JPH077638A - Ghost removal device - Google Patents

Abstract

PURPOSE:To eliminate malfunction owing to the errorneous detection of preceding ghost, etc., and to precisely detect/remove ghost in a wider range by eliminating a 2T pulse generated from the trailing edge part of a reference signal and the ghost component of the 2T pulse in a waveform converter. CONSTITUTION:When y(n) being the output of the waveform converter 1-4 passes through a low pass filter 1-9, the high pass component of a GCR pulse in the signal y(n) is suppressed, and a sinx/x waveform is changed into a 2T pulse waveform. A position adjuster 1-10 in a next stage is a delay circuit delaying a signal from the low pass filter 1-9. The GCR pulse becomes the 2T pulse and the ghost of the 2T pulse by the processing of the low pass filter 1-9 and the position adjuster 1-10. A differential signal y(n) from the waveform converter 1-4 and a signal from the position adjuster 1-10 are supplied to an adder 1-11. They are added and synthesized and it becomes output y0(n). Then, the 2T pulse generated from the trailing edge part of a GCR waveform, and the ghost are eliminated.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a ghost eliminating device suitable for various video equipment such as a television (TV) receiver and a video tape recorder (VTR). The present invention particularly aims to provide a ghost removing device capable of preventing a malfunction in a reception environment in which a front ghost exists and performing a stable ghost removing operation in a wider range than conventional. .

[0002]

2. Description of the Related Art FIG. 2 shows a configuration example of a conventional ghost removing device. The ZF method (Zero Forcing Methode) is a typical algorithm of such a ghost removing apparatus.

For video signals of TV broadcasting, Japanese Patent Application No. 1-69179
, The reference signal for ghost cancellation (G
(CR signal) and the GCR waveform of FIG.
The 0 pedestal waveform in (b) is inserted in an 8-field sequence. (Insert on the sending side)

FIGS. 4 (a) to 4 (h) are layout diagrams of GCR signals from the first field (F1) to the eighth field (F8), which are periodic signals having 8 fields as one cycle. When the difference between the signals of FIG. 6A and FIG. 6B separated by 4 fields from these signal arrangements is taken, the horizontal sync signal and the burst signal are erased (in the same phase between the fields separated by 4 fields). Therefore, the GCR waveform as shown in FIG. 6C can be extracted.

The GCR waveform is a bar pulse having a width of about 44.7 μsec. This GCR waveform is 1 / (4 fsc)
If the difference processing is performed at an interval of (where fsc is a color subcarrier frequency), a reference pulse for ghost cancellation, that is, a GCR pulse as shown in FIG. 6D can be obtained from the leading edge portion thereof. As shown in FIG. 5A, the GCR pulse is fm (= about 4 MHz) which is the frequency occupied band of the video signal.
It has a flat characteristic. FIG. 6 (e) shows the delay time T
FIG. 6 is a waveform diagram of a GCR pulse and its ghost when there is an in-phase ghost of g. Figure 6 shows the ghost detection time range.
In (d) and (e), t between t1 and t2
It is in the range of 1 + t2. The values of t1 and t2 are

[0006]

[Equation 1]

Is set to the value of

The NTSC system video signal received by the antenna is supplied to the input line L1 in FIG.
In this video signal, as described above, as shown in FIGS.
The GCR signal of 4 is inserted in an 8-field sequence as shown in FIGS. The insertion period of the GCR signal is 18H and 281H.

In the following processing system, description will be made mainly on the assumption of digital signal processing. Therefore, it is assumed that the input signal is AD (analog-digital) converted. Along with this, the final output signal from the line L2 is to be DA (digital-analog) converted thereafter. The value normally used as the sampling frequency fs at this time is 4 fsc.

The input signal is first supplied to the filter section 2-1 which is a transversal filter. FIG. 3 shows a specific circuit example. 3-A in FIG. 3 is an FIR filter, and 3-B is an IIR filter. The signal that has passed through these two filters is output from the line L2. FIR
The filter 3-A has a delay time T having the same value as the sampling period.
(See the following formula)

[0011]

[Equation 2]

A delay circuit group 3-1 in which 2M delay circuits (M is a positive integer value) are connected in series and the input / output signals of each delay circuit are supplied and given an arbitrary coefficient value. It is composed of a 2M + 1 coefficient setting circuit group 3-2 to be amplified and a combiner 3-3 to add and combine the 2M + 1 signals amplified here. The output of this synthesizer is supplied to the next IIR filter 3-B.

The IIR filter 3-B includes a subtractor 3-4 and
It is composed of a FIR filter 3-C and a delay circuit 3-5 which delays the output signal of the subtractor 3-4 by the time MT and supplies it to the FIR filter 3-C. FIR filter 3
-C, like the FIR filter 3-A, has N-1 (N
Is a delay circuit group 3-6 having an integer value greater than 1), N coefficient setting circuit groups 3-7, and a combiner 3-8.

The structure shown in FIG. 3 can be expressed by the following equation representing frequency characteristics.

[0015]

[Equation 3]

Ga (f) is the characteristic of the FIR filter 3-A which is responsible for waveform equalization and proximity ghost removal, and Gc
(F) is a characteristic of the IIR filter 3-B that takes charge of normal ghost and long ghost removal. Also, Gb
(F) is a composite characteristic of the delay circuit 3-5 and the FIR filter 3-C.

The output signal of the filter section 2-1 is output from the line L2 as a final output signal, and is also supplied to the waveform capturing device 2-2.

In the waveform capturing device 2-2, the GCR signal portion is fetched for each field from the input video signal (see FIG. 4) and stored in a memory or the like, for example, 4 fields apart in FIG. 6 (a). 6C and the 0 pedestal waveform of FIG. 6B are taken to extract the GCR waveform of FIG. 6C.

In the next synchronous adder 2-3, in order to reduce the noise component contained in the fetched signal, the GCR waveforms are added and averaged while aligning the positions of the rising slopes of the GCR waveforms.

The waveform converter 2-4 is a difference circuit and is shown in FIG.
A difference signal y including a GCR pulse as shown in FIG. 6D is obtained by performing difference processing on a GCR waveform as shown in FIG. 6C at an interval of T = 1 / (4fsc) (where fsc is a color subcarrier frequency).
(N) (where n is the number of the data string, y
(N) means y (nT). Hereinafter, the same expression method is used. ) Is working. As described above, when a ghost exists, the result is as shown in FIG.

The reference waveform generator 2-6 produces the original GCR pulse waveform without distortion, that is, the reference waveform s (n). Actually, the method is such that the data is written in the ROM or the like in advance and is sequentially read by the clock pulse. 6F shows the reference waveform, which in this case is the same waveform as the GCR pulse of FIG. 6D.

In the subtractor 2-5, the difference between the difference signal y (n) from the waveform converter 2-4 and the reference waveform s (n) from the reference waveform generator 2-6 is obtained, and the error signal e ( n) is obtained.

[0023]

[Equation 4]

The coefficient former 2-7 is a circuit block for performing coefficient updating processing based on the ZF method, and is given by the following equation.

[0025]

[Equation 5]

A new coefficient value is calculated. Adaptive control can also be performed by changing the magnification α according to the state of the signal.

In this way, when the video signal input from the line L1 is filtered and output from the line L2, the GCR signal portion included in the video signal is taken out, the GCR pulse and the ghost are detected, and the reference waveform is obtained. The function of the conventional ghost removing apparatus based on the ZF method is to obtain an error signal from the difference between the difference and the coefficient value of the filter unit 2-1 and to successively reduce the amount of ghost.

[0028]

The drawback of the prior art device shown in FIG. 2 comes to the surface in the presence of a pre-ghost. Consider an example in which a positive ghost exists in the differential signal y (n) including the GCR pulse before the time Tg of the GCR pulse as shown in FIG. 7 (h) (the GCR pulse starts from the leading edge of the GCR waveform). Generated). At the time rear of this waveform diagram, 2 generated from the trailing edge of the GCR waveform.
There is a negative T pulse. The 2T pulse itself does not fall within the ghost detection range because it is at t = t3 (> t2), but in the case of FIG. 7 (h), the ghost of the 2T pulse is at time tg before time t3 by time Tg. = T4
Since it appears in (<t2), it completely falls within the ghost detection range. Ghost is an unnecessary component for original ghost detection. Therefore, if the ghost removal processing is advanced in the false detection state of detecting the ghost, the time t = t4
If there is a ghost in the image, a process for deleting it is performed, and as a result, a malfunction in which the ghost is added is caused.

Furthermore, the ghost elimination algorithm has Z
Since a successive coefficient correction method such as the F method is used, even if it starts from the state shown in FIG. 7 (h) at the beginning, in the process of suppressing the ghost, as shown in FIG. As becomes smaller, the time Tg
A grandchild ghost (10) occurs before. The 2T pulse ghost becomes smaller as shown in (11) along with the generation of the grandchild ghost (10), and at the same time, the grandchild ghost (12) is generated and should not be originally detected in addition to the originally required ghost data, (10). Detects (11) and (12) that do not happen as ghosts. Therefore, the coefficient value corresponding to the erroneous detection of the ghosts (11) and (12) is set in the transversal filter, and the conventional device has a problem that the ghost is added reversely.

It is an object of the present invention to provide a ghost removing device capable of eliminating the erroneous operation due to the erroneous detection of the previous ghost and the like as in the conventional example, and detecting and removing the ghost accurately in a wider range.

[0031]

In order to solve the above problems, the present invention provides

The input signal is the first signal,

A transversal filter which is supplied with the first signal and outputs a second signal;

The second signal is supplied, and the regularly inserted ghost eliminating reference signal is fetched.
Waveform capture device that outputs as a signal of

A synchronous adder which is supplied with the third signal and outputs a fourth signal obtained by adding and averaging the third signal a predetermined number of times in order to reduce noise components,

A waveform converter to which the fourth signal is supplied and which outputs a fifth signal including a reference pulse obtained by differentially processing the fourth signal at a time interval T;

A trailing edge canceller for receiving the fifth signal and outputting a sixth signal;

The sixth signal for comparing with the sixth signal
A reference waveform generator that outputs a reference waveform synchronized with the signal of

A subtracter for subtracting the reference waveform from the sixth signal to output a seventh signal which is an error signal,

Based on the seventh signal, the coefficient set in the transversal filter is obtained by sequentially calculating and correcting,
A ghost elimination device comprising a coefficient former that sequentially updates the coefficients of the transversal filter,

The trailing edge canceller is

A low-pass filter supplied with the fifth signal,

A position adjuster for delaying the output signal of the low-pass filter,

In the waveform converter, the 2T pulse generated from the trailing edge of the reference signal and the ghost component of the 2T pulse are eliminated from the fifth signal and the output signal of the position adjuster. The present invention provides a ghost elimination device characterized by comprising a synthesizer which synthesizes and outputs the sixth signal.

[0045]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a diagram showing an embodiment of the ghost removing device of the present invention. FIG. 3 is a block diagram of the filter unit,
FIG. 8 is a block diagram of a digital filter used for a low-pass filter or the like. 4 to 7 are operation explanatory diagrams. In the operation explanatory diagram, a simplified simulated expression method is also used for convenience. Further, even when a digital circuit is taken as a specific circuit example, the signal waveform may be shown as an analog waveform in order to make the explanation of the operation easier to understand.

In FIG. 1, 1-1 is a filter section, 1-
2 is a waveform capture device, 1-3 is a synchronous adder, 1-4 is a waveform converter, 1-5 is a subtractor, 1-6 is a reference waveform generator, 1-
Reference numeral 7 is a coefficient former, and 1-8 is a trailing edge canceller which is a characteristic part of the present embodiment.

It should be noted that, for convenience of explanation, a signal delay due to a processing time of each circuit itself, and a delay circuit or the like usually used only for simply correcting the delay will be omitted unless necessary for explanation. To do.

The input signal applied to the line L1 handled by this ghost eliminating device is a video signal of a television of the NTSC broadcasting system whose frequency band is up to fm (about 4 MHz) as shown in FIG. 5 (a).

The functions of blocks 1-1 to 1-7 in FIG. 1 are exactly the same as the functions of blocks 2-1 to 2-7 in FIG. The characteristic trailing edge canceller 1-8 will be mainly described.

The trailing edge canceller 1-8 is a low pass filter 1-9.
And a position adjuster 1-10 and an adder 1-11. The low-pass filter 1-9 is shown in FIG.
It is a low-pass filter having the frequency characteristic shown in (b), and its transfer function Gl (f) is given by the following equation.

[0051]

[Equation 6]

Difference signal y which is the output of the waveform converter 1-4
When (n) passes through this low pass filter 1-9, the signal y
In the GCR pulse in (n), the high frequency component is suppressed, and si
The nx / x waveform changes to a 2T pulse waveform. To construct the low-pass filter 1-9, a transversal filter as shown in FIG. 8 is used. Block 4-1 of FIG.
Is composed of 16 delay circuits (block numbers 1 to 16) having a delay time T (see Formula 2) connected in cascade, and an input signal is delayed by time T and output. The output of each delay circuit is the block 4-2 (block number -8 to 8) of the next stage.
The coefficient values are set in such a manner that the 2T characteristic (FIG. 5B) is obtained, and the nine signals multiplied by these coefficients are added in the next block 4-3. It is output after being added and synthesized.

Returning to FIG. 1, the position adjuster 1-1 at the next stage.
Reference numeral 0 is a delay circuit that delays the signal from the low-pass filter 1-9 by the time Td represented by the following equation.

[0054]

[Equation 7]

By the processing of the low pass filter 1-9 and the position adjuster 1-10, the difference signal y (n) of FIG.
As shown in (i), the GCR pulse becomes a 2T pulse and the ghost becomes a ghost.

The adder 1-11 is supplied with the difference signal y (n) from the waveform converter 1-4 and the signal from the position adjuster 1-10, and is added and synthesized to produce an output yo (n). . The signals shown in FIGS. 7 (h) and 7 (i) are added and combined to obtain a signal as shown in FIG. 7 (j). 7 (h) y (n) and FIG.
The difference between (j) and yo (n) is that yo (n) has y.
2 generated from the trailing edge of the GCR waveform, which was in (n)
This is the point where the T pulse and its ghost are erased.

Therefore, in the subtractor 1-5 at the next stage shown in FIG. 1, the difference signal yo (n) from the adder 1-11 and the reference waveform s (n) from the reference waveform generator 1-6 are used. The difference is obtained, and the error signal e (n) as in the following equation is obtained.

[0058]

[Equation 8]

This expression differs from the expression 4 cited in the description of the conventional example, and corresponds to the waveform shown in FIG. In the illustrated waveform diagram, the influence of the trailing edge of the GCR waveform disappears, and only the original ghost component remains. Therefore, even when the coefficient updater 1-7 at the next stage performs the coefficient updating process shown in Formula 5, a malfunction in the conventional example, that is, a ghost formed from the trailing edge of the GCR waveform is detected and detected. It is possible to avoid making a mistake such as setting a coefficient value to be deleted. This is the purpose and purpose of this embodiment.

Here, as described in the conventional example, when the successive coefficient correction method such as the ZF method is used for the ghost elimination algorithm, the ghost is suppressed even if the state shown in FIG. 7 (h) is started. In the process of making, as the ghost becomes smaller as shown in Fig. 7 (m), a grandchild ghost (10) occurs further before time Tg, and the ghost of 2T pulse becomes smaller as shown in (11) accordingly. At the same time, the grandchild ghost (12) occurs. However, due to the action of the trailing edge canceller 1-8 according to the present embodiment, the 2T pulse in FIG. 7 (m) and its ghost (11) and grandchild ghost (1
The waveform obtained as the error signal e (n) by canceling out with 2) is only the ghost formed by the GCR pulse and the grandchild ghost (10) as shown in FIG. 7 (n). The generation of such a grandchild ghost is also one of the causes that the filter unit that processes the previous ghost is a part of the FIR filter configuration. Further, a more complicated ghost may be formed in the process of the successive coefficient correction algorithm. However, since the erroneous operation (excessive ghost detection and erroneous coefficient setting operation accompanying it) can be eliminated by the trailing edge cancellation processing of the present proposal, the ghost component is finally erased.

As a result of the improvement of the malfunction problem, the range in which the front ghost and the delayed ghost can be stably and surely removed is widened as compared with the conventional case, and the ghost detection range is wider than the range of Mathematical Expression 1. The range can be set. The value of t1 in equation 1 is -10 μsec to -20 μsec
The value of t2 can be expanded to 50 μsec to 60 μsec as required.

[0062]

As described above, the ghost removing device of the present invention has the following effects.

(A) When the ghost is detected, the influence of the trailing edge of the GCR waveform can be eliminated, and the malfunction of the conventional example at the time of the coefficient updating process, that is, the unnecessary ghost formed from the trailing edge of the GCR waveform is detected. It is possible to prevent a malfunction such as adding a new ghost, and remove the ghost accurately.

(B) ZF as an algorithm for removing ghosts
In the case of using the iterative correction method using a transversal filter as represented by the method, a ghost at the trailing edge of the GCR waveform and further a ghost thereof are generated.
In the present invention, these two detection-unnecessary ghosts can be eliminated and only the originally required ghost component can be detected, so that a malfunction as in the conventional example can be prevented.

(C) As a result of the improvement in the malfunction problem described above, the range in which the forward ghost and the delayed ghost can be accurately dealt with is widened and the range in which the ghost can be reliably removed can be set in a wider range than in the past.

[Brief description of drawings]

FIG. 1 is a block diagram of an embodiment.

FIG. 2 is a block diagram of a conventional example.

FIG. 3 is a configuration diagram of a filter unit.

FIG. 4 is an operation explanatory diagram of a conventional example and an example.

FIG. 5 is an operation explanatory diagram of a conventional example and an example.

FIG. 6 is an operation explanatory diagram of a conventional example and an example.

FIG. 7 is an operation explanatory diagram of a conventional example and an example.

FIG. 8 is a configuration diagram of a low pass filter.

[Explanation of symbols]

 1-1 Filter unit 1-2 Waveform capturer 1-3 Synchronous adder 1-4 Waveform converter 1-5 Subtractor 1-6 Reference waveform generator 1-7 Coefficient former 1-8 Trailing edge canceller

Claims (2)

[Claims] 1. A transversal filter in which an input signal is a first signal, the first signal is supplied and a second signal is output, and the second signal is supplied and regularly inserted. A waveform capturer that captures a reference signal for ghost removal and outputs it as a third signal, and the third signal is supplied, and the third signal is added and averaged a predetermined number of times to reduce noise components. A synchronous adder that outputs a processed fourth signal, and a fifth signal that is supplied with the fourth signal and that includes a reference pulse obtained by differentially processing the fourth signal at a time interval T A waveform converter, a trailing edge canceller that is supplied with the fifth signal and outputs a sixth signal, and a reference waveform synchronized with the sixth signal for comparison with the sixth signal. A reference waveform generator for outputting, and the reference signal from the sixth signal A subtracter that outputs a seventh signal that is an error signal by subtracting the shape, and a coefficient that is set in the transversal filter based on the seventh signal, is sequentially calculated and corrected to obtain the coefficient of the transversal filter. A ghost removing device comprising: a coefficient forming unit for sequentially updating the low-pass filter, wherein the trailing-edge canceller includes a low-pass filter to which the fifth signal is supplied; A position adjuster that delays an output signal, a 2T pulse generated from the trailing edge of the reference signal in the waveform converter, and a ghost component of the 2T pulse, the fifth signal and the output signal of the position adjuster. And a combiner for outputting the sixth signal, and a ghost removing device. 2. The ghost elimination device according to claim 1, wherein the low-pass filter has a 2T characteristic, and the position adjuster is a delay circuit having a delay time of about 160 cycles of the color subcarrier. A ghost removal device.