JPH05183778A - Circuit for compensating video signal - Google Patents

Abstract

PURPOSE:To provide a video signal compensating circuit where exact compensation is operated to a video signal by specially reducing the errorneous detection of a ringing component concerning the video signal compensating circuit which is appropriate to various kinds of video signal processors such as VTR, a picture processor, an encoder for broadcasting in addition to a television receiver. CONSTITUTION:The front side part and rear side part of an edge are independently detected by a front edge detecting part 6-1 and a rear edge detecting part 6-2. Two detecting signals are amplitude-limited by non-linear amplitude limiters 6-5 and 6-6 and, then, a max. value selecting circuit 6-8 obtains a max. amplitude value and a dividing circuit 6-9 obtains an amplitude ratio. The optimum mixture rate of a signal with LPF 1-2 and an original signal is decided in a matrix circuit 6-10 in accordance with the two values. Thus, an output signal where exact compensation is operated is obtained.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a video signal compensating circuit suitable for various video signal processing devices such as television receivers, VTRs, image processing devices, broadcast encoders and the like. It is an object of the present invention to provide a video signal compensating circuit which can reduce the erroneous detection of the ringing component and perform an accurate compensating operation for the video signal.

[0002]

2. Description of the Related Art An example of a conventional video signal compensating circuit is shown in FIG. The input video signal is given to L1-1. The input video signal is band-limited in some way in accordance with the transmission path, and ringing as shown in FIG. 8 occurs at the rising edge portion and the falling edge portion thereof, deteriorating the image quality.

The input video signal from L1-1 is a block 1-
1 edge detector, block 1-2 low pass filter (LP
F) and the delay line (DL) of blocks 1-3. The edge detection section 1-1 is usually composed of a high-pass filter and outputs a signal obtained by differentiating the input video signal. Ringing occurs in a portion where the change in the video signal is large and a lot of high frequency components are contained, such as the rising and falling edges of the signal and both sides of the pulse. Therefore, block 1-1
Is a high-pass filter, such a ringing occurrence portion (edge portion) can be detected. Edge detector 1-
The output of 1 is given to the nonlinear amplitude limiters 1-4.

The LPF of the block 1-2 limits the ringing component of the input video signal to limit the band at a frequency lower than the band limit given to the input video signal. Further, the characteristic of the LPF1-2 near the cutoff frequency is a roll-off characteristic in which ringing is less likely to occur. Therefore, the signal given from the LPF1-2 to the DL1-5 is a signal with very little ringing, but on the other hand, since the band is limited at a lower frequency, the signal (high frequency component) of the fine portion of the image is It is a lost signal.
DL1-3 and DL1-5 adjust only the delay amount without applying any limitation to the band of the input video signal.

The outputs of DL1-5 and DL1-3 are given to a mixing circuit composed of a multiplier of block 1-6, a multiplier of block 1-7, and an adder of block 1-8. , Are mixed and then output from L1-2 to the outside. At the time of this mixing, the delay amount from the input of each signal given to the mixing circuit needs to be equal, and the delay amounts of DL1-3 and DL1-5 are adjusted to satisfy the condition.

It is the non-linear amplitude limiter 1-4 that gives the mixing ratio to this mixing circuit. The non-linear amplitude limiter 1-4 supplies the detection signal k to the multiplier 1-6, and determines the mixing rate k of the input signal passing through the LPF1-2 and DL1-5. The non-linear amplitude limiter 1-4 also supplies the detection signal k to the multiplier 1-7, and determines the mixing ratio (1-k) of the input signal that has passed through the DL1-3.

Since ringing is generally generated in the edge portion detected by the edge detecting section 1-1, LPF1-2 which is a low frequency component which does not include ringing in that portion.
It is sufficient to suppress the ringing by mixing a large amount of the signal from the original signal. However, if the band limitation is applied to the edge itself of the original signal, the fine signal at the contour portion of the image is lost and the image quality is rather deteriorated. Therefore, the edge detector 1-
For the large amplitude part of the edge part detected in 1 that is considered as the edge itself, the nonlinear amplitude limiter 1-4 does not output (that is, detect it so that the original signal is mixed a lot). A limitation such as signal k = 0) is added to the non-linear amplitude limiter 1-4. FIG. 9 shows an example of the input / output amplitude characteristics of the nonlinear amplitude limiter 1-4 that realizes this limitation.

As described above, in the conventional video signal compensating circuit, in the portion where ringing is generated, a large amount of the signal whose band is limited by the roll-off characteristic by LPF1-2 is mixed in order to remove the ringing. On the contrary, a large amount of the original signal is mixed in the portion where ringing does not occur in order to preserve the fine component of the image. As a result, a signal in which only the ringing component is removed appears in the output signal.

FIG. 10 shows an example of the signal waveform of each part of this video signal compensation circuit. 10A is an input signal waveform, FIG. 10B is an LPF output waveform, FIG. 10C is an edge detector output waveform, and FIG. 10D is a nonlinear amplitude limiter output waveform. (E) is an output signal waveform.

Although the above circuit can be realized by either analog or digital, the digital circuit is more suitable in consideration of the accurate adjustment of the delay amount and the complicated response of the amplitude limiter.

[0011]

In the configuration example according to the prior art, when a plurality of edges are present close to each other in the input signal, the edge detection section (detection filter by the high-pass filter) 1-
Even if there is a ringing part in the center of the impulse response of 1, the edges themselves overlap in the peripheral part. Therefore, in the output of the detection filter, a large amplitude similar to that when the edge itself is detected appears, and the detection output is limited by the nonlinear amplitude limiter 1-4, resulting in an erroneous detection state. The ringing cannot be removed. This state is shown in FIGS.

Next, consider the case where the input signal is a sine wave having a medium amplitude. The frequency of this sine wave is in the region where it is detected by the detection filter (edge detection section 1-1). The medium amplitude of this sine wave is the amplitude of the amplitude limiter 1-4 in the latter stage. It is assumed that it is not suppressed. In this case, since the sine wave is erroneously detected as a ringing component, the signal obtained by low-pass filtering the input signal becomes the output signal, and the sine wave is lost. This state is shown in FIGS.
As described above, the conventional example has a problem that erroneous detection of ringing occurs under some conditions and an accurate compensation operation cannot be performed.

The problem to be solved by the present invention is to provide a video signal compensating circuit capable of reducing false detection of a ringing component and performing an accurate compensating operation for a video signal. There is a point.

[0014]

In order to solve the above-mentioned problems, the present invention provides a detection circuit for detecting an edge portion of an input video signal and outputting a detection signal, and a mixing ratio according to the detection signal. And a mixing circuit for adaptively mixing the input video signal and a signal obtained by passing the input video signal through a low-pass filter, which occurs at rising and falling edge portions of the input video signal. A video signal compensation circuit for removing a ringing component, wherein the detection circuit has a first detection filter having an impulse response that is relatively advanced with respect to the supplied input video signal, and the supplied input video signal. A second detection filter having an impulse response that is relatively delayed with respect to the signal; a first nonlinear amplitude limiter that limits the amplitude with respect to the output signal of the first filter; and the second filter. A second non-linear amplitude limiter for limiting the amplitude of the output signal of the digital filter and the output signals of the first and second non-linear amplitude limiters, and a maximum value for selecting the maximum amplitude value from the two signals. A selection circuit, and the first and second
The non-linear amplitude limiter output signal is supplied, a division circuit for obtaining the amplitude ratio of two signals, the maximum amplitude value and the amplitude ratio are supplied, and the detection signal is output according to the two values. An image signal compensating circuit comprising a matrix circuit is provided.

[0015]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS In the present invention, two detection filters, a filter having an impulse response that is advanced relative to an input signal and a filter having an impulse response that is delayed, are prepared. Further, a non-linear amplitude limiter is also prepared independently for each detection filter, and in the matrix circuit in the subsequent stage, the mixing ratio is determined by the maximum amplitude and the amplitude ratio of each detection signal of the two non-linear amplitude limiters. By preparing the two detection filters independently, it becomes possible to detect where the edge and the ringing are temporally located with respect to the input signal. Further, the non-linear amplitude limiter connected to the two detection filters and working independently can reduce the influence of noise and reduce the dynamic range of the circuit thereafter, so that the increase in the circuit scale can be suppressed. Furthermore, by using the maximum amplitude and ratio of the two detection signals when setting the mixing ratio in the matrix circuit, the mixing ratio that is finely adapted to the waveform of the video signal can be determined, and false detection of ringing components is reduced. it can.

FIG. 1 shows an embodiment of a video signal compensation circuit according to the present invention. The same parts as those in the conventional example are designated by the same reference numerals. The input video signal from the line L6-1 is supplied to the front edge detecting section 6-1, the rear edge detecting section 6-2, the LPF1-2, and the delay line (DL) 1-5. The impulse response of the front edge detection section (first detection filter) 6-1 is shown in FIG.
As shown in (a), the leading edge detection unit 6-1 detects only the leading edge of the input signal and the leading edge in time. Although the impulse response is sinusoidal, this period is the impulse response (s) due to the band limitation performed before the input signal is input to L6-1.
inx / x), the leading edge detection unit for a continuous burst signal having the same frequency as the ringing.
The sensitivity of 6-1 can be increased.

Rear edge detecting section (second detection filter) 6-
As shown in FIG. 2B, the impulse response 2 has only a component relatively delayed with respect to the input signal, and the trailing edge detector 6-2 detects an edge delayed in time. This impulse response is also sinusoidal, but the cycle is the cycle of the impulse response (sinx / x) due to the band limitation applied before the input signal is input to L6-1, as in the case of the leading edge detection. This makes it possible to increase the sensitivity of the trailing edge detection unit 6-2 for a continuous burst signal having the same frequency as the ringing.

In order to realize an impulse response that is advanced in time, it is necessary to perform a process of responding before the signal is input, but in this case, it is a relative process.
By adjusting the delay amount of L1-3, 1-5, it is possible to realize a relatively advanced response to the delay of the original signal.

The LPF1-2 performs band limitation at a frequency lower than the band limitation given to the input video signal in order to suppress the ringing component of the input video signal. Further, the characteristic of the LPF1-2 near the cutoff frequency is a roll-off characteristic in which ringing hardly occurs. Therefore,
The output signal of the LPF1-2 is a signal with very little ringing, but on the other hand, the signal in the fine portion of the image is lost because the band is limited at a lower frequency.

The outputs of DL1-3 and 1-5 are multiplied by multipliers 6-11 and 6-
12 and is supplied to the mixing circuit composed of the adder 6-13,
After being mixed, it is output to the outside from the line L6-2. At the time of this mixing, the delay amount from the arrival of each signal applied to the mixing circuit needs to be the delay amount at which the above-described impulse response for the front edge detection is realized, and DL1 The delay amount of -3,1-5 is adjusted to meet the condition.

The detection circuit for generating the detection signal for giving the mixing ratio to the mixing circuit includes a front edge detection section 6-1 and a rear edge detection section.
6-2, Non-linear amplitude limiter 6-5, 6-6, Maximum value selection circuit 6-8
, A division circuit 6-9, and a matrix circuit 6-10. The output signal of the front edge detection unit 6-1 is supplied to the non-linear amplitude limiter 6-5, and the output signal of the rear edge detection unit 6-2 is supplied to the non-linear amplitude limiter 6-6. Non-linear amplitude limiter
The maximum value selection circuit 6-8 is applied to each output signal of 6-5 and 6-6.
, Is supplied to the division circuit 6-9. Maximum value selection circuit 6-8,
The respective output signals of the division circuit 6-9 are supplied to the matrix circuit 6-10, and the detection signals giving the mixing ratio from the matrix circuit 6-10 are supplied to the multipliers 6-11 and 6-12.

The input / output amplitude characteristics of the non-linear amplitude limiters 6-5 and 6-6 are shown in FIG. 3 (a). In the amplitude limiters 6-5 and 6-6, the polarity of the input signal is meaningless, so the absolute value of the input signal is first taken. Next, a dead zone, a linear zone, and a saturation zone are provided for the input, and a signal corresponding to these is output. The dead zone is
It is a region for preventing erroneous detection due to noise, the linear region is a region for outputting an edge detection result according to an input, and the saturation region is a region for reducing the dynamic range of the subsequent stage.

The maximum value selection circuit 6-8 outputs the larger output value of the two output values of the non-linear amplitude limiters 6-5 and 6-6. In the division circuit 6-9, the nonlinear amplitude limiter 6-5, 6-6
The amplitude ratio of the two output values of is output. Non-linear amplitude limiter
If the output values of 6-5 and 6-6 are x and y, the amplitude ratio is max.
It is calculated by (x, y) / min (x, y), and the amplitude ratio is 1 or more. In the matrix circuit 6-10, the input signal and the LPF1-
2. Determine the mixing ratio k with the signal from 2. The relationship between the amplitude ratio, the maximum amplitude value, and the mixing ratio k is shown in FIG.
Is the shaded area in the figure). First, when the amplitude ratio is close to 1, the front side (front edge detection area portion) and the back side (rear edge detection area portion) of the original signal include the same high frequency components and include ringing components. Assuming that the signal has a continuous burst shape close to a non-sinusoidal wave, the mixing rate k is increased so that a large amount of high frequency components are allowed to pass. Next, when the amplitude ratio is large and the maximum amplitude value is also large, the high-frequency component changes between the front side and the rear side of the original signal, but the included component is not a ringing component and has a complicated shape with large front and rear. Assuming that the signal is a signal, the mixing ratio k is increased so that the original signal is allowed to pass through. When the amplitude ratio is large and the maximum amplitude value is small, a large edge or pulse is included on one side of the original signal,
Assuming that the original signal contains a ringing component, the mixing rate k
Lower and suppress ringing.

As described above, in the video signal compensating circuit of this embodiment, in the portion where ringing occurs, many signals are subjected to band limitation of the roll-off characteristic by LPF1-2 in order to remove it. On the contrary, a large amount of the original signal is mixed in the portion where ringing does not occur in order to preserve the fine component of the image. As a result, a signal in which only the ringing component is removed appears in the output signal.

FIG. 4 shows an example of the signal waveform of each part of the video signal compensating circuit. 4 (a) is an input signal waveform, FIG. 4 (b) is an LPF output waveform, FIG. 4 (c) is a front edge detection unit output waveform, and FIG. 4 (d) is a rear edge detection unit output waveform. Figure (e) shows the output waveform of the non-linear amplitude limiter connected to the front edge detector, (f) shows the output waveform of the non-linear amplitude limiter connected to the rear edge detector, and (g) shows the maximum. The output waveform of the value selection circuit, the figure (h) is the amplitude ratio output waveform of the division circuit, the figure (i) is the output waveform (mixing ratio) of the matrix circuit, and the figure (j) is the output signal waveform. As shown in (i) of the same figure, the mixing ratio k is 0 at the ringing occurrence part of the original signal, and is 1 at the other parts. As a result, as shown in (j) of the same figure, the ringing component is Only the output signal is properly removed.

Further, the circuit of this embodiment can accurately distinguish ringing due to an approaching pulse from a continuous burst signal, which in the prior art often cannot be accurately distinguished due to erroneous detection. This is possible by reducing the false detection by previously optimizing the width of the impulse response of the detection filter, the slope of the linear region of the amplitude limiter, and the matrix of the maximum amplitude value and the amplitude ratio. Figure 5
An example in which ringing due to an approaching pulse is accurately detected as ringing without being erroneously detected as a continuous burst signal is shown. As shown in (g) of the same figure, the mixing ratio is 0 in the ringing occurrence part of the original signal and is 1 in other parts, and as a result, as shown in (h) of the figure, only the ringing component is The output signal is accurately removed. Further, FIG. 6 shows an example in which a continuous burst signal is accurately detected as a burst signal without being erroneously detected as ringing due to an approaching pulse. As shown in (g) of the same figure, the mixing ratio is always 1, and as a result, as shown in (h) of the same figure, an output signal having the original waveform of the signal is obtained.

In this video signal compensating circuit, since the detection filter exists independently for the front and rear edges, and the mixing ratio control is changed according to the pattern of the maximum amplitude value and the amplitude ratio, each input signal is detected. It is possible to perform processing suitable for the state of.

Although the above circuit can be realized by either analog or digital as in the conventional example, it is more suitable to be realized by a digital circuit for the same reason as in the conventional example. Further, in the configuration according to the example of the present invention, the number of detection filters and the non-linear amplitude limiter is increased as compared with the configuration according to the prior art, and the matrix circuit is added. However, the impulse response of each detection filter need only have an impulse response that is either an advanced time or a delayed time with respect to the input signal. Compared with, the overall filter scale does not increase much. Further, the non-linear amplitude limiter, the maximum value selection circuit, the division circuit, and the matrix circuit can be realized by one small ROM in a lump when they are constituted by digital circuits, so that the overall hardware scale is increased. Can be small.

[0029]

As described above, in the video signal compensating circuit according to the present invention, the edge and the pulse component of the front side and the rear side of the original signal are independently taken out to detect and remove the ringing component more accurately. Therefore, the compensation operation of the video signal can be performed accurately.

[Brief description of drawings]

FIG. 1 is a diagram showing an example.

FIG. 2 is a diagram showing an impulse response of an edge detection unit.

FIG. 3 is an operation explanatory diagram of a non-linear amplitude limiter and a matrix circuit.

FIG. 4 is a diagram showing an example of a signal waveform of each part.

FIG. 5 is a diagram illustrating a processing example for ringing due to an approaching pulse.

FIG. 6 is a diagram showing a processing example for a continuous burst signal.

FIG. 7 is a diagram showing a conventional example.

FIG. 8 is a diagram showing an example of an input signal.

FIG. 9 is a diagram illustrating the operation of the nonlinear amplitude limiter.

FIG. 10 is a diagram showing an example of a signal waveform of each part.

FIG. 11 is a diagram illustrating a processing example for adjacent edges and continuous bursts.

[Explanation of symbols]

 1-2 LPF (Low-pass filter) 1-3,1-5 DL (Delay line) 6-1 Leading edge detector (first detection filter) 6-2 Rear edge detector (second detection filter) 6-5,6-6 Non-linear amplitude limiter 6-8 Maximum value selection circuit 6-9 Division circuit 6-10 Matrix circuit 6-11,6-12 Multiplier 6-13 Adder

Claims (1)

[Claims] 1. A detection circuit which detects an edge portion of an input video signal and outputs a detection signal, a mixing ratio is determined according to the detection signal, and the input video signal and the input video signal are low-pass filtered. A video signal compensating circuit for removing a ringing component generated at rising and falling edge portions of an input video signal, the detection circuit comprising: A first detection filter having an impulse response relatively advanced with respect to the supplied input video signal, and a second detection filter having an impulse response relatively delayed with respect to the supplied input video signal A filter, a first non-linear amplitude limiter that limits the amplitude of the output signal of the first filter, and a second non-linear amplitude that limits the amplitude of the output signal of the second filter A limiter, a maximum value selection circuit that is supplied with the output signals of the first and second nonlinear amplitude limiters, and selects a maximum amplitude value from the two signals; and a first and second nonlinear amplitude limiter It comprises a division circuit which is supplied with an output signal and obtains an amplitude ratio of two signals, and a matrix circuit which is supplied with the maximum amplitude value and the amplitude ratio and which outputs the detection signal according to the two values. Video signal compensation circuit characterized by.