JPH0488754A - Ghost elimination device - Google Patents

Abstract

PURPOSE:To eliminate ghost by stopping tentatively tap correction of a transversal filter against a large ghost fluctuation in a short time and initializing the tap against a permanent fluctuation. CONSTITUTION:A digitized signal is inputted to a transversal filter 2. An input signal and an output signal of the filter 2 are fetched in waveform memories 4A, 4B by the DMA operation for a period of a signal generated by a synchronizing signal resulting synchronization separating an input video signal. The fetched signals are subject to field sequence processing by a CPU 5 and a difference between the result and a ghost reference signal is calculated. The CPU 5 calculates the absolute total sum of the tap correction per one sequential equalization of the filter 2 and stores the result to a tap correction quantity total sum memory 6. Then a flag representing ghost fluctuation is cleared. Then the 2nd sequential equalization is proceeded by fetching the signal again. When a prescribed value exceeds 2, the tap coefficient of the filter 2 is cleared and ghost elimination is retried from the beginning.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to a ghost removal device that performs ghost removal using a ghost removal reference (C; CR) signal.

Conventional Technology First generation HDTV broadcasting, which aims to achieve high image quality while maintaining compatibility with current television systems, is about to begin, and ghost removal is attracting a lot of attention. Among the ghost removal performance requirements, there are items such as image quality evaluation after improvement and removal time.In other words, this means how quickly the ghost can be removed by reducing the amount of ghost remaining after removal. become.

An example of a conventional ghost removal device is "Television Society Technical Report RE80-6.p p, 9-14, 1972.
There is a ghost canceller reported in February 2016.

This detects ghosts by using the differential signal of the leading edge of the vertical synchronization signal unique to television signals as a reference waveform, and performs correlation calculations on the time axis using the detected ghost signals to perform transformer detection. Ghosts are removed by sequentially correcting the tap coefficients of the versal filter.

In addition, as a ghost removal device using OCR signals,
Television Television Society Technical Report ROF Ta2-6, P
There is a ghost canceller reported in P. 3l-36J. This uses a transversal filter in the ghost canceller like the ghost canceller described above, but the input and output of the transversal filter are taken into the CPU via memory, synchronized, and inter-fields are processed according to the transmission sequence. Perform all ghost removal calculations including processing in .

An example of a conventional ghost removal device will be described below with reference to the drawings. FIG. 3 is a schematic block diagram showing the configuration of a conventional ghost removal device. In Figure 3, 1
3 is CPU, 10 is transversal filter, 9 is A
11 is a D/A converter, and 12A and 12B are waveform memories.

The operation of the ghost removal device configured as above will be explained. The input video signal is sent to the A/D converter 9
The signals are A/D converted and input to the transversal filter 10 and the waveform memory 12A, respectively. The input and output of the transversal filter 10 are connected to the waveform memory 12.
It is input to the CPU 13 via A and 12B. In the first generation HDTV broadcast, the WR shown in Figures 4(a) and (c)
The B signal and 0 pedestal signal are sent in the same horizontal period in a sequence that goes around in 8 fields: WRB signal → 0 pedestal signal → WRB signal → 0 pedestal signal → O pedestal signal => WRB signal → 0 pedestal signal => WRB signal. Ru. By performing the calculation shown in equation 1 below on the signals of these 8 fields, the signal shown in Figure 4 (C3) can be obtained. However, Fn (n = 1 to 8) is the signal of the nth field. Hereinafter, the processing between fields according to the sending sequence as shown in the first equation will be referred to as the processing between fields according to the field sequence as field sequence processing.

F-1/4 ((Fl-F5)+(F6-F2)+(F
3-F T)+(F B-F4))...
・(1) In general, the following ghost removal calculation is performed using the signal shown in FIG. 4(d) obtained by subtracting the signal in FIG. 4(C) by one clock as the reference signal for ghost detection. MSE (Mean 5quare Er
ror ) method or ZF (Zero Forcin)
g) method, etc., which sequentially correct on the time axis according to a certain algorithm to finally find the optimal tap coefficient. If the output signal of the transversal filter is (Yk), the reference signal is (Rkl), the difference signal between the output signal of the transversal filter and the reference signal (Ekl), and the total number of taps is M+N+1, then the nth tap coefficient of the transversal filter is C(i)la+ is modified based on the following equation 2 in the MSE method and equation 3 in the ZF method. However, αβ is a coefficient for determining the amount of correction.

(Ci) ""' = (C4l "'-α・ΣYk-4
-EK... (2) After performing the can processing, the second
This series of processing, in which the tap coefficients are repeatedly corrected by calculating the formula or the third formula, is performed by software.
In ghost detection, processing is repeated until the amount of residual ghost becomes sufficiently small.

Problems to be Solved by the Invention In view of the above-mentioned problems, the present invention provides stable removal time even when ghosts fluctuate in a short period of time or when the state of occurrence of ghosts changes over time over a long period of time. This provides a short ghost removal device.

Ghosts generally rarely change over a short period of time, but it is known that the state of a ghost changes over a long period of time. A method of intermittently performing control at fixed time intervals has been used to ensure stability of control over long periods of time (Kibu Publication No. 16072/1983), but it is difficult to predict when flutter will occur due to aircraft. Equations (2) and (3) are used when the situation of ghosts is unpredictable and changes in a short period of time.
If control is performed according to the formula, the tap gain of the transversal filter may increase extremely, resulting in a divergent state.

Means for Solving the Problems In order to solve the above problems, the ghost removal device of the present invention includes a method that compares the sum of absolute values of tap correction amounts for each round of successive equalization with the sum of absolute values of tap correction amounts for the previous round. However, if the ratio exceeds a certain value, the tap correction operation of the actual transversal filter is stopped, and after a predetermined number of times, the absolute value sum of the tap correction amount is calculated, and the tap correction operation of the actual transversal filter is stopped. This is a means of comparing the total absolute value of the tap correction amount immediately before the time, and is configured to be able to determine whether the ghost has caused a short-term change or a large constant change.

Operation The present invention has the above-mentioned structure. The transversal filter's tap correction operation is temporarily stopped in response to large ghost fluctuations in a short period of time, such as during two evenings, to prevent unstable operation, and the transversal filter is initialized in response to large constant fluctuations. Since ghost removal is performed, the removal operation can be performed stably and quickly.

Embodiment Hereinafter, a ghost removal device according to an embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a schematic block diagram of the circuit configuration of a ghost removal device in an embodiment of the present invention. In FIG. 1, l is an A/D converter, 2 is a transversal filter, 3 is a D/A converter, 4A and 4B are waveform memories, 5 is a CPU, 6 is a tap correction amount summation memory, and 7 is a comparison controller. Memory 8 is a number counter.

The operation of the ghost removal device configured as described above will be described below with reference to FIG. 2.

An input video signal is input to the A/D converter l.

Here, the A/D converted and digitized signal is input to the transversal filter 2. The input and output signals of the transversal filter 2 are 18H and 281H generated from synchronous signals obtained by synchronously separating the input video signal.
During the period of the signal indicating , the waveform memory 4 is
It is taken into A and 4B.

The captured signal is subjected to field sequence processing by the CPU 5, the difference with the ghost reference signal is calculated, and the arithmetic processing of equation (3) is performed. The absolute sum of the correction amounts is: (4) N-th sequential equalization i: Tap N.

m: Calculate according to the total number of taps and store the value in the tap correction amount summation memory 6. At the first time of sequential equalization, the (3)
According to the result of the expression, a flag indicating a first-order ghost variation that modifies the tap coefficient of the transversal filter is cleared. Thereafter, the signal is taken in again and the second successive equalization operation is started. Sequence processing, ghost detection, Part (3)
Perform the tap correction calculation of the formula, perform the calculation of the formula (4), and then calculate SAM (n+1)/SAM (n)-D with the contents of the tap correction amount total memory 6.
5] is calculated.

When the value of D is smaller than the predetermined value X (preferably around 2), the tap coefficient is corrected according to equation (3). If it is large, it is determined that the ghost is changing, and the flag indicating that the ghost is changing is set to "1".At the same time, the contents of the tap correction amount summation memory 6 are transferred to the comparison value memory 7 and set as a comparison value PSAM.
The number counter 8 is incremented, and the process moves to a capture operation. When the ghost fluctuation flag is "l", only the total tap correction amount is calculated according to equation (4), and the value of the zero count counter 8 is monitored to repeat the operation of incrementing the count counter 8, and the value is When the predetermined value K is reached, calculate SAM/PSAM, and if the value does not exceed X, at that point, the tap coefficient of the transversal filter 2 is corrected based on the result of equation (3), and the ghost removal operation is performed. Calculate SAM/PSAM when the predetermined number of times is reached, and if the value exceeds X, calculate only the total tap correction amount according to equation (4), and increment the number counter When the value of the counter 8 reaches 2K, SAM/PSAM is calculated again, and if the value exceeds X, the tap coefficient of the transversal filter 2 is multiplied by 17, and the first Redo the ghost removal operation.

Effects of the Invention As described above, according to the present invention, the sum of the absolute values of the tap correction amounts is compared with the sum of the absolute values of the immediately preceding tap correction amounts,
If the ratio exceeds a certain value, the tap correction operation of the actual transversal filter is stopped, and after a predetermined number of times, the absolute value sum of the amount of correction is calculated, and the tap correction operation of the actual transversal filter is stopped. Ghost fluctuations are determined by comparing the sum of absolute values of tap correction amounts immediately before the time, and the tap correction operation of the transversal filter is temporarily stopped in case of large ghost fluctuations in a short period of time due to flutter, etc. This prevents unstable operation, and in the case of large constant fluctuations, ghost removal is performed after initialization, making it possible to perform ghost removal that is stable and can respond to constant ghost changes. Become.

[Brief explanation of the drawing]

FIG. 1 is a schematic block diagram of a ghost removal device according to an embodiment of the present invention, FIG. 2 is a flowchart showing the processing flow of the ghost removal device of the present invention, and FIG. 3 is a block diagram of a conventional ghost removal device. 4 are for explaining the basic ghost removal operation, (a) is a signal waveform diagram of the WRB signal, (b) is an O pedestal signal waveform diagram, (C
) is the signal waveform diagram after field sequence processing, (
d) is a one-clock difference signal waveform diagram from (C). 1...A/D converter, 2...Transfer filter, 4A, 4B...Waveform memory, 5...CPU, 6... ...Tap correction amount total memory, 7...Comparison value memory, 8...
...Number of times counter. Name of agent: Patent attorney Shigetaka Awano (1 person) Figure 1 Figure 1

Claims (2)

[Claims] (1) In a device that uses a GCR signal as a reference signal for ghost removal and performs ghost removal by modifying the tap coefficients of a digital transversal filter by successive equalization, calculation of the sum of the absolute values of the tap correction amount per time is performed. means for storing and memorizing the sum; and means for calculating the ratio of the sum of the absolute values of the tap correction amounts at the previous correction in the sequential equalization to the sum of the absolute values of the tap correction amounts at the current correction. , If the value of the ratio is larger than the predetermined value, the total absolute value of the previous tap correction amount is separately stored as a comparison value, and the actual correction operation of the tap coefficient of the transversal filter is stopped, and then the correction operation of the actual tap coefficient of the transversal filter is stopped, and then the correction operation is performed a predetermined number of times. means to calculate only the sum of absolute values of the tap correction amount and calculate the ratio with the comparison value after a predetermined number of times; and when the ratio value is smaller than the predetermined value, the actual tap coefficient of the transversal filter The tap coefficient restarts the correction operation, and when the tap coefficient is larger than the predetermined value, the ratio with the comparison value is calculated again after a predetermined number of times have elapsed, and when the tap coefficient is larger than the predetermined value again, the tap coefficient is initialized. Ghost removal device. (2) The ghost removal device according to claim 1, wherein each of the means is performed by software processing by a CPU.