JPH0457576A - Ghost elimination device - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed Description of the Invention] [Object of the Invention] (Industrial Application Field) The present invention relates to a ghost removal device, and more particularly, to a ghost removal device that enables stable operation even when a fluttery ghost is mixed in. Regarding.

(Prior Art) In recent years, in television broadcasting, a GCR (ghost cancellation reference) signal has been superimposed and transmitted as a reference signal for ghost removal. G
The CR signal is composed of a 5inx/x bar waveform and a pedestal waveform, and is inserted in an 8-field sequence.The ghost removal device extracts this OCR signal from the television signal, performs ghost I/removal, waveform, etc. I am trying to do this.

FIG. 5 is a block diagram showing such a conventional ghost removal device, and FIG. 6 is a flowchart for explaining its operation.

An input video signal that has been subjected to ghost interference or the like is input to an input terminal 1. An OCR signal is inserted into this input video signal. The input video signal is converted into a digital signal by an A/D converter 2, and then input to a transversal filter (hereinafter referred to as TF) 3 for waveform equalization, where the waveform is equalized.

First, in step S1 in FIG. 6, waveform capture is performed. The digital video signal from the A/D converter 2 is input to the memory 4, and the memory 4 extracts the GCR signal included in the input video signal at the timing of the timing signal from the timing circuit 5. On the other hand, the video signal output from the TF 3 is converted into an analog signal by the D/A converter 6 and output from the output terminal 7.
The memory 8 is also provided. Memory 8 is timing circuit 5
The GCR signal (sinx/x bar waveform) included in the output video signal is extracted by calculating an 8-field sequence at the timing of the timing signal from. Memory 4, 8
The signals read from the circuits are respectively applied to difference circuits 9 and 10 for difference, and an input/output difference signal (sjnx/x waveform) is obtained. The input difference signal is given to the tap coefficient control port i13, and the output difference signal is given to the subtraction circuit 11.

In the next step S2, an error meter 9 is performed. The reference signal generator 12 generates a reference signal having the same waveform as the GCR signal (sin x / It is given to 11. The subtraction circuit 11 subtracts this reference signal and the output difference signal to obtain an error signal and outputs it to the tap coefficient control circuit 13.

In the next step S3, a correlation calculation is performed. The tap coefficient control circuit 13 performs a correlation calculation between the input difference signal and the error signal to obtain a new tap coefficient.

In the next step S4, the tap coefficient from the tap coefficient control circuit 13 is given to TF3, and the tap coefficient of TF3 is corrected. TF3 performs waveform equalization of the input video signal based on the new tap coefficients and outputs an output video signal.

Thereafter, steps S1 to S4 are repeated. Through these steps S1 to S4, tap coefficients are generated based on the magnitude of the error signal, that is, tap coefficients such that the error signal converges to 0, and the input video signal is waveform-equalized. In this way, an output video signal from which ghosts have been removed is output to the output terminal 7.

Note that the correlation calculation, tap coefficient correction, etc. are controlled by the CPU 14. CPU 14 is address bus 1
7. It is connected to the program ROM 15 and the work RAM 16 via the data bus 18 and the control bus 19, and executes various processes using the work RAM 16 based on the programs stored in the program ROM 15.

In this way, the tap coefficients are determined by the tap coefficient correction calculations in steps S1 to S4 described above, that is, the goth I.
It is corrected and updated at predetermined time intervals by component extraction calculations, correlation calculations, and the like. When the variation in the ghost amount is relatively small, the tap coefficient correction calculation can follow the ghost variation amount, and a tap coefficient corresponding to the ghost is created. However, when the amount of ghost fluctuates rapidly, such as with flutter ghosts, the tap coefficient correction calculation cannot follow the amount of ghost fluctuation, making correction of the tap coefficient unstable and stopping the ghost removal operation. There was a problem with this.

(Problem to be Solved by the Invention) As described above, in the conventional ghost removal device described above, when a fluttery ghost is mixed in the input video signal, correction of the tap coefficient becomes unstable and the ghost removal operation becomes unstable. There was a problem that it would stop.

The present invention has been made in view of such problems, and includes:
It is an object of the present invention to provide a ghost removal device that can stably obtain tap coefficients even when flutter ghosts are mixed in.

[Structure of the Invention] (Means for Solving the Problems) A ghost removal device according to the present invention includes a transversal filter that performs waveform equalization based on tap coefficients, and a transversal filter that is included in an input/output video signal of the transversal filter. and a tap coefficient correction calculation means for correcting the tap coefficient by calculation using a reference signal and an internal reference signal from a reference signal generator, wherein the ghost removal apparatus includes a tap coefficient correction calculation means for correcting the tap coefficient by calculation using a reference signal and an internal reference signal from a reference signal generator. a first memory for extracting and storing a reference signal included in the input video signal; and a first memory for extracting and storing a reference signal included in the input video signal, the data from the first memory can be transferred, and the data is stored after ghost removal is stabilized; a second memory that provides data to the tap coefficient correction calculation means; a third memory that extracts and stores the reference signal included in the output video signal at predetermined intervals; and the first and second memories. a subtraction circuit that calculates a difference value between the stored data;
subtracting means for subtracting a reference signal from the first memory; and a subtracting means for subtracting the reference signal from the first memory, and comparing the difference value with a predetermined value to judge the slowness or speed of ghost fluctuation and transfer the data from the first memory to the second memory. control, and if it is determined that the ghost fluctuation is slow, the data in the first memory is transferred to the second memory.
If it is determined that the ghost fluctuation is rapid, the data in the second memory is directly transferred to the tap coefficient correction calculation means without being transferred to the first memory, and the data is transferred to the tap coefficient correction calculation means. and control means for providing an output to the tap coefficient correction calculation means.

(Operation) In the present invention, the reference signal extracted by the first memory is provided to the tap coefficient correction calculation means until the ghost removal operation becomes stable. Once the ghost removal operation has stabilized, the second memory captures the initial reference signal from the input video signal. The first memory always takes in the reference signal from the input video signal, and transfers the data of the taken in reference signal to the second memory based on instructions from the control means. Based on the difference value between the first and second memories obtained by the subtraction circuit, the control means determines whether the ghost change is slow or rapid. If it is determined that the ghost fluctuation is slow,
The control means reads out the reference signal data transferred from the first memory from the second memory and supplies it to the tap coefficient correction calculation means. As a result, ghost removal similar to the conventional method is performed.

If it is determined that the ghost fluctuation is rapid, the control means transfers the data from the first memory to the second memory without transferring the data from the first memory.
The stable reference signal data stored in the memory before the sudden ghost fluctuation is given to the tap coefficient correction calculation means. This stabilizes the calculation by the tap coefficient correction calculation means. Further, in this case, the control means provides the subtraction value between the difference value from the subtraction means and the data in the third memory to the tap coefficient correction calculation means, so that the variation of the reference signal included in the output video signal is canceled out.

(Example) Hereinafter, an example of the present invention will be described in detail based on the drawings. FIG. 1 is a block diagram showing an embodiment of a ghost removal device according to the present invention. In FIG. 1, the same components as in FIG. 5 are given the same reference numerals.

An input video signal is input to the input terminal 1.

This input video signal is applied to TF3 via A/D converter 2. TF3 is composed of a delay device, a multiplier, and an adder (not shown), and it equalizes the waveform of the input video signal based on the tap coefficient given to the multiplier and converts it into a D/A.
The signal is output to a converter 6 and a memory 8. D
/A converter 6 converts the digital output video signal into an analog signal and outputs it to output terminal 7.

In this embodiment, the input video signal from A/D converter 2 is also provided to memory 20.21. The memory 20 extracts and stores the GCR signal (sinx/x bar waveform) included in the input video signal every predetermined period by performing an 8-field sequence operation at the timing of the timing signal, and outputs the stored data. . In addition,
The memory 20 is configured to transfer stored data to the memory 21 according to instructions from the CPU 14, which will be described later. The timing circuit 5 is controlled by the CPU 14 to output a timing signal synchronized with the input video signal. The memory 21 is connected to the CPU 1 as described later.
4, when it is determined that the tap coefficient is stable, the OCR signal included in the input video signal is extracted at the timing of the timing signal. Furthermore, the memory 21 is the memory 20
It stores and outputs the data transferred from GCR (No. 3).

The memory 8 extracts and stores the GCR signal (sinx/x bar waveform) included in the output video signal in accordance with the timing signal from the timing circuit 5, as in the conventional case.

The data read from the memory 20.21 is sent to the subtraction circuit 2.
3 and also to the differential circuit 9 via the selector 22. The subtraction circuit 23 subtracts the output data of the memory 20.21 to obtain a difference value Δ and outputs it to the selector 24. The selector 22 is controlled by the timing signal of the timing circuit 5, and the memory 20
．． 21 is selectively applied to the differential circuit 9. That is, the selector 22 selects the memory 20 immediately after power is turned on and immediately after channel switching, and selects the memory 21 when the tap coefficient becomes stable. On the other hand, the selector 24 is controlled by the timing signal to convert the difference value Δ from the subtraction circuit 23 into CP tJ
14, and if the CPU 14 determines that a sudden ghost change has occurred, it also outputs the difference value Δ to the subtraction circuit 25. Subtraction diagram 1
The i125 subtracts this difference value Δ and the GCR signal from the memory 8 and outputs the result to the difference circuit 10.

The difference circuits 9 and 10 each perform a difference calculation to output an input difference signal and an output difference signal.

The input difference signal is given to the tap coefficient control circuit 13, and the output difference signal is given to the subtraction circuit 11. Subtraction circuit 11
subtracts the reference signal from the reference signal generator 12 and the output difference signal to obtain an error signal and outputs it to the tap coefficient control circuit 13. The tap coefficient control circuit 13 calculates a new tap coefficient by calculating the correlation between the input difference signal and the error signal, and outputs it to the TF3. Note that the CPU 14 has an address bus 17 and a data bus 1.
8 and the program RO1 via the control bus 19.
5 and the work RAM 16 as in the conventional case.

Next, control of the selectors 22 and 24 by the CPU 14 will be explained with reference to FIGS. 2 and 3. Figures 2 and 3
The figure is a graph showing the fluctuation of the amount of ghost, with time on the horizontal axis and gain on the vertical axis. FIG. 3 is an enlarged view of the phase of FIG. 2 near 90 degrees or 270 degrees on the time axis.

As mentioned above, the GCR signal is inserted in an 8-field sequence, and a GCR signal is inserted every 8 fields (8/60 seconds).
A difference signal of the CR signal waveform is obtained. Assuming that the fluctuation of the ghost shown in Fig. 2 is represented by a sin curve, 8
The ghost phase change θ[deΩ] during the field period can be expressed by the following equation (1).

360deg: θde (1 = 60 seconds: 8/60 seconds,
In the sine curve, as shown in FIG. 3, the amount of change per unit time is maximum at phases of 90 degrees and 270 degrees. The amount of change α in gain in a minute period of 8760 seconds (the amount of change in phase is 0.8 [deg]) near 90 degrees or 270 degrees can be approximately expressed by the following equation (2).

37 dB... (2) That is, in the relatively slow ghost fluctuation shown by the sine curve, the maximum value of the amount of change in gain during the 8-field period is α - 37 dB as shown in equation (2) above. .

The CPU 14 compares the difference value Δ transferred from the selector 24 with α in equation (2) above. If the difference value Δ is smaller than α, it is determined that the ghost fluctuation is relatively slow, the tap coefficient correction calculation is determined to be able to follow the ghost fluctuation, and the same operation as before is specified. do. That is, in this case, the CPU 14 causes the selector 22 to select the memory 21 after the tap coefficients are stabilized, and also causes the data in the memory 20 to be sequentially transferred to the memory 21 and provided to the difference circuit 9. Further, the CPU 14 controls the selector 24 to select O so that the difference value Δ is not provided to the subtraction circuit 25.

Conversely, if the difference value Δ is larger than α, the CPU 14
It is determined that the ghost fluctuation is relatively rapid, and that the tap coefficient correction calculation cannot follow it using the same operation as the conventional one. In this case, the CPU 14 causes the selector 22 to select the memory 21 after the tap coefficients are stabilized, and provides the data in the memory 21 as is to the difference circuit 9. Furthermore, since a sudden change in the input OCR signal also appears in the output OCR signal, the amount of change in the output OCR signal Δ
In order to cancel the difference, the CPU 14 controls the selector 24 to provide the difference value Δ to the subtraction circuit 25.

Next, the operation of the ghost removal apparatus configured as described above will be explained with reference to flowchart 1 in FIG. 4.

First, when the power is turned on or the channel is switched, the CPU 14 causes the selector 22 to select memory 2°, and causes the selector 24 to stop supplying the difference value Δ to the subtraction circuit 25. As a result, the same operation as before is performed. That is, the input video signal is converted into a digital signal by the A/D converter 2, and then the ghost signal is extracted in the memory 20.

The output of the memory 20 is applied to the difference circuit 9 via the selector 22, and the input difference signal is applied to the tap coefficient control circuit 13. - On the other hand, the GCR signal included in the output video signal from TF 3 is extracted by memory 8, and an output difference signal is obtained by difference circuit 10 and given to subtraction circuit 11. The subtraction circuit 11 is a reference signal generator 12
The reference signal and the output difference signal are subtracted to obtain an error signal and provided to the tap correction control circuit 13. The tap correction control circuit 13 performs a correlation calculation between the input difference signal and the error signal, generates a new tap coefficient, and supplies it to the TF3. Thereby, TF3 performs waveform equalization of the input video signal.

In this way, when the ghost component is reduced and the tap coefficient is stabilized, the CPU 14 sets the initial value O to the number of loops n in step S5, and the memory 21 in the next step S6.
The input OCR signal included in the input video signal is captured in the memory 8, and the output OCR signal included in the output video signal is captured in the memory 8.
Capture the CR signal. CPU 14 is selector 22
selects the memory 21, and the data of the GCR signal stored in the memory 21 is applied to the difference circuit 9 to obtain an input difference signal.

In the next step S7, the number of loops n is increased by 1, and then in step S8, the GCR is stored in the memory 20°8.
Capture the signal. Next, in step S9, the memory 20
．． Compare the output data of No.21. The outputs of the memories 20 and 21 are given to a subtraction circuit 23, and the subtraction circuit 23 calculates the difference between the two and sends the difference value Δ to the CP via the selector 24.
Output to U14. The CPU 14 performs the next step S10.
It is determined whether the difference value Δ is larger than α. If the difference value Δ is smaller than α, it is determined that the amount of ghost variation is relatively small, and the data in the memory 20 is transferred to the memory 21 in the next step S11. Next, in step S12, the difference value Δ is set to 0, and then the next step S1
Move to 3. In step S13, CP LJ 14
makes the selector 22 select the memory 21, and the data transferred from the memory 20 to the memory 21 is given to the difference circuit 9 to obtain an input difference signal. In addition, step S
12, the difference value Δ is set to 0, so step S
In step S13, the G and CR signals from the memory 8 and the difference value Δ obtained in step S9 are not added.

Steps S14 to S17 are normal equalization loops,
Using the input difference signal, output difference signal, and reference signal output from the difference diagrams B9 and B10, error calculation and correlation calculation are performed to obtain tap coefficients and provide them to TF3.

In this way, TF3 performs waveform equalization using new tap coefficients. Then, the process returns to step S7 and repeats the process. The memory 20 takes in the OCR signal for each loop, and the data in the memory 20 is applied to the difference circuit 9 via the memory 21 and the selector 22, and the same operation as in the prior art is performed.

On the other hand, if the difference value Δ is larger than α in step S10, the CPU 14 determines that the amount of ghost variation is relatively large, and moves the process to step S13.

That is, the data from the previous loop is stored in the memory 21, and this data is provided to the difference circuit 9 via the selector 22. Further, the difference value Δ is given to the subtraction circuit 25 via the selector 24 and added to the data in the memory 8. As a result, the output G for the previous loop
Fluctuations in the CR signal are canceled out. In this way, difference diagram 19
10 outputs an input/output difference signal based on the OCR signal from the previous loop. Subsequent steps S14 to S1
The operation of step 7 is the same as last time.

As described above, in this embodiment, by determining the ghost variation amount (difference value Δ) of the input video signal and comparing it with α in equation (2), it is possible to determine whether the tap coefficient correction calculation can follow the ghost variation. I am deciding whether or not to do so. If the ghost fluctuations are relatively slow, the same operation as before is instructed, and if the ghost fluctuations are rapid, the OCR signal from the previous loop is used to stabilize the state before the fluctuation. Therefore, it is possible to always perform stable tap coefficient correction calculations regardless of the amount of ghost fluctuation. Therefore, even if flutter ghosts are mixed in,
The ghost l removal operation never stops.

[Effects of the Invention] As explained above, according to the present invention, even when flutter ghosts are mixed, tap coefficients can be stably obtained.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of a ghost removal device according to the present invention, and FIGS. 2 and 3 are a block diagram of a CPU 1 in FIG.
4 is a flow chart for explaining the operation of the embodiment, FIG. 5 is a block diagram showing a conventional ghost removal device, and FIG. 6 is a graph for explaining the operation of the conventional example. This is a flowchart for 1...Input terminal, 3...TF, 8.20.21...
- Memory, 9.10... Differential circuit, 11.23... Subtraction circuit, 12... Reference signal generator, 13 Tap coefficient control circuit, 14... CPU, 22
, 24...Selector, 25...Subtraction circuit. Figure 2 Figure 6

Claims (1)

[Claims] A transversal filter that performs waveform equalization based on tap coefficients, a reference signal included in the input/output video signal of this transversal filter, and an operation using an internal reference signal from a reference signal generator. A ghost removal device comprising a tap coefficient correction calculation means for correcting the tap coefficient, the first memory extracting and storing a reference signal included in the input video signal at predetermined intervals; The included reference signal is extracted and stored, and the first
a second memory to which data from the memory can be transferred and which supplies the stored data to the tap coefficient correction calculation means after ghost removal has stabilized; and a second memory which extracts a reference signal included in the output video signal at predetermined intervals. a subtraction circuit that calculates a difference value between the data stored in the first and second memories; and a subtraction circuit that subtracts the difference value from the reference signal from the third memory. means, determining whether the ghost fluctuation is slow or slow by comparing the difference value with a predetermined value, controlling the transfer of data from the first memory to the second memory, and determining that the ghost fluctuation is slow. If it is determined that the ghost change is rapid, the data in the first memory is transferred to the second memory, and if it is determined that the ghost fluctuation is rapid, the data in the second memory is transferred without transferring the first memory. 1. A ghost removal device comprising: a control means that applies the output of the addition means to the tap coefficient correction calculation means as it is, and also applies the output of the addition means to the tap coefficient correction calculation means.