JPH03289270A - Ghost elimination circuit - Google Patents

Abstract

PURPOSE:To attain high speed processing by setting a value in response to a time series data representing a filtered reception ghost elimination reference signal to a cyclic filter in a relevant time relation. CONSTITUTION:An FIR filter coefficient h(k) is calculated and the coefficient is set to an FIR filter 10 and a sin x/x signal subject to ghost interference, that is, a signal x(k) passes through the FIR filter 10. Then a time series data representing an inverse of a signal x' (k) is set to an IIR filter 20 in a relevant time relation as the filter coefficient. As a result, a part of the signal y(k) corresponding to the signal x'(k) is cancelled and an output signal o(k) of the IIR filter 20 is expressed as o(k) r(k). Thus, the filter coefficient is set in a short time.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to a ghost removal circuit for removing ghost disturbances in a television system, and more specifically, to a method for improving filter coefficients in a filter in such a circuit. This invention relates to a ghost removal circuit that can be set in a short time.

[Conventional technology]

A circuit has been proposed that uses a ghost removal reference signal (hereinafter referred to as a GCR signal) having a rising waveform of sin x/x to remove the ghost component from a received television signal to obtain high image quality. ing.

This type of ghost removal circuit generally consists of a cascade connection of an FIR filter (non-recursive filter), which mainly removes nearby ghosts, and an IIR filter (recursive filter), which mainly removes normal ghosts. That's what happens.

Conventionally, in a ghost removal circuit with such a configuration,
The received GCR signal and the GC stored in the circuit
FIR is executed by a microprocessor etc. based on the R signal.
Each filter coefficient (so-called,
The tap coefficients of the filters are calculated and set in these filters, and then the received television signal is passed through these filters to remove its ghost components. In this case, l'l
The filter coefficients for the R filter can be calculated as follows based on the received GCR signal.

For example, the frequency characteristics of the received GCR signal are determined by fast Fourier transform (PFT), and the frequency characteristics obtained by dividing the frequency characteristics of the known OCR signal by the determined frequency characteristics are subjected to inverse FFT and the filter coefficients are seek. Further, as another calculation method, there is a method using the least square error method. That is, the filter coefficients are calculated using the least square error method based on the waveform of the received GCR signal and the waveform of the known GCR signal. Furthermore, the filter coefficients for the IIR filter can also be calculated using the same method as described above. In this case, it is necessary to calculate several tens of filter coefficients for the FIR filter and several hundred filter coefficients for the IIR filter to achieve satisfactory ghost removal. For this reason, even though the filter coefficients for the PIR filter can be calculated in a relatively short time, it takes a lot of time to calculate the filter coefficients for the IIR filter, and this increases the speed of this type of ghost removal circuit. This had become an obstacle.

[Problem to be solved by the invention]

Therefore, an object of the present invention is to provide a ghost removal circuit having an FIR filter and an IIR filter that operates faster, that is, can set the filter coefficients of these filters in a shorter time. It's about doing.

[Means for Solving the Problems and Their Effects] A ghost removal circuit according to the present invention is formed by cascading an acyclic filter and a cyclic filter, and stores a received ghost removal reference signal in a human video signal. a first memory; a second memory storing a predetermined ghost removal reference signal; a received ghost removal reference signal in the first memory;
Based on the ghost removal reference signal in the memory of
calculation means for calculating a first filter coefficient group for the acyclic filter and setting the first filter coefficient group for the acyclic filter; filter means for obtaining a filtered received ghost removal reference signal by passing it through the acyclic filter in which a group is set; and a value corresponding to time series data representing the filtered received ghost removal reference signal; The present invention is characterized by comprising a setting means for setting a second filter coefficient group in the recursive filter in a corresponding time relationship.

According to such a configuration, the filter coefficients of the recursive filter are set by using the values corresponding to the time series data of the output of the acyclic filter as they are without performing any complicated calculations, so it is extremely simple. Completes in a short time.

Further, another configuration example of the ghost removal circuit according to the present invention is as follows:
In a ghost removal circuit formed by cascading an acyclic filter and a cyclic filter, a first memory stores a received ghost removal reference signal in an input video signal, and a second memory stores a predetermined ghost removal reference signal. a memory, a received ghost removal reference signal in the first memory and a second
a first calculation means for calculating a first group of filter coefficients for the acyclic filter based on a ghost removal reference signal in the memory of and setting the first group of filter coefficients for the acyclic filter; , a second method for calculating, based on the received ghost removal reference signal and the first filter coefficient group, an output of the acyclic filter to which the first filter coefficient group is set for the received ghost removal reference signal; and a value corresponding to the time series data representing the calculated output above,
The present invention is characterized by comprising a setting means for setting a second filter coefficient group in the recursive filter in a corresponding time relationship.

Even with this configuration, the filter coefficients of the recursive filter can be set in an extremely short time because the values corresponding to the time series data of the output of the acyclic filter are used as they are without any complicated calculations. Complete.

〔Example〕

First, before explaining the embodiments of the ghost removal circuit according to the present invention, the principle of operation of the present invention will be explained.

Figure 2 shows F in the ghost removal circuit according to the present invention.
It shows a cascade connection relationship between an IR filter and an ITR filter, and the connection configuration of such two filters itself is known. In this case, the FIR filter 10 is a transversal filter (TF) 11, while the IIR filter 20 is a transversal filter (TF) in the return path.
21, and the output of the FIR filter 10 is input to the adder 22.

In the above configuration, the fill coefficient for the FIR filter 10 (hereinafter referred to as FIR filter coefficient) is calculated by a method similar to the conventional method. That is, sin X/X in the ghost-impaired GCR signal provided as input signal.
Signal time series information X (k) and reference sin X/X signal time series information r (k) stored in advance in the circuit
), each FIR filter coefficient is calculated using the least squares method. These calculated FlR filter coefficients are respectively set in the transversal filter 11 in the FIR filter 10. Next, the output signal y(k) for the signal X(k) from the PIR filter 10 whose filter coefficients have been set in this manner is obtained, and the time series information y(k) is subjected to a special calculation. Zuni,
That is, each part of this time series data is set as a filter coefficient in the IIR filter 20 with the corresponding time relationship.

The reason why the ghost disturbance can be removed by setting the output signal y (k) of the FIR filter 10 as the coefficient of the IIR filter 20 in this way will be explained below.

First, to simplify the explanation, ghost removal is performed on a sinusoid that has suffered a ghost disturbance, for example, the waveform shown in FIG.
Assume that the filter operation is to reproduce a reference sin X/X signal r(k) whose waveform is shown in FIG. 3(a) from the X/X signal X(k). Note that the above signal r(k>
The frequency characteristic R(w) of is, as shown in FIG. 3(b),
It has a characteristic that it is flat from DC to a frequency Wc of 4.2 MHz, and rapidly attenuates to zero at frequencies higher than that.

As shown in FIG. 2, the signal x (k) includes a front ghost on the upstream side of the reference point ○ of the signal, and a delayed ghost on the downstream side thereof. Then, each filter coefficient h (k) of the FIR filter 10 is determined by this signal x (k
) and the part corresponding to the side lobe of the reference signal r(k), that is, the region indicated by P-Q in FIG. 2, the signal x(k) in this region is
The signal y(k) after passing through the FIR filter 10 is y(k) = r(k) (P≦k
≦ B') ・Set so that (1) is satisfied.

In this case, the output y (k) of FIR filter 0 becomes. Furthermore, from the above equation (1), the FIR filter coefficients need to have the following relationship.

[R-[H][X]
... (3) Here, [R] = [r(P) r(P+1)-r(Q)]
[)1] = [h(P) h(P+1)...
h (0) ]. Therefore, the FIR filter coefficient h (k) is [H] = [R][Xdo1
...It can be obtained by calculating (4). Here, [X]-' is the inverse matrix of [X].

However, since calculation of the inverse matrix of [X] is very complicated, the FIR filter coefficients are usually determined by an adaptive method such as the mean square error method or an algorithm using FFT.

In this way, after calculating the FIR filter coefficient h (k) and setting these coefficients in the FIR filter 10, the sin Pass through filter 10. In this case, the output signal y (k
The part corresponding to the region PQ of > is the reference signal r (
It should be substantially equal to k). Therefore, the signal y(k)
is y(k) = x”(k) + r(k) + X
” (k) − (5) Or, when passing through an IIR filter 20 with layered wavenumber characteristics, the output 0(W) of this filter 20 is 0(w) =
Y(w) −X' (w)0(w) −(7
)It can be expressed as. Therefore, it can be expressed as follows, and its waveform is shown in FIG. here,
The signal X''(k) represents the residual ghost in the region upstream from point P, and the signal x'(k) represents the residual ghost in the region downstream from point Q.

Then, the inverted signal -x' (k) of the above signal x' (k)
) is set as a filter coefficient in the IIR filter 20 in a corresponding time relationship. As a result,
Since the portion of the signal y(k) in FIG. 4 that corresponds to the signal (k)
'= r(k) - (6).

The ghost removal operation described above can be proven as follows.

The output Y (w) of the FIR filter 10 is -X' (w)
It can be expressed as (8). Here, as mentioned earlier, FI
The filter coefficient of the R filter 10 is the front ghost area ○-
Since the setting is made to cover up to P, .Here, as shown in Fig. 3(b), the frequency characteristic R(w) is 1 within the frequency band, so it can be expressed by dividing it into two regions as follows. .

Here, wc is the frequency at which R(w) starts to decrease from 1.

Therefore, the frequency portion OL(w) lower than the frequency wc (w < We) at the output 0(w) is 1+
X'L(w) 1+ X'L(w) = 1 = RL(w)
-...(11). Thus, the frequency portion below the frequency wc at 0(w) is equally flat as the low frequency portion at R(w>).

Also, the frequency wc (we
≦W) The frequency part L+(w) is Ri(w)
X'g(w>.Here,X'i(w) I
t, X' (w) Sokuchi residual: I-- strike signal x' (
This is the high frequency part of the frequency characteristic of k). This (1
2) Formula shows that the smaller X'B(w) is, the more 0-(w) is R
It shows that it approaches II(w).

Therefore, unless the received signal contains extremely large ghosts, Ox(w) can be considered to be substantially the same as R++(w).

Therefore, according to the method of setting each filter coefficient as described above, the low frequency part of the output 0(w) of the coast removal circuit is the same as the low frequency part of R(w), and 0(w) The high frequency part of R(w) is also very close to the high frequency part of R(w).

Embodiment Using Ordinary CPU Next, an embodiment of the ghost removal circuit according to the present invention constructed using an ordinary microprocessor will be described with reference to FIG.

In FIG. 6, F
The IR filter 10 and the IIR filter 20 have the same configuration as that shown in FIG. A switch 31 is provided between the input terminal of the transversal filter 11 constituting the PIR filter 10 and the signal input terminal 30 of the ghost removal circuit, and this input terminal 30 receives signals sampled at a predetermined sampling frequency. At the same time, digitized received video signals are sequentially supplied. The switch 31 is configured, for example, as a multiplexer, and under the control of a microprocessor 32 (to be described later), outputs either the video signal from the input terminal 30 or the output of a buffer memory 33 having a storage capacity of one line of video signals. P.I.
The output of the transversal filter 11 is supplied to the input terminal of the R filter 1 (l).The output of the transversal filter 11 is also supplied to the adder 22 in the IIR filter 20, and is also supplied to the waveform acquisition terminal having a storage capacity of at least one line of video signal. memory 3
4 is also supplied.

The output end of this waveform acquisition memory 34 is connected to the input bus of the microprocessor 32. Connected to the microprocessor 32 are a ROM 35 that stores programs for calculation and control, reference data, etc., and a RAM 36 for temporarily storing intermediate data and the like. Further, the output bus of the microprocessor 32 is connected to the input terminal of the buffer memory 33 and to each filter coefficient input terminal of the transversal filters 11 and 21. Note that the video signal obtained at the signal output terminal 37 of the ghost removal circuit connected to the output end of the IIR filter 20 is converted into an analog signal, for example, and then supplied to an image display circuit (not shown) or the like.

In the above configuration, the microprocessor 32 first
For example, just before the 18th horizontal period (18H or 281H) of each field, the switch 31 is set to the position shown in the figure, and the transversal filter 11 is assigned a predetermined low-pass filter stored in advance in the ROM 35 as a filter coefficient for each stage. Set the filter coefficients for each. As a result, the video signal of, for example, 18H or 281H inputted to the filter 11 via the terminal 30 is outputted from the filter in a form in which its high frequency noise component has been removed, and is stored in the waveform acquisition memory 34. In this case, the filter coefficients of the filter 11 are set to four values of "0" instead of the above-mentioned low-pass filter coefficients, so that the video signal supplied to the terminal 30 is passed through the transversal filter 11 in its original form. Waveform acquisition memory 3
4 may be stored. Next, the microprocessor 32 returns the filter coefficients of the filter 11 to their previous values, and then reads out the contents of the memory 34 and stores them in the RAM 36 at a predetermined timing. The microprocessor 32 executes the above operation according to the 8-field sequence method of the BT^ (Broadcast Technology Development Council) standard, and based on the data obtained as a result in the RAM 36, the GCR signal in the input video signal is (time series information)
Play. Then the microprocessor 32
By performing the necessary processing on the R signal and then differentiating it, we obtain a sin
n x/X signal time series x(k), is reproduced and stored in RAM.
Temporarily stored in 36. Next, the microprocessor 32
is the above signal x (k) in RAM 36 and ROM 35
The reference sin as shown in FIG. 3(a) is stored in advance in
The filter coefficient h (k) for the FIR filter is calculated by the least square error method based on the time series r (k) of the x/x signal.
Calculate.

Next, the microprocessor 32 reads the signal ×(k) stored in the RAM 36 and stores it in the buffer memory 33, and also applies the filter coefficient h(k) to each of the transversal filters 11 during the vertical retrace period. At the same time as setting the transversal filter 11, the switch 31 is switched so that the signal X(k) in the buffer memory 33 is supplied to the transversal filter 11. In this case, the above signal x
The output from the transversal filter 1 for (k), ie, the signal y(k) as shown in FIG. 4, is stored in the waveform acquisition memory 34. When the above operation is completed, the microprocessor 32 returns the switch 31 to the illustrated state.

Next, the microprocessor 32 controls the waveform acquisition memory 34.
, the part of the signal y(k) falling from point QJ and J, that is, the signal X(
k), and extract the residual ghost region, expressed as
'(k) The signs of each position of the time series are inverted and set in each stage of the transversal filter 21 in a corresponding time relationship.

According to the above configuration, the signal output terminal 37
As shown in FIG. 5, the output o(k) of the circuit obtained from the circuit is almost the same as the reference sin It will be removed.

Another embodiment of the ghost removal circuit according to the present invention will be described with reference to an embodiment using a digital signal processor.

Figure 7 shows a digital signal processor (hereinafter referred to as DSP).
This figure shows the configuration of a ghost removal circuit according to the present invention, which aims to achieve higher speed by using the following. In this figure, FIR
The configuration itself in which the filter 40 and the IIR filter 50 are connected in cascade is the same as the configuration shown in FIG. In this case, the analog video signal input to the signal input terminal 60 is sequentially converted into, for example, 8-bit digital data at a predetermined sampling period by the A/D converter 61 and is input to the input terminal 40a of the FIR filter 40. Supplied. Further, from the output terminal 50b of the IIR filter 50, a filtered video signal of, for example, 8 bits is sequentially supplied to a D/A converter 62, and from this D/A converter is supplied in analog form to an output terminal 63. . Further, an input end of a line memory 64 having a capacity sufficient to store a video signal within one horizontal line period is connected between the output end 40b of the FIR filter 40 and the input end 50a of the IIR filter 50. The output of the memory is connected to the input bus of the DSP 65. DSP 65 is, for example, 7M5320C25 manufactured by Texas Instruments (this DSP is
It is extremely fast as it can perform multiplication, addition, and data fetching simultaneously in a single cycle, and has a chip-on-ROM for storing programs and data. RAM used to store data and as a working area 6
6 is connected. Further, the output bus of this DSP 65 is connected to filter coefficient input terminals of the FIR filter 40 and the IIR filter 50, etc.

As shown in FIG. 8, the FIR' filter 40 includes a variable delay element 42 that delays the input video signal by a time set by the DSP 65 and outputs the delayed signal, and an output of the variable delay element and the input video signal. A transduction filter 43 receives signals at terminals a and b, respectively.
-0, and three transversal filters 43-1, 43-2, and 43-3 connected in cascade to this transversal filter 43-0 via their output terminals C and d. The above transformer filter 43-
0.43-1.43-2.43-3 each have a similar configuration, and the detailed configuration of one of them is shown in FIG.

In FIG. 9, the input terminal a (8-bit input) of the transversal filter has a one-clock delay element 4.
4 input terminals are connected, and the output terminals of this delay element are connected to 16 multipliers 46-0, 46-1, 46-2, . . . 461.
5 and is also connected to the output terminal C. Each of the above multipliers 46-0.46-1゜46-
2, . . 46-15 are supplied with filter coefficients from the DSP 65 as multiplication coefficients. Furthermore, between the other input terminal b (16-bit input) and the other output terminal d of this transversal filter, there are 17 one-clock delay elements 47-0.47-1.47-2, . . .・47-1
5.4716 adders 48-0, 48-1 between them
．． 48-2, . . . 48-15 are inserted in a cascade connection. Further, the multiplier 46-0.46-1.
Each output of 46-2, . . . 46-15 is transmitted to each corresponding adder 48-0.
8-15, respectively.

Returning again to FIG. 8, the output terminal 40b1 of the FIR filter 40, that is, the output terminal d1 of the transversal filter 43-3 is connected to the input terminal of the line memory 64, and also to the input terminal 50a of the IIR filter 50. ing. This input terminal 50a is connected to the input terminal of a transversal filter 51-0, and seven transversal filters 51-1, 51-2, . . . 51-7 are connected in series to this transversal filter 51-0. It is connected. In this case, the transversal filter 51
-0.51-L Each of 51-2,...51-7 is
It has a configuration similar to that shown in FIG. The output terminal d of the transversal filter 51-7 is
Limiter 5 that limits 16-bit information to 8-bit information
2 to the output terminal 50b of the IIR filter 50. Further, the output of the limiter 52 is transmitted through variable delay elements 53-0.53-1.53-2, . . . 53-
7 through the transversal filters 51-0, respectively.
The signals are fed back to the input terminals a of 51-1, 51-2, . . . 51-7, respectively. In this case, the variable delay elements 53-0.53-1.53-2,...53-
7, a delay amount is set by the DSP 65, respectively.

Next, the operation of this embodiment having the above configuration will be explained with reference to the flowchart shown in FIG.

When a falling edge of the vertical synchronization signal of an odd field in the received video signal is detected by a circuit not shown,
The program shown in FIG. 10 is started by the DSP 65,
First, in block 100, initialization processing is performed.

In this block 100, a value "0" is set as a filter coefficient in the FIR filter 40 and the IIR filter 50, respectively. That is, the DSP 65 includes transversal filters 43-0 to 43-3 and 51-0.
Each of the multipliers 46-0 to 46- in the multipliers 46-0 to 51-7
The value "0" is set to 15 as a multiplication coefficient. Next, in block 101, a counter CNT is set to "0", and then the process proceeds to block 102. In block 102, the low-pass filter coefficient LPF (k) previously stored in the chip-on ROU of the DSP 65 is set as a filter coefficient in the FIR filter 40 immediately before the 18th horizontal period of the field. Next, in block 103, the data stored in the line memory 64, that is, the G from which high-frequency noise has been removed by the filter 40 based on the above-mentioned low-pass filter coefficient LPF (k).
The CR signal (or pedestal signal) is stored in RAM 66. Then, at block 104, the FIR
Each filter coefficient of the filter 40 is returned to the value immediately before the block 102, and "1'" is added to the counter CNT. Next, in block 105, it is determined whether or not the counter CNT has reached, for example, a multiple of "8""8N". If it has not reached it, the process returns to block 102, and if it has reached it, the process proceeds to the next block 106. move on. That is, in blocks 102 to 105 described above, for example, BT
In order to apply the A-standard 8-field sequence method, GCR signals and pedestal signals for consecutive multiple fields of 8 are collected. Next, in block 106, the GCR signal and pedestal signal collected as described above are processed based on the 8-field sequence method to obtain an averaged GCR signal, and further,
The GCR signal obtained in this way and one of the GCR signals
Take the difference from the signal shifted by the clock (i.e.,
(differentiating the GCR signal) to obtain a sin x/X signal x (k) that has suffered ghost damage. The signal×(k) thus determined is stored in the RAM 66.

Next, in block 107, the time series information r (k) of the reference sin X/X signal stored in advance in the chip-on ROM of the DSP 65 and the above-mentioned signal The filter coefficients h(k) for the FIR filter 40 (64 values in this example, and hereinafter these filter coefficients are expressed as fi
r(k)) are calculated and temporarily stored in the RAM 66. Proceeding next to block 108, x(
When the signal x' (k) is manually applied to an FIR filter in which a filter coefficient of fir(k) is set, the signal x' (k) is output from the FIR filter.
) is calculated by calculation. That is, the DSP 65 calculates the following for each k after point Q in FIG. x' (k) obtained in this way
is temporarily stored in RAM 66. Next block 109
In this step, the DSP 65 obtains a predetermined number (eight in this example) of peaks from the above x' (k), and calculates a predetermined number (16 in this example) of x' consecutively centering on each of these peaks. (k) (These data become filter coefficients of the IIR filter 5o, and will be expressed as i ir (k) below). Then, the DSP 65 extracts each of the 16 pieces of data 1ir (k).
) are stored in the RAM 66. In addition, the DSP 65 calculates the I/O value based on each time from the 0 point in FIG.
Each delay amount d1y7, dly6, dly for IR filter
5, ..., find the old yO, and store each of these delay amounts in RAM.
66.

Next, in block 110, each of the filter coefficients fir(k) and 1ir(k) calculated as described above and the delay amount dly? , dly6, dly5, ..., dly
O in a predetermined horizontal period within the vertical emission line erasure period.
The IR filter 40 and the IIR filter 50 are set respectively. That is, the filter coefficient f ir (k) is determined by the multiplier 46 of each transversal filter 43-0 to 43-3.
-0 to 46-15 are respectively set as multiplication coefficients. Further, the delay amounts dly7 to dtyo are set to the variable delay elements 53-7 to 53-0, respectively, and the values obtained by inverting the sign of each data of the filter coefficient 1ir(k) are set to the respective transversal filters 517 to 51-0. Multiplier 4
6-0 to 46-15 are respectively set as multiplication coefficients.

When the above processing is completed, this program will block 1
01, and thereafter the same processing as described above is repeated.

Note that in the embodiment shown in FIG.
4 is provided on the output side of the FIR filter 40. However, when calculating the output of the FIR filter 40 in which the filter coefficient for ghost removal is set by a program as in the above embodiment, this line memory may be provided on the output side of the FIR filter 40. A line memory 64' shown by a dotted line in the figure is provided on the input side of the FIR filter 40, and the receiving GC is provided on the upstream side of the FIR filter 40.
The waveform of the R signal may also be captured. However, in this case, the FIR filter 40 cannot be used to perform other filter processing on the received GCR signal. Therefore, if the waveform of the received GCR signal is to be taken in after its high frequency noise components are removed, it is desirable to provide the line memory 64 downstream of the FIR filter 40.

[Brief explanation of drawings]

FIG. 1 is a block diagram for explaining the operating principle of the ghost removal circuit according to the present invention, FIG. 2 is a waveform diagram of a received ghost removal reference signal that has been affected by ghost interference, and FIG. A waveform diagram of the ghost removal reference signal stored in advance in the memory of the ghost removal circuit according to the invention; FIG. 3(b) is a frequency characteristic diagram of the ghost removal reference signal; FIG. FIG. 5 is a waveform diagram of the received ghost removal reference signal output from the ghost removal circuit according to the present invention, and FIG. 6 is a waveform diagram of the received ghost removal reference signal according to the embodiment of the present invention. FIG. 7 is a block diagram showing the configuration of another embodiment of the ghost removal circuit according to the present invention. FIG. 8 is a detailed configuration of the acyclic filter and the cyclic filter in the same embodiment. FIG. 9 is a block diagram showing the detailed configuration of the transversal filter that constitutes these filters, and FIG.
8th I! 1 is a flowchart for explaining the operation of the embodiment of FIG.

Claims (1)

[Claims] 1. A ghost removal circuit configured by cascading an acyclic filter and a cyclic filter, comprising: a first memory that stores a received ghost removal reference signal in an input video signal; a second memory storing a ghost removal reference signal; and a first memory for the acyclic filter based on the received ghost removal reference signal in the first memory and the ghost removal reference signal in the second memory. calculation means for calculating a filter coefficient group and setting the first filter coefficient group in the acyclic filter; and calculating means for calculating the first filter coefficient group in the acyclic filter; a filter means for obtaining a filtered received ghost removal reference signal by passing the filtered received ghost removal reference signal through the cyclic filter; A ghost removal circuit comprising: a setting means for setting a second filter coefficient group to a second filter coefficient group. 2. The ghost removal circuit according to claim 1, wherein the first memory includes a waveform capture memory for capturing a predetermined number of received ghost removal reference signals and at least one waveform capture memory for capturing a predetermined number of received ghost removal reference signals. and a buffer memory for storing one received ghost removal reference signal that has been averaged, and the filter means converts the input video signal and the received ghost removal reference signal in the buffer memory into the acyclic filter. 1. A ghost removal circuit comprising switch means for selectively supplying a signal to a type filter. 3. In the ghost removal circuit according to claim 2, the waveform capture memory is provided on the output side of the acyclic filter, and when the received ghost removal reference signal is captured into the waveform capture memory, the acyclic filter The filter includes a third filter such that the acyclic filter has predetermined low-pass filter characteristics.
A ghost removal circuit characterized in that a group of filter coefficients are set. 4. In a ghost removal circuit configured by cascading an acyclic filter and a cyclic filter, the first memory stores a received ghost removal reference signal in an input video signal, and a predetermined ghost removal reference signal is stored. calculating a first filter coefficient group for the acyclic filter based on the received ghost removal reference signal in the first memory and the ghost removal reference signal in the second memory; and a first calculation means for setting the first filter coefficient group to the acyclic filter, and a first calculation means for setting the first filter coefficient group to the first filter coefficient group based on the received ghost removal reference signal and the first filter coefficient group. a second calculation means for calculating an output of the acyclic filter with respect to the received ghost removal reference signal; A ghost removal circuit comprising: setting means for setting a second filter coefficient group in the recursive filter in a time relationship. 5. The ghost removal circuit according to claim 4, wherein the first memory is provided on the input side of the acyclic filter. 6. The ghost removal circuit according to claim 4, wherein the first memory is provided on the output side of the acyclic filter, and when the received ghost removal reference signal is taken into the first memory, the acyclic A ghost removal circuit characterized in that a third filter coefficient group is set in the filter so that the acyclic filter has predetermined low-pass filter characteristics.