JP2542242B2 - Decorator - Google Patents

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a decoder for re-sampling a video signal that has been subjected to multiple sub-sampling and then subjected to predetermined processing to obtain a reproduced video signal.

“Conventional technology” Conventionally, MUSE (Multiple Sub-Nyq), which is a multi-sub-sample transmission method using offset sub-sampling, has been used as a method for band-compressing and transmitting a wide-band high-quality video signal.
uist Sampling Encoding) method has been proposed.

In this method, samples of an input video signal are rearranged and transmitted as sample values of analog signals, and the decoder on the receiving side reassembles the samples.

FIG. 5 shows a sampling pattern of the input video signal on the transmission side, which is a sub-sampling pattern that makes one cycle in two frames. , 4n + 2nd, 4n + 3rd, 4th
The 4n + 4th position (n = 0,1,2, ...) indicates the position where subsampling is performed, and the x mark indicates the position where subsampling is not performed in any field. In the figure,
d indicates a sampling period, and the sampling frequency 1 / d is, for example, 48.6 MHz.

In addition, as described above, the signal that is subsampled and transmitted is resampled at a frequency of 16.2 MHz on the receiving side, and then the following processing is performed to reproduce a video signal. That is, in the case of a still image, the video signal is reproduced by synthesizing the samples of the past three fields transmitted sequentially and the samples transmitted in the current field. On the other hand, in the case of a moving image, multi-line blurring occurs if the same processing as in the case of a still image is performed, so that the video signal is reproduced only by the samples transmitted in the current field.

[Problems to be Solved by the Invention] FIG. 6 shows a basic configuration of such a sampled value transmission system, in which 30 is a transmission means, 40 is a transmission line, and 50 is a transmission line.
Is a receiving means.

Since the characteristic of the transmission line 40 can be approximated to the characteristic of the low-pass filter, if there is a sampled signal at point A in FIG. 6 as shown in FIG. 7, at point B in FIG. The signal has a sinX / X waveform as shown in FIG.

Since the signal shown in FIG. 8 is restored to the signal shown in FIG. 7, the resampling phase in the decoder on the receiving side becomes important. In other words, if resampling is performed with the correct phase as shown by * in FIG.
It is returned to the signal shown in the figure. However, if the resampling phase shifts as shown by * in FIG. 10, the signal is not returned to the signal shown in FIG. 7, and ringing occurs. As is clear from the figure, this ringing has a repetition period near 8.1 MHz.

This ringing deteriorates the quality of the reproduced image because the reproduced video signal finally obtained by the above-mentioned processing remains.

The above-mentioned resampling phase changes due to mismatching of the transmission line 40, temperature characteristics of devices, elements, and the like.

The present invention has been made in view of the above points, and an object thereof is to reduce ringing and prevent deterioration of image quality of a reproduced image.

"Means for Solving the Problem" The present invention reduces ringing occurring in a signal in a decoder that resamples a video signal that has been subjected to multiplex subsampling and then performs predetermined processing to obtain a reproduced video signal. A ringing reducer that
This ringing reducer has a low-pass filter and a ringing discriminator that discriminates ringing.
The low pass filter is inserted in the signal line only when the ringing discriminator determines that the reproduced video signal has ringing.

The ringing discriminator uses the amplitude of the difference between the sampled signals and the amplitude of the difference between the sign and the difference signal for the ringing discrimination.

Further, in the case where the signal line includes the horizontal contour emphasizing circuit, the ringing reducer is arranged before the horizontal contour emphasizing circuit.

[Operation] In the above-described configuration, the signal line is provided with the ringing reducer that reduces the ringing that occurs,
The ringing of the finally obtained reproduced video signal can be significantly reduced. Further, since only the signal that matches the ringing condition is reduced by the low-pass filter by the ringing discriminator, the influence on other signals is reduced.

Further, the ringing discriminator can accurately discriminate the ringing by using the amplitude of the difference between the sampled signals and the amplitude of the difference between the sign and the difference signal.

Further, in the case where the horizontal contour emphasizing circuit is provided, the ringing can be efficiently reduced by arranging the ringing reducer before the horizontal contour emphasizing circuit.

[Embodiment] An embodiment of the present invention will be described below with reference to the drawings.

In this example, the signal line of a decoder (not shown) that resamples the transmitted video signal and then performs a predetermined process to obtain a reproduced video signal by the MUSE method is used to shift the resampling phase as described above. In order to reduce the ringing generated by the ringing reducer 100 as shown in FIG. For example, it is arranged before the horizontal contour enhancement circuit for increasing the sharpness.

In the figure, 101 is an input terminal, and this input terminal 1
The input signal supplied to 01 is supplied to the fixed terminal on the a side of the changeover switch 102 which constitutes the selector. Further, the input signal supplied to the input terminal 101 is supplied to the fixed terminal on the b side of the changeover switch 102 via the low-pass filter 103 having the characteristic of attenuating the signal of the frequency of 8.1 MHz or higher.

Further, the input signal supplied to the input terminal 101 is supplied to the ringing discriminator 104. The switching signal of the ringing discriminator 104 controls the switching of the changeover switch 102. That is, the ringing discriminator 104
When the ringing is discriminated by, it is connected to the b side, and in other cases, it is connected to the a side.

Further, the output terminal 105 is led out from the changeover switch 102.

In the above configuration, when ringing is not discriminated by the ringing discriminator 104, the changeover switch 102 is connected to the a side, and the input signal supplied to the input terminal 101 is led to the output terminal 105 via the changeover switch 102. However, when ringing is discriminated, the changeover switch 102 is connected to the b side, the ringing of the input signal supplied to the input terminal 101 is attenuated by the low-pass filter 103, and then the output terminal 1 is output via the changeover switch 102.
Derived in 05. Therefore, a signal with reduced ringing is output to output terminal 105.

Next, the position of the signal used in the ringing discriminator 104 will be described with reference to FIG. In the figure,
When the current signal for determining whether to pass through the low-pass filter 103 is S00, signals up to two clocks before and after the current signal are S-20, S-10, S10, and S20, respectively, and the current signal S00.
On the other hand, the signal one horizontal period (1H) before is S01. In addition, the difference SX0-S (X-1) 0 between each signal is D-20, D from the left.
-10, D10, D20, similarly, the difference between S01 and S00 is D01, and the difference between D-20 and D20 is W1.

By the way, the frequency of the final signal processing in the reproduction of the video signal transmitted by the MUSE method is 48.6 MHz.
The reason why 5 signals are used in the horizontal direction as shown in the figure is as follows. 3 to 48.6MHz
It shows the 8.1MHz signal when sampled at the frequency of.

Comparing the point at which the polarity of the slope to be filtered changes and before and after that, the cross-correlation of ,,, is low when all seven signals within the 1 / 8.1MHz period are used. There are many conditions for determining that all signals are 8.1 MHz. Therefore, the circuit scale becomes large.

Next, with the five signals before and after the point where the polarity changes, the discrimination frequency spreads below 12.2MHz,
Since the cross-correlation is high in and,
It is relatively easy to determine that the signal is below MHz. In addition,
, Has a low correlation with, and cannot be determined to be a signal of 12.2MHz or less under the same conditions as ,, but the waveform of does not easily occur as a ringing waveform, and contains a lot of high frequency components at the poles. Therefore, it does not matter so much even if it is not distinguished from ringing.

Also, at the point where the polarity changes and the three signals before and after that,
It is difficult to determine only the ringing because there is little information available.

From the above, in this example, five signals are used in the horizontal direction. Needless to say, more signals may be used.

Now, the conditions for discriminating ringing when using five signals in this way are set as follows.

[Condition I] 1 at the center of 5 signals (1 / 12.2 MHz range)
The polarity of the slope changes only by one.

[Condition II] The absolute values of the difference signals D-10 and D10 are both below the reference value.

[Condition III] The absolute value of the difference signal W1 is greater than or equal to the reference value.

[Condition IV] The absolute value of the difference signal D01 is less than or equal to the reference value.

 Next, each condition will be described.

Condition I is for detecting a signal of 12.2 MHz or less. As a result, a signal having a component exceeding 12.2 MHz is not passed through the low-pass filter 103, so that sharpness of image quality is not deteriorated.

The condition II is for determining ringing.
That is, the impulse response of the low-pass filter (characteristic of the transmission line) is approximated to sinx / x. Since the amplitude of the sinx / x transition period (corresponding to ringing) does not exceed a certain value, it can be used as a ringing discriminating condition by appropriately selecting a reference value. As a result, one having an amplitude change larger than the reference value is not selected. The overshoot or undershoot, which is the first ringing for the signal, has a large change in amplitude, but since the meaning of outline enhancement appears visually,
The reference value may be set to exclude this.

Condition III is also for determining ringing. In other words, ringing is always a damping phenomenon if one chooses a positive or negative direction along the time axis. Therefore, the absolute values of the difference signals D-20 and D20 are always different. That is, the absolute value of the difference signal W1 does not become zero. Therefore, by properly selecting the reference value, it is possible to use it as the ringing discrimination condition. As a result, when the input is a sine wave or the like, it does not pass through the low pass filter 103.

Condition IV is also for determining ringing.
Since the ringing appears as a vertical line on the screen, the vertical correlation is high. That is, the absolute value of the difference signal D01 is quite small. Therefore, by properly selecting the reference value, it is possible to use it as the ringing discrimination condition.

FIG. 4 shows a specific configuration diagram of the ringing discriminator 104. In the figure, the input signal supplied to the input terminal 3 is directly supplied to the subtractor 12 and
It is supplied to the subtractor 12 via the delay circuit 8 having a delay time (1 / 48.6 MHz) of a clock. Further, the output signal of the delay circuit 8 is directly supplied to the subtractor 13 and, at the same time, is passed through the delay circuit 9 having a delay time of one clock, and the subtractor 13
Is supplied to. Further, the output signal of the delay circuit 9 is a subtractor
It is directly supplied to the subtractor 14 via the delay circuit 10 having a delay time of one clock. Also,
The output signal of the delay circuit 10 is directly supplied to the subtractor 15 and is also supplied to the subtractor 15 via the delay circuit 11 having a delay time of one clock. In this case, assuming that the input signal supplied to the input terminal 3 is S20, the delay circuits 8, 9, 10 and 1
The output signals of 1 are signals S10, S00, S-10 and S-, respectively.
20 and the output signals of the subtractors 12, 13, 14 and 15 correspond to the difference signals D20, -D10, D-10 and -D-20, respectively.

Also, the sign bit of the output signals of the subtracters 12 and 13 is
It is supplied to the input side of the exclusive OR circuit 16, and the sign bits of the output signals of the subtracters 13 and 14 are supplied to the input side of the exclusive NOR circuit 17, and further,
The sign bit of the output signals of the subtractors 14 and 15 is supplied to the input side of the exclusive OR circuit 18. Then, the output signals of the exclusive OR circuit 16, the exclusive NOR circuit 17, and the exclusive OR circuit 18 are supplied to the input side of the AND circuit 19, and this AND circuit is supplied.
The output signal of 19 is supplied to the input side of the AND circuit 28.

In this case, the exclusive OR circuit 16, the exclusive NOR circuit 17, the exclusive OR circuit 18, and the AND circuit 19 form a determination unit for the condition I (shown by a broken line). Here, the signals S-20, S-20, S00, and S00, S10, S
When the polarities of the slopes are the same in the section of 20, the output signals of the exclusive OR circuits 16 and 18 are high level “1” respectively.
Becomes The output signal of the exclusive NOR circuit 17 becomes a high level "1" when the polarity of the inclination changes in the section of the signals S-10, S00, S10. Therefore, the output signal of the AND circuit 19 becomes the high level "1" only when the condition I is satisfied.

The output signals of the subtracters 13 and 14 are converted into absolute values by the absolute value conversion circuits 21 and 22, respectively, and then supplied to the comparators 24 and 25 to be compared with the reference value. The comparators 24 and 25 output a high-level "1" signal when the output signals of the absolute value conversion circuits 21 and 22 are equal to or lower than the reference value, and a low-level "0" signal otherwise. Is output. The output signals of the comparators 24 and 25 are supplied to the input side of the AND circuit 27, and the output signal of the AND circuit 27 is supplied to the input side of the AND circuit 28.

In this case, the absolute value conversion circuits 21 and 22, the comparators 24 and 25, and the AND circuit 27 form a determination unit for condition II. here,
When the absolute values of the difference signals D-10 and D10 are equal to or less than the reference value, the output signals of the comparators 24 and 25 are at high level "1". Therefore, the output signal of the AND circuit 27 becomes the high level "1" only when the condition II is satisfied.

The output signals of the subtracters 12 and 15 are supplied to the subtractor 20, and the output signal of the subtractor 20 is converted into an absolute value by the absolute value conversion circuit 23 and then supplied to the comparator 26 to be compared with the reference value. It The comparator 26 outputs a high level “1” signal when the value of the output signal of the absolute value conversion circuit 23 is equal to or greater than the reference value, and outputs a low level “0” signal otherwise. . Then, the output signal of the comparator 26 is supplied to the input side of the AND circuit 28. In this case, the absolute value conversion circuit 23 and the comparator 26 form a determination unit for condition III. Here, when the absolute value of the difference signal W1 is greater than or equal to the reference value, the comparator
The output signal of 26 becomes high level "1". Therefore, the condition
Only when III is satisfied, the output signal of the comparator 26 becomes the high level "1".

The output signal of the delay circuit 9 is directly supplied to the subtractor 5 and is also supplied to the subtractor 5 via the delay circuit 4 having a delay time of one horizontal period (1H). In this case, when the output signal of the delay circuit 9 corresponds to the signal S00, the output signal of the delay circuit 4 corresponds to the signal S01, and thus the output signal of the subtractor 5 corresponds to the difference signal D01.

The output signal of the subtractor 5 is converted into an absolute value by the absolute value conversion circuit 6 and then supplied to the comparator 7 to be compared with the reference value. The comparator 7 outputs a high level "1" signal when the output signal of the absolute value conversion circuit 6 is equal to or lower than the reference value, and outputs a low level "0" signal otherwise. Then, the output signal of the comparator 7 is supplied to the input side of the AND circuit 28. In this case, the absolute value conversion circuit 6 and the comparator 7 form a determination unit for condition IV. Here, when the absolute value of the difference signal D01 is less than or equal to the reference value, the output signal of the comparator 7 becomes a high level "1". Therefore, the condition
Only when the IV is satisfied, the output signal of the comparator 7 becomes the high level "1".

Thus, when all the conditions I to IV are satisfied, the output signal of the AND circuit 28 becomes high level "1", and otherwise low level "0". The output signal of the AND circuit 28 is supplied as a control signal to the changeover switch 102 described above. The changeover switch 102 is connected to the a side and the b side when the output signal of the AND circuit 28 is at the low level "0" and the high level "1", respectively. Therefore, as described above, the ringing discriminator 10
When the ringing is not determined by 4, the changeover switch 102 is connected to the side a, while when the ringing is determined, the changeover switch 102 is connected to the side b.

As described above, according to this example, the ringing reducer 100 that reduces the ringing caused by the phase shift of resampling is arranged in the signal line, so that the reproduced video signal finally obtained does not include ringing. it can.

In addition, since the ringing discriminator 104 reduces only the signal that matches the ringing condition by the low-pass filter 103, the influence on other signals is reduced. Therefore, according to this example, it is possible to reduce deterioration of image quality.

Further, since the ringing discriminator 104 uses the amplitude of the difference between the sampled signals and the amplitude of the difference between the sign and the difference signal, the ringing can be discriminated accurately.

Further, since the ringing reducer 100 is arranged in the preceding stage of the horizontal contour emphasizing circuit, ringing can be efficiently reduced.

Although the ringing caused by the phase shift of the resampling is mentioned in the embodiment described above,
As is clear from the circuit diagram in the figure, the ringing that occurs is processed to reduce it, and the cause of ringing is not necessarily limited to the case where it occurs due to the phase shift of resampling. Not.

[Advantages of the Invention] According to the present invention described above, since the ringing reducer that reduces the ringing that occurs is arranged in the signal line, it is possible to greatly reduce the ringing of the reproduced video signal that is finally obtained. it can. Further, since only the signal that matches the ringing condition is reduced by the low-pass filter by the ringing discriminator, the influence on other signals is reduced. Therefore, the deterioration of the image quality can be reduced.

Further, the ringing discriminator can accurately discriminate ringing by using the amplitude of the difference between the sampled signals and the amplitude of the difference between the sign and the difference signal.

Further, in the case where the horizontal contour emphasizing circuit is provided, the ringing can be efficiently reduced by disposing the ringing reducer in the preceding stage.

[Brief description of drawings]

FIG. 1 is a block diagram of a ringing reducer, FIGS. 2 and 3 are diagrams for explaining a ringing discriminator, FIG. 4 is a block diagram of a ringing discriminator, and FIGS. 5 to 10 are explanations of ringing. FIG. 100 …… Ringing reducer 101 …… Input terminal 102 …… Changeover switch 103 …… Low pass filter 104 …… Ringing discriminator 105 …… Output terminal

Front Page Continuation (72) Inventor Toshiro Omura 1-10-11 Kinuta, Setagaya-ku, Tokyo Inside Broadcasting Research Laboratories of Japan Broadcasting Corporation (56) Reference JP-A-63-141465 (JP, A)

Claims (1)

(57) [Claims] 1. An original video signal is sampled at a predetermined sampling frequency, converted into a sub-sampling frequency lower than the predetermined sampling frequency and transmitted, and then resampled, and then subjected to a predetermined process. In the decoder to obtain the reproduced video signal, the signal line is provided with a ringing reducer for reducing ringing, and the ringing reducer is a low-pass filter that transmits only signals of the sub-sampling frequency or lower,
A ringing discriminating means for discriminating ringing occurring at the predetermined sampling frequency period, and the low-pass filter is inserted in the signal line only when the ringing discriminating means discriminates that the reproduced video signal has the ringing. A decoder characterized by being performed.