by adopting the configuration shown in FIG. 20, the circuit can be simplified.

In the ghost removal device 147 shown in FIG. 20, the output signal of the A/D conversion circuit 600 is input to the variable delay line 141, but as in the ghost removal device 151 shown in FIG. A similar effect can be obtained by inputting the variable delay line 141 to the variable delay line 141.

In the ghost removal device 147 or 151 shown in FIG. 20 or 22, the variable delay line 141 and the transversal filter 601 are configured with digital circuits, but the same effect can be obtained even if these are configured with analog circuits. There is no need to explain what happens.

FIG. 23 shows an embodiment of a television receiver equipped with a ghost removal device according to the present invention. In the same figure, 1
81 is RF input terminal, 182 is tuner, 183 is IF
(Intermediate frequency)/detection circuit, 184 is a ghost removal device,
185 is a luminance signal/chrominance signal separation circuit (hereinafter abbreviated as Y/C separation circuit), 186 is a color demodulation circuit, 187 is a color synchronization circuit, 188 is a matrix circuit, and 189 is a display device.

RF double signal tuner 182 inputted from RF input terminal 181. After being tuned and detected by the IF/detection circuit 183, the signal is input to the ghost removal device 184. The output signal of the ghost removal device 184 is sent to the Y/C separation circuit 1851.
Color demodulation circuit 186° color synchronization circuit 187. The matrix circuit 188 converts the signals into RGB signals, and the display device 189
is supplied to

FIG. 24 shows an embodiment of a television tuner equipped with a ghost removal device according to the present invention.

In the figure, blocks with the same numbers as in FIG. 23 indicate the same functional blocks. Further, 191 is a video signal output.

FIG. 25 shows an embodiment of a VTR (video tape recorder) equipped with a ghost removal device according to the present invention. In the figure, blocks with the same numbers as in FIG. 23 are the same functional blocks. , 201 is a video input terminal, 202 is an input signal selection circuit, 203 is a video signal recording circuit, 204 is a video head, 205 is a video signal reproducing circuit, 206 is an RF converter, 207 is a video output terminal,
208 is an RF output terminal.

The operation of the VTR shown in the figure will be briefly explained below.

The signal selection circuit 202 selects either the signal supplied from the video input terminal 201 or the output signal of the ghost removal device 184, and the selected signal is sent to the video signal recording circuit 203. The video head 204 records on the magnetic tape. The video signal reproducing circuit 205 is connected to the video head 2
04 converts the signal read from the magnetic tape into a video signal and supplies it to the video output terminal 207. RF
Converter 206 converts the output of video signal reproduction circuit 205 into an RF signal and supplies it to RF output terminal 208 .

In the above explanation, the ghost was removed using a transversal filter, but the same effect can be obtained by creating a signal that cancels the ghost using multiple variable gain amplifiers and variable delay lines. There is no need to explain it.

Further, it goes without saying that the ghost removal device according to the present invention can also be incorporated into a VTR with a built-in tuner to obtain the same effect.

〔Effect of the invention〕

According to the present invention, since detection and removal of ghost components are performed at a stage before image detection, ghost components can be removed with good performance regardless of the detection method. In addition, when detecting and removing ghost components after video detection, it is possible to determine not only the delay time and magnitude but also the phase as parameters of the ghost component.
This allows for more detailed ghost removal.

Furthermore, as the target signal for ghost component detection is not limited to the synchronization signal, any signal (picture information) of the television signal can be used, so it is possible to speed up the ghost component removal cycle, and as a result, This has the advantage that the time required to remove ghost components can be significantly shortened.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of the ghost removal device of the present invention, FIGS. 2 and 2A are block diagrams showing details of the cepstrum analysis circuit in FIG. 1, and FIG. 3 is a block diagram showing an embodiment of the ghost removal device of the present invention. FIGS. 4 and 5 are block diagrams showing other embodiments of the ghost removal device according to the present invention, and FIG. 6 is a block diagram showing an embodiment of a television receiver equipped with the ghost removal device according to the present invention. FIG. 7 is a block diagram showing another embodiment of the ghost removal device according to the present invention. FIG. 8 is a block diagram showing another embodiment of the television receiver equipped with the ghost removal device according to the present invention. FIG. 9 is a block diagram showing still another embodiment of the apparatus, FIG. 9 is a block diagram showing still another embodiment of the television receiver equipped with the ghost removal device according to the present invention, and FIG. 10 shows the operation of the transversal filter. Illustration of the principle, No. 10A
The figure is a waveform diagram showing the signal waveform of the main part in FIG. 10, and FIG. 11 is a block diagram showing a conventional example of a ghost removal device.
FIG. 11A is a waveform diagram showing the signal waveform of the main part in FIG. 11, FIG. 12 is an explanatory diagram showing the principle of detecting delay time by cepstrum analysis, and FIG. 13 is a graph showing the relationship between ghost amplitude and cepstrum. Fig. 14(a) is a characteristic diagram showing the vestigial sideband method, Fig. 14(b) is a characteristic diagram of the IF filter, Fig. 15 is a characteristic diagram showing the power spectrum of a video signal including a ghost component, and Fig. 16 Figure 17 is an explanatory diagram of the detection output of quadrature ghosts, Figure 17 is a characteristic diagram showing the cepstrum of in-phase and quadrature ghosts, Figure 18 is a characteristic diagram showing an example of determining the cepstrum of an image signal to which in-phase ghosts are added, and Figure 19 FIG. 20 is a block diagram showing still another embodiment of the ghost removal device of the present invention, and FIG.
FIG. 1 is a block diagram showing a configuration example of a transversal filter, FIG. 22 is a block diagram showing still another embodiment of the ghost removal device of the present invention, and FIG. 23 is a television equipped with the ghost removal device of the present invention. FIG. 24 is a block diagram showing an embodiment of a receiver, FIG. 24 is a block diagram showing an embodiment of a tuner equipped with a ghost removal device according to the present invention, and FIG. 25 is a block diagram showing an embodiment of a VTR equipped with a ghost removal device according to the present invention. It is a block diagram showing an example. Explanation of symbols 104.501, 601...Transversal filter, 502,600.../D conversion circuit, 503...
Waveform memory, 504... Cepstrum analysis circuit, 50
5... Subtraction circuit, 506... Tap gain memory, 5
07.602...D/A conversion circuit, 510,511.
800.810...Ghost removal device, 202°70
0.2100.2200...Frequency conversion circuit, 203
．． 701...Intermediate frequency amplification circuit, 702, 703...
・Switch circuit, 204...Video detection circuit, 2000
．． 2030... Fourier transform circuit, 2010... Power spectrum calculation circuit, 2020... Logarithm calculation circuit, 2040... Tap gain correction amount calculation circuit Representative Patent Attorney Akio Namiki Figure 1 Figure 3?
Old Fig. 10 Fig. 1 OA Fig. 11 Fig. 1 (J5 Fig. 11A 0 Fig. 12 (a) Coast t...
Cepstrum) Fig. 14 - 1,25 MHz O4, 2 MHz (a) Retraction tilt/ far-far method b) IF field special production Fig. 15 (a) Same-age self-ghost (b) Orthogonal boost (1) O~5MHz (2)2-3M
Hz 16th diagram, general axis, group 7i axis, cylindrical axis: 1I (a) lshi size'i (b) 1.25MH2 or less (C) 1.25MH2ea ("Fl)C,
TI? I IM Figure 19 7--------- Drops-----
7 -------J Figure 21 so+

Claims (1)

[Claims] 1. A transversal filter (501) that receives and outputs a video signal, and a tap gain memory (501) that stores the tap gain of each tap amplifier included in the filter.
6), a cepstrum analysis means (504) for detecting a parameter of a ghost component of picture information included in the video signal by performing cepstrum analysis on the video signal that has passed the filter; means (505) for modifying tap gain data stored in a gain memory; and removing ghost components from the video signal passed through the filter by controlling the gain of each tap amplifier in the filter in accordance with the modified data. A ghost removal device comprising: means for removing a ghost. 2. In the ghost removal device according to claim 1, the cepstrum analysis means includes a first Fourier transform circuit (1002) that performs Fourier transform on an input signal, and a A power spectrum calculation circuit (1003) that receives a Fourier-transformed output signal and calculates its power spectrum, and a power spectrum calculation circuit (1003) that receives a Fourier-transformed output signal and calculates its power spectrum, and the result of logarithmic calculation of the power spectrum signal from the power spectrum calculation circuit, or the power before the logarithm calculation. A ghost removal device comprising: a second Fourier transform circuit (1005) that receives a spectrum signal and performs Fourier transform. 3. In the ghost removal device according to claim 1, the input video signal is a video signal at an IF (intermediate frequency signal) stage before image detection, and cepstrum analysis is performed on the video signal at the IF stage. A ghost removal device characterized in that a ghost component is removed by removing a ghost component. 4. The ghost removal apparatus according to claim 3, wherein the video signal at the IF stage is converted to a frequency conversion circuit (700).
After frequency conversion is performed, the IF after the frequency conversion is
A ghost removal device is characterized in that it performs cepstrum analysis on a video signal at various stages to remove ghost components. 5. The ghost removal device according to claim 1, wherein the input video signal is a video signal after image detection, and the ghost component is removed by performing cepstral analysis on the video signal. Ghost removal device. 6. Ghost detection means for detecting a ghost component contained in an input video signal, and ghost component removal for removing the ghost component contained therein from the video signal using the detected ghost component. means and
In the ghost removing device, when the video signal is in an IF stage before image detection, the ghost detecting means performs cepstral analysis on the video signal in the IF stage to detect the video signal. By detecting the delay time and magnitude of a ghost component of picture information included in a signal with respect to the picture information, and when the video signal is a video signal after video detection, performing cepstral analysis on the video signal, In addition to the delay time and magnitude, it includes a cepstrum analysis circuit that detects the phase of the ghost component (the amount of phase shift of the ghost component with respect to the phase of the picture information as a direct wave), and detects the detected delay time and magnitude. A ghost removal device characterized in that a ghost component is removed using phase and phase. 7. The ghost removal apparatus according to claim 6, wherein the cepstrum analysis circuit includes a first Fourier transform circuit (1
002), a power spectrum calculation circuit (1003) that receives the output signal of the first Fourier transform circuit, calculates and outputs its power spectrum, and receives the output signal of the power spectrum calculation circuit and calculates the logarithm thereof. A logarithm calculation circuit (1004) that calculates and outputs the output signal, and a second Fourier transform circuit (100) that receives the output signal of the logarithm calculation circuit, performs Fourier transform or inverse Fourier transform on it, and outputs
5, 1005A). 8. The ghost removal device according to claim 7, wherein the cepstrum analysis circuit includes an averaging circuit (100
6) A ghost removal device further comprising: 9. The ghost removal device according to claim 6, wherein the ghost detection means includes a memory (503) that stores the output signal from the ghost component removal means, and a memory (503) that performs cepstrum analysis of the signal waveform read from the memory. and a cepstrum analysis circuit (504), and the ghost component removal means includes a transversal filter (601) that receives a video signal and outputs it, and a tap gain of the transversal filter according to the detection result by the ghost detection means. Tap gain changing means (50
5, 506). 10. The ghost removal device according to claim 6, wherein the ghost detection means includes a memory (503) that stores the output signal from the ghost component removal means, and a memory (503) that performs cepstrum analysis of the signal waveform read from the memory. a cepstrum analysis circuit (504), and the ghost component removal means includes a transversal filter (60).
1), a subtraction circuit (602) that takes the difference between the input video signal and the output signal of the transversal filter and uses it as an input signal of the transversal filter, and a subtraction circuit (602) that calculates the difference between the input video signal and the output signal of the transversal filter, and a tap gain of the transversal filter that is calculated by the ghost detection means. A ghost removal device comprising: tap gain changing means (505, 506) that changes according to the detection result. 11. In the ghost removal device according to claim 6, the ghost detection means includes a memory (503) for storing the output signal from the ghost component removal means, and a cepstrum for performing cepstrum analysis of the signal waveform read from the memory. an analysis circuit (504), and the ghost component removal means includes a variable delay circuit (141) that delays and outputs the input video signal according to the delay time that is the analysis result from the cepstrum analysis circuit; A transversal filter (601) supplied with the output of the circuit
a to 601c), a subtraction circuit (144) that performs subtraction between the output signal of the transversal filter and the video signal before being delayed by the variable delay circuit, and outputs the result as an output; and the cepstral analysis. tap gain changing means (50
5a to 505c, 506a to 506c). 12. The ghost removal device according to claim 6, wherein the ghost detection means includes a memory (503) that stores the output signal from the ghost component removal means, and a cepstrum that performs cepstrum analysis of the signal waveform read from the memory. an analysis circuit (504), and the ghost component removal means includes a variable delay circuit (141) that delays and outputs the output video signal according to the delay time that is the analysis result from the cepstrum analysis circuit; A transversal filter (601) supplied with the output of the circuit
a to 601c), and a subtraction circuit (144) that performs subtraction between the output signal of the transversal filter and the video signal before being delayed by the variable delay circuit and supplies the result to the variable delay circuit as its input. ), and tap gain changing means (506a to 506c, 505a to 5
05c). A ghost removal device characterized by comprising: 13. The ghost removal device according to claim 11 or 12, wherein the variable delay circuit (141) has a configuration capable of outputting a plurality of delayed signals, each of which is delayed by an independently controllable delay amount and output. The transversal filter consists of a plurality of transversal filters each supplied with the video signal delayed by a different delay amount in the delay circuit, and the outputs of the plurality of transversal filters are mutually connected. A ghost removal device further comprising an addition circuit (143) that adds the information to the subtraction circuit and then supplies it to the subtraction circuit. 14. A ghost detection device comprising a memory for storing an input video signal, and a ghost component detection means for detecting a ghost component from a video signal waveform read from the memory, wherein the ghost component detection means detects a cepstrum of the video signal waveform. A ghost detection device characterized by comprising a cepstrum analysis circuit that performs analysis. 15. A channel selection circuit (182) that selects a channel based on the received television RF signal, an IF/detection circuit (183) that converts the output of the channel selection circuit into a video signal, and an IF/detection circuit (183) that converts the output of the IF/detection circuit into a video signal. In a television receiver comprising a video signal processing circuit (185 to 188) for converting into RGB (three primary color signals) and a display device (189), a ghost component is removed from the output signal of the IF/detection circuit to produce the video. A television receiver comprising the ghost removal device according to claim 3 as a ghost removal device (184) that supplies the signal processing circuit. 16. A television tuner comprising a channel selection circuit (182) that selects a channel based on a received television RF signal, and an IF/detection circuit (183) that converts the output of the channel selection circuit into a video signal, wherein the IF - A television tuner comprising the ghost removal device according to claim 3 as a ghost removal device (184) for removing ghost components from the output signal of the detection circuit. 17. A channel selection circuit (182) that selects a channel based on the received television RF signal, an IF/detection circuit (183) that converts the output of the channel selection circuit into a video signal, and an output signal of the IF/detection circuit. In a video tape recorder comprising a video signal recording circuit (203) for recording a video signal and a video signal reproducing circuit (205) for reproducing the video signal recorded in the recording circuit, ghost components are detected from the output signal of the IF/detection circuit. 4. A video tape recorder comprising the ghost removing device according to claim 3, as a ghost removing device (184) for removing the image signal and supplying the removed signal to the video signal recording circuit.