JPH10248027A - S/n improving device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To set an optimum shift amount automatically at all times by employing a DC shift type noise reducer. SOLUTION: An original input signal A is fed to a reference signal generating means 4, in which a reference signal is generated by predicting it that no noise is superimposed on the input signal. The reference signal B and a delayed original input signal A' are fed to a shift direction discrimination means 11, in which a DC shift direction of the original input signal A' is discriminated, a DC shift means 6 applies DC shift to the original input signal A' in a direction in response to a shift direction signal C by an amount in response to a shift amount control signal I. A blanking extract means 12 and a noise amount detection means 13 detect a noise amount H from an output signal E of the DC shift means 6 and a shift amount control means 14 generates a shift amount control signal I in response to the increase/decrease in the noise amount H.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an SN ratio improving apparatus, and more particularly, to a DC shift type noise reducer which realizes noise reduction by level shifting (DC shifting) an input signal on which noise is superimposed. The present invention relates to an SN ratio improving device.

[0002]

2. Description of the Related Art A noise reducer for improving an SN ratio of an input signal such as a video signal is described in, for example, "Television Society Journal", VOL. 48 No. 12 (1994) pp. 1553-1564.
As described in a paper by Izumi et al. In "Proposal of DC Additive Noise Reducer and Development of Noise Reducer for MUSE Decoder Moving Image Processing" and Japanese Patent Application Laid-Open No. 7-250264, a MUSE type high-vision receiver (hereinafter referred to as M
A noise reducer for a USE decoder) is known, and this noise reducer is also called a DC shift type noise reducer.

FIG. 8 is a block diagram showing a conventional SN ratio improving apparatus using such a DC shift type noise reducer, wherein 1 is an input terminal, 2 is a decoding means, 3 is a delay means,
4 is a reference signal generating means, 5 is a level comparing means,
6 is a DC shift means, 7 is an output terminal, and 8 is an input terminal.

In FIG. 1, a MUSE signal is input from an input terminal 1, and the MUSE signal is decoded by a decoding means 2 into an original wideband HDTV video input signal. Hereinafter, the decoded Hi-Vision video signal is referred to as an original input signal A. Usually, a noise component is superimposed on the original input signal A.

The original input signal A is supplied to a reference signal generating means 4 for predicting an ideal ihavision video signal when noise is not superimposed on the original input signal A,
The video signal obtained by this prediction is output as a reference signal B. As the reference signal generating means 4, for example, a median filter is used. The median filters are sequentially arranged adjacent to each other in the horizontal and / or vertical direction of the screen, as described in the above-mentioned known document. The level of the output sample point is an intermediate level or an average level among the levels of a plurality of sample points (pixels), and the level with a higher priority is used as the level of the output sample point. , A reference signal B in which a waveform whose level changes, such as an edge of the original input signal, is stored.

[0006] The reference signal B generated from the reference signal generating means 4 is supplied to the level comparing means 5. The level comparing means 5 includes the decoding means 2
Is supplied by the delay means 3 after being delayed by the delay means 3 for a time corresponding to the delay time of the reference signal generation means 4. The level comparing means 5 compares the level of the supplied reference signal B with the level of the delayed original input signal A ', and outputs a signal indicating the shift direction of the DC level of the original input signal A' in accordance with the magnitude relationship between the levels. (Less than,
A shift direction signal C) is generated.

The original input signal A 'delayed by the delay means 3
Is also supplied to the DC shift means 6 and the sequential D
The C level is shifted in the shift direction indicated by the shift direction signal C output from the level comparison means 5 by the shift amount represented by the shift amount control signal D input from the input terminal 8. The shift amount control signal D indicates the reception C / N of the MUSE signal input from the input terminal 1, and sets the DC shift amount of the original input signal A according to the reception C / N. is there.

In the DC shift means 6, specifically, the DC shift amount indicated by the shift amount control signal D input from the input terminal 8 in accordance with the shift direction indicated by the shift direction signal C from the level comparing means 5 Is added to or subtracted from the original input signal A ′. Accordingly, the original input signal A 'has the sequential level of the reference signal B
The DC level is shifted so as to approach the level of the input signal A ′, and the noise component superimposed on the original input signal A ′ is compressed by the DC shift. Further, a waveform whose level fluctuates, such as the edge of the original input signal A ', is stored.

In this manner, the DC shift means 6 obtains a video signal E in which the noise component is compressed and is equivalent to the original input signal A equivalently, and outputs it from the output terminal 7.

Here, the median filter used as the reference signal generating means 4 functions as a kind of LPF for flattening the change of the signal, and has a characteristic of preserving the edge of the signal. The signal has too much distortion to be used as a video signal as it is. However, the output signal of this median filter is only used for facilitating the identification of the noise component superimposed on the original input signal A, and there is no particular problem with such use.

[0011]

The above-described conventional DC shift type noise reducer requires a user to set a DC shift amount by looking at a screen or to set a data indicating a relationship between the noise amount and the DC shift amount in advance. A table is stored in a ROM or the like, the amount of noise of the input signal is detected, a DC shift amount corresponding to the detected noise amount is obtained from this data table, and the original input signal is determined according to the DC shift amount. This is to perform so-called feed-forward control for shifting the DC level.

For this reason, it is necessary to determine the optimum DC shift amount for each system and provide an optimum value in advance, and it is not always possible to set the optimum DC shift amount.
In such a case, if the DC shift amount is too large with respect to the noise amount, adverse effects such as deterioration of S / N will occur.

For example, let N be the amount of noise contained in the original input signal A 'and D be the amount of DC shift at this time.
When it is intended to set the DC shift amount D so that D = N / 2 and to compress the noise amount N to 1/2, this DC shift amount can be set accurately. If D << N / 2 is set, the DC shift amount of the video signal is too small with respect to the noise amount, the noise component is hardly compressed, and the SN ratio improvement effect is hardly obtained. If D >> N / 2, the level shift of the original input signal A ′ is excessively performed, and the DC fluctuation due to the DC shift processing is added to the noise, and conversely, SN The ratio will be degraded.

An object of the present invention is to provide an SN improvement apparatus which solves such a problem and enables always and automatically setting an optimum DC shift amount.

[0015]

SUMMARY OF THE INVENTION In order to achieve the above object, the present invention provides a DC signal which reduces the amount of noise by shifting the level of an input video signal according to the amount of noise contained therein. In an SN improvement device using a shift-type noise reducer, the amount of noise is detected from a video signal obtained by subjecting the input video signal to DC shift and subjected to a noise reduction process.

According to this configuration, the DC shift is performed by feedback control, whereby the optimum DC shift amount with the noise amount optimally reduced is always automatically set.

[0017]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 shows a first embodiment of an SN ratio improving apparatus according to the present invention.
9 and 10 are input terminals, 11 is a shift direction determining means, 12 is a blanking extracting means, 13 is a noise amount detecting means, 14 is a shift amount controlling means, and 15 is an adding means. , 16 are shift amount setting means, and portions and signals corresponding to those in FIG.

In FIG. 1, shift direction determining means 11 is shown.
Are the original input signal A 'from the delay means 3 and the reference signal generation means 4 in the same manner as the level comparison means 5 in FIG.
From the reference signal B, and forms and outputs a shift direction signal C indicating the DC shift direction of the original input signal A ′ according to the magnitude relationship between the original input signal A ′ and the reference signal B. When the levels are equal, a shift inhibit signal C 'is generated.

The DC shift means 6 comprises an adding means 15 and a shift amount setting means 16. Each time the shift amount control signal 14 is supplied from the shift amount control unit 14, the shift amount setting unit 16 increases or decreases the level value by a unit amount, and outputs the shift direction indicated by the shift direction signal C from the shift direction determination unit 11. , A DC signal J having a positive or negative value is generated. This DC signal J is supplied to the adding means 15 and added to the original input signal A 'from the delay means 3, whereby the DC level of the original input signal A' is shifted. When the shift direction judging means 11 outputs the shift prohibition signal C ', the shift amount setting means 16 prohibits the output of the DC signal J, whereby the original input signal A' passes through the adding means 15 as it is.

Here, when the level of the original input signal A 'is larger than that of the reference signal B (A'> B), the shift direction signal C is converted by the DC shift means 6 into the signal D of the original input signal A '.
This is a signal of code "1" instructing to lower (subtract) the C level, whereby the DC signal J output from the shift amount setting means 16 becomes a negative value, and the original input signal A ' Is subtracted by this DC signal J, and A′−
The video signal E in which the noise component of J is compressed is obtained. If the level of the original input signal A 'is smaller than that of the reference signal B (A'<B), the DC direction of the shift direction signal C is increased (added) by the DC shift means 6. This signal is a signal of a designated code "0", whereby the DC signal J outputted from the shift amount setting means 16 becomes a positive value, and the original input signal A 'is added by the adding means 15 by this DC signal J. , A ′ + J are obtained. When the level of the original input signal A 'is equal to the level of the reference signal B (A' = B)
The shift prohibition signal C ′
Is output.

The shift prohibition signal C 'is used to prohibit the output of the DC signal J from the shift amount setting means 16, but the DC signal J may be set to a value of zero. . Also, a switch means 17 is provided in parallel with the adder means 15, and the switch means 17 is turned ON only when the shift prohibition signal C 'is output, so that the original input signal A' is Therefore, the output may be performed by bypassing the adding means 15.

The video signal E output from the DC shift means 6 is output from the output terminal 7 and also supplied to the blanking extraction means 12 and specified by the blanking timing signal F input from the input terminal 10. Is extracted. The extraction period is a period during which the level in the video signal is constant, such as a vertical blanking period of the video signal.

The signal G extracted by the blanking extraction means 12 is supplied to a noise amount detection means 13 and the original level of the signal G is set as a reference level, and the reference level is subtracted from the signal G. , A random small-amplitude noise component superimposed on the signal G is extracted. Then, the noise amount H is detected by integrating the absolute value of the extracted noise component for a certain period or detecting the peak value of the noise component.

Each time the blanking timing signal F is input from the input terminal 10, the shift amount control means 14 takes in the noise amount H from the noise detection means 13 and generates a shift amount control signal I. Here, the shift amount control signal I is determined according to the noise amount H previously acquired, the shift amount control signal I output immediately before (previous) and the noise amount H acquired from the current noise amount detecting means 13. This is for changing the level of the DC signal J generated and output from the shift amount setting means 16 by a unit amount in an optimal direction.

As described above, in the first embodiment, the noise reduction processing is performed by the feedback control, whereby the video signal E obtained at the output terminal 7 is always and automatically obtained. The control operation is performed so that the remaining noise component is reduced. Accordingly, the control operation for most effectively reducing noise is automatically performed without setting the relationship between the noise amount and the DC shift amount as in the above-described conventional technology, and an optimum noise reduction effect is obtained. Can be.

FIG. 2 shows the shift amount control means 14 in FIG.
FIG. 18 is a block diagram showing a specific example of the first embodiment.
21 is a change direction comparing means, 22 is a control signal generating means, 2
Reference numeral 3 denotes an output terminal, 29 denotes an input terminal, and 30 denotes a switch. Signals corresponding to those in FIG.

In the figure, a noise amount H output from the noise amount detecting means 13 (FIG. 1) is input from an input terminal 18 and a blanking timing signal F input from an input terminal 10 (FIG. 1) to an input terminal 29. At this timing, the signal is supplied to the level comparing means 20 via the switch 30 and is also supplied to and stored in the pre-noise amount storing means 19. The preceding noise amount storage means 19 is for delaying the supplied noise amount H by one period of the blanking timing signal F. The noise amount H input from the input terminal 18 via the switch 30 is compared with the previously input noise amount H ′ from the previous noise amount storage unit 19 by the level comparison unit 20 to detect the level of noise. It is determined whether the noise amount H detected this time has increased or decreased with respect to the noise amount H ′.

The size comparison means 20 determines the result of the comparison,
When the input noise amount H is larger than the immediately preceding noise amount H ′, it is determined that the noise amount has increased, and the determination signal K of “0” is determined.
Is output, and when the input noise amount H is smaller than the immediately preceding noise amount H ′, it is determined that the noise amount has decreased, and the determination signal K of “1” is output.

The determination signal K is supplied to the change direction comparing means 21. The determination signal K is used to determine whether or not it matches the shift amount control signal I generated by the previous input of the noise amount output from the control signal generating means 22. Done. In response to the determination result L, the control signal generating means 22 generates a new shift amount control signal I, which is supplied from the output terminal 23 to the shift amount setting means 16 in FIG.

Here, when the shift amount control signal I is to increase the level of the DC signal J generated by the shift amount setting means 16 in FIG. 1, it is set to "0", and when it is to be decreased, it is set to "0". If 1 ”is set, the control signal generating means 22
As shown in FIG. 3, when this shift amount control signal I matches the determination result signal L of the change direction comparing means 21,
A new shift amount control signal I of "1" for decreasing the level of the DC signal J is generated, and when the shift amount control signal I and the determination result signal L of the change direction comparing means 21 do not match, DC A new shift amount control signal I of "0" for increasing the level of the signal J is generated.

Therefore, in FIG. 1, when the noise amount in the original input signal A 'is larger than the DC shift amount (ie, the absolute value of the level of the DC signal J), the DC shift amount is increased. It can be expected that the amount of noise will decrease. Therefore, the shift amount control means 14 generates a shift amount control signal I for increasing the DC shift amount by a unit amount. In this case, in FIG. 2, the shift amount control signal I
Is set to “0”, the input noise amount H decreases with respect to the previous noise amount H ′, and the determination output K of the magnitude comparing means 20 becomes “1”. Therefore, from FIG. The control signal generating means 22 generates a new shift amount control signal I for further increasing the level as "0".

As described above, when the level of the DC signal J is sequentially increased by a unit amount and then changes beyond the optimum value of the DC shift amount of the original input signal A ′, the shift amount becomes noise as the original input signal A ′. ', And the amount of noise increases. In this case, the DC shift amount of the original input signal A 'must be reduced.

In this case, in FIG. 2, the input noise amount H becomes larger than the previous noise amount H ', and the judgment result signal K of the magnitude comparing means 20 becomes "0". At this time, since the shift amount control signal I is "0", the change direction comparing means 21 determines that the two inputs match, thereby outputting from the control signal generating means 22 as shown in FIG. The shift amount control signal I is a signal of “1” that reduces the level of the DC signal J by a unit amount. As a result,
The decision result signal K of the magnitude comparing means 20 becomes "1". Thereafter, when the DC shift amount of the original input signal A 'exceeds the optimum value again and the decision result signal K of the magnitude comparing means 20 becomes "0", the shift is performed. The quantity control signal I becomes "0" and increases the level of the DC signal J.

By operating as described above, the DC shift amount of the original input signal A 'is set to an optimum value or a value close to the optimum value, and optimum noise suppression is performed.

FIG. 4 is a flowchart showing the above processing operation.

In the figure, first, when the power is turned on and the system is started, or when the channel is switched (step 401), the DC shift amount is initialized (step 402). This is to set the DC signal J in the shift amount setting means 16 in FIG. 1 to 0, and when the noise amount is not detected, such as at the time of channel switching or system startup, a DC shift larger than the noise amount is performed. This is to prevent the SN ratio from being significantly deteriorated due to the setting of the amount. At this time, the pre-noise amount storage means 19 in FIG. 2 is also cleared to zero.

Next, an initial noise amount is detected (step 403). This is the DC shift amount (DC signal J)
Is to detect the amount of noise of the video signal E output from the DC shift means 6 when is zero. This noise amount is equivalent to the noise amount of the original input signal A and is stored in the pre-noise amount storage means 19 in FIG.

Then, the initial shift amount is controlled (step 404). This is to generate a shift amount control signal I of "0" from the control signal generating means 22 in FIG. 2, and thereby the level of the DC signal J is increased by the unit amount in the shift amount setting means 16 in FIG. And adding means 15
In, the DC level of the original input signal A ′ is increased or decreased (shifted) by this unit amount (step 405). At this time, the noise amount of the video signal E is detected by the blanking extracting means 12 and the noise amount detecting means 13 in FIG. 1 as described above (step 406). First, the pre-noise amount storage means 1
9, the initial noise amount stored in step 9 is set as the previous noise amount H ', and these magnitude comparison judgments are performed (step 40).
7) A new shift amount control signal I is generated from the determination result and the like as described above.

In this case, since the shift amount control signal I is "0" to increase the DC shift amount, when the input noise amount H becomes larger than the initial noise amount H 'by the DC shift, the DC shift amount is increased. To reduce
A new shift amount control signal I of "1" is generated (step 408). Conversely, when the input noise amount H is smaller than the initial noise amount H ′, a newly generated shift amount control signal I is generated in order to further increase the DC shift amount.
Is also set to "0" (step 409). As a result, FIG.
The level of the DC signal J output from the shift amount setting means 16 in the step (1) increases by a unit amount.

The pre-noise amount storage means 19 in FIG.
Is rewritten with the input noise amount H, and the process returns to step 405 to perform the same processing operation with the new shift amount control signal I. When the shift amount control signal I is “1” and the level of the DC signal J is reduced by the unit amount, and the input noise amount H is smaller than the previous noise amount H ′ (step 407), The new shift amount control signal I is kept at "1" (step 408), and if the input noise amount H is larger than the previous noise amount H '(step 407), the new shift amount control signal I is changed to "1". It is set to "0" (step 409).

Thereafter, the operation from step 405 is performed every time a new shift amount control signal I is generated, whereby the DC shift amount of the original input signal A 'is set to an optimum value or a value close to the optimum value. .

FIG. 5 shows the shift amount control means 14 in FIG.
24 is a block diagram showing another specific example, wherein 24 is a difference detecting means, 25 is a gate signal generating means, and 26 is a gate means, and the portions and signals corresponding to those in FIG. Is omitted.

In the first embodiment, the DC shift amount is always increased or decreased by a unit amount. For this reason, even if the DC shift amount approaches the optimum value, if there is a difference between the current noise amount H and the previous noise amount H ′,
The increase / decrease of the DC shift amount is repeated. Therefore, even in a state where the noise component of the original input signal A is most effectively suppressed, the DC level of the obtained video signal E is constantly increased and decreased, and the displayed image flickers and the image quality is reduced. Will deteriorate. This specific example shown in FIG. 5 is designed to reduce such image quality fluctuation.

In FIG. 5, the difference M between the input noise amount H and the previous noise amount H 'stored in the previous noise amount storage means 19 is detected by the difference detection means 24, and this difference M is set in advance. When it is within a sufficiently narrow range, the gate signal generating means 25 generates the gate signal N.

On the other hand, the control signal generating means 22 and the output terminal 2
The gate means 26 is turned off during the signal period of the gate signal N to cut off the connection between the control signal generating means 22 and the output terminal 23. It is not supplied to the shift amount setting means 16 in FIG. Accordingly, the DC signal J output from the period shift amount setting means 16 is fixed at the level immediately before this period.

As described above, in this specific example, even if the noise amount changes, the difference between the input noise amount H and the previous noise amount is sufficiently small, and the original input signal A 'is almost the optimum DC shift amount. When the noise component is compressed and the noise component is compressed, the above state is set as a dead zone against a change in the noise amount, and the DC shift amount of the original input signal A ′ is fixed. As a result, even when the original input signal A ′ is DC-shifted with almost the optimum DC shift amount, the level of the video signal fluctuates due to the change in the DC shift amount, thereby preventing the image quality from deteriorating due to the image flickering. Can be.

Even if the gate means 26 is turned off, the shift amount control signal I output from the control signal generation means 22 is supplied to the change direction comparison means 21.
The change direction comparing means 21 is always operated normally.

FIG. 6 shows the shift amount control means 14 in FIG.
6 is a flowchart showing the operation of the first embodiment when the specific example shown in FIG. 5 is used.

This operation is basically the same as the operation shown in FIG. 4, except that step 601 of the operation by the difference detecting means 24, the gate signal generating means 25 and the gate means 26 in FIG. 5 is added. This is different from the operation shown in FIG.

That is, in FIG. 6, between the steps 406 and 407, the noise amount detecting means 13 in FIG.
Thus, a step 601 for determining whether or not the difference between the currently detected noise amount and the previously acquired noise amount is within the predetermined range is added. When the difference is within the predetermined range, step 406
When it is out of the range, the process proceeds to step 407.

Thus, it is possible to prevent the DC shift amount from being excessively changed in accordance with the repetition of the minute change of the noise amount, and to cause the flicker caused by the change of the SN ratio of the video signal due to the repetition of the increase and decrease of the DC shift amount. Can be reduced.

FIG. 7 shows a second embodiment of the SN ratio improving apparatus according to the present invention.
FIG. 27 is a block diagram showing an embodiment of the present invention, wherein 27 is a shift amount table means, and 28 is an input terminal. Portions corresponding to those in FIG.

In the figure, the shift amount table means 27
Stores a table (shift amount table) of values near the optimal DC shift amount (initial DC shift amount) for each noise amount included in the original input signal A ′. Further, a start-up signal P is input from the input terminal 28 when the system is started by switching channels or turning on power. From this shift amount table means 27, an initial DC shift amount Q corresponding to the noise amount H sequentially output from the noise amount detection means 13 is read out and output. , The initial DC shift amount Q can be set.

Next, the operation of the second embodiment will be described.

When the start-up signal P is input from the input terminal 28 at the time of channel switching or system start-up, first, the shift amount setting means 16 sets the DC shift amount to 0 (DC signal J =
0) is set, and the original input signal A 'is passed through the adding means 15 without being DC-shifted. When the noise amount H (that is, the noise amount of the original input signal A ′) at this time is detected by the noise amount detection means 13, the noise amount H is
The shift amount control signal 14 is supplied to the shift amount control means 14 to generate the shift amount control signal I, as described above. At the same time, the shift amount control signal I is also supplied to the shift amount table means 27, so that the initial shift amount Q is read. At this time, since the start-up signal P is input from the input terminal 28, the shift amount setting means 16
Instead of the shift amount control signal I from the controller, an initial shift amount Q from the shift amount table means 27 is fetched, and a DC signal J based on this and the shift direction signal C from the shift direction determination means 11 is generated. Therefore, the original input signal A ′ is
The adding means 15 performs DC shift by the initial shift amount Q.

After that, when the rising signal P from the input terminal 28 is released, the shift amount setting means 16 takes in the shift amount control signal I from the shift amount control means 14, and the set initial value is set. The same operation as in the first embodiment is performed in which the shift amount Q is increased or decreased by a unit amount for each shift amount control signal I.

As described above, in the second embodiment, as the initial operation, the DC shift operation is started by the DC shift means 6 from a value near the optimum shift amount, so that the shift to the optimum shift amount is quickly performed. Therefore, when the channel is switched or the system is activated, a video signal with an improved SN is obtained immediately.

[0059]

As described above, according to the present invention,
Feedback control is performed to control the DC shift amount of the input video signal for reducing the noise in accordance with a change in the noise amount detected from the input video signal whose noise amount has been reduced by the DC shift. Therefore, the optimum DC shift amount is always automatically set, and the SN ratio is improved most efficiently.

According to the present invention, an optimum DC shift amount corresponding to the noise amount in the input video signal is prepared in advance,
Since the DC shift of the input video signal can be performed using this, the shift amount of the DC shift process can be reduced by using the optimum DC shift amount at the time of power-on or channel switching, for example. Optimization can be expedited.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a first embodiment of an SN ratio improving apparatus according to the present invention.

FIG. 2 is a block diagram showing a specific example of a shift amount control unit in FIG.

3 is a diagram for explaining types of shift amount control signals formed in the specific example shown in FIG. 2;

FIG. 4 is a flowchart showing the operation of the first embodiment shown in FIG. 1 when the specific example shown in FIG. 2 is used as the shift amount control means.

FIG. 5 is a block diagram showing another specific example of the shift amount control means in FIG. 1;

6 is a flowchart showing the operation of the first embodiment shown in FIG. 1 when the specific example shown in FIG. 5 is used as the shift amount control means.

FIG. 7 is a block diagram showing a second embodiment of the SN ratio improving apparatus according to the present invention.

FIG. 8 is a block diagram showing an example of a conventional SN ratio improving device.

[Explanation of symbols]

 Reference Signs List 3 delay means 4 reference signal generation means 6 DC shift means 7 output terminal 9, 10 input terminal 11 shift direction determination means 12 blanking extraction means 13 noise amount detection means 14 shift amount control means 15 addition means 16 shift amount setting means 17 switch Means 18 Input terminal 19 Pre-noise amount storage means 20 Size comparison means 21 Change direction comparison means 22 Control signal generation means 23 Output terminal 24 Difference detection means 25 Gate signal generation means 26 Gate means 27 Shift amount table means 28, 29 Input terminal 30 switch

────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] March 5, 1997

[Procedure amendment 1]

[Document name to be amended] Drawing

[Correction target item name] Fig. 5

[Correction method] Change

[Correction contents]

FIG. 5

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Mitsuo Nakajima 292 Yoshida-cho, Totsuka-ku, Yokohama-shi, Kanagawa Prefecture Inside the multimedia system development headquarters of Hitachi, Ltd. 292 Hitachi Multimedia Systems Development Headquarters Co., Ltd. (72) Inventor Yuichi Ninomiya 1-10-11 Kinuta, Setagaya-ku, Tokyo Inside Japan Broadcasting Corporation Broadcasting Research Institute (72) Inventor Yoshinori Izumi Setagaya-ku, Tokyo 1-10-11 Kinuta, Japan Broadcasting Corporation Broadcasting Research Institute (72) Inventor Masahide Naemura 1-10-11 Kinuta, Setagaya-ku, Tokyo Japan Broadcasting Research Institute, Japan Broadcasting Research Institute (72) Inventor Jun Fukuda Tokyo 1-10-11 Kinuta, Setagaya-ku Japan Broadcasting Corporation Broadcasting Research Institute (72) Inventor Koichi Yamaguchi Tokyo Setagaya Kinuta chome No. 10, No. 11, Japan Broadcasting Association broadcasting technology within the Institute (72) inventor Koshi Seiichi Shibuya-ku, Tokyo Jinnan chome No. 2 No. 1 Japan Broadcasting Association of Broadcasting Center in

Claims (5)

[Claims] 1. A means for generating a noise-free video signal predicted from the input video signal and outputting it as a reference signal, and comparing the input video signal with the reference signal to perform a DC shift of the input video signal. Means for determining a direction, means for generating a control signal for determining a DC shift amount of the input video signal, and means for DC-shifting the input video signal in the DC shift direction by a DC shift amount determined by the control signal. In the SN ratio improving apparatus, the control signal indicates a noise amount included in a video signal obtained by DC-shifting the input signal, and the DC shift amount of the input video signal is determined according to the noise amount. An SN ratio improving device characterized by the above-mentioned. 2. The SN ratio improving device according to claim 1, wherein the noise amount indicated by the control signal is a noise amount in a blanking period portion of the DC-shifted video signal. 3. The SN ratio improving device according to claim 1, wherein the DC shift amount is controlled according to the increase or decrease of the noise amount. 4. The signal-to-noise ratio according to claim 3, wherein the DC shift amount is fixed to a value immediately before the period when the amount of change in the noise amount is within a predetermined range. Improvement device. 5. The method according to claim 1, wherein an optimum D shift is estimated as the DC shift amount at the time of starting.
An SN ratio improving device, wherein an initial DC shift amount having a value near the C shift amount is provided.